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18 EXPERIMENTAL_CERTNAMES is withdrawn.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.
100 TL/01 When running the test suite, the README says that variables such as
101 no_msglog_check are global and can be placed anywhere in a specific
102 test's script, however it was observed that placement needed to be near
103 the beginning for it to behave that way. Changed the runtest perl
104 script to read through the entire script once to detect and set these
105 variables, reset to the beginning of the script, and then run through
106 the script parsing/test process like normal.
108 TL/02 The BSD's have an arc4random API. One of the functions to induce
109 adding randomness was arc4random_stir(), but it has been removed in
110 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
111 function when detected.
113 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
114 cause callback expansion.
116 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
117 syntax errors in an expansion can be treated as a string instead of
118 logging or causing an error, due to the internal use of bool_lax
119 instead of bool when processing it.
121 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
122 server certificates when making smtp deliveries.
124 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
126 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
128 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
130 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
131 Merged patch from Sebastian Wiedenroth.
133 JH/05 Fix results-pipe from transport process. Several recipients, combined
134 with certificate use, exposed issues where response data items split
135 over buffer boundaries were not parsed properly. This eventually
136 resulted in duplicates being sent. This issue only became common enough
137 to notice due to the introduction of conection certificate information,
138 the item size being so much larger. Found and fixed by Wolfgang Breyha.
140 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
141 size buffer was used, resulting in syntax errors when an expansion
144 JH/07 Add support for directories of certificates when compiled with a GnuTLS
145 version 3.3.6 or later.
147 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
148 is EXPERIMENTAL_EVENT, the main-configuration and transport options
149 both become "event_action", the variables become $event_name, $event_data
150 and $event_defer_errno. There is a new variable $verify_mode, usable in
151 routers, transports and related events. The tls:cert event is now also
152 raised for inbound connections, if the main configuration event_action
155 TL/06 In test suite, disable OCSP for old versions of openssl which contained
156 early OCSP support, but no stapling (appears to be less than 1.0.0).
158 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
159 server certificate names available under the smtp transport option
160 "tls_verify_cert_hostname" now do not permit multi-component wildcard
163 JH/10 Time-related extraction expansions from certificates now use the main
164 option "timezone" setting for output formatting, and are consistent
165 between OpenSSL and GnuTLS compilations. Bug 1541.
167 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
168 encoded parameter in the incoming message. Bug 1558.
170 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
171 include certificate info, eximon was claiming there were spoolfile
174 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
176 JH/14 Log delivery-related information more consistently, using the sequence
177 "H=<name> [<ip>]" wherever possible.
179 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
180 are problematic for Debian distribution, omit them from the release
183 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
185 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
188 JH/17 Fix a null-indirection in certextract expansions when a nondefault
189 output list separator was used.
194 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
195 checkers that were complaining about end of non-void function with no
198 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
199 This was a regression intruduced in 4.83 by another bugfix.
201 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
203 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
204 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
210 TF/01 Correctly close the server side of TLS when forking for delivery.
212 When a message was received over SMTP with TLS, Exim failed to clear up
213 the incoming connection properly after forking off the child process to
214 deliver the message. In some situations the subsequent outgoing
215 delivery connection happened to have the same fd number as the incoming
216 connection previously had. Exim would try to use TLS and fail, logging
217 a "Bad file descriptor" error.
219 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
220 utilities have not been installed.
222 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
223 temporary space as the ACL may create new global variables.
225 TL/01 LDAP support uses per connection or global context settings, depending
226 upon the detected version of the libraries at build time.
228 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
229 to extract and use the src ip:port in logging and expansions as if it
230 were a direct connection from the outside internet. PPv2 support was
231 updated based on HAProxy spec change in May 2014.
233 JH/02 Add ${listextract {number}{list}{success}{fail}}.
235 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
236 Properly escape header and check for NULL return.
238 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
241 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
243 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
244 characters in header names, implemented as a verify condition.
245 Contributed by Michael Fischer v. Mollard.
247 TL/05 Rename SPF condition results err_perm and err_temp to standardized
248 results permerror and temperror. Previous values are deprecated but
249 still accepted. In a future release, err_perm and err_temp will be
250 completely removed, which will be a backward incompatibility if the
251 ACL tests for either of these two old results. Patch contributed by
252 user bes-internal on the mailing list.
254 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
256 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
257 selectors, in both main and reject logs.
259 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
262 JH/07 Add malware type "sock" for talking to simple daemon.
264 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
266 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
267 routers/transports under cutthrough routing.
269 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
270 numbers. Touch up "bool" conditional to keep the same definition.
272 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
274 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
276 JH/12 Expand items in router/transport headers_add or headers_remove lists
277 individually rather than the list as a whole. Bug 1452.
279 Required for reasonable handling of multiple headers_ options when
280 they may be empty; requires that headers_remove items with embedded
281 colons must have them doubled (or the list-separator changed).
283 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
284 view the policy declared in the DMARC record. Currently, $dmarc_status
285 is a combined value of both the record presence and the result of the
288 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
290 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
291 dnslookup router and the smtp transport (applying to the forward
294 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
295 of ldap servers used for a specific lookup. Patch provided by Heiko
298 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
299 New variable $lookup_dnssec_authenticated for observability.
301 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
302 Patch submitted by Lars Timman.
304 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
306 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
307 Requires trusted mode and valid format message id, aborts otherwise.
308 Patch contributed by Heiko Schlichting.
310 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
311 certextract with support for various fields. Bug 1358.
313 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
314 is requested by default, modifiable by smtp transport option
317 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
318 operate on certificate variables to give certificate fingerprints
319 Also new ${sha256:cert_variable}.
321 JH/23 The PRDR feature is moved from being Experimental into the mainline.
323 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
324 Christian Aistleitner.
326 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
328 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
329 file. Patch from Wolfgang Breyha.
331 JH/25 Expand the coverage of the delivery $host and $host_address to
332 client authenticators run in verify callout. Bug 1476.
334 JH/26 Port service names are now accepted for tls_on_connect_ports, to
335 align with daemon_smtp_ports. Bug 72.
337 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
338 support and error reporting did not work properly.
340 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
341 and is readable. Patch from Andrew Colin Kissa.
343 TL/14 Enhance documentation of ${run expansion and how it parses the
344 commandline after expansion, particularly in the case when an
345 unquoted variable expansion results in an empty value.
347 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
349 PP/02 Fix internal collision of T_APL on systems which support RFC3123
350 by renaming away from it. Addresses GH issue 15, reported by
353 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
355 TL/15 SECURITY: prevent double expansion in math comparison functions
356 (can expand unsanitized data). Not remotely exploitable.
363 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
365 PP/02 Make -n do something, by making it not do something.
366 When combined with -bP, the name of an option is not output.
368 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
371 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
372 $sender_host_name and config options to manage this, and basic check
375 PP/05 DSCP support for outbound connections and control modifier for inbound.
377 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
378 (Only plugin which currently uses this is kerberos4, which nobody should
379 be using, but we should make it available and other future plugins might
380 conceivably use it, even though it would break NAT; stuff *should* be
381 using channel bindings instead).
383 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
384 name; added for Sendmail compatibility; requires admin caller.
385 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
386 just ignore it); requires trusted caller.
387 Also parse but ignore: -Ac -Am -X<logfile>
390 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
392 TL/02 Add +smtp_confirmation as a default logging option.
394 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
395 Patch by Magnus Holmgren from 2007-02-20.
397 TL/04 Bugzilla 1281 - Spec typo.
398 Bugzilla 1283 - Spec typo.
399 Bugzilla 1290 - Spec grammar fixes.
401 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
403 TL/06 Add Experimental DMARC support using libopendmarc libraries.
405 TL/07 Fix an out of order global option causing a segfault. Reported to dev
406 mailing list by by Dmitry Isaikin.
408 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
410 JH/02 Support "G" suffix to numbers in ${if comparisons.
412 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
414 NM/01 Bugzilla 1197 - Spec typo
415 Bugzilla 1196 - Spec examples corrections
417 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
419 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
420 gnutls_enable_pkcs11, but renamed to more accurately indicate its
423 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
424 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
426 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
427 "acl {{name}{arg}...}", and optional args on acl condition
430 JH/05 Permit multiple router/transport headers_add/remove lines.
432 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
434 JH/07 Avoid using a waiting database for a single-message-only transport.
435 Performance patch from Paul Fisher. Bugzilla 1262.
437 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
440 JH/09 Add $headers_added variable, with content from use of ACL modifier
441 add_header (but not yet added to the message). Bugzilla 199.
443 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
444 Pulled from Bugzilla 817 by Wolfgang Breyha.
446 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
448 (nb: this is the same fix as in Exim 4.80.1)
450 JH/11 Add A= logging on delivery lines, and a client_set_id option on
453 JH/12 Add optional authenticated_sender logging to A= and a log_selector
456 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
458 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
459 advertise SMTP AUTH mechanism to us, instead of a generic
460 protocol violation error. Also, make Exim more robust to bad
461 data from the Dovecot auth socket.
463 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
465 When a queue runner is handling a message, Exim first routes the
466 recipient addresses, during which it prunes them based on the retry
467 hints database. After that it attempts to deliver the message to
468 any remaining recipients. It then updates the hints database using
471 So if a recipient address works intermittently, it can get repeatedly
472 deferred at routing time. The retry hints record remains fresh so the
473 address never reaches the final cutoff time.
475 This is a fairly common occurrence when a user is bumping up against
476 their storage quota. Exim had some logic in its local delivery code
477 to deal with this. However it did not apply to per-recipient defers
478 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
480 This change adds a proper retry rule check during routing so that the
481 final cutoff time is checked against the message's age. We only do
482 this check if there is an address retry record and there is not a
483 domain retry record; this implies that previous attempts to handle
484 the address had the retry_use_local_parts option turned on. We use
485 this as an approximation for the destination being like a local
486 delivery, as in LMTP.
488 I suspect this new check makes the old local delivery cutoff check
489 redundant, but I have not verified this so I left the code in place.
491 TF/02 Correct gecos expansion when From: is a prefix of the username.
493 Test 0254 submits a message to Exim with the header
497 When I ran the test suite under the user fanf2, Exim expanded
498 the header to contain my full name, whereas it should have added
499 a Resent-Sender: header. It erroneously treats any prefix of the
500 username as equal to the username.
502 This change corrects that bug.
504 GF/01 DCC debug and logging tidyup
505 Error conditions log to paniclog rather than rejectlog.
506 Debug lines prefixed by "DCC: " to remove any ambiguity.
508 TF/03 Avoid unnecessary rebuilds of lookup-related code.
510 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
511 Bug spotted by Jeremy Harris; was flawed since initial commit.
512 Would have resulted in OCSP responses post-SNI triggering an Exim
513 NULL dereference and crash.
515 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
517 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
518 Bug detection, analysis and fix by Samuel Thibault.
519 Bugzilla 1331, Debian bug #698092.
521 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
523 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
524 Server implementation by Todd Lyons, client by JH.
525 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
526 config variable "prdr_enable" controls whether the server
527 advertises the facility. If the client requests PRDR a new
528 acl_data_smtp_prdr ACL is called once for each recipient, after
529 the body content is received and before the acl_smtp_data ACL.
530 The client is controlled by bolth of: a hosts_try_prdr option
531 on the smtp transport, and the server advertisement.
532 Default client logging of deliveries and rejections involving
533 PRDR are flagged with the string "PRDR".
535 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
536 fclose(). Diagnosis by Todd Lyons.
538 PP/17 Update configure.default to handle IPv6 localhost better.
539 Patch by Alain Williams (plus minor tweaks).
542 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
543 This is now consistent with GnuTLS, and is now documented: the
544 previous undocumented portable approach to treating the option as
545 unset was to force an expansion failure. That still works, and
546 an empty string is now equivalent.
548 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
549 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
550 not performing validation itself.
552 PP/20 Added force_command boolean option to pipe transport.
553 Patch from Nick Koston, of cPanel Inc.
555 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
558 TF/04 Added udpsend ACL modifer and hexquote expansion operator
560 PP/21 Fix eximon continuous updating with timestamped log-files.
561 Broken in a format-string cleanup in 4.80, missed when I repaired the
562 other false fix of the same issue.
563 Report and fix from Heiko Schlichting.
566 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
567 Report from Prashanth Katuri.
569 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
570 It's SecureTransport, so affects any MacOS clients which use the
571 system-integrated TLS libraries, including email clients.
573 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
574 using a MIME ACL for non-SMTP local injection.
575 Report and assistance in diagnosis by Warren Baker.
577 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
579 JH/16 Fix comparisons for 64b. Bugzilla 1385.
581 TL/09 Add expansion variable $authenticated_fail_id to keep track of
582 last id that failed so it may be referenced in subsequent ACL's.
584 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
587 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
588 ldap library initialization, allowing self-signed CA's to be
589 used. Also properly sets require_cert option later in code by
590 using NULL (global ldap config) instead of ldap handle (per
591 session). Bug diagnosis and testing by alxgomz.
593 TL/12 Enhanced documentation in the ratelimit.pl script provided in
594 the src/util/ subdirectory.
596 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
597 renamed to Transport Post Delivery Action by Jeremy Harris, as
600 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
601 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
602 redis_servers = needs to be configured which will be used by the redis
603 lookup. Patch from Warren Baker, of The Packet Hub.
605 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
607 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
608 hostname or reverse DNS when processing a host list. Used suggestions
609 from multiple comments on this bug.
611 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
613 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
614 Missed a few lines, added it to make the runtest require no keyboard
617 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
618 contains upper case chars. Make router use caseful_local_part.
620 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
621 support when GnuTLS has been built with p11-kit.
627 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
629 This, or similar/improved, will also be change PP/11 of 4.82.
635 PP/01 Handle short writes when writing local log-files.
636 In practice, only affects FreeBSD (8 onwards).
637 Bugzilla 1053, with thanks to Dmitry Isaikin.
639 NM/01 Bugzilla 949 - Documentation tweak
641 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
644 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
646 PP/02 Implemented gsasl authenticator.
648 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
650 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
651 `pkg-config foo` for cflags/libs.
653 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
654 with rest of GSASL and with heimdal_gssapi.
656 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
657 `pkg-config foo` for cflags/libs for the TLS implementation.
659 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
660 properties get this fed in as external SSF. A number of robustness
661 and debugging improvements to the cyrus_sasl authenticator.
663 PP/08 cyrus_sasl server now expands the server_realm option.
665 PP/09 Bugzilla 1214 - Log authentication information in reject log.
666 Patch by Jeremy Harris.
668 PP/10 Added dbmjz lookup type.
670 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
672 PP/12 MAIL args handles TAB as well as SP, for better interop with
673 non-compliant senders.
674 Analysis and variant patch by Todd Lyons.
676 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
677 Bug report from Lars Müller <lars@samba.org> (via SUSE),
678 Patch from Dirk Mueller <dmueller@suse.com>
680 PP/13 tls_peerdn now print-escaped for spool files.
681 Observed some $tls_peerdn in wild which contained \n, which resulted
682 in spool file corruption.
684 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
685 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
686 or write after TLS renegotiation, which otherwise led to messages
689 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
690 as a tracking header (ie: a signed header comes before the signature).
691 Patch from Wolfgang Breyha.
693 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
694 comma-sep list; embedded commas doubled.
696 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
698 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
700 Report and patch from Dmitry Banschikov.
702 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
703 Removed SSL_clear() after SSL_new() which led to protocol negotiation
704 failures. We appear to now support TLS1.1+ with Exim.
706 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
707 lets Exim select keys and certificates based upon TLS SNI from client.
708 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
709 before an outbound SMTP session. New log_selector, +tls_sni.
711 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
712 NULL dereference. Report and patch from Alun Jones.
714 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
715 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
716 Not seeing resolver debug output on NetBSD, but suspect this is a
717 resolver implementation change.
719 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
720 Left warnings. Added "eximon gdb" invocation mode.
722 PP/21 Defaulting "accept_8bitmime" to true, not false.
724 PP/22 Added -bw for inetd wait mode support.
726 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
727 locate the relevant includes and libraries. Made this the default.
729 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
730 Bugzilla 1246, report and most of solution from Tomasz Kusy.
732 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
733 This may cause build issues on older platforms.
735 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
736 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
737 gnutls_require_mac & gnutls_require_protocols (no longer supported).
738 Added SNI support via GnuTLS too.
739 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
741 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
743 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
744 Applied second patch from Janne, implementing suggestion to default
745 multiple-strings-in-record handling to match SPF spec.
747 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
749 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
750 read-only, out of scope).
751 Patch from Wolfgang Breyha, report from Stuart Northfield.
753 PP/29 Fix three issues highlighted by clang analyser static analysis.
754 Only crash-plausible issue would require the Cambridge-specific
755 iplookup router and a misconfiguration.
756 Report from Marcin Mirosław.
758 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
760 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
761 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
762 As part of this, removing so much warning spew let me fix some minor
763 real issues in debug logging.
765 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
766 assignment on my part. Fixed.
768 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
769 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
770 Janne Snabb (who went above and beyond: thank you).
772 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
773 string otherwise requires a connection and a bunch more work and it's
774 relatively easy to get wrong. Should also expose TLS library linkage
777 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
778 64-bit ${eval} (JH/03).
780 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
781 GNU libc to support some of the 64-bit stuff, should not lead to
782 conflicts. Defined before os.h is pulled in, so if a given platform
783 needs to override this, it can.
785 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
786 protection layer was required, which is not implemented.
787 Bugzilla 1254, patch from Wolfgang Breyha.
789 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
790 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
791 tls_dhparam take prime identifiers. Also unbreak combination of
792 OpenSSL+DH_params+TLSSNI.
794 PP/39 Disable SSLv2 by default in OpenSSL support.
800 PP/01 Solaris build fix for Oracle's LDAP libraries.
801 Bugzilla 1109, patch from Stephen Usher.
803 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
805 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
808 TF/02 Fix a couple more cases where we did not log the error message
809 when unlink() failed. See also change 4.74-TF/03.
811 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
812 lock up or crash if it happened to be inside a call to libc when it
813 got a SIGUSR1 from exiwhat.
815 The SIGUSR1 handler appends the current process status to the process
816 log which is later printed by exiwhat. It used to use the general
817 purpose logging code to do this, but several functions it calls are
818 not safe for signals.
820 The new output code in the SIGUSR1 handler is specific to the process
821 log, and simple enough that it's easy to inspect for signal safety.
822 Removing some special cases also simplifies the general logging code.
823 Removing the spurious timestamps from the process log simplifies
826 TF/04 Improved ratelimit ACL condition.
828 The /noupdate option has been deprecated in favour of /readonly which
829 has clearer semantics. The /leaky, /strict, and /readonly update modes
830 are mutually exclusive. The update mode is no longer included in the
831 database key; it just determines when the database is updated. (This
832 means that when you upgrde Exim will forget old rate measurements.)
834 Exim now checks that the per_* options are used with an update mode that
835 makes sense for the current ACL. For example, when Exim is processing a
836 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
837 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
838 must specify per_mail/readonly. If you omit the update mode it defaults to
839 /leaky where that makes sense (as before) or /readonly where required.
841 The /noupdate option is now undocumented but still supported for
842 backwards compatibility. It is equivalent to /readonly except that in
843 ACLs where /readonly is required you may specify /leaky/noupdate or
844 /strict/noupdate which are treated the same as /readonly.
846 A useful new feature is the /count= option. This is a generalization
847 of the per_byte option, so that you can measure the throughput of other
848 aggregate values. For example, the per_byte option is now equivalent
849 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
851 The per_rcpt option has been generalized using the /count= mechanism
852 (though it's more complicated than the per_byte equivalence). When it is
853 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
854 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
855 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
856 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
857 Note that using per_rcpt with a non-readonly update mode in more than
858 one ACL will cause the recipients to be double-counted. (The per_mail
859 and per_byte options don't have this problem.)
861 The handling of very low rates has changed slightly. If the computed rate
862 is less than the event's count (usually one) then this event is the first
863 after a long gap. In this case the rate is set to the same as this event's
864 count, so that the first message of a spam run is counted properly.
866 The major new feature is a mechanism for counting the rate of unique
867 events. The new per_addr option counts the number of different
868 recipients that someone has sent messages to in the last time period. It
869 behaves like per_rcpt if all the recipient addresses are different, but
870 duplicate recipient addresses do not increase the measured rate. Like
871 the /count= option this is a general mechanism, so the per_addr option
872 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
873 example, measure the rate that a client uses different sender addresses
874 with the options per_mail/unique=$sender_address. There are further
875 details in the main documentation.
877 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
879 TF/06 Removed a few PCRE remnants.
881 TF/07 Automatically extract Exim's version number from tags in the git
882 repository when doing development or release builds.
884 PP/02 Raise smtp_cmd_buffer_size to 16kB.
885 Bugzilla 879. Patch from Paul Fisher.
887 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
888 Heavily based on revision 40f9a89a from Simon Arlott's tree.
891 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
893 PP/05 Variable $av_failed, true if the AV scanner deferred.
894 Bugzilla 1078. Patch from John Horne.
896 PP/06 Stop make process more reliably on build failure.
897 Bugzilla 1087. Patch from Heiko Schlittermann.
899 PP/07 Make maildir_use_size_file an _expandable_ boolean.
900 Bugzilla 1089. Patch from Heiko Schlittermann.
902 PP/08 Handle ${run} returning more data than OS pipe buffer size.
903 Bugzilla 1131. Patch from Holger Weiß.
905 PP/09 Handle IPv6 addresses with SPF.
906 Bugzilla 860. Patch from Wolfgang Breyha.
908 PP/10 GnuTLS: support TLS 1.2 & 1.1.
910 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
913 PP/11 match_* no longer expand right-hand-side by default.
914 New compile-time build option, EXPAND_LISTMATCH_RHS.
915 New expansion conditions, "inlist", "inlisti".
917 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
919 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
921 PP/14 fix log_write() format string regression from TF/03.
922 Bugzilla 1152. Patch from Dmitry Isaikin.
928 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
930 PP/02 Harmonised TLS library version reporting; only show if debugging.
931 Layout now matches that introduced for other libraries in 4.74 PP/03.
933 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
935 PP/04 New "dns_use_edns0" global option.
937 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
940 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
941 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
943 TK/01 Updated PolarSSL code to 0.14.2.
944 Bugzilla 1097. Patch from Andreas Metzler.
946 PP/07 Catch divide-by-zero in ${eval:...}.
949 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
952 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
953 format-string attack -- SECURITY: remote arbitrary code execution.
955 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
956 time unintentionally subject to list matching rules, letting the header
957 cause arbitrary Exim lookups (of items which can occur in lists, *not*
958 arbitrary string expansion). This allowed for information disclosure.
960 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
961 INT_MIN/-1 -- value coerced to INT_MAX.
967 NM/01 Workround for PCRE version dependency in version reporting
970 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
971 This fixes portability to compilers other than gcc, notably
972 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
974 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
975 makefiles for portability to HP-UX and POSIX correctness.
977 PP/01 Permit LOOKUP_foo enabling on the make command-line.
978 Also via indented variable definition in the Makefile.
979 (Debugging by Oliver Heesakkers).
981 PP/02 Restore caching of spamd results with expanded spamd_address.
982 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
984 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
985 Improves build reliability. Fix from: Frank Elsner
987 NM/02 Fix wide character breakage in the rfc2047 coding
988 Fixes bug 1064. Patch from Andrey N. Oktyabrski
990 NM/03 Allow underscore in dnslist lookups
991 Fixes bug 1026. Patch from Graeme Fowler
993 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
994 Code patches from Adam Ciarcinski of NetBSD.
996 NM/04 Fixed exiqgrep to cope with mailq missing size issue
999 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1000 is logged, to avoid truncation. Patch from John Horne.
1002 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1003 Patch from Jakob Hirsch.
1005 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1006 SQL string expansion failure details.
1007 Patch from Andrey Oktyabrski.
1009 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1010 Patch from Simon Arlott.
1012 PP/09 New lookups functionality failed to compile on old gcc which rejects
1013 extern declarations in function scope.
1014 Patch from Oliver Fleischmann
1016 PP/10 Use sig_atomic_t for flags set from signal handlers.
1017 Check getgroups() return and improve debugging.
1018 Fixed developed for diagnosis in bug 927 (which turned out to be
1021 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1022 Patch from Mark Zealey.
1024 PP/12 Bugzilla 1056: Improved spamd server selection.
1025 Patch from Mark Zealey.
1027 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1028 Based on patch from Heiko Schlittermann.
1030 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1031 Patch from Uwe Doering, sign-off by Michael Haardt.
1033 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1034 variable declaration deep within a block. Bug and patch from
1037 PP/15 lookups-Makefile IRIX compatibilty coercion.
1039 PP/16 Make DISABLE_DKIM build knob functional.
1041 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1042 Patch by Simon Arlott
1044 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1045 variable argument macros. Our copy now differs from upstream.
1051 TF/01 Failure to get a lock on a hints database can have serious
1052 consequences so log it to the panic log.
1054 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1055 controlled using the smtp_confirmation log selector.
1057 TF/03 Include the error message when we fail to unlink a spool file.
1059 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1060 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1061 for maintaining out-of-tree patches for some time.
1063 PP/01 Bugzilla 139: Documentation and portability issues.
1064 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1065 Handle per-OS dynamic-module compilation flags.
1067 PP/02 Let /dev/null have normal permissions.
1068 The 4.73 fixes were a little too stringent and complained about the
1069 permissions on /dev/null. Exempt it from some checks.
1070 Reported by Andreas M. Kirchwitz.
1072 PP/03 Report version information for many libraries, including
1073 Exim version information for dynamically loaded libraries. Created
1074 version.h, now support a version extension string for distributors
1075 who patch heavily. Dynamic module ABI change.
1077 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1078 privilege escalation vulnerability whereby the Exim run-time user
1079 can cause root to append content of the attacker's choosing to
1082 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1085 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1086 If dropping privileges for untrusted macros, we disabled normal logging
1087 on the basis that it would fail; for the Exim run-time user, this is not
1088 the case, and it resulted in successful deliveries going unlogged.
1089 Fixed. Reported by Andreas Metzler.
1095 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1096 only prepend for the Resent-* case. Fixes regression introduced in
1097 Exim 4.70 by NM/22 for Bugzilla 607.
1099 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1100 increasing numbers of administrators be bitten by this.
1102 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1104 PP/03 Bugzilla 994: added openssl_options main configuration option.
1106 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1108 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1110 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1112 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1113 without a peer certificate, leading to a segfault because of an
1114 assumption that peers always have certificates. Be a little more
1115 paranoid. Problem reported by Martin Tscholak.
1117 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1118 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1119 NB: ClamAV planning to remove STREAM in "middle of 2010".
1120 CL also introduces -bmalware, various -d+acl logging additions and
1121 more caution in buffer sizes.
1123 PP/09 Implemented reverse_ip expansion operator.
1125 PP/10 Bugzilla 937: provide a "debug" ACL control.
1127 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1129 PP/12 Bugzilla 973: Implement --version.
1131 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1133 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1135 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1137 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1138 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1139 ignore trailing whitespace.
1141 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1143 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1146 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1147 Notification from Dr Andrew Aitchison.
1149 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1150 ExtendedDetectionInfo response format.
1151 Notification from John Horne.
1153 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1156 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1157 XSL and documented dependency on system catalogs, with examples of how
1160 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1163 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1164 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1165 configuration file which is writeable by the Exim user or group.
1167 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1168 of configuration files to cover files specified with the -C option if
1169 they are going to be used with root privileges, not just the default
1172 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1173 option (effectively making it always true).
1175 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1176 files to be used while preserving root privileges.
1178 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1179 that rogue child processes cannot use them.
1181 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1182 run-time user, instead of root.
1184 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1185 Exim run-time user without dropping privileges.
1187 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1188 result string, instead of calling string_vformat() twice with the same
1191 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1192 for other users. Others should always drop root privileges if they use
1193 -C on the command line, even for a whitelisted configure file.
1195 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1197 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1203 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1204 $data_path, and $header_path variables; fixed documentation bugs and
1207 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1208 exipick to access non-standard spools, including the "frozen" queue
1211 NM/01 Bugzilla 965: Support mysql stored procedures.
1212 Patch from Alain Williams
1214 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1216 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1217 Patch from Andreas Metzler
1219 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1220 Patch from Kirill Miazine
1222 NM/05 Bugzilla 671: Added umask to procmail example.
1224 JJ/03 installed exipick 20100323.0, fixing doc bug
1226 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1227 directory. Notification and patch from Dan Rosenberg.
1229 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1231 TK/02 Improve log output when DKIM signing operation fails.
1233 MH/01 Treat the transport option dkim_domain as a colon separated
1234 list, not as a single string, and sign the message with each element,
1235 omitting multiple occurences of the same signer.
1237 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1238 Bugzilla 985, 986. Patch by Simon Arlott
1240 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1241 Patch by Simon Arlott
1243 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1244 MBX locking. Notification from Dan Rosenberg.
1250 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1252 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1254 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1256 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1258 NM/04 Bugzilla 915: Flag broken perl installation during build.
1264 TK/01 Added patch by Johannes Berg that expands the main option
1265 "spamd_address" if it starts with a dollar sign.
1267 TK/02 Write list of recipients to X-Envelope-Sender header when building
1268 the mbox-format spool file for content scanning (suggested by Jakob
1271 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1272 (http://www.dcc-servers.net/) support via dccifd. Activated by
1273 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1275 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1276 by Mark Daniel Reidel <mr@df.eu>.
1278 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1279 When building exim an external PCRE library is now needed -
1280 PCRE is a system library on the majority of modern systems.
1281 See entry on PCRE_LIBS in EDITME file.
1283 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1284 conversation. Added nologin parameter to request.
1285 Patch contributed by Kirill Miazine.
1287 TF/01 Do not log submission mode rewrites if they do not change the address.
1289 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1291 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1292 log files in place. Contributed by Roberto Lima.
1294 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1296 TF/03 Bugzilla 615: When checking the local_parts router precondition
1297 after a local_part_suffix or local_part_prefix option, Exim now
1298 does not use the address's named list lookup cache, since this
1299 contains cached lookups for the whole local part.
1301 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1302 Robert Millan. Documentation is in experimental-spec.txt.
1304 TF/04 Bugzilla 668: Fix parallel build (make -j).
1306 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1308 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1309 Patch provided by Jan Srzednicki.
1311 TF/05 Leading white space used to be stripped from $spam_report which
1312 wrecked the formatting. Now it is preserved.
1314 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1315 that they are available at delivery time.
1317 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1319 TF/08 TLS error reporting now respects the incoming_interface and
1320 incoming_port log selectors.
1322 TF/09 Produce a more useful error message if an SMTP transport's hosts
1323 setting expands to an empty string.
1325 NM/06 Bugzilla 744: EXPN did not work under TLS.
1326 Patch provided by Phil Pennock.
1328 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1329 Patch provided by Richard Godbee.
1331 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1332 acl_smtp_notquit, added index entry.
1334 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1335 Patch provided by Eugene Bujak.
1337 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1338 accept(). Patch provided by Maxim Dounin.
1340 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1341 Patch provided by Phil Pennock.
1343 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1345 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1346 Patch provided by Brad "anomie" Jorsch.
1348 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1349 Patch provided by Dean Brooks.
1351 TK/05 Add native DKIM support (does not depend on external libraries).
1353 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1354 Patch provided by Graeme Fowler.
1356 NM/16 Bugzilla 851: Documentation example syntax fix.
1358 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1360 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1363 NM/19 Bugzilla 745: TLS version reporting.
1364 Patch provided by Phil Pennock.
1366 NM/20 Bugzilla 167: bool: condition support.
1367 Patch provided by Phil Pennock.
1369 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1370 clients. Patch provided by Phil Pennock.
1372 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1373 Patch provided by Brad "anomie" Jorsch.
1375 NM/23 Bugzilla 687: Fix misparses in eximstats.
1376 Patch provided by Heiko Schlittermann.
1378 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1379 Patch provided by Heiko Schlittermann.
1381 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1382 plus update to original patch.
1384 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1386 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1387 Patch provided by David Brownlee.
1389 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1391 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1393 NM/30 Bugzilla 888: TLS documentation bugfixes.
1395 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1397 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1398 Unlike the original bugzilla I have changed all shell scripts in src tree.
1400 NM/33 Bugzilla 898: Transport filter timeout fix.
1401 Patch by Todd Rinaldo.
1403 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1404 Patch by Serge Demonchaux.
1406 NM/35 Bugzilla 39: Base64 decode bug fixes.
1407 Patch by Jakob Hirsch.
1409 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1411 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1413 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1415 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1421 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1422 ALT-N's libdkim that I have put here:
1423 http://duncanthrax.net/exim-experimental/
1425 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1426 were called 'true' and it seems that C99 defines that as a reserved
1427 keyword to be used with 'bool' variable types. That means you could
1428 not include C99-style headers which use bools without triggering
1429 build errors in sieve.c.
1431 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1432 as mailq or other aliases. Changed the --help handling significantly
1433 to do whats expected. exim_usage() emits usage/help information.
1435 SC/01 Added the -bylocaldomain option to eximstats.
1437 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1439 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1441 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1447 PH/01 Another patch from the Sieve maintainer.
1449 PH/02 When an IPv6 address is converted to a string for single-key lookup
1450 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1451 dots are used instead of colons so that keys in lsearch files need not
1452 contain colons. This was done some time before quoting was made available
1453 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1454 (notated using the quote facility) so as to distinguish them from IPv4
1455 keys. This meant that lookups for IP addresses in host lists did not work
1456 for iplsearch lookups.
1458 This has been fixed by arranging for IPv6 addresses to be expressed with
1459 colons if the lookup type is iplsearch. This is not incompatible, because
1460 previously such lookups could never work.
1462 The situation is now rather anomolous, since one *can* have colons in
1463 ordinary lsearch keys. However, making the change in all cases is
1464 incompatible and would probably break a number of configurations.
1466 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1469 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1470 conversion specification without a maximum field width, thereby enabling
1471 a rogue spamd server to cause a buffer overflow. While nobody in their
1472 right mind would setup Exim to query an untrusted spamd server, an
1473 attacker that gains access to a server running spamd could potentially
1474 exploit this vulnerability to run arbitrary code as the Exim user.
1476 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1477 $primary_hostname instead of what libspf2 thinks the hosts name is.
1479 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1480 a directory entry by the name of the lookup key. Previously, if a
1481 symlink pointed to a non-existing file or a file in a directory that
1482 Exim lacked permissions to read, a lookup for a key matching that
1483 symlink would fail. Now it is enough that a matching directory entry
1484 exists, symlink or not. (Bugzilla 503.)
1486 PH/03 The body_linecount and body_zerocount variables are now exported in the
1489 PH/04 Added the $dnslist_matched variable.
1491 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1492 This means they are set thereafter only if the connection becomes
1495 PH/06 Added the client_condition to authenticators so that some can be skipped
1496 by clients under certain conditions.
1498 PH/07 The error message for a badly-placed control=no_multiline_responses left
1499 "_responses" off the end of the name.
1501 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1503 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1504 (without spaces) instead of just copying the configuration text.
1506 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1508 PH/11 Added $max_received_linelength.
1510 PH/12 Added +ignore_defer and +include_defer to host lists.
1512 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1513 way in which PCRE > 7.0 is built.
1515 PH/14 Implemented queue_only_load_latch.
1517 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1518 MAIL command. The effect was to mangle the value on 64-bit systems.
1520 PH/16 Another patch from the Sieve maintainer.
1522 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1524 PH/18 If a system quota error occurred while trying to create the file for
1525 a maildir delivery, the message "Mailbox is full" was not appended to the
1526 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1527 only to a quota excession during the actual writing of the file.
1529 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1530 characters?) which causes problems in log lines. The DN values are now
1531 passed through string_printing() before being added to log lines.
1533 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1534 and InterBase are left for another time.)
1536 PH/21 Added message_body_newlines option.
1538 PH/22 Guard against possible overflow in moan_check_errorcopy().
1540 PH/23 POSIX allows open() to be a macro; guard against that.
1542 PH/24 If the recipient of an error message contained an @ in the local part
1543 (suitably quoted, of course), incorrect values were put in $domain and
1544 $local_part during the evaluation of errors_copy.
1550 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1551 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1554 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1555 issue a MAIL command.
1557 PH/02 In an ACL statement such as
1559 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1561 if a client was not listed at all, or was listed with a value other than
1562 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1563 the condition was not true (as it should be), so access was not denied.
1564 The bug was that the ! inversion was incorrectly passed on to the second
1565 item. This has been fixed.
1567 PH/03 Added additional dnslists conditions == and =& which are different from
1568 = and & when the dns lookup returns more than one IP address.
1570 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1571 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1573 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1574 FSYNC, which compiles an option called disable_fsync that allows for
1575 bypassing fsync(). The documentation is heavily laced with warnings.
1577 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1579 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1580 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1581 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1582 including adding "make clean"; (3) Added -fPIC when compiling the test
1583 dynamically loaded module, to get rid of a warning.
1585 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1586 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1587 = 0s. The bug is otherwise harmless.
1589 PH/07 There was a bug in the dovecot authenticator such that the value of
1590 $auth1 could be overwritten, and so not correctly preserved, after a
1591 successful authentication. This usually meant that the value preserved by
1592 the server_setid option was incorrect.
1594 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1596 PH/09 Installed PCRE release 7.0.
1598 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1599 run for batched SMTP input. It is now run at the start of every message
1600 in the batch. While fixing this I discovered that the process information
1601 (output by running exiwhat) was not always getting set for -bs and -bS
1602 input. This is fixed, and it now also says "batched" for BSMTP.
1604 PH/11 Added control=no_pipelining.
1606 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1607 patch, slightly modified), and move the expansion of helo_data till after
1608 the connection is made in the smtp transport (so it can use these
1611 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1613 PH/14 Added log_selector = +pid.
1615 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1617 PH/16 Add ${if forany and ${if forall.
1619 PH/17 Added dsn_from option to vary the From: line in DSNs.
1621 PH/18 Flush SMTP output before performing a callout, unless control =
1622 no_callout_flush is set.
1624 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1625 was true (the default) a successful delivery failed to delete the retry
1626 item, thus causing premature timeout of the address. The bug is now
1629 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1631 PH/21 Long custom messages for fakedefer and fakereject are now split up
1632 into multiline reponses in the same way that messages for "deny" and
1633 other ACL rejections are.
1635 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1636 with slight modification.
1638 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1639 draft, changing the syntax and factoring some duplicate code".
1641 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1642 for deliveries of the second and subsequent messages over the same SMTP
1645 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1646 ${reduce, with only minor "tidies".
1648 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1650 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1651 expansion side effects.
1653 PH/27 When a message times out after an over-quota error from an Exim-imposed
1654 quota, the bounce message says "mailbox is full". This message was not
1655 being given when it was a system quota that was exceeded. It now should
1658 MH/03 Made $recipients available in local_scan(). local_scan() already has
1659 better access to the recipient list through recipients_list[], but
1660 $recipients can be useful in postmaster-provided expansion strings.
1662 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1663 in the case of a MAIL command with additional options following the
1664 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1665 were accidentally chopped off.
1667 PH/29 SMTP synchronization checks are implemented when a command is read -
1668 there is a check that no more input is waiting when there shouldn't be
1669 any. However, for some commands, a delay in an ACL can mean that it is
1670 some time before the response is written. In this time, more input might
1671 arrive, invalidly. So now there are extra checks after an ACL has run for
1672 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1673 pipelining has not been advertised.
1675 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1677 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1678 returned in the SMTP response when smtp_return_error_details was false.
1679 This has been fixed.
1681 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1682 instead of the C I/O that was originally supplied, because problems were
1683 reported on Solaris.
1685 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1686 Exim which did not show up earlier: it was assuming that a call to
1687 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1688 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1689 was a macro that became an assignment, so it seemed to work. This has
1690 changed to a proper function call with a void return, hence the compile
1691 error. Exim's code has been fixed.
1693 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1696 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1698 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1700 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1701 because it turns out to be too big or there is a timeout) there is no
1702 "Completed" line in the log. When some messages of this type were
1703 selected by exigrep, they were listed as "not completed". Others were
1704 picked up by some special patterns. I have improved the selection
1705 criteria to be more general.
1707 PH/38 The host_find_failed option in the manualroute router can now be set
1708 to "ignore", to completely ignore a host whose IP address cannot be
1709 found. If all hosts are ignored, the behaviour is controlled by the new
1710 host_all_ignored option.
1712 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1713 homing or because of multiple MX records with /mx) generated more than
1714 one IP address, and the following item turned out to be the local host,
1715 all the secondary addresses of the first item were incorrectly removed
1716 from the list, along with the local host and any following hosts (which
1717 is what is supposed to happen).
1719 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1720 whoever called Exim into the -H file. In the case of the daemon it was
1721 behaving confusingly. When first started, it used values for whoever
1722 started the daemon, but after a SIGHUP it used the Exim user (because it
1723 calls itself on a restart). I have changed the code so that it now always
1726 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1727 message are rejected with the same error (e.g. no authentication or bad
1728 sender address), and a DATA command is nevertheless sent (as can happen
1729 with PIPELINING or a stupid MUA), the error message that was given to the
1730 RCPT commands is included in the rejection of the DATA command. This is
1731 intended to be helpful for MUAs that show only the final error to their
1734 PH/42 Another patch from the Sieve maintainer.
1736 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1737 Eximstats - Fixed some broken HTML links and added missing column headers
1739 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1740 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1742 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1744 PH/43 Yet another patch from the Sieve maintainer.
1746 PH/44 I found a way to check for a TCP/IP connection going away before sending
1747 the response to the final '.' that terminates a message, but only in the
1748 case where the client has not sent further data following the '.'
1749 (unfortunately, this is allowed). However, in many cases there won't be
1750 any further data because there won't be any more messages to send. A call
1751 to select() can be used: if it shows that the input is "ready", there is
1752 either input waiting, or the socket has been closed. An attempt to read
1753 the next input character can distinguish the two cases. Previously, Exim
1754 would have sent an OK response which the client would never have see.
1755 This could lead to message repetition. This fix should cure that, at
1756 least in a lot of common cases.
1758 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1759 advertised in response to EHLO.
1765 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1766 fixed by 4.65/MH/01 (is this a record?) are fixed:
1768 (i) An empty string was always treated as zero by the numeric comparison
1769 operators. This behaviour has been restored.
1771 (ii) It is documented that the numeric comparison operators always treat
1772 their arguments as decimal numbers. This was broken in that numbers
1773 starting with 0 were being interpreted as octal.
1775 While fixing these problems I realized that there was another issue that
1776 hadn't been noticed. Values of message_size_limit (both the global option
1777 and the transport option) were treated as octal if they started with 0.
1778 The documentation was vague. These values are now always treated as
1779 decimal, and I will make that clear in the documentation.
1785 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1786 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1789 MH/01 Don't check that the operands of numeric comparison operators are
1790 integers when their expansion is in "skipping" mode (fixes bug
1791 introduced by 4.64-PH/07).
1793 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1794 child addresses, Exim now panics and dies. Previously, because the count
1795 is held in a short int, deliveries were likely to be lost. As such a
1796 large number of recipients for a single message is ridiculous
1797 (performance will be very, very poor), I have chosen to impose a limit
1798 rather than extend the field.
1804 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1805 leftover -K file (the existence of which was triggered by #402).
1806 While we were at it, introduced process PID as part of the -K
1807 filename. This should rule out race conditions when creating
1810 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1811 processing considerably. Previous code took too long for large mails,
1812 triggering a timeout which in turn triggers #401.
1814 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1815 in the DK code in transports.c. sendfile() is not really portable,
1816 hence the _LINUX specificness.
1818 TF/01 In the add_headers option to the mail command in an Exim filter,
1819 there was a bug that Exim would claim a syntax error in any
1820 header after the first one which had an odd number of characters
1823 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1824 callout verification, Exim cached a "reject" for the entire domain. This
1825 is correct for most verifications, but it is not correct for a recipient
1826 verification with use_sender or use_postmaster set, because in that case
1827 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1828 case of MAIL FROM:<> rejection from other early rejections (e.g.
1829 rejection of HELO). When verifying a recipient using a non-null MAIL
1830 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1831 Whatever the result of the callout, the value of the domain cache is
1832 left unchanged (for any other kind of callout, getting as far as trying
1833 RCPT means that the domain itself is ok).
1835 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1838 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1839 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1840 session, the daemon ignores ECONNECT errors and logs others; it now
1841 ignores EPIPE as well.
1843 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1844 (quoted-printable decoding).
1846 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1847 later the small subsequent patch to fix an introduced bug.
1849 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1851 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1853 PH/08 An error is now given if message_size_limit is specified negative.
1855 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1856 to be given (somewhat) arbitrary names.
1858 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1861 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1862 miscellaneous code fixes
1864 PH/10 Added the log_reject_target ACL modifier to specify where to log
1867 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1868 hostname. This is wrong, because it relates to the incoming message (and
1869 probably the interface on which it is arriving) and not to the outgoing
1870 callout (which could be using a different interface). This has been
1871 changed to use the value of the helo_data option from the smtp transport
1872 instead - this is what is used when a message is actually being sent. If
1873 there is no remote transport (possible with a router that sets up host
1874 addresses), $smtp_active_hostname is used.
1876 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1877 tweaks were necessary in order to get it to work (see also 21 below):
1878 (a) The code assumed that strncpy() returns a negative number on buffer
1879 overflow, which isn't the case. Replaced with Exim's string_format()
1881 (b) There were several signed/unsigned issues. I just did the minimum
1882 hacking in of casts. There is scope for a larger refactoring.
1883 (c) The code used strcasecmp() which is not a standard C function.
1884 Replaced with Exim's strcmpic() function.
1885 (d) The code set only $1; it now sets $auth1 as well.
1886 (e) A simple test gave the error "authentication client didn't specify
1887 service in request". It would seem that Dovecot has changed its
1888 interface. Fortunately there's a specification; I followed it and
1889 changed what the client sends and it appears to be working now.
1891 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1894 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1895 address A is aliased to B and C, where B exists and C does not. Without
1896 -v the output is "A verified" because verification stops after a
1897 successful redirection if more than one address is generated. However,
1898 with -v the child addresses are also verified. Exim was outputting "A
1899 failed to verify" and then showing the successful verification for C,
1900 with its parentage. It now outputs "B failed to verify", showing B's
1901 parentage before showing the successful verification of C.
1903 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1904 look up a TXT record in a specific list after matching in a combined
1907 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1908 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1909 they consult the DNS. I had assumed they would set it the way they
1910 wanted; and indeed my experiments on Linux seem to show that in some
1911 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1912 To be on the safe side, however, I have now made the interface to
1913 host_find_byname() similar to host_find_bydns(), with an argument
1914 containing the DNS resolver options. The host_find_byname() function now
1915 sets these options at its start, just as host_find_bydns() does. The smtp
1916 transport options dns_qualify_single and dns_search_parents are passed to
1917 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1918 of host_find_byname() use the default settings of RES_DEFNAMES
1919 (qualify_single) but not RES_DNSRCH (search_parents).
1921 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1922 spool_read_header() do less string testing, by means of a preliminary
1923 switch on the second character of optional "-foo" lines. (This is
1924 overdue, caused by the large number of possibilities that now exist.
1925 Originally there were few.) While I was there, I also converted the
1926 str(n)cmp tests so they don't re-test the leading "-" and the first
1927 character, in the hope this might squeeze out yet more improvement.
1929 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1930 flag allowing group syntax was set by the header_syntax check but not
1931 turned off, possible causing trouble later; (2) The flag was not being
1932 set at all for the header_verify test, causing "group"-style headers to
1933 be rejected. I have now set it in this case, and also caused header_
1934 verify to ignore an empty address taken from a group. While doing this, I
1935 came across some other cases where the code for allowing group syntax
1936 while scanning a header line wasn't quite right (mostly, not resetting
1937 the flag correctly in the right place). These bugs could have caused
1938 trouble for malformed header lines. I hope it is now all correct.
1940 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1941 with the "reply" argument non-NULL. The code, however (which originally
1942 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1943 but it didn't always do it. This confused somebody who was copying the
1944 code for some other use. I have removed all the tests.
1946 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1947 feature that was used to support insecure browsers during the U.S. crypto
1948 embargo. It requires special client support, and Exim is probably the
1949 only MTA that supported it -- and would never use it because real RSA is
1950 always available. This code has been removed, because it had the bad
1951 effect of slowing Exim down by computing (never used) parameters for the
1952 RSA_EXPORT functionality.
1954 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1955 authenticator to fail if there's a tab character in the incoming data
1956 (there should never be unless someone is messing about, as it's supposed
1957 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1958 if the connection is using TLS or if the remote IP is the same as the
1959 local IP, and the "valid-client-cert option" if a client certificate has
1962 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1963 authenticators. This can be used for authorization after authentication
1964 succeeds. (In the case of plaintext, it servers for both authentication
1967 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1968 if any retry times were supplied.
1970 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1971 connection, where there is no client IP address to check. In this
1972 situation, the verify now always succeeds.
1974 PH/25 Applied John Jetmore's -Mset patch.
1976 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1978 PH/27 In a string expansion for a processed (not raw) header when multiple
1979 headers of the same name were present, leading whitespace was being
1980 removed from all of them, but trailing whitespace was being removed only
1981 from the last one. Now trailing whitespace is removed from each header
1982 before concatenation. Completely empty headers in a concatenation (as
1983 before) are ignored.
1985 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1986 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1988 PH/29 [Removed. This was a change that I later backed out, and forgot to
1989 correct the ChangeLog entry (that I had efficiently created) before
1990 committing the later change.]
1992 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1993 address errors (4xx response to RCPT) over the same connection as other
1994 messages routed to the same hosts. Such deliveries are always "forced",
1995 so retry times are not inspected. This resulted in far too many retries
1996 for the affected addresses. The effect occurred only when there were more
1997 hosts than the hosts_max_try setting in the smtp transport when it had
1998 the 4xx errors. Those hosts that it had tried were not added to the list
1999 of hosts for which the message was waiting, so if all were tried, there
2000 was no problem. Two fixes have been applied:
2002 (i) If there are any address or message errors in an SMTP delivery, none
2003 of the hosts (tried or untried) are now added to the list of hosts
2004 for which the message is waiting, so the message should not be a
2005 candidate for sending over the same connection that was used for a
2006 successful delivery of some other message. This seems entirely
2007 reasonable: after all the message is NOT "waiting for some host".
2008 This is so "obvious" that I'm not sure why it wasn't done
2009 previously. Hope I haven't missed anything, but it can't do any
2010 harm, as the worst effect is to miss an optimization.
2012 (ii) If, despite (i), such a delivery is accidentally attempted, the
2013 routing retry time is respected, so at least it doesn't keep
2014 hammering the server.
2016 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2017 in ${readsocket because some servers need this prod.
2019 PH/32 Added some extra debug output when updating a wait-xxx database.
2021 PH/33 The hint "could be header name not terminated by colon", which has been
2022 given for certain expansion errors for a long time, was not being given
2023 for the ${if def:h_colon_omitted{... case.
2025 PH/34 The spec says: "With one important exception, whenever a domain list is
2026 being scanned, $domain contains the subject domain." There was at least
2027 one case where this was not true.
2029 PH/35 The error "getsockname() failed: connection reset by peer" was being
2030 written to the panic log as well as the main log, but it isn't really
2031 panic-worthy as it just means the connection died rather early on. I have
2032 removed the panic log writing for the ECONNRESET error when getsockname()
2035 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2036 runs only) independently of the message's sender address. This meant
2037 that, if the 4xx error was in fact related to the sender, a different
2038 message to the same recipient with a different sender could confuse
2039 things. In particualar, this can happen when sending to a greylisting
2040 server, but other circumstances could also provoke similar problems.
2041 I have changed the default so that the retry time for these errors is now
2042 based a combination of the sender and recipient addresses. This change
2043 can be overridden by setting address_retry_include_sender=false in the
2046 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2047 remote server are returned as part of bounce messages. This was not
2048 happening for LMTP over a pipe (the lmtp transport), but now it is the
2049 same for both kinds of LMTP.
2051 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2052 in header lines that were in fact CNAMEs. This is no longer the case.
2054 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2055 and queue runs started by the daemon processed all messages. This has
2056 been fixed so that -R and -S can now usefully be given with -q<time>.
2058 PH/40 Import PCRE release 6.7 (fixes some bugs).
2060 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2062 PH/42 Give an error if -q is specified more than once.
2064 PH/43 Renamed the variables $interface_address and $interface_port as
2065 $received_ip_address and $received_port, to make it clear that these
2066 values apply to message reception, and not to the outgoing interface when
2067 a message is delivered. (The old names remain recognized, of course.)
2069 PH/44 There was no timeout on the connect() call when using a Unix domain
2070 socket in the ${readsocket expansion. There now is.
2072 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2073 be meaningful with "accept".
2075 SC/01 Eximstats V1.43
2076 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2078 SC/02 Eximstats V1.44
2079 Use a glob alias rather than an array ref in the generated
2080 parser. This improves both readability and performance.
2082 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2083 Collect SpamAssassin and rejection statistics.
2084 Don't display local sender or destination tables unless
2085 there is data to show.
2086 Added average volumes into the top table text output.
2088 SC/04 Eximstats V1.46
2089 Collect data on the number of addresses (recipients)
2090 as well as the number of messages.
2092 SC/05 Eximstats V1.47
2093 Added 'Message too big' to the list of mail rejection
2094 reasons (thanks to Marco Gaiarin).
2096 SC/06 Eximstats V1.48
2097 Mainlog lines which have GMT offsets and are too short to
2098 have a flag are now skipped.
2100 SC/07 Eximstats V1.49 (Alain Williams)
2101 Added the -emptyok flag.
2103 SC/08 Eximstats V1.50
2104 Fixes for obtaining the IP address from reject messages.
2106 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2107 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2108 whitesspace changes from 4.64-PH/27
2110 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2113 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2114 are found, allow negative numbers in numeric criteria)
2116 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2118 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2119 to match changes made in 4.64-PH/43
2121 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2123 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2124 "domain" is actually an IP address, and force a failure. This locks out
2125 those revolvers/nameservers that support "A-for-A" lookups, in
2126 contravention of the specifications.
2128 PH/48 When a host name was looked up from an IP address, and the subsequent
2129 forward lookup of the name timed out, the host name was left in
2130 $sender_host_name, contrary to the specification.
2132 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2133 restricted to single-key lookups, Exim was not diagnosing an error if
2134 * or *@ was used with a query-style lookup.
2136 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2138 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2139 long ago, but noone interested enough thought of it. Let's just say that
2140 the "1.1" means that there are some new functions that weren't there at
2141 some point in the past.
2143 PH/51 Error processing for expansion failure of helo_data from an smtp
2144 transport during callout processing was broken.
2146 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2147 tested/used via the -bh/-bhc/-bs options.
2149 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2150 bug, fixed in subsequent PCRE releases).
2152 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2153 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2155 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2161 SC/01 Use a glob alias rather than an array ref in eximstats generated
2162 parser. This improves both readability and performance.
2164 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2165 Don't display local sender or destination tables in eximstats unless
2166 there is data to show.
2167 Added average volumes into the eximstats top table text output.
2169 SC/03 Collect data on the number of addresses (recipients) as well
2170 as the number of messages in eximstats.
2172 TF/01 Correct an error in the documentation for the redirect router. Exim
2173 does (usually) call initgroups() when daemonizing.
2175 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2176 with consistent privilege compared to when running as a daemon.
2178 TF/03 Note in the spec that $authenticated_id is not set for local
2179 submissions from trusted users.
2181 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2182 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2184 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2185 by adding some example configuration directives to the default
2186 configuration file. A little bit of work is required to uncomment the
2187 directives and define how usernames and passwords are checked, but
2188 there is now a framework to start from.
2190 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2191 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2192 without this. I don't know how relevant this is to other LDAP libraries.
2194 PH/02 Add the verb name to the "unknown ACL verb" error.
2196 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2198 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2200 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2201 directory not expanded when it should be if an expanded home directory
2202 was set for the address (which is overridden by the transport).
2204 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2207 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2208 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2209 because it is too late at that time, and has no effect.
2211 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2212 security issue with \' (bugzilla #107). I could not use the
2213 PQescapeStringConn() function, because it needs a PGconn value as one of
2216 PH/08 When testing addresses using -bt, indicate those final addresses that
2217 are duplicates that would not cause an additional delivery. At least one
2218 person was confused, thinking that -bt output corresponded to deliveries.
2219 (Suppressing duplicates isn't a good idea as you lose the information
2220 about possibly different redirections that led to the duplicates.)
2222 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2223 systems where poll() doesn't work, in particular OS X.
2225 PH/10 Added more information to debugging output for retry time not reached.
2227 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2228 operations in malware.c.
2230 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2233 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2234 syslog_duplication set false, log lines that would normally be written
2235 both the the main log and to the reject log were not written to syslog at
2238 PH/14 In the default configuration, change the use of "message" in ACL warn
2239 statements to "add_header".
2241 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2242 not followed by a command (e.g. "seen endif").
2244 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2245 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2248 PH/17 Added extra conditions to the default value of delay_warning_condition
2252 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2253 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2254 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2257 The Auto-Submitted: and various List- headers are standardised, whereas I
2258 don't think Precedence: ever was.
2260 PH/18 Refactored debugging code in route_finduser() to show more information,
2261 in particular, the error code if getpwnam() issues one.
2263 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2264 This is apparently needed in addition to the PH/07 change above to avoid
2265 any possible encoding problems.
2267 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2268 but not after initializing Perl.
2270 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2271 output them only if debugging. By default they are written stderr,
2272 apparently, which is not desirable.
2274 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2277 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2280 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2282 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2283 authenticated or an ident call has been made. Suppress the default
2284 values for $authenticated_id and $authenticated_sender (but permit -oMai
2285 and -oMas) when testing with -bh.
2287 PH/24 Re-jigged the order of the tests in the default configuration so that the
2288 tests for valid domains and recipients precede the DNS black list and CSA
2289 tests, on the grounds that those ones are more expensive.
2291 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2292 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2293 This bug exists in every version of Exim that I still have, right back to
2296 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2297 However, an attempt to turn on case-sensitivity in a regex key by
2298 including (?-i) didn't work because the subject string was already
2299 lowercased, and the effects were non-intuitive. It turns out that a
2300 one-line patch can be used to allow (?-i) to work as expected.
2306 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2307 other effects) broke the use of negated acl sub-conditions.
2309 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2312 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2313 "Deny" causes Exim to reject the incoming connection with a 554 error.
2314 Unfortunately, if there is a major crisis, such as a disk failure,
2315 tcp-wrappers gives "deny", whereas what one would like would be some
2316 kind of temporary error. A kludge has been added to help with this.
2317 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2318 554 error is used if errno is still zero or contains ENOENT (which occurs
2319 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2322 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2324 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2325 errors. Otherwise a message that provokes a temporary error (when other
2326 messages do not) can cause a whole host to time out.
2328 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2329 addresses were routed directly to files or pipes from a redirect router.
2330 File deliveries just didn't batch; pipe deliveries might have suffered
2333 PH/06 A failure to get a lock for a hints database would erroneously always say
2334 "Failed to get write lock", even when it was really a read lock.
2336 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2337 of 0600. This has been changed to use the value of the lockfile_mode
2338 option (which defaults to 0600).
2340 PH/08 Applied small patch from the Sieve maintainer.
2342 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2343 folder from quota calculations, a direct delivery into this folder messed
2344 up the contents of the maildirsize file. This was because the regex was
2345 used only to exclude .Trash (or whatever) when the size of the mailbox
2346 was calculated. There was no check that a delivery was happening into an
2347 excluded directory. This bug has been fixed by ignoring all quota
2348 processing for deliveries into excluded directories.
2350 PH/10 Added the maildirfolder_create_regex option to appendfile.
2356 PH/01 The code for finding all the local interface addresses on a FreeBSD
2357 system running IPv6 was broken. This may well have applied to all BSD
2358 systems, as well as to others that have similar system calls. The broken
2359 code found IPv4 interfaces correctly, but gave incorrect values for the
2360 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2361 that it would not match correctly against @[] and not recognize the IPv6
2364 PH/02 The ipliteral router was not recognizing addresses of the form user@
2365 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2367 PH/03 Added disable_ipv6.
2369 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2370 decoded form, because it is most often used to construct To: headers
2371 lines in autoreplies, and the decoded form may well be syntactically
2372 invalid. However, $reply_address has leading white space removed, and all
2373 newlines turned into spaces so that the autoreply transport does not
2376 PH/05 If group was specified without a user on a router, and no group or user
2377 was specified on a transport, the group from the router was ignored.
2379 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2380 for visible compile-time settings that can be used to change these
2381 numbers, for those that want even more. Backwards compatibility with old
2382 spool files has been maintained. However, going back to a previous Exim
2383 release will lost any variables that are in spool files.
2385 PH/07 Two small changes when running in the test harness: increase delay when
2386 passing a TCP/IP connection to a new process, in case the original
2387 process has to generate a bounce, and remove special handling of
2388 127.0.0.2 (sic), which is no longer necessary.
2390 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2391 be the same on different OS.
2393 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2396 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2397 whether --show-vars was specified or not
2399 JJ/02 exipick: Added support for new ACL variable spool format introduced
2402 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2403 syntactically invalid From: or Reply-to: line, and a filter used this to
2404 generate an autoreply, and therefore failed to obtain an address for the
2405 autoreply, Exim could try to deliver to a non-existent relative file
2406 name, causing unrelated and misleading errors. What now happens is that
2407 it logs this as a hard delivery error, but does not attempt to create a
2410 PH/11 The exinext utility has a -C option for testing purposes, but although
2411 the given file was scanned by exinext itself; it wasn't being passed on
2412 when Exim was called.
2414 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2415 an end-of-file indication when reading a command response.
2417 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2418 compiled. In many other places in Exim, IPv6 addresses are always
2419 recognized, so I have changed this. It also means that IPv4 domain
2420 literals of the form [IPV4:n.n.n.n] are now always recognized.
2422 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2423 used if the router is not running as root, for example, when verifying at
2424 ACL time, or when using -bh. The debugging output from this situation was
2425 non-existent - all you got was a failure to exec. I have made two
2428 (a) Failures to set uid/gid, the current directory, or a process leader
2429 in a subprocess such as that created by queryprogram now generate
2430 suitable debugging ouput when -d is set.
2432 (b) The queryprogram router detects when it is not running as root,
2433 outputs suitable debugging information if -d is set, and then runs
2434 the subprocess without attempting to change uid/gid.
2436 PH/15 Minor change to Makefile for building test_host (undocumented testing
2439 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2440 additional section of a DNS packet that returns MX or SRV records.
2441 Instead, it always explicitly searches for A/AAAA records. This avoids
2442 major problems that occur when a DNS server includes only records of one
2443 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2444 fixed another bug: if SRV records were looked up and the corresponding
2445 address records were *not* found in the additional section, the port
2446 values from the SRV records were lost.
2448 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2449 using the correct key (the original address) when searching the retry
2450 rules in order to find which one to use for generating the retry hint.
2452 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2453 adding the default one. Similarly, if it contains a Reply-To: header, the
2454 errors_reply_to option, if set, is not used.
2456 PH/19 When calculating a retry time, Exim used to measure the "time since
2457 failure" by looking at the "first failed" field in the retry record. Now
2458 it does not use this if it is later than than the arrival time of the
2459 message. Instead it uses the arrival time. This makes for better
2460 behaviour in cases where some deliveries succeed, thus re-setting the
2461 "first failed" field. An example is a quota failure for a huge message
2462 when small messages continue to be delivered. Without this change, the
2463 "time since failure" will always be short, possible causing more frequent
2464 delivery attempts for the huge message than are intended.
2465 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2467 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2468 $1, $2, $3) because the numerical variables can be reset during some
2469 expansion items (e.g. "match"), thereby losing the authentication data.
2471 PH/21 Make -bV show the size of off_t variables so that the test suite can
2472 decide whether to run tests for quotas > 2G.
2474 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2475 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2476 filecount value is greater than 2G or if a quota value is greater than 2G
2477 on a system where the size of off_t is not greater than 4, a panic error
2480 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2481 never match. The debug and -bh output now contains an explicit error
2482 message indicating a malformed IPv4 address or mask.
2484 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2485 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2486 PH/23 above applies.
2488 PH/25 Do not write to syslog when running in the test harness. The only
2489 occasion when this arises is a failure to open the main or panic logs
2490 (for which there is an explicit test).
2492 PH/26 Added the /no_tell option to "control=freeze".
2494 PH/27 If a host name lookup failed very early in a connection, for example, if
2495 the IP address matched host_lookup and the reverse lookup yielded a name
2496 that did not have a forward lookup, an error message of the form "no IP
2497 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2498 could be logged. Now it outputs the IP address instead of "NULL".
2500 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2501 allows the sender and the authenticated sender to be set when
2502 submitting a message from within Exim. Since child_open_exim() is
2503 documented for local_scan(), the new function should be too.
2505 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2506 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2507 results in an empty string is now treated as unset.
2509 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2511 PH/31 Added sender_verify_fail logging option.
2513 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2514 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2515 tidied the source and removed it altogether.
2517 PH/33 When a queue run was abandoned because the load average was too high, a
2518 log line was always written; now it is written only if the queue_run log
2519 selector is set. In addition, the log line for abandonment now contains
2520 information about the queue run such as the pid. This is always present
2521 in "start" and "stop" lines but was omitted from the "abandon" line.
2523 PH/34 Omit spaces between a header name and the colon in the error message that
2524 is given when verify = headers_syntax fails (if there are lots of them,
2525 the message gets confusing).
2527 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2528 names, as there are now some PTR records that contain slashes. This check
2529 is only to protect against broken name servers that fall over on strange
2530 characters, so the fact that it applies to all lookups doesn't matter.
2532 PH/36 Now that the new test suite is complete, we can remove some of the
2533 special code in Exim that was needed for the old test suite. For example,
2534 sorting DNS records because real resolvers return them in an arbitrary
2535 order. The new test suite's fake resolver always returns records in the
2538 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2539 bounces) except when queue_only is set, to avoid logging races between
2540 the different processes.
2542 PH/38 Panic-die if .include specifies a non-absolute path.
2544 PH/39 A tweak to the "H" retry rule from its user.
2546 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2547 a label. They prevented compilation on older perls.
2549 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2550 a warning to be raised on newish perls.
2552 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2553 on queue. Changes to match documented behaviour of showing count of
2554 messages matching specified criteria.
2556 PH/40 Changed the default ident timeout from 30s to 5s.
2558 PH/41 Added support for the use of login_cap features, on those BSD systems
2559 that have them, for controlling the resources used by pipe deliveries.
2561 PH/42 The content-scanning code uses fopen() to create files in which to put
2562 message data. Previously it was not paying any attention to the mode of
2563 the files. Exim runs with umask(0) because the rest of the code creates
2564 files with open(), and sets the required mode explicitly. Thus, these
2565 files were ending up world-writeable. This was not a big issue, because,
2566 being within the spool directory, they were not world-accessible. I have
2567 created a function called modefopen, which takes an additional mode
2568 argument. It sets umask(777), creates the file, chmods it to the required
2569 mode, then resets the umask. All the relevant calls to fopen() in the
2570 content scanning code have been changed to use this function.
2572 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2573 to 24 hours. This avoids potential overflow problems when processing G
2574 and H retry rules. I suspect nobody ever tinkers with this value.
2576 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2578 PH/45 When the plaintext authenticator is running as a client, the server's
2579 challenges are checked to ensure they are valid base64 strings. By
2580 default, the authentication attempt is cancelled if an invalid string is
2581 received. Setting client_ignore_invalid_base64 true ignores these errors.
2582 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2583 they are received. Thus, the responses can be made to depend on the
2584 challenges. If an invalid string is ignored, an empty string is placed in
2587 PH/46 Messages that are created by the autoreply transport now contains a
2588 References: header, in accordance with RFCs 2822 and 3834.
2590 PH/47 Added authenticated_sender_force to the smtp transport.
2592 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2594 PH/49 Installed latest patch from the Sieve maintainer.
2596 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2597 was also set, the appendfile transport was unnecessarily scanning a
2598 directory of message files (e.g. for maildir delivery) to find the count
2599 of files (along with the size), even though it did not need this
2600 information. It now does the scan only if it needs to find either the
2601 size of the count of files.
2603 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2605 PH/52 Two bugs concerned with error handling when the smtp transport is
2608 (i) Exim was not creating retry information for temporary errors given
2609 for individual recipients after the DATA command when the smtp transport
2610 was used in LMTP mode. This meant that they could be retried too
2611 frequently, and not timed out correctly.
2613 (ii) Exim was setting the flag that allows error details to be returned
2614 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2615 recipients that were returned after the DATA command.
2617 PH/53 This is related to PH/52, but is more general: for any failing address,
2618 when detailed error information was permitted to be returned to the
2619 sender, but the error was temporary, then after the final timeout, only
2620 "retry timeout exceeded" was returned. Now it returns the full error as
2621 well as "retry timeout exceeded".
2623 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2624 do this, and (what is worse) MTAs that accept it.
2626 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2627 will now be deprecated.
2629 PH/56 New os.c-cygwin from the Cygwin maintainer.
2631 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2632 formats (previously only available in exim formats via -bpr, -bpru,
2633 and -bpra. Now also available in native and exiqgrep formats)
2635 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2636 with very large, slow to parse queues
2638 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2640 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2642 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2643 responses to authentication challenges, though it was showing the
2644 challenges; (ii) I've removed the CR characters from the debug output for
2647 PH/58 Allow for the insertion of a newline as well as a space when a string
2648 is turned into more than one encoded-word during RFC 2047 encoding. The
2649 Sieve code now uses this.
2651 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2652 data_4xx, lost_connection, tls_required.
2654 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2655 message was being sent as an SMTP response.
2657 PH/61 Add -l and -k options to exicyclog.
2659 PH/62 When verifying, if an address was redirected to one new address, so that
2660 verification continued, and the new address failed or deferred after
2661 having set something in $address_data, the value of $address_data was not
2662 passed back to the ACL. This was different to the case when no
2663 redirection occurred. The value is now passed back in both cases.
2665 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2666 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2667 use login_cap.h, so on its own it isn't the distinguishing feature. The
2668 new name refers directly to the setclassresources() function.
2670 PH/65 Added configuration files for NetBSD3.
2672 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2674 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2675 is preferred over IPv4.
2677 PH/68 The bounce_return_message and bounce_return_body options were not being
2678 honoured for bounces generated during the reception of non-SMTP messages.
2679 In particular, this applied to messages rejected by the ACL. This bug has
2680 been fixed. However, if bounce_return_message is true and bounce_return_
2681 body is false, the headers that are returned for a non-SMTP message
2682 include only those that have been read before the error was detected.
2683 (In the case of an ACL rejection, they have all been read.)
2685 PH/69 The HTML version of the specification is now built in a directory called
2686 spec_html instead of spec.html, because the latter looks like a path with
2687 a MIME-type, and this confuses some software.
2689 PH/70 Catch two compiler warnings in sieve.c.
2691 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2692 function verify_get_ident() calls ip_connect() to connect a socket, but
2693 if the "connect()" function timed out, ip_connect() used to close the
2694 socket. However, verify_get_ident() also closes the socket later, and in
2695 between Exim writes to the log, which may get opened at this point. When
2696 the socket was closed in ip_connect(), the log could get the same file
2697 descriptor number as the socket. This naturally causes chaos. The fix is
2698 not to close the socket in ip_connect(); the socket should be closed by
2699 the function that creates it. There was only one place in the code where
2700 this was missing, in the iplookup router, which I don't think anybody now
2701 uses, but I've fixed it anyway.
2703 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2704 well as to direct DNS lookups. Otherwise the handling of names in host
2705 lists is inconsistent and therefore confusing.
2711 PH/01 Two changes to the default runtime configuration:
2713 (1) Move the checks for relay_from_hosts and authenticated clients from
2714 after to before the (commented out) DNS black list checks.
2716 (2) Add control=submission to the relay_from_hosts and authenticated
2717 clients checks, on the grounds that messages accepted by these
2718 statements are most likely to be submissions.
2720 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2722 (1) Generate an error if the third argument for the ${prvs expansion is
2725 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2728 (3) Reset the variables that are obtained from the first argument of
2729 ${prvscheck and used in the second argument before leaving the code,
2730 because their memory is reclaimed, so using them afterwards may do
2733 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2734 one (it's much easier than Tom thought :-).
2736 (5) Because of (4), we can now allow for the use of $prvscheck_result
2737 inside the third argument.
2739 PH/03 For some reason, the default setting of PATH when running a command from
2740 a pipe transport was just "/usr/bin". I have changed it to
2743 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2744 anything to be listed in the output from -bV.
2746 PH/05 When a filter generated an autoreply, the entire To: header line was
2747 quoted in the delivery log line, like this:
2749 => >A.N.Other <ano@some.domain> <original@ddress> ...
2751 This has been changed so that it extracts the operative address. There
2752 may be more than one such address. If so, they are comma-separated, like
2755 => >ano@some.domain,ona@other.domain <original@ddress> ...
2757 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2758 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2759 not being looked up in the rDNS to get a host name, Exim was showing the
2760 IP address twice in Received: lines, even though the IP addresses were
2761 identical. For example:
2763 Received: from [1.2.3.4] (helo=[1.2.3.4])
2765 However, if the real host name was known, it was omitting the HELO data
2766 if it matched the actual IP address. This has been tidied up so that it
2767 doesn't show the same IP address twice.
2769 PH/07 When both +timestamp and +memory debugging was on, the value given by
2770 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2771 called by the debug printing, thereby overwriting the timestamp buffer.
2772 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2774 PH/08 When the original message was included in an autoreply transport, it
2775 always said "this is a copy of the message, including all the headers",
2776 even if body_only or headers_only was set. It now gives an appropriate
2779 PH/09 Applied a patch from the Sieve maintainer which:
2781 o fixes some comments
2782 o adds the (disabled) notify extension core
2783 o adds some debug output for the result of if/elsif tests
2784 o points to the current vacation draft in the documentation
2785 and documents the missing references header update
2789 o fixes a bug in processing the envelope test (when testing
2790 multiple envelope elements, the last element determinted the
2793 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2794 Electronic Mail") by including:
2796 Auto-submitted: auto-generated
2798 in the messages that it generates (bounce messages and others, such as
2799 warnings). In the case of bounce messages for non-SMTP mesages, there was
2800 also a typo: it was using "Auto_submitted" (underscore instead of
2801 hyphen). Since every message generated by Exim is necessarily in response
2802 to another message, thes have all been changed to:
2804 Auto-Submitted: auto-replied
2806 in accordance with these statements in the RFC:
2808 The auto-replied keyword:
2810 - SHOULD be used on messages sent in direct response to another
2811 message by an automatic process,
2813 - MUST NOT be used on manually-generated messages,
2815 - MAY be used on Delivery Status Notifications (DSNs) and Message
2816 Disposition Notifications (MDNs),
2818 - MUST NOT be used on messages generated by automatic or periodic
2819 processes, except for messages which are automatic responses to
2822 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2823 to the default Received: header definition.
2825 PH/12 Added log selector acl_warn_skipped (default on).
2827 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2828 variables because (a) they are in the wrong storage pool and (b) even if
2829 they were copied, it wouldn't work properly because of the caching.
2831 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2832 checking when decoding. Apparently there are clients that generate
2833 overlong encoded strings. Why am I not surprised?
2835 PH/15 If the first argument of "${if match_address" was not empty, but did not
2836 contain an "@" character, Exim crashed. Now it writes a panic log message
2837 and treats the condition as false.
2839 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2841 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2842 extension "envelope-auth". The code is finished and in agreement with
2843 other implementations, but there is no documentation so far and in fact,
2844 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2845 not changing the active code.
2847 Print executed "if" and "elsif" statements when debugging is used. This
2848 helps a great deal to understand what a filter does.
2850 Document more things not specified clearly in RFC3028. I had all this
2851 sorted out, when out of a sudden new issues came to my mind. Oops."
2853 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2856 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2857 canonical form (as documented). However, after a host name lookup from
2858 the IP address, check_host() was doing a simple string comparison with
2859 addresses acquired from the DNS when checking that the found name did
2860 have the original IP as one of its addresses. Since any found IPv6
2861 addresses are likely to be in abbreviated form, the comparison could
2862 fail. Luckily, there already exists a function for doing the comparison
2863 by converting both addresses to binary, so now that is used instead of
2864 the text comparison.
2866 PH/20 There was another similar case to PH/19, when a complete host name was
2867 given in a host list; looking up its IP address could give an abbreviated
2868 form, whereas the current host's name might or might not be abbreviated.
2869 The same fix has been applied.
2875 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2876 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2879 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2880 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2882 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2884 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2885 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2886 Also, for "Auto-Submitted", treat anything other than "no" as
2887 non-personal, in accordance with RFC 3834. (Previously it treated
2888 anything starting "auto-" as non-personal.)
2890 TF/01 The control=submission/name=... option had a problem with syntax
2891 errors if the name included a slash character. The /name= option
2892 now slurps the rest of the string, so it can include any characters
2893 but it must come last in the list of options (after /sender_retain
2896 PH/05 Some modifications to the interface to the fake nameserver for the new
2904 TK/01 Added the "success_on_redirect" address verification option. See
2905 NewStuff for rationale and an example.
2907 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2909 PH/02 Patch to exigrep to allow it to work on syslog lines.
2911 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2912 fread() to skip over the body file's header line, because in Cygwin the
2913 header line is locked and is inaccessible.
2915 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2916 co-exist for some time) to make it clear that it is the Exim ID that is
2917 referenced, not the Message-ID: header line.
2919 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2920 string_format() function, because snprintf() does not exist on all
2923 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2924 ${stat: expansion item.
2926 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2927 protocol synchronization error", to keep the pedants happy.
2929 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2930 well as for IRIX systems, when gcc is being used. See the host.c source
2933 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2935 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2938 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2939 transport and to the smtp transport in LMTP mode.
2941 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2943 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2944 run a filter in a subprocess. This could lead to confusion in subsequent
2945 lookups in the parent process. There should also be a search_tidyup() at
2946 the end of the subprocess.
2948 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2949 only if the host matched helo_try_verify_hosts, which caused the
2950 verification to occur when the EHLO/HELO command was issued. The ACL just
2951 tested the remembered result. Now, if a previous verification attempt has
2952 not happened, "verify = helo" does it there and then.
2954 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2956 TK/03 Fix log output including CR from clamd.
2958 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2959 exist provoked a memory error which could cause a segfault.
2961 PH/15 Installed PCRE 6.2
2963 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2965 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2966 of the problem. Specifically, suggested +O2 rather than +O1 for the
2969 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2971 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2972 record, the port was not correctly logged when the outgoing_port log
2973 selector was set (it logged the transort's default port).
2975 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2976 fallback_hosts, and "hosts" in the smtp transport.
2978 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2979 host errors such as "Connection refused".
2981 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2982 authentication with radiusclient 0.4.9:
2984 - Error returned from rc_read_config was caught wrongly
2985 - Username/password not passed on to radius server due to wrong length.
2987 The presumption is that some radiusclient API changes for 4.51/PH/17
2988 were not taken care of correctly. The code is still untested by me (my
2989 Linux distribution still has 0.3.2 of radiusclient), but it was
2990 contributed by a Radius user.
2992 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2993 expanding the "port" option of the smtp transport.
2995 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2996 while reading a MIME header. Thanks to Tom Hughes for a patch.
2998 PH/24 Include config.h inside local_scan.h so that configuration settings are
3001 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3002 that in an ACL for RCPT (for example), you can examine exactly what was
3005 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3006 commands, but it was not correctly comparing the address with the actual
3007 client host address. Thus, it would show the EHLO address in Received:
3008 header lines when this was not necessary.
3010 PH/27 Added the % operator to ${eval:}.
3012 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3013 it should be ignoring failures (because with -C, for example, it has lost
3014 privilege). It wasn't ignoring creation failures other than "already
3017 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3020 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3021 because an input line was too long, either on its own, or by virtue of
3022 too many continuations, the temporary file was not being removed, and the
3023 return code was incorrect.
3025 PH/31 Missing "BOOL" in function definition in filtertest.c.
3027 PH/32 Applied Sieve patches from the maintainer.
3029 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3031 PH/33 Added "verify = not_blind".
3033 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3034 Local/Makefile (with some defaults set). These are used in built scripts
3035 such as exicyclog, but they have never been used in the exim_install
3036 script (though there are many overriding facilities there). I have
3037 arranged that the exim_install script now takes note of these two
3040 PH/35 Installed configuration files for Dragonfly.
3042 PH/36 When a locally submitted message by a trusted user did not contain a
3043 From: header, and the sender address was obtained from -f or from an SMTP
3044 MAIL command, and the trusted user did not use -F to supply a sender
3045 name, $originator_name was incorrectly used when constructing a From:
3046 header. Furthermore, $originator_name was used for submission mode
3047 messages from external hosts without From: headers in a similar way,
3048 which is clearly wrong.
3050 PH/37 Added control=suppress_local_fixups.
3052 PH/38 When log_selector = +received_sender was set, and the addition of the
3053 sender made the log line's construction buffer exactly full, or one byte
3054 less than full, an overflow happened when the terminating "\n" was
3057 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3058 when the result of a list match is failure because a DNS lookup failed.
3060 PH/40 RM_COMMAND is now used in the building process.
3062 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3063 the "build-* directories that it finds.
3065 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3066 address in a domain literal was a prefix of an interface address.
3068 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3069 when verifying a sender address, unless rewrite_headers is false.
3071 PH/44 Wrote a long comment about why errors_to addresses are verified as
3072 recipients, not senders.
3074 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3075 the ratelimit ACL was added.
3077 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3079 PH/46 Added extra information about PostgreSQL errors to the error string.
3081 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3082 avoiding the need to install special zones in a real server. This is
3083 backwards compatible; if it can't find the fake resolver, it drops back.
3084 Thus, both old and new test suites can be run.
3086 TF/02 Added util/ratelimit.pl
3088 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3089 clock is set back in time.
3091 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3092 Candler <B.Candler@pobox.com>.
3094 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3095 OK for addresses that are the result of redirections.
3097 PH/48 A number of further additions for the benefit of the new test suite,
3098 including a fake gethostbyname() that interfaces to the fake DNS resolver
3101 TF/06 The fix for widen_domains has also been applied to qualify_single and
3102 search_parents which are the other dnslookup options that can cause
3105 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3108 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3110 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3111 service" if it encountered an MX record with an empty target hostname.
3112 The message is now "an MX or SRV record indicated no SMTP service".
3114 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3115 if the DNS of the sending site is misconfigured. This is quite a
3116 common situation. This change restores the behaviour of treating a
3117 helo verification defer as a failure.
3119 PH/51 If self=fail was set on a router, the bounce message did not include the
3120 actual error message.
3126 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3128 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3129 command itself ended in error, the underlying message about the transport
3130 filter timeout was being overwritten with the pipe command error. Now the
3131 underlying error message should be appended to the second error message.
3133 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3135 PH/02 Reduce the amount of output that "make" produces by default. Full output
3136 can still be requested.
3138 PH/03 The warning log line about a condition test deferring for a "warn" verb
3139 was being output only once per connection, rather than after each
3140 occurrence (because it was using the same function as for successful
3141 "warn" verbs). This seems wrong, so I have changed it.
3143 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3144 it should not have, which might have caused a crash in the right
3145 circumstances, but probably never did.
3147 PH/04 Installed a modified version of Tony Finch's patch to make submission
3148 mode fix the return path as well as the Sender: header line, and to
3149 add a /name= option so that you can make the user's friendly name appear
3152 TF/03 Added the control = fakedefer ACL modifier.
3154 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3155 Mark Lowes for thorough testing.
3157 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3159 TK/03 Merged latest SRS patch from Miles Wilton.
3161 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3162 with the definition in sysexits.h (which is #included earlier).
3163 Fortunately, Exim does not actually use EX_OK. The code used to try to
3164 preserve the sysexits.h value, by assumimg that macro definitions were
3165 scanned for macro replacements. I have been disabused of this notion,
3166 so now the code just undefines EX_OK before #including unistd.h.
3168 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3169 in the smtp transport. When a block could not be written in a single
3170 write() function, the timeout was being re-applied to each part-write.
3171 This seems wrong - if the receiver was accepting one byte at a time it
3172 would take for ever. The timeout is now adjusted when this happens. It
3173 doesn't have to be particularly precise.
3175 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3176 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3178 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3179 without a domain if the check to <postmaster@domain> fails.
3181 SC/01 Eximstats: added -xls and the ability to specify output files
3182 (patch written by Frank Heydlauf).
3184 SC/02 Eximstats: use FileHandles for outputing results.
3186 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3188 SC/04 Eximstats: fixed display of large numbers with -nvr option
3190 SC/05 Eximstats: fixed merging of reports with empty tables.
3192 SC/06 Eximstats: added the -include_original_destination flag
3194 SC/07 Eximstats: removed tabs and trailing whitespace.
3196 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3198 TK/06 MBOX spool code: Add real "From " MBOX separator line
3199 so the .eml file is really in mbox format (even though
3200 most programs do not really care). Patch from Alex Miller.
3202 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3203 The latter is generated from $received_to and is only set if the
3204 message has one envelope recipient. SA can use these headers,
3205 obviously out-of-the-box. Patch from Alex Miller.
3207 PH/08 The ${def test on a variable was returning false if the variable's
3208 value was "0", contrary to what the specification has always said!
3209 The result should be true unless the variable is empty.
3211 PH/09 The syntax error of a character other than { following "${if
3212 def:variable_name" (after optional whitespace) was not being diagnosed.
3213 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3214 accidental colon was present, for example, could give incorrect results.
3216 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3217 result is used (not including appendfile, where other changes are about
3220 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3221 This involved changing a lot of size variables from int to off_t. It
3222 should work with maildirs and everything.
3224 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3225 spamd dying while we are connected to it.
3227 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3230 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3231 function for BDB 4.3.
3233 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3235 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3236 the sources for PCRE have been re-arranged and more files are now
3239 PH/15 The code I had for printing potentially long long variables in PH/11
3240 above was not the best (it lost precision). The length of off_t variables
3241 is now inspected at build time, and an appropriate printing format (%ld
3242 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3243 to be "long long int" or "long int". This is needed for the internal
3244 formatting function string_vformat().
3246 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3247 the configuration file to be ":syslog", then the script "guesses" where
3248 the logs files are, rather than using the compiled in default. In our
3249 case the guess is not the same as the compiled default, so the script
3250 suddenly stopped working when I started to use syslog. The patch checks
3251 to see if log_file_path is "". If so, it attempts to read it from exim
3252 with no configuration file to get the compiled in version, before it
3253 falls back to the previous guessing code."
3255 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3256 implementing BATV in an Exim configuration. See NewStuff for the gory
3259 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3260 Makefile that are specific to HP-UX.
3262 PH/18 If the "use_postmaster" option was set for a recipient callout together
3263 with the "random" option, the postmaster address was used as the MAIL
3264 FROM address for the random test, but not for the subsequent recipient
3265 test. It is now used for both.
3267 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3268 patch removes a few documentation additions to RFC 3028, because the
3269 latest draft now contains them. It adds the new en;ascii-case comparator
3270 and a new error check for 8bit text in MIME parts. Comparator and
3271 require names are now matched exactly. I enabled the subaddress
3272 extension, but it is not well tested yet (read: it works for me)."
3274 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3275 rework some of the code of TK/09 above to avoid the hardwired use of
3276 "%lld" and "long long". Replaced the call to snprintf() with a call to
3279 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3280 records point to non-existent hosts", "retry timeout exceeded", and
3281 "retry time not reached for any host after a long failure period".
3283 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3284 experimental DomainKeys support:
3286 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3287 (2) On an error such as an illegally used "control", the wrong name for
3288 the control was given.
3290 These problems did NOT occur unless DomainKeys support was compiled.
3292 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3294 PH/24 Added ${if match_ip condition.
3296 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3297 databases so that it will be absolutely obvious if a crash occurs in the
3298 DB library. This is a regular occurrence (often caused by mis-matched
3301 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3302 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3303 picked up on a user's system that detects such things. There doesn't seem
3304 to be a gcc warning option for this - only an attribute that has to be
3305 put on the function's prototype. It seems that in Fedora Core 4 they have
3306 set this on a number of new functions. No doubt there will be more in due
3309 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3310 specify a transport. However, if an address that was verified by such a
3311 router was the subject of a callout, Exim crashed because it tried to
3312 read the rcpt_include_affixes from the non-existent transport. Now it
3313 just assumes that the setting of that option is false. This bug was
3314 introduced by 4.51/PH/31.
3316 PH/28 Changed -d+all to exclude +memory, because that information is very
3317 rarely of interest, but it makes the output a lot bigger. People tend to
3318 do -d+all out of habit.
3320 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3321 code in os-type was giving problems when libc.so lives in lib64, like on
3324 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3325 aren't the modern standard, and it seems that some systems' include files
3326 don't always have them. Exim was already checking for some of the newer
3327 ones like T_AAAA, and defining it itself. I've added checks for all the
3328 record types that Exim uses.
3330 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3331 not automatically generating a new one, as it is supposed to. This
3332 prevented TLS from working. If the file did exist, but contained invalid
3333 data, a new version was generated, as expected. It was only the case of a
3334 non-existent file that was broken.
3336 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3337 with a change in libdomainkeys > 0.64.
3339 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3340 from DNS. If the selector record carries the flag, it now has
3341 precedence over the domain-wide flag.
3343 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3345 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3346 the use of a port name that isn't defined in /etc/services) occurred, the
3347 message was deferred as in a normal delivery, and thus remained on the
3348 spool, instead of being failed because of the mua_wrapper setting. This
3349 is now fixed, and I tidied up some of the mua_wrapper messages at the
3352 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3353 the messages in a hash of arrays rather than using individual hashes.
3354 This is a bit cleaner and results in dramatic memory savings, albeit
3355 at a slight CPU cost.
3357 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3358 as requested by Marc Sherman.
3360 SC/10 Eximstats: added histograms for user specified patterns as requested
3363 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3365 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3366 fopen() in the content-scanning modules that did not already have it.
3372 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3373 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3375 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3377 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3379 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3380 to test Sieve filters that use "vacation".
3382 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3383 that changes the way the GnuTLS parameters are stored in the cache file.
3384 The new format can be generated externally. For backward compatibility,
3385 if the data in the cache doesn't make sense, Exim assumes it has read an
3386 old-format file, and it generates new data and writes a new file. This
3387 means that you can't go back to an older release without removing the
3390 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3391 work if there are any delivery delays because "one_time" forces the
3392 parent to be marked "delivered", so its unseen clone is never tried
3393 again. For this reason, Exim now forbids the simultaneous setting of
3396 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3397 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3398 entry if you want to know the details. The fix, however, neglected to
3399 consider the case when local delivery batching is involved. The test for
3400 "previously delivered" was not happening when checking to see if an
3401 address could be batched with a previous (undelivered) one; under
3402 certain circumstances this could lead to multiple deliveries to the same
3405 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3406 in its include files, and this causes problems building Exim.
3408 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3409 header_syntax) but Exim was just ignoring anything given after a slash.
3410 In particular, this caused confusion with an attempt to use "verify =
3411 reverse_host_lookup/defer_ok". An error is now given when options are
3412 supplied for verify items that do not have them. (Maybe reverse_host_
3413 lookup should have a defer_ok option, but that's a different point.)
3415 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3416 defined by RFC 821) to 2048, because there were problems with some AUTH
3417 commands, and RFC 1869 says the size should be increased for extended
3418 SMTP commands that take arguments.
3420 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3423 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3424 "unknown" error; now it says that the functionality isn't in the binary.
3426 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3427 an address' error message when a string expansion fails (syntax or
3428 whatever). Otherwise the password may appear in the log. Following change
3429 PH/42 below, there is no longer a chance of it appearing in a bounce
3432 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3434 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3435 of its addresses was ever tried. (Bugzilla bug #2.)
3437 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3438 the result incorrectly in the debug output. (It correctly added a newline
3439 to what was transported.)
3441 TF/01 Added $received_time.
3443 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3444 commented out examples of how to interface to a virus scanner and to
3445 SpamAssassin. Also added commented examples of av_scanner and
3446 spamd_address settings.
3448 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3449 and controls are allowed in which ACLs. There were a couple of minor
3450 errors. Some of the entries in the conditions table (which is a table of
3451 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3452 negation of where the condition IS allowed.
3454 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3456 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3457 header file does not have a version number, so I've had to invent a new
3458 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3459 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3460 radiusclient), but it was contributed by a Radius user.
3462 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3463 files or directories, for OpenSSL.
3465 PH/19 When an Exim process that is running as root has to create an Exim log
3466 file, it does so in a subprocess that runs as exim:exim so as to get the
3467 ownership right at creation (otherwise, other Exim processes might see
3468 the file with the wrong ownership). There was no test for failure of this
3469 fork() call, which would lead to the process getting stuck as it waited
3470 for a non-existent subprocess. Forks do occasionally fail when resources
3471 run out. I reviewed all the other calls to fork(); they all seem to check
3474 PH/20 When checking for unexpected SMTP input at connect time (before writing
3475 the banner), Exim was not dealing correctly with a non-positive return
3476 from the read() function. If the client had disconnected by this time,
3477 the result was a log entry for a synchronization error with an empty
3478 string after "input=" when read() returned zero. If read() returned -1
3479 (an event I could not check), uninitialized data bytes were printed.
3480 There were reports of junk text (parts of files, etc) appearing after
3483 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3485 PH/22 Added support for macro redefinition, and (re)definition in between
3486 driver and ACL definitions.
3488 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3489 forgetting to use the resulting value; it was using the unexpanded value.
3491 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3492 hadn't been configured. The fix is from Juergen Kreileder, who
3493 understands it better than I do:
3495 "Here's what I see happening with three configured cyrus_sasl
3496 authenticators configured (plain, login, cram-md5):
3498 On startup auth_cyrus_sasl_init() gets called for each of these.
3499 This means three calls to sasl_listmech() without a specified mech_list.
3500 => SASL tests which mechs of all available mechs actually work
3501 => three warnings about OTP not working
3502 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3504 With the patch, sasl_listmech() also gets called three times. But now
3505 SASL's mech_list option is set to the server_mech specified in the the
3506 authenticator. Or in other words, the answer from sasl_listmech()
3507 gets limited to just the mech you're testing for (which is different
3509 => the return list contains just 'plain' or 'login', 'cram-md5' or
3510 nothing depending on the value of ob->server_mech.
3512 I've just tested the patch: Authentication still works fine,
3513 unavailable mechs specified in the exim configuration are still
3514 caught, and the auth.log warnings about OTP are gone."
3516 PH/25 When debugging is enabled, the contents of the command line are added
3517 to the debugging output, even when log_selector=+arguments is not
3520 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3521 answer is "GNU", and only if the return is "GNU/something" is the answer
3524 PH/27 $acl_verify_message is now set immediately after the failure of a
3525 verification in an ACL, and so is available in subsequent modifiers. In
3526 particular, the message can be preserved by coding like this:
3528 warn !verify = sender
3529 set acl_m0 = $acl_verify_message
3531 Previously, $acl_verify_message was set only while expanding "message"
3532 and "log_message" when a very denied access.
3534 PH/28 Modified OS/os.c-Linux with
3536 -#ifndef OS_LOAD_AVERAGE
3537 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3539 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3540 nomenclature these days.)
3542 PH/29 Installed patch from the Sieve maintainer that adds the options
3543 sieve_useraddress and sieve_subaddress to the redirect router.
3545 PH/30 In these circumstances:
3546 . Two addresses routed to the same list of hosts;
3547 . First host does not offer TLS;
3548 . First host accepts first address;
3549 . First host gives temporary error to second address;
3550 . Second host offers TLS and a TLS session is established;
3551 . Second host accepts second address.
3552 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3553 and peerdn, if requested) that were in fact used only for the second
3556 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3557 attention to any local part prefix or suffix that was matched by the
3558 router that accepted the address. It now behaves in the same way as it
3559 does for delivery: the affixes are removed from the local part unless
3560 rcpt_include_affixes is set on the transport.
3562 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3563 timeout during the DATA phase of an incoming message.
3565 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3566 applied a patch sanctioned by the Sieve maintainer.
3568 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3569 the uid or gid is negative. A case of a negative gid caused this to be
3570 noticed. The fix allows for either to be negative.
3572 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3573 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3576 PH/36 The change PH/12 above was broken. Fixed it.
3578 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3579 the grounds that routing the same address twice would always produce the
3580 same answer. This might have been true once, but it is certainly no
3581 longer true now. Routing a child address may depend on the previous
3582 routing that produced that child. Some complicated redirection strategies
3583 went wrong when messages had multiple recipients, and made Exim's
3584 behaviour dependent on the order in which the addresses were given.
3586 I have moved the duplicate checking until after the routing is complete.
3587 Exim scans the addresses that are assigned to local and remote
3588 transports, and removes any duplicates. This means that more work will be
3589 done, as duplicates will always all be routed, but duplicates are
3590 presumably rare, so I don't expect this is of any significance.
3592 For deliveries to pipes, files, and autoreplies, the duplicate checking
3593 still happens during the routing process, since they are not going to be
3596 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3597 It corrects a timeout issue with spamd. This is Ian's comment: "The
3598 background is that sometimes spamd either never reads data from a
3599 connection it has accepted, or it never writes response data. The exiscan
3600 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3601 blindly assumes that writes won't block so it may never time out."
3603 PH/39 Allow G after quota size as well as K and M.
3605 PH/40 The value set for $authenticated_id in an authenticator may not contain
3606 binary zeroes or newlines because the value is written to log lines and
3607 to spool files. There was no check on this. Now the value is run through
3608 the string_printing() function so that such characters are converted to
3609 printable escape sequences.
3611 PH/41 $message_linecount is a new variable that contains the total number of
3612 lines in the message. Compare $body_linecount, which is the count for the
3615 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3616 bounce and delay warning messages, except in certain special cases, which
3619 (a) An SMTP error message from a remote host;
3620 (b) A message specified in a :fail: redirection;
3621 (c) A message specified in a "fail" command in a system filter;
3622 (d) A message specified in a FAIL return from the queryprogram router;
3623 (e) A message specified by the cannot_route_message router option.
3625 In these cases only, Exim does include the error details in bounce and
3626 warning messages. There are also a few cases where bland messages such
3627 as "unrouteable address" or "local delivery error" are given.
3629 PH/43 $value is now also set for the "else" part of a ${run expansion.
3631 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3632 being worked on, but at least Exim now implements the latest version to
3635 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3636 process to complete was treated as a delivery failure, a timeout while
3637 writing the message to the pipe was logged, but erroneously treated as a
3638 successful delivery. Such timeouts include transport filter timeouts. For
3639 consistency with the overall process timeout, these timeouts are now
3640 treated as errors, giving rise to delivery failures by default. However,
3641 there is now a new Boolean option for the pipe transport called
3642 timeout_defer, which, if set TRUE, converts the failures into defers for
3643 both kinds of timeout. A transport filter timeout is now identified in
3646 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3647 systems where "make" and "gmake" are different, calling "gmake" at top
3648 level broke things. I've arranged for the value of $(MAKE) to be passed
3649 from the Makefile to this script so that it can call the same version of
3653 A note about Exim versions 4.44 and 4.50
3654 ----------------------------------------
3656 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3657 changes of various kinds. As a consequence, a big documentation update was
3658 needed. This delayed the release for rather longer than seemed good, especially
3659 in the light of a couple of (minor) security issues. Therefore, the changes
3660 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3661 release. So 4.44 and 4.50 are in effect two different branches that both start
3664 I have left the 4.50 change log unchanged; it contains all the changes since
3665 4.43. The change log for 4.44 is below; many of its items are identical to
3666 those for 4.50. This seems to be the most sensible way to preserve the
3667 historical information.
3673 1. Minor wording change to the doc/README.SIEVE file.
3675 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3676 computation of the current number of files was incorrect.
3678 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3679 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3682 4. Give more explanation in the error message when the command for a transport
3683 filter fails to execute.
3685 5. There are several places where Exim runs a non-Exim command in a
3686 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3687 was being done only for the command run by the queryprogram router. It is
3688 now done for all such subprocesses. The other cases are: ${run, transport
3689 filters, and the commands run by the lmtp and pipe transports.
3691 6. Added CONFIGURE_GROUP build-time option.
3693 7. Some older OS have a limit of 256 on the maximum number of file
3694 descriptors. Exim was using setrlimit() to set 1000 as a large value
3695 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3696 systems. I've change it so that if it can't get 1000, it tries for 256.
3698 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3699 was an oversight, and furthermore, ever since the addition of extra
3700 controls (e.g. 4.43/32), the checks on when to allow different forms of
3701 "control" were broken. There should now be diagnostics for all cases when a
3702 control that does not make sense is encountered.
3704 9. Added the /retain_sender option to "control=submission".
3706 10. $recipients is now available in the predata ACL (oversight).
3708 11. Tidy the search cache before the fork to do a delivery from a message
3709 received from the command line. Otherwise the child will trigger a lookup
3710 failure and thereby defer the delivery if it tries to use (for example) a
3711 cached ldap connection that the parent has called unbind on.
3713 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3714 of $address_data from the recipient verification was clobbered by the
3715 sender verification.
3717 13. The value of address_data from a sender verification is now available in
3718 $sender_address_data in subsequent conditions in the ACL statement.
3720 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3722 15. Added a new option "connect=<time>" to callout options, to set a different
3725 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3726 was its contents. (It was OK if the option was not defined at all.)
3728 17. A "Completed" log line is now written for messages that are removed from
3729 the spool by the -Mrm option.
3731 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3732 information about exactly what failed.
3734 19. Added -dd to debug only the daemon process.
3736 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3737 handles timeouts, both on the server side and network timeouts. Renamed the
3738 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3740 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3741 It is now set to "smtps".
3743 22. $host_address is now set to the target address during the checking of
3744 ignore_target_hosts.
3746 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3747 being passed; this would have caused $sender_host_name to have been used if
3748 matching the list had actually called for a host name (not very likely,
3749 since this list is usually IP addresses). A host name is now passed as
3752 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3753 code that specifies a non-restarting handler (typically sigaction() in
3754 modern systems) in an attempt to fix a rare and obscure crash bug.
3756 25. Narrowed the window for a race in the daemon that could cause it to ignore
3757 SIGCHLD signals. This is not a major problem, because they are used only to
3758 wake it up if nothing else does.
3760 26. A malformed maildirsize file could cause Exim to calculate negative values
3761 for the mailbox size or file count. Odd effects could occur as a result.
3762 The maildirsize information is now recalculated if the size or filecount
3765 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3766 support for a long time. Removed HAVE_SYS_VFS_H.
3768 28. Installed the latest version of exipick from John Jetmore.
3770 29. In an address list, if the pattern was not a regular expression, an empty
3771 subject address (from a bounce message) matched only if the pattern was an
3772 empty string. Non-empty patterns were not even tested. This was the wrong
3773 because it is perfectly reasonable to use an empty address as part of a
3774 database query. An empty address is now tested by patterns that are
3775 lookups. However, all the other forms of pattern expect the subject to
3776 contain a local part and a domain, and therefore, for them, an empty
3777 address still always fails if the pattern is not itself empty.
3779 30. Exim went into a mad DNS loop when attempting to do a callout where the
3780 host was specified on an smtp transport, and looking it up yielded more
3781 than one IP address.
3783 31. Re-factored the code for checking spool and log partition space into a
3784 function that finds that data and another that does the check. The former
3785 is then used to implement four new variables: $spool_space, $log_space,
3786 $spool_inodes, and $log_inodes.
3788 32. The RFC2047 encoding function was originally intended for short strings
3789 such as real names; it was not keeping to the 75-character limit for
3790 encoded words that the RFC imposes. It now respects the limit, and
3791 generates multiple encoded words if necessary. To be on the safe side, I
3792 have increased the buffer size for the ${rfc2047: expansion operator from
3795 33. It is now permitted to omit both strings after an "if" condition; if the
3796 condition is true, the result is "true". As before, when the second string
3797 is omitted, a false condition yields an empty string. This makes it less
3798 cumbersome to write custom ACL and router conditions.
3800 34. Failure to deliver a bounce message always caused it to be frozen, even if
3801 there was an errors_to setting on the router. The errors_to setting is now
3804 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3805 canonical form (fully expanded) before being placed in
3806 $sender_host_address.
3808 36. The table in the code that translates DNS record types into text (T_A to
3809 "A" for instance) was missing entries for NS and CNAME. It is just possible
3810 that this could have caused confusion if both these types were looked up
3811 for the same domain, because the text type is used as part of Exim's
3812 per-process caching. But the chance of anyone hitting this buglet seems
3815 37. The dnsdb lookup has been extended in a number of ways.
3817 (1) There is a new type, "zns", which walks up the domain tree until it
3818 finds some nameserver records. It should be used with care.
3820 (2) There is a new type, "mxh", which is like "mx" except that it returns
3821 just the host names, not the priorities.
3823 (3) It is now possible to give a list of domains (or IP addresses) to be
3824 looked up. The behaviour when one of the lookups defers can be
3825 controlled by a keyword.
3827 (4) It is now possible to specify the separator character for use when
3828 multiple records are returned.
3830 38. The dnslists ACL condition has been extended: it is now possible to supply
3831 a list of IP addresses and/or domains to be looked up in a particular DNS
3834 39. Added log_selector=+queue_time_overall.
3836 40. When running the queue in the test harness, wait just a tad after forking a
3837 delivery process, to get repeatability of debugging output.
3839 41. Include certificate and key file names in error message when GnuTLS fails
3840 to set them up, because the GnuTLS error message doesn't include the name
3841 of the failing file when there is a problem reading it.
3843 42. Allow both -bf and -bF in the same test run.
3845 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3847 44. The "Exiscan patch" is now merged into the mainline Exim source.
3849 45. Sometimes the final signoff response after QUIT could fail to get
3850 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3851 < 0 before doing a fflush(). This bug looks as though it goes back to the
3852 introduction of TLS in release 3.20, but "sometimes" must have been rare
3853 because the tests only now provoked it.
3855 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3856 (this can affect the format of dates).
3858 47. exim_tidydb, when checking for the continued existence of a message for
3859 which it has found a message-specific retry record, was not finding
3860 messages that were in split spool directories. Consequently, it was
3861 deleting retry records that should have stayed in existence.
3863 48. Steve fixed some bugs in eximstats.
3865 49. The SPA authentication driver was not abandoning authentication and moving
3866 on to the next authenticator when an expansion was forced to fail,
3867 contradicting the general specification for all authenticators. Instead it
3868 was generating a temporary error. It now behaves as specified.
3870 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3871 (the order specifies the preference for clients). The order is now AES256,
3872 AES128, 3DES, ARCFOUR128.
3874 51. Small patch to Sieve code - explicitly set From: when generating an
3877 52. Exim crashed if a remote delivery caused a very long error message to be
3878 recorded - for instance if somebody sent an entire SpamAssassin report back
3879 as a large number of 550 error lines. This bug was coincidentally fixed by
3880 increasing the size of one of Exim's internal buffers (big_buffer) that
3881 happened as part of the Exiscan merge. However, to be on the safe side, I
3882 have made the code more robust (and fixed the comments that describe what
3885 53. Now that there can be additional text after "Completed" in log lines (if
3886 the queue_time_overall log selector is set), a one-byte patch to exigrep
3887 was needed to allow it to recognize "Completed" as not the last thing in
3890 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3891 patch that reportedly fixes this has been added. I am not expert enough to
3892 create a test for it. This is what the patch creator wrote:
3894 "I found a little strange behaviour of ldap code when working with
3895 Windows 2003 AD Domain, where users was placed in more than one
3896 Organization Units. When I tried to give exim partial DN, the exit code
3897 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3898 But simultaneously result of request was absolutely normal ldap result,
3899 so I produce this patch..."
3901 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3902 REFERENCE, so I have modified the code to exclude the patch when that macro
3905 55. Some experimental protocols are using DNS PTR records for new purposes. The
3906 keys for these records are domain names, not reversed IP addresses. The
3907 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3908 leaves it alone. Component reversal etc. now happens only for IP addresses.
3911 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3913 57. Double the size of the debug message buffer (to 2048) so that more of very
3914 long debug lines gets shown.
3916 58. The exicyclog utility now does better if the number of log files to keep
3917 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3919 59. Two changes related to the smtp_active_hostname option:
3921 (1) $smtp_active_hostname is now available as a variable.
3922 (2) The default for smtp_banner uses $smtp_active_hostname instead
3923 of $primary_hostname.
3925 60. The host_aton() function is supposed to be passed a string that is known
3926 to be a valid IP address. However, in the case of IPv6 addresses, it was
3927 not checking this. This is a hostage to fortune. Exim now panics and dies
3928 if the condition is not met. A case was found where this could be provoked
3929 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3930 components; fortuitously, this particular loophole had already been fixed
3931 by change 4.50/55 above.
3933 If there are any other similar loopholes, the new check in host_aton()
3934 itself should stop them being exploited. The report I received stated that
3935 data on the command line could provoke the exploit when Exim was running as
3936 exim, but did not say which command line option was involved. All I could
3937 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3938 running as the user.
3941 61. There was a buffer overflow vulnerability in the SPA authentication code
3942 (which came originally from the Samba project). I have added a test to the
3943 spa_base64_to_bits() function which I hope fixes it.
3946 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3947 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3949 63. The daemon start-up calls getloadavg() while still root for those OS that
3950 need the first call to be done as root, but it missed one case: when
3951 deliver_queue_load_max is set with deliver_drop_privilege. This is
3952 necessary for the benefit of the queue runner, because there is no re-exec
3953 when deliver_drop_privilege is set.
3955 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3956 This has been fixed.
3958 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3959 was in use, was not putting the data itself into the right store pool;
3960 consequently, it could be overwritten for a subsequent message in the same
3961 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3964 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3966 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3967 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3968 the function were treating the return as a boolean value, which happened to
3969 work because 0=false and not-0=true, but is not correct code.
3971 68. The host_aton() function was not handling scoped IPv6 addresses (those
3972 with, for example, "%eth0" on the end) correctly.
3974 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3975 negated items (that is, ~something) in unsigned ints. Some compilers
3976 apparently mutter when there is no cast.
3978 70. If an address verification called from an ACL failed, and did not produce a
3979 user-specific message (i.e. there was only a "system" message), nothing was
3980 put in $acl_verify_message. In this situation, it now puts the system
3983 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3984 session; change 4.31/43 added the unwanted input to the log line - except
3985 that it did not do this in the start of session case. It now does.
3987 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3988 This is wrong and can cause the other end to generate a synchronization
3989 error if it is another Exim or anything else that does the synchronization
3990 check. A QUIT command is no longer sent after a timeout.
3992 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3993 during host lookups.
3995 74. The defer_ok option of callout verification was not working if it was used
3996 when verifying addresses in header lines, that is, for this case:
3998 verify = header_sender/callout=defer_ok
4000 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4001 those file descriptors could be used for SMTP connections. If anything
4002 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4003 could be sent to the SMTP client, causing chaos. The daemon now opens
4004 stdin, stdout, and stderr to /dev/null when it puts itself into the
4007 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4008 log by default. The user can override this with suitable Perl magic.
4010 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4011 the log message when discard triggers, was not working for the DATA ACL or
4012 for the non-SMTP ACL.
4014 78. Error message wording change in sieve.c.
4016 79. If smtp_accept_max_per_host was set, the number of connections could be
4017 restricted to fewer than expected, because the daemon was trying to set up
4018 a new connection before checking whether the processes handling previous
4019 connections had finished. The check for completed processes is now done
4020 earlier. On busy systems, this bug wouldn't be noticed because something
4021 else would have woken the daemon, and it would have reaped the completed
4024 80. If a message was submitted locally by a user whose login name contained one
4025 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4026 It caused a spool format error. I have fixed the spool reading code. A
4027 related problem was that the "from" clause in the Received: line became
4028 illegal because of the space(s). It is now covered by ${quote_local_part.
4030 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4033 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4035 83. Patch from Sieve maintainer to fix unterminated string problem in
4036 "vacation" handling.
4038 84. Some minor changes to the Linux configuration files to help with other
4039 OS variants using glibc.
4041 85. One more patch for Sieve to update vacation handling to latest spec.
4044 ----------------------------------------------------
4045 See the note above about the 4.44 and 4.50 releases.
4046 ----------------------------------------------------
4052 1. Change 4.43/35 introduced a bug that caused file counts to be
4053 incorrectly computed when quota_filecount was set in an appendfile
4056 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4057 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4060 3. Give more explanation in the error message when the command for a transport
4061 filter fails to execute.
4063 4. There are several places where Exim runs a non-Exim command in a
4064 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4065 was being done only for the command run by the queryprogram router. It is
4066 now done for all such subprocesses. The other cases are: ${run, transport
4067 filters, and the commands run by the lmtp and pipe transports.
4069 5. Some older OS have a limit of 256 on the maximum number of file
4070 descriptors. Exim was using setrlimit() to set 1000 as a large value
4071 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4072 systems. I've change it so that if it can't get 1000, it tries for 256.
4074 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4075 was an oversight, and furthermore, ever since the addition of extra
4076 controls (e.g. 4.43/32), the checks on when to allow different forms of
4077 "control" were broken. There should now be diagnostics for all cases when a
4078 control that does not make sense is encountered.
4080 7. $recipients is now available in the predata ACL (oversight).
4082 8. Tidy the search cache before the fork to do a delivery from a message
4083 received from the command line. Otherwise the child will trigger a lookup
4084 failure and thereby defer the delivery if it tries to use (for example) a
4085 cached ldap connection that the parent has called unbind on.
4087 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4088 of $address_data from the recipient verification was clobbered by the
4089 sender verification.
4091 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4092 was its contents. (It was OK if the option was not defined at all.)
4094 11. A "Completed" log line is now written for messages that are removed from
4095 the spool by the -Mrm option.
4097 12. $host_address is now set to the target address during the checking of
4098 ignore_target_hosts.
4100 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4101 being passed; this would have caused $sender_host_name to have been used if
4102 matching the list had actually called for a host name (not very likely,
4103 since this list is usually IP addresses). A host name is now passed as
4106 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4107 code that specifies a non-restarting handler (typically sigaction() in
4108 modern systems) in an attempt to fix a rare and obscure crash bug.
4110 15. Narrowed the window for a race in the daemon that could cause it to ignore
4111 SIGCHLD signals. This is not a major problem, because they are used only to
4112 wake it up if nothing else does.
4114 16. A malformed maildirsize file could cause Exim to calculate negative values
4115 for the mailbox size or file count. Odd effects could occur as a result.
4116 The maildirsize information is now recalculated if the size or filecount
4119 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4120 support for a long time. Removed HAVE_SYS_VFS_H.
4122 18. Updated exipick to current release from John Jetmore.
4124 19. Allow an empty sender to be matched against a lookup in an address list.
4125 Previously the only cases considered were a regular expression, or an
4128 20. Exim went into a mad DNS lookup loop when doing a callout where the
4129 host was specified on the transport, if the DNS lookup yielded more than
4132 21. The RFC2047 encoding function was originally intended for short strings
4133 such as real names; it was not keeping to the 75-character limit for
4134 encoded words that the RFC imposes. It now respects the limit, and
4135 generates multiple encoded words if necessary. To be on the safe side, I
4136 have increased the buffer size for the ${rfc2047: expansion operator from
4139 22. Failure to deliver a bounce message always caused it to be frozen, even if
4140 there was an errors_to setting on the router. The errors_to setting is now
4143 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4144 canonical form (fully expanded) before being placed in
4145 $sender_host_address.
4147 24. Updated eximstats to version 1.33
4149 25. Include certificate and key file names in error message when GnuTLS fails
4150 to set them up, because the GnuTLS error message doesn't include the name
4151 of the failing file when there is a problem reading it.
4153 26. Expand error message when OpenSSL has problems setting up cert/key files.
4156 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4157 (this can affect the format of dates).
4159 28. exim_tidydb, when checking for the continued existence of a message for
4160 which it has found a message-specific retry record, was not finding
4161 messages that were in split spool directories. Consequently, it was
4162 deleting retry records that should have stayed in existence.
4164 29. eximstats updated to version 1.35
4165 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4166 1.35 - bugfix such that pie charts by volume are generated correctly
4168 30. The SPA authentication driver was not abandoning authentication and moving
4169 on to the next authenticator when an expansion was forced to fail,
4170 contradicting the general specification for all authenticators. Instead it
4171 was generating a temporary error. It now behaves as specified.
4173 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4174 (the order specifies the preference for clients). The order is now AES256,
4175 AES128, 3DES, ARCFOUR128.
4177 31. Small patch to Sieve code - explicitly set From: when generating an
4180 32. Exim crashed if a remote delivery caused a very long error message to be
4181 recorded - for instance if somebody sent an entire SpamAssassin report back
4182 as a large number of 550 error lines. This bug was coincidentally fixed by
4183 increasing the size of one of Exim's internal buffers (big_buffer) that
4184 happened as part of the Exiscan merge. However, to be on the safe side, I
4185 have made the code more robust (and fixed the comments that describe what
4188 33. Some experimental protocols are using DNS PTR records for new purposes. The
4189 keys for these records are domain names, not reversed IP addresses. The
4190 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4191 leaves it alone. Component reversal etc. now happens only for IP addresses.
4194 34. The host_aton() function is supposed to be passed a string that is known
4195 to be a valid IP address. However, in the case of IPv6 addresses, it was
4196 not checking this. This is a hostage to fortune. Exim now panics and dies
4197 if the condition is not met. A case was found where this could be provoked
4198 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4199 components; fortuitously, this particular loophole had already been fixed
4200 by change 4.50/55 or 4.44/33 above.
4202 If there are any other similar loopholes, the new check in host_aton()
4203 itself should stop them being exploited. The report I received stated that
4204 data on the command line could provoke the exploit when Exim was running as
4205 exim, but did not say which command line option was involved. All I could
4206 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4207 running as the user.
4210 35. There was a buffer overflow vulnerability in the SPA authentication code
4211 (which came originally from the Samba project). I have added a test to the
4212 spa_base64_to_bits() function which I hope fixes it.
4215 36. The daemon start-up calls getloadavg() while still root for those OS that
4216 need the first call to be done as root, but it missed one case: when
4217 deliver_queue_load_max is set with deliver_drop_privilege. This is
4218 necessary for the benefit of the queue runner, because there is no re-exec
4219 when deliver_drop_privilege is set.
4221 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4222 was in use, was not putting the data itself into the right store pool;
4223 consequently, it could be overwritten for a subsequent message in the same
4224 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4227 38. Sometimes the final signoff response after QUIT could fail to get
4228 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4229 < 0 before doing a fflush(). This bug looks as though it goes back to the
4230 introduction of TLS in release 3.20, but "sometimes" must have been rare
4231 because the tests only now provoked it.
4237 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4238 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4239 Now it returns an integer. A number of calls were still expecting a T/F
4240 return. Fortuitously, in all cases, the tests worked in OK situations,
4241 which is the norm. However, things would have gone wrong on any write
4242 failures on the smtp file descriptor. This function is used when sending
4243 messages over SMTP and also when doing verify callouts.
4245 2. When Exim is called to do synchronous delivery of a locally submitted
4246 message (the -odf or -odi options), it no longer closes stderr before doing
4249 3. Implemented the mua_wrapper option.
4251 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4253 5. Implemented the functions header_remove(), header_testname(),
4254 header_add_at_position(), and receive_remove_recipient(), and exported them
4257 6. If an ACL "warn" statement specified the addition of headers, Exim already
4258 inserted X-ACL-Warn: at the start if there was no header name. However, it
4259 was not making this test for the second and subsequent header lines if
4260 there were newlines in the string. This meant that an invalid header could
4261 be inserted if Exim was badly configured.
4263 7. Allow an ACL "warn" statement to add header lines at the start or after all
4264 the Received: headers, as well as at the end.
4266 8. Added the rcpt_4xx retry error code.
4268 9. Added postmaster_mailfrom=xxx to callout verification option.
4270 10. Added mailfrom=xxxx to the callout verification option, for verify=
4273 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4274 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4276 12. Inserted some casts to stop certain compilers warning when using pointer
4277 differences as field lengths or precisions in printf-type calls (mostly
4278 affecting debugging statements).
4280 13. Added optional readline() support for -be (dynamically loaded).
4282 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4283 same clock tick as a message's arrival, so that its received time was the
4284 same as the "first fail" time on the retry record, and that message
4285 remained on the queue past the ultimate address timeout, every queue runner
4286 would try a delivery (because it was past the ultimate address timeout) but
4287 after another failure, the ultimate address timeout, which should have then
4288 bounced the address, did not kick in. This was a "< instead of <=" error;
4289 in most cases the first failure would have been in the next clock tick
4290 after the received time, and all would be well.
4292 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4293 being recognized when the domain list was tested by the match_domain
4294 condition in an expansion string.
4296 16. Added the ${str2b64: operator.
4298 17. Exim was always calling setrlimit() to set a large limit for the number of
4299 processes, without checking whether the existing limit was already
4300 adequate. (It did check for the limit on file descriptors.) Furthermore,
4301 errors from getrlimit() and setrlimit() were being ignored. Now they are
4302 logged to the main and panic logs, but Exim does carry on, to try to do its
4303 job under whatever limits there are.
4305 18. Imported PCRE 5.0.
4307 19. Trivial typo in log message " temporarily refused connection" (the leading
4310 20. If the log selector return_path_on_delivery was set and an address was
4311 redirected to /dev/null, the delivery process crashed because it assumed
4312 that a return path would always be set for a "successful" delivery. In this
4313 case, the whole delivery is bypassed as an optimization, and therefore no
4316 21. Internal re-arrangement: the function for sending a challenge and reading
4317 a response while authentication was assuming a zero-terminated challenge
4318 string. It's now changed to take a pointer and a length, to allow for
4319 binary data in such strings.
4321 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4323 23. Exim was not respecting finduser_retries when seeking the login of the
4324 uid under which it was called; it was always trying 10 times. (The default
4325 setting of finduser_retries is zero.) Also, it was sleeping after the final
4326 failure, which is pointless.
4328 24. Implemented tls_on_connect_ports.
4330 25. Implemented acl_smtp_predata.
4332 26. If the domain in control=submission is set empty, Exim assumes that the
4333 authenticated id is a complete email address when it generates From: or
4334 Sender: header lines.
4336 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4337 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4338 chown and chgrp in /bin and hostname in /usr/bin.
4340 28. Exim was keeping the "process log" file open after each use, just as it
4341 does for the main log. This opens the possibility of it remaining open for
4342 long periods when the USR1 signal hits a daemon. Occasional processlog
4343 errors were reported, that could have been caused by this. Anyway, it seems
4344 much more sensible not to leave this file open at all, so that is what now
4347 29. The long-running daemon process does not normally write to the log once it
4348 has entered its main loop, and it closes the log before doing so. This is
4349 so that log files can straightforwardly be renamed and moved. However,
4350 there are a couple of unusual error situations where the daemon does write
4351 log entries, and I had neglected to close the log afterwards.
4353 30. The text of an SMTP error response that was received during a remote
4354 delivery was being truncated at 512 bytes. This is too short for some of
4355 the long messages that one sometimes sees. I've increased the limit to
4358 31. It is now possible to make retry rules that apply only when a message has a
4359 specific sender, in particular, an empty sender.
4361 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4362 it possible to be selective about when SMTP synchronization is enforced.
4364 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4366 32. Implemented hosts_connection_nolog.
4368 33. Added an ACL for QUIT.
4370 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4373 35. Added mailbox_size and mailbox_filecount to appendfile.
4375 36. Added control = no_multiline_responses to ACLs.
4377 37. There was a bug in the logic of the code that waits for the clock to tick
4378 in the case where the clock went backwards by a substantial amount such
4379 that the microsecond fraction of "now" was more than the microsecond
4380 fraction of "then" (but the whole seconds number was less).
4382 38. Added support for the libradius Radius client library this is found on
4383 FreeBSD (previously only the radiusclient library was supported).
4389 1. When certain lookups returned multiple values in the form name=value, the
4390 quoting of the values was not always being done properly. Specifically:
4391 (a) If the value started with a double quote, but contained no whitespace,
4393 (b) If the value contained whitespace other than a space character (i.e.
4394 tabs or newlines or carriage returns) it was not quoted.
4395 This fix has been applied to the mysql and pgsql lookups by writing a
4396 separate quoting function and calling it from the lookup code. The fix
4397 should probably also be applied to nisplus, ibase and oracle lookups, but
4398 since I cannot test any of those, I have not disturbed their existing code.
4400 2. A hit in the callout cache for a specific address caused a log line with no
4401 reason for rejecting RCPT. Now it says "Previous (cached) callout
4402 verification failure".
4404 3. There was an off-by-one bug in the queryprogram router. An over-long
4405 return line was truncated at 256 instead of 255 characters, thereby
4406 overflowing its buffer with the terminating zero. As well as fixing this, I
4407 have increased the buffer size to 1024 (and made a note to document this).
4409 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4410 when Exim is waiting for an SMTP response from a remote server, Exim
4411 restarts its select() call on the socket, thereby resetting its timeout.
4412 This is not a problem when such interrupts are rare. Somebody set up a cron
4413 job to run exiwhat every 2 minutes, which is less than the normal select()
4414 timeout (5 or 10 minutes). This meant that the select() timeout never
4415 kicked in because it was always reset. I have fixed this by comparing the
4416 time when an interrupt arrives with the time at the start of the first call
4417 to select(). If more time than the timeout has elapsed, the interrupt is
4418 treated as a timeout.
4420 5. Some internal re-factoring in preparation for the addition of Sieve
4421 extensions (by MH). In particular, the "personal" test is moved to a
4422 separate function, and given an option for scanning Cc: and Bcc: (which is
4423 not set for Exim filters).
4425 6. When Exim created an email address using the login of the caller as the
4426 local part (e.g. when creating a From: or Sender: header line), it was not
4427 quoting the local part when it contained special characters such as @.
4429 7. Installed new OpenBSD configuration files.
4431 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4432 try to make them clearer.
4434 9. Callout options, other than the timeout value, were being ignored when
4435 verifying sender addresses in header lines. For example, when using
4437 verify = header_sender/callout=no_cache
4439 the cache was (incorrectly) being used.
4441 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4442 only the Cygwin environment.
4444 11. When return_path_on_delivery was set as a log selector, if different remote
4445 addresses in the same message used different return paths and parallel
4446 remote delivery occurred, the wrong values would sometimes be logged.
4447 (Whenever a remote delivery process finished, the return path value from
4448 the most recently started remote delivery process was logged.)
4450 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4451 lines when AUTH and/or TLS are in use. This is the "received protocol"
4452 field. Exim used to use "asmtp" for authenticated SMTP, without any
4453 indication (in the protocol name) for TLS use. Now it follows the RFC and
4454 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4455 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4456 names appear in log lines as well as in Received: header lines.
4458 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4459 extensions, and comparison tests, and to fix some bugs.
4461 14. Changes to the "personal" filter test:
4463 (1) The test was buggy in that it was just doing the equivalent of
4464 "contains" tests on header lines. For example, if a user's address was
4465 anne@some.where, the "personal" test would incorrectly be true for
4467 To: susanne@some.where
4469 This test is now done by extracting each address from the header in turn,
4470 and checking the entire address. Other tests that are part of "personal"
4471 are now done using regular expressions (for example, to check local parts
4472 of addresses in From: header lines).
4474 (2) The list of non-personal local parts in From: addresses has been
4475 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4476 taken from the Sieve specification recommendations.
4478 (3) If the message contains any header line starting with "List-" it is
4479 treated as non-personal.
4481 (4) The test for "circular" in the Subject: header line has been removed
4482 because it now seems ill-conceived.
4484 15. Minor typos in src/EDITME comments corrected.
4486 16. Installed latest exipick from John Jetmore.
4488 17. If headers_add on a router specified a text string that was too long for
4489 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4490 of string_sprintf() is now avoided.
4492 18. $message_body_size was not set (it was always zero) when running the DATA
4493 ACL and the local_scan() function.
4495 19. For the "mail" command in an Exim filter, no default was being set for
4496 the once_repeat time, causing a random time value to be used if "once" was
4497 specified. (If the value happened to be <= 0, no repeat happened.) The
4498 default is now 0s, meaning "never repeat". The "vacation" command was OK
4499 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4500 (I found it when inspecting the code).
4502 20. There is now an overall timeout for performing a callout verification. It
4503 defaults to 4 times the callout timeout, which applies to individual SMTP
4504 commands during the callout. The overall timeout applies when there is more
4505 than one host that can be tried. The timeout is checked before trying the
4506 next host. This prevents very long delays if there are a large number of
4507 hosts and all are timing out (e.g. when the network connections are timing
4508 out). The value of the overall timeout can be changed by specifying an
4509 additional sub-option for "callout", called "maxwait". For example:
4511 verify = sender/callout=5s,maxwait=20s
4513 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4514 to the end before writing, but this should make it even safer).
4516 22. Exim was forgetting that it had advertised PIPELINING for the second and
4517 subsequent messages on an SMTP connection. It was also not resetting its
4518 memory on STARTTLS and an internal HELO.
4520 23. When Exim logs an SMTP synchronization error within a session, it now
4521 records whether PIPELINING has been advertised or not.
4523 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4524 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4525 rather than long int.
4527 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4529 26. Added the never_mail option to autoreply.
4535 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4536 crash if the getsockname() call failed; this can happen if a connection is
4537 closed very soon after it is established. The problem was simply in the
4538 order in which certain operations were done, causing Exim to try to write
4539 to the SMTP stream before it had set up the file descriptor. The bug has
4540 been fixed by making things happen in the correct order.
4546 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4547 before the connection was closed, thus losing the rejection response.
4549 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4550 some early Solaris releases, but causes trouble in current releases where
4551 socklen_t is defined.
4553 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4556 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4559 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4560 that releasing the top of it at the end releases what was used for sub-
4561 expansions (unless the block got too big). However, discard this block if
4562 the first thing is a variable or header, so that we can use its block when
4563 it is dynamic (useful for very large $message_headers, for example).
4565 6. Lookups now cache *every* query, not just the most recent. A new, separate
4566 store pool is used for this. It can be recovered when all lookup caches are
4567 flushed. Lookups now release memory at the end of their result strings.
4568 This has involved some general refactoring of the lookup sources.
4570 7. Some code has been added to the store_xxx() functions to reduce the amount
4571 of flapping under certain conditions.
4573 8. log_incoming_interface used to affect only the <= reception log lines. Now
4574 it causes the local interface and port to be added to several more SMTP log
4575 lines, for example "SMTP connection from", and rejection lines.
4577 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4579 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4581 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4582 could be overwritten at the end of the current message (or the start of a
4583 new message if it was set in a HELO ACL). The value is now preserved for
4584 the duration of the SMTP connection.
4586 12. If a transport had a headers_rewrite setting, and a matching header line
4587 contained an unqualified address, that address was qualified, even if it
4588 did not match any rewriting rules. The underlying bug was that the values
4589 of the flags that permit the existence of unqualified sender and recipient
4590 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4591 non-local messages, and by -bnq for local messages) were not being
4592 preserved with the message after it was received.
4594 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4595 "next input=" as part of the text comprising the host identity instead of
4596 the correct text. The code was using the same buffer for two different
4597 strings. However, depending on which order the printing function evaluated
4598 its arguments, the bug did not always show up. Under Linux, for example, my
4599 test suite worked just fine.
4601 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4602 doesn't work with some older versions of Perl. It has been changed to "my",
4603 which in any case is probably the better facility to use.
4605 15. A really picky compiler found some instances of statements for creating
4606 error messages that either had too many or two few arguments for the format
4609 16. The size of the buffer for calls to the DNS resolver has been increased
4610 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4611 for addresses that have a lot of PTR records. This alleviates a problem; it
4612 does not fully solve it.
4614 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4615 buffer now truncates the list and logs the incident, which is the same
4616 action as happens when Exim is looking up a host name and its aliases.
4617 Previously in this situation something unpredictable would happen;
4618 sometimes it was "internal error: store_reset failed".
4620 18. If a server dropped the connection unexpectedly when an Exim client was
4621 using GnuTLS and trying to read a response, the client delivery process
4622 crashed while trying to generate an error log message.
4624 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4625 string, for example:
4627 warn message = H1: something\nH2: something
4629 the text was added as a single header line from Exim's point of view
4630 though it ended up OK in the delivered message. However, searching for the
4631 second and subsequent header lines using $h_h2: did not work. This has been
4632 fixed. Similarly, if a system filter added multiple headers in this way,
4633 the routers could not see them.
4635 20. Expanded the error message when iplsearch is called with an invalid key to
4636 suggest using net-iplsearch in a host list.
4638 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4639 ACLs are no longer actually imposed (and a message to that effect is
4642 22. If a "gecos" field in a passwd entry contained escaped characters, in
4643 particular, if it contained a \" sequence, Exim got it wrong when building
4644 a From: or a Sender: header from that name. A second bug also caused
4645 incorrect handling when an unquoted " was present following a character
4646 that needed quoting.
4648 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4649 was not being matched caselessly.
4651 24. Arranged for all hyphens in the exim.8 source to be escaped with
4654 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4655 properly. Recipient callout cache records were still being keyed to include
4656 the sender, even when use_sender was set false. This led to far more
4657 callouts that were necessary. The sender is no longer included in the key
4658 when use_sender is false.
4660 26. Added "control = submission" modifier to ACLs.
4662 27. Added the ${base62d: operator to decode base 62 numbers.
4664 28. dnsdb lookups can now access SRV records.
4666 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4667 the configuration file.
4669 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4670 (-v) mode. This makes the output for a verbose queue run more intelligible.
4672 31. Added a use_postmaster feature to recipient callouts.
4674 32. Added the $body_zerocount variable, containing the number of binary zero
4675 bytes in the message body.
4677 33. The time of last modification of the "new" subdirectory is now used as the
4678 "mailbox time last read" when there is a quota error for a maildir
4681 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4683 35. Added +ignore_unknown as a special item in host lists.
4685 36. Code for decoding IPv6 addresses in host lists is now included, even if
4686 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4687 address was recognized as an IP address, but was then not correctly decoded
4688 into binary, causing unexpected and incorrect effects when compared with
4695 1. Very minor rewording of debugging text in manualroute to say "list of
4696 hosts" instead of "hostlist".
4698 2. If verify=header_syntax was set, and a header line with an unqualified
4699 address (no domain) and a large number of spaces between the end of the
4700 name and the colon was received, the reception process suffered a buffer
4701 overflow, and (when I tested it) crashed. This was caused by some obsolete
4702 code that should have been removed. The fix is to remove it!
4704 3. When running in the test harness, delay a bit after writing a bounce
4705 message to get a bit more predictability in the log output.
4707 4. Added a call to search_tidyup() just before forking a reception process. In
4708 theory, someone could use a lookup in the expansion of smtp_accept_max_
4709 per_host which, without the tidyup, could leave open a database connection.
4711 5. Added the variables $recipient_data and $sender_data which get set from a
4712 lookup success in an ACL "recipients" or "senders" condition, or a router
4713 "senders" option, similar to $domain_data and $local_part_data.
4715 6. Moved the writing of debug_print from before to after the "senders" test
4718 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4719 problems for message scanning, either using a data ACL, or using
4720 local_scan() because the Received: header was not generated till after they
4721 were called (in order to set the time as the time of reception completion).
4722 I have revised the way this works. The header is now generated after the
4723 body is received, but before the ACL or local_scan() are called. After they
4724 are run, the timestamp in the header is updated.
4730 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4731 before starting a queue runner without re-exec. This happened only when
4732 deliver_drop_privilege was set or when the Exim user was set to root. The
4733 effect of the bug was that timeouts during subsequent deliveries caused
4734 crashes instead of being properly handled. The handler is now left at its
4735 default (and expected) setting.
4737 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4738 message, again when deliver_drop_privilege is set or Exim is run as root.
4739 The bug described in (1) was not present in this case, but the tidying up
4740 of the other signals was missing. I have made the two cases consistent.
4742 3. The ignore_target_hosts setting on a manualroute router was being ignored
4743 for hosts that were looked up using the /MX notation.
4745 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4748 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4749 operated on the sender address. After changing the $sender_address to <>
4750 for the sender address verify, Exim was re-instated it as the original
4751 (before rewriting) address, but remembering that it had rewritten it, so it
4752 wasn't rewriting it again. This bug also had the effect of breaking the
4753 sender address verification caching when the sender address was rewritten.
4755 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4756 This has been changed so that if the ip literal address matches
4757 ignore_target_hosts, the router declines.
4759 7. Added expansion conditions match_domain, match_address, and match_local_
4760 part (NOT match_host).
4762 8. The placeholder for the Received: header didn't have a length field set.
4764 9. Added code to Exim itself and to exim_lock to test for a specific race
4765 condition that could lead to file corruption when using MBX delivery. The
4766 issue is with the lockfile that is created in /tmp. If this file is removed
4767 after a process has opened it but before that process has acquired a lock,
4768 there is the potential for a second process to recreate the file and also
4769 acquire a lock. This could lead to two Exim processes writing to the file
4770 at the same time. The added code performs the same test as UW imapd; it
4771 checks after acquiring the lock that its file descriptor still refers to
4772 the same named file.
4774 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4775 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4778 11. Added the smtp_active_hostname option. If used, this will typically be made
4779 to depend on the incoming interface address. Because $interface_address is
4780 not set up until the daemon has forked a reception process, error responses
4781 that can happen earlier (such as "too many connections") no longer contain
4784 12. If an expansion in a condition on a "warn" statement fails because a lookup
4785 defers, the "warn" statement is abandoned, and the next ACL statement is
4786 processed. Previously this caused the whole ACL to be aborted.
4788 13. Added the iplsearch lookup type.
4790 14. Added ident_timeout as a log selector.
4792 15. Added tls_certificate_verified as a log selector.
4794 16. Added a global option tls_require_ciphers (compare the smtp transport
4795 option of the same name). This controls incoming TLS connections.
4797 17. I finally figured out how to make tls_require_ciphers do a similar thing
4798 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4799 before starting the TLS session.
4801 18. Tabs are now shown as \t in -bP output.
4803 19. If the log selector return_path_on_delivery was set, Exim crashed when
4804 bouncing a message because it had too many Received: header lines.
4806 20. If two routers both had headers_remove settings, and the first one included
4807 a superfluous trailing colon, the final name in the first list and the
4808 first name in the second list were incorrectly joined into one item (with a
4809 colon in the middle).
4815 1. Added -C and -D options to the exinext utility, mainly to make it easier
4816 to include in the automated testing, but these could be helpful when
4817 multiple configurations are in use.
4819 2. The exinext utility was not formatting the output nicely when there was
4820 an alternate port involved in the retry record key, nor when there was a
4821 message id as well (for retries that were specific to a specific message
4822 and a specific host). It was also confused by IPv6 addresses, because of
4823 the additional colons they contain. I have fixed the IPv4 problem, and
4824 patched it up to do a reasonable job for IPv6.
4826 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4827 delivery, the log line now contains "pipelined" if PIPELINING was used.
4829 4. An SMTP transport process used to panic and die if the bind() call to set
4830 an explicit outgoing interface failed. This has been changed; it is now
4831 treated in the same way as a connect() failure.
4833 5. A reference to $sender_host_name in the part of a conditional expansion
4834 that was being skipped was still causing a DNS lookup. This no longer
4837 6. The def: expansion condition was not recognizing references to header lines
4838 that used bh_ and bheader_.
4840 7. Added the _cache feature to named lists.
4842 8. The code for checking quota_filecount in the appendfile transport was
4843 allowing one more file than it should have been.
4845 9. For compatibility with Sendmail, the command line option
4853 and sets the incoming protocol and host name (for trusted callers). The
4854 host name and its colon can be omitted when only the protocol is to be set.
4855 Note the Exim already has two private options, -pd and -ps, that refer to
4856 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4857 "s", but I don't think that's a major issue.
4859 10. A number of refactoring changes to the code, none of which should affect
4862 (a) The number of logging options was getting close to filling up the
4863 32-bit word that was used as a bit map. I have split them into two classes:
4864 those that are passed in the argument to log_write(), and those that are
4865 only ever tested independently outside of that function. These are now in
4866 separate 32-bit words, so there is plenty of room for expansion again.
4867 There is no change in the user interface or the logging behaviour.
4869 (b) When building, for example, log lines, the code previously used a
4870 macro that called string_cat() twice, in order to add two strings. This is
4871 not really sufficiently general. Furthermore, there was one instance where
4872 it was actually wrong because one of the argument was used twice, and in
4873 one call a function was used. (As it happened, calling the function twice
4874 did not affect the overall behaviour.) The macro has been replaced by a
4875 function that can join an arbitrary number of extra strings onto a growing
4878 (c) The code for expansion conditions now uses a table and a binary chop
4879 instead of a serial search (which was left over from when there were very
4880 few conditions). Also, it now recognizes conditions like "pam" even when
4881 the relevant support is not compiled in: a suitably worded error message is
4882 given if an attempt is made to use such a condition.
4884 11. Added ${time_interval:xxxxx}.
4886 12. A bug was causing one of the ddress fields not to be passed back correctly
4887 from remote delivery subprocesses. The field in question was not being
4888 subsequently used, so this caused to problems in practice.
4890 13. Added new log selectors queue_time and deliver_time.
4892 14. Might have fixed a bug in maildirsizefile handling that threw up
4893 "unexpected character" debug warnings, and recalculated the data
4894 unnecessarily. In any case, I expanded the warning message to give more
4897 15. Added the message "Restricted characters in address" to the statements in
4898 the default ACL that block characters like @ and % in local parts.
4900 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4901 Three changes have been made:
4903 (a) There was a serious bug; a negative response to MAIL caused the whole
4904 recipient domain to be cached as invalid, thereby blocking all messages
4905 to all local parts at the same domain, from all senders. This bug has
4906 been fixed. The domain is no longer cached after a negative response to
4907 MAIL if the sender used is not empty.
4909 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4912 (c) A new callout option, "use_sender" has been added for people who want
4913 the modified behaviour.
4919 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4922 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4923 indeed breaks things for older releases.
4925 3. Added additional logging to the case where there is a problem reading data
4926 from a filter that is running in a subprocess using a pipe, in order to
4927 try to track down a specific problem.
4929 4. Testing facility fudge: when running in the test harness and attempting
4930 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4931 getting "No route to host". Convert this to a timeout.
4933 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4936 6. Some OS don't have socklen_t but use size_t instead. This affects the
4937 fifth argument of getsockopt() amongst other things. This is now
4938 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4939 can be set for individual OS. I have set it for SunOS5, OSF1, and
4940 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4941 some earlier ones do not.
4943 7. Change 4.30/15 was not doing the test caselessly.
4945 8. The standard form for an IPv6 address literal was being rejected by address
4946 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4947 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4948 this, as well as the form without the "IPv6" on the front (but only when
4949 address literals are enabled, of course).
4951 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4953 10. Exim crashed if a message with an empty sender address specified by -f
4954 encountered a router with an errors_to setting. This could be provoked only
4955 by a command such as
4959 where an empty string was supplied; "<>" did not hit this bug.
4961 11. Installed PCRE release 4.5.
4963 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4964 remained set. It is now erased.
4966 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4967 times from message ids (which are base 36 rather than the normal 62).
4969 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4970 were being counted as actual protocol errors, and logged if the log
4971 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4972 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4973 and DATA following a set of rejected RCPTs do not count as protocol errors.
4974 In other words, Exim assumes they were pipelined, though this may not
4975 actually be the case. Of course, in all cases the client gets an
4976 appropriate error code.
4978 15. If a lookup fails in an ACL condition, a message about the failure may
4979 be available; it is used if testing the ACL cannot continue, because most
4980 such messages specify what the cause of the deferral is. However, some
4981 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4982 that caused an old message to be retained and used if a later statement
4983 caused a defer, replacing the real cause of the deferral.
4985 16. If an IP address had so many PTR records that the DNS lookup buffer
4986 was not large enough to hold them, Exim could crash while trying to process
4987 the truncated data. It now detects and logs this case.
4989 17. Further to 4.21/58, another change has been made: if (and only if) the
4990 first line of a message (the first header line) ends with CRLF, a bare LF
4991 in a subsequent header line has a space inserted after it, so as not to
4992 terminate the header.
4994 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4995 unnecessarily, used atoi() instead of strtol(), and didn't check the
4996 termination when getting file sizes from file names by regex.
4998 19. Completely re-implemented the support for maildirsize files, in the light
4999 of a number of problems with the previous contributed implementation
5000 (4.30/29). In particular:
5002 . If the quota is zero, the maildirsize file is maintained, but no quota is
5005 . If the maildir directory does not exist, it is created before any attempt
5006 to write a maildirsize file.
5008 . The quota value in the file is just a cache; if the quota is changed in
5009 the transport, the new value overrides.
5011 . A regular expression is available for excluding directories from the
5014 20. The autoreply transport checks the characters in options that define the
5015 message's headers; it allows continued headers, but it was checking with
5016 isspace() after an embedded newline instead of explicitly looking for a
5019 21. If all the "regular" hosts to which an address was routed had passed their
5020 expiry times, and had not reached their retry times, the address was
5021 bounced, even if fallback hosts were defined. Now Exim should go on to try
5024 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5025 equivalent code in the SMTP transport. Some hosts send humungous responses
5026 to HELO/EHLO, more than 1024 it seems.
5028 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5029 gives compiler warnings in some environments. I've now done it "properly",
5032 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5033 (because of problems with the built-in one) was declared to return uschar *
5034 instead of char *, causing compiler failure.
5036 25. Fixed a file descriptor leak when processing alias/forward files.
5038 26. Fixed a minor format string issue in dbfn.c.
5040 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5042 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5043 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5046 29. When a sender address is verified, it is cached, to save repeating the test
5047 when there is more than one recipient in a message. However, when the
5048 verification involves a callout, it is possible for different callout
5049 options to be set for different recipients. It is too complicated to keep
5050 track of this in the cache, so now Exim always runs a verification when a
5051 callout is required, relying on the callout cache for the optimization.
5052 The overhead is duplication of the address routing, but this should not be
5055 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5056 to be verified with callout=postmaster, and the main callout worked but the
5057 postmaster check failed, the verification correctly failed. However, if a
5058 subsequent RCPT command asked for sender verification *without* the
5059 postmaster check, incorrect caching caused this verification also to fail,
5062 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5063 it was not caching the DNS options (qualify_single, search_parents) that
5064 were used when the lookup failed. A subsequent lookup with different
5065 options therefore always gave the same answer, though there were cases
5066 where it should not have. (Example: a "domains = !$mx_any" option on a
5067 dnslookup router: the "domains" option is always processed without any
5068 widening, but the router might have qualify_single set.) Now Exim uses the
5069 cached value only when the same options are set.
5071 32. Added John Jetmore's "exipick" utility to the distribution.
5073 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5074 than a timeout (e.g. a certificate is required, and is not provided), an
5075 Exim server now closes the connection immediately. Previously it waited for
5076 the client to close - but if the client is SSL, it seems that they each
5077 wait for each other, leading to a delay before one of them times out.
5079 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5080 maintained 0.8.x compatibility because I don't think many are using it, and
5081 it is clearly obsolete.
5083 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5086 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5087 client certificate was expired. A simple patch fixes this, though I don't
5088 understand the full logic of why the verify callback is called multiple
5091 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5092 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5093 which causes problems with some clients (such as the Certicom SSL Plus
5094 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5095 disables the coutermeasure allowing Eudora to connect."
5097 38. Exim was not checking that a write() to a log file succeeded. This could
5098 lead to Bad Things if a log got too big, in particular if it hit a file
5099 size limit. Exim now panics and dies if it cannot write to a log file, just
5100 as it does if it cannot open a log file.
5102 39. Modified OS/Makefile-Linux so that it now contains
5104 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5106 The two -D definitions ensure that Exim is compiled with large file
5107 support, which makes it possible to handle log files that are bigger than
5110 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5111 instance) a domain was checked against a named list that involved a lookup,
5112 causing $domain_data to be set, then another domain was checked against the
5113 same list, then the first domain was re-checked, the value of $domain_data
5114 after the final check could be wrong. In particular, if the second check
5115 failed, it could be set empty. This bug probably also applied to
5118 41. The strip_trailing_dot option was not being applied to the address given
5119 with the -f command-line option.
5121 42. The code for reading a message's header from the spool was incrementing
5122 $received_count, but never initializing it. This meant that the value was
5123 incorrect (doubled) while delivering a message in the same process in which
5124 it was received. In the most common configuration of Exim, this never
5125 happens - a fresh exec is done - but it can happen when
5126 deliver_drop_privilege is set.
5128 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5129 it now includes up to 150 characters of the unexpected data in the log
5132 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5133 and building data strings. The size of both of these buffers was 10 000
5134 bytes - far larger than anybody would *ever* want, thought I. Needless to
5135 say, somebody hit the limit. I have increased the maximum line length to
5136 20 000 and the maximum data length of concatenated lines to 100 000. I have
5137 also fixed two bugs, because there was no checking on these buffers. Tsk,
5138 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5139 buffer is too small.
5141 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5142 lsearch lookups. Now it does.
5144 46. When parsing a route_list item in a manualroute router, a fixed-length
5145 buffer was used for the list of hosts. I made this 1024 bytes long,
5146 thinking that nobody would ever have a list of hosts that long. Wrong.
5147 Somebody had a whole pile of complicated expansion conditions, and the
5148 string was silently truncated, leading to an expansion error. It turns out
5149 that it is easier to change to an unlimited length (owing to other changes
5150 that have happened since this code was originally written) than to build
5151 structure for giving a limitation error. The length of the item that
5152 expands into the list of hosts is now unlimited.
5154 47. The lsearch lookup could not handle data where the length of text line was
5155 more than 4095 characters. Such lines were truncated, leading to shortened
5156 data being returned. It should now handle lines of any length.
5158 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5159 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5162 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5163 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5164 no longer happens in comment lines. A list of replacements is now placed
5165 at the head of all of the source files, except those whose only change is
5166 to replace PERL_COMMAND in the very first #! line.
5168 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5169 messages on the queue, with a bottom-up merge sort, using code contributed
5170 by Michael Haardt. This should make operations like -bp somewhat faster on
5171 large queues. It won't affect queue runners, except when queue_run_in_order
5174 51. Installed eximstats 1.31 in the distribution.
5176 52. Added support for SRV lookups to the dnslookup router.
5178 53. If an ACL referred to $message_body or $message_body_end, the value was not
5179 reset for any messages that followed in the same SMTP session.
5181 54. The store-handling optimization for building very long strings was not
5182 differentiating between the different store pools. I don't think this
5183 actually made any difference in practice, but I've tidied it.
5185 55. While running the routers to verify a sender address, $sender_address
5186 was still set to the sender address. This is wrong, because when routing to
5187 send a bounce to the sender, it would be empty. Therefore, I have changed
5188 it so that, while verifying a sender address, $sender_address is set to <>.
5189 (There is no change to what happens when verifying a recipient address.)
5191 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5192 target A or AAAA records (if not already returned) without resetting the
5193 qualify_single or search_parents options of the DNS resolver. These are
5194 inappropriate in this case because the targets of MX and SRV records must
5195 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5196 target that, when qualified, matched something in the local domain. These
5197 two options are now turned off when doing these lookups.
5199 57. It seems that at least some releases of Reiserfs (which does not have the
5200 concept of a fixed number of inodes) returns zero and not -1 for the
5201 number of available inodes. This interacted badly with check_spool_inodes,
5202 which assumed that -1 was the "no such thing" setting. What I have done is
5203 to check that the total number of inodes is greater than zero before doing
5204 the test of how many are available.
5206 58. When a "warn" ACL statement has a log_message modifier, the message is
5207 remembered, and not repeated. This is to avoid a lot of repetition when a
5208 message has many recipients that cause the same warning to be written.
5209 Howewer, Exim was preserving the list of already written lines for an
5210 entire SMTP session, which doesn't seem right. The memory is now reset if a
5211 new message is started.
5213 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5214 debugging output unless log_rewrite was also set.
5216 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5217 of (int)(handle) when we know that handle contains (void *)(-1).
5219 61. The Exim daemon panic-logs an error return when it closes the incoming
5220 connection. However "connection reset by peer" seems to be common, and
5221 isn't really an error worthy of noting specially, so that particular error
5224 62. When Exim is trying to find all the local interfaces, it used to panic and
5225 die if the ioctl to get the interface flags failed. However, it seems that
5226 on at least one OS (Solaris 9) it is possible to have an interface that is
5227 included in the list of interfaces, but for which you get a failure error
5228 for this call. This happens when the interface is not "plumbed" into a
5229 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5230 failure of the "get flags" call assumes that the interface is down.
5232 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5233 makes life easier for people who are doing arithmetic on fields extracted
5234 from dates, where you often get leading zeros that should not be
5235 interpreted as octal.
5237 64. Added qualify_domain to the redirect router, to override the global
5240 65. If a pathologically long header line contained very many addresses (the
5241 report of this problem mentioned 10 000) and each of them was rewritten,
5242 Exim could use up a very large amount of memory. (It kept on making new
5243 copies of the header line as it rewrote, and never released the old ones.)
5244 At the expense of a bit more processing, the header rewriting function has
5245 been changed so that it no longer eats memory in this way.
5247 66. The generation of the Received: header has been moved from the time that a
5248 message starts to be received, to the time that it finishes. The timestamp
5249 in the Received: header should now be very close to that of the <= log
5250 line. There are two side-effects of this change:
5252 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5253 logged header lines no longer include the local Received: line, because
5254 it has not yet been created. The same applies to a copy of the message
5255 that is returned to a non-SMTP sender when a message is rejected.
5257 (b) When a filter file is tested using -bf, no additional Received: header
5258 is added to the test message. After some thought, I decided that this
5261 This change does not affect the value of $received_for. It is still set
5262 after address rewriting, but before local_scan() is called.
5264 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5266 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5267 gave an unhelpful panic error message, and a defer error. I have managed to
5268 change this behaviour so that it now rejects any supplied certificate,
5269 which seems right, as the list of acceptable certificates is empty.
5271 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5272 gave an unhelpful defer error. I have not managed to make this reject any
5273 supplied certificates, but the error message it gives is "no certificate
5274 supplied", which is not helpful.
5276 70. exigrep's output now also includes lines that are not associated with any
5277 message, but which match the given pattern. Implemented by a patch from
5278 Martin Sluka, which also tidied up the Perl a bit.
5280 71. Recipient callout verification, like sender verification, was using <> in
5281 the MAIL FROM command. This isn't really the right thing, since the actual
5282 sender may affect whether the remote host accepts the recipient or not. I
5283 have changed it to use the actual sender in the callout; this means that
5284 the cache record is now keyed on a recipient/sender pair, not just the
5285 recipient address. There doesn't seem to be a real danger of callout loops,
5286 since a callout by the remote host to check the sender would use <>.
5287 [SEE ABOVE: changed after hitting problems.]
5289 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5290 temporary errors. However, in the case of such a code being given after
5291 the end of a data transmission (i.e. after ".") Exim was failing to write
5292 a retry record for the message. (Yes, there was some broken host that was
5293 actually sending 8xx at this point.)
5295 73. An unknown lookup type in a host list could cause Exim to panic-die when
5296 the list was checked. (An example that provoked this was putting <; in the
5297 middle of a list instead of at the start.) If this happened during a DATA
5298 ACL check, a -D file could be left lying around. This kind of configuration
5299 error no longer causes Exim to die; instead it causes a defer errror. The
5300 incident is still logged to the main and panic logs.
5302 74. Buglet left over from Exim 3 conversion. The message "too many messages
5303 in one connection" was written to the rejectlog but not the mainlog, except
5304 when address rewriting (yes!) was being logged.
5306 75. Added write_rejectlog option.
5308 76. When a system filter was run not as root (that is, when system_filter_user
5309 was set), the values of the $n variables were not being returned to the
5310 main process; thus, they were not subsequently available in the $sn
5313 77. Added +return_path_on_delivery log selector.
5315 78. A connection timeout was being treated differently from recipients deferred
5316 when testing hosts_max_try with a message that was older than the host's
5317 retry timeout. (The host should not be counted, thus allowing all hosts to
5318 be tried at least once before bouncing.) This may have been the cause of an
5319 occasionally reported bug whereby a message would remain on the queue
5320 longer than the retry timeout, but would be bounced if a delivery was
5321 forced. I say "may" because I never totally pinned down the problem;
5322 setting up timeout/retry tests is difficult. See also the next item.
5324 79. The ultimate address timeout was not being applied to errors that involved
5325 a combination of host plus message (for example, a timeout on a MAIL
5326 command). When an address resolved to a number of possible hosts, and they
5327 were not all tried for each delivery (e.g. because of hosts_max_try), a
5328 message could remain on the queue longer than the retry timeout.
5330 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5333 81. Fixed an obscure SMTP outgoing bug which required at least the following
5334 conditions: (a) there was another message waiting for the same server;
5335 (b) the server returned 5xx to all RCPT commands in the first message so
5336 that the message was not completed; (c) the server dropped the connection
5337 or gave a negative response to the RSET that Exim sends to abort the
5338 transaction. The observed case was a dropped connection after DATA that had
5339 been sent in pipelining mode. That is, the server had advertised PIPELINING
5340 but was not implementing it correctly. The effect of the bug was incorrect
5341 behaviour, such as trying another host, and this could lead to a crash.
5347 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5348 and daemon.c were passed as pointers to ints; they should have been
5349 pointers to socklen_t variables (which are typically unsigned ints).
5351 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5354 3. Fixed a really odd bug that affected only the testing scheme; patching a
5355 certain fixed string in the binary changed the value of another string that
5356 happened to be identical to the end of the original first string.
5358 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5359 name", it returns that address as the IP address. On some operating
5360 systems (e.g. Solaris), it also passes back the IP address string as the
5361 "host name". However, on others (e.g. Linux), it passes back an empty
5362 string. Exim wasn't checking for this, and was changing the host name to an
5363 empty string, assuming it had been canonicized.
5365 5. Although rare, it is permitted to have more than one PTR record for a given
5366 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5367 all the names associated with an address, because they do in Solaris.
5368 However, it seems that they do not in Linux for data that comes from the
5369 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5370 I found this out when I moved to a new Linux workstation and tried to run
5371 the Exim test suite.
5373 To get round this problem I have changed the code so that it now does its
5374 own call to the DNS to look up PTR records when searching for a host name.
5375 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5376 addresses that are only in /etc/hosts are still found.
5378 This behaviour is, however, controlled by an option called host_lookup_
5379 order, which defaults to "bydns:byaddr". If people want to use the other
5380 order, or indeed, just use one or the other means of lookup, they can
5381 specify it in this variable.
5383 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5384 some operating systems, this comes back from gethostbyaddr() as an empty
5385 string, and this is what Exim used to test for. However, it seems that in
5386 other systems, "." is yielded. Exim now tests for this case too.
5388 7. The values of check_spool_space and check_log_space are now held internally
5389 as a number of kilobytes instead of an absolute number of bytes. If a
5390 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5391 kilobyte. This means that much larger values can be stored.
5393 8. Exim monitor: an attempt to get the action menu when not actually pointing
5394 at a message produces an empty menu entitled "No message selected". This
5395 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5396 no entries in it ("Shell widget menu has zero width and/or height"). So I
5397 have added a single, blank menu entry in this case.
5399 9. Added ${quote_local_part.
5401 10. MIME decoding is now applied to the contents of Subject: header lines when
5404 11. Now that a reference to $sender_host_address automatically causes a reverse
5405 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5406 host lookup before query-style lookups in lists that might use this
5407 variable. This has therefore been abolished, and the "net-" prefix is no
5408 longer necessary for query-style lookups.
5410 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5411 have been a typo for LFLAGS, so it has been changed.
5413 13. The install script calls Exim with "-C /dev/null" in order to find the
5414 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5415 to be output. Howeve, since Exim outputs its version number before the
5416 error, it didn't break the script. It just looked ugly. I fixed this by
5417 always allowing "-C /dev/null" if the caller is root.
5419 14. Ignore overlarge ACL variable number when reading spool file - insurance
5420 against a later release with more variables having written the file.
5422 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5423 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5424 this, as well as the form without the "IPv6" on the front.
5426 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5427 OS/Makefile-Darwin file.
5429 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5430 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5432 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5433 variable. However, some OS use ints for inodes. I've added cast to long int
5434 to get rid of the compiler warning.
5436 19. I had forgotten to lock out "/../" in configuration file names when
5437 ALT_CONFIG_PREFIX was set.
5439 20. Routers used for verification do not need to specify transports. However,
5440 if such a router generated a host list, and callout was configured, Exim
5441 crashed, because it could not find a port number from the (non-existent)
5442 transport. It now assumes port 25 in this circumstance.
5444 21. Added the -t option to exigrep.
5446 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5447 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5448 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5449 them is set, LOOKUP_LSEARCH is forced.
5451 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5453 24. Added sender and host information to the "rejected by local_scan()" log
5454 line; previously there was no indication of these.
5456 25. Added .include_if_exists.
5458 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5459 Linux. It turns out that not all file systems support this. Apparently some
5460 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5461 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5462 not supported on the file descriptor, is now ignored when Exim is trying to
5463 sync a directory. This applies only to Linux.
5465 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5467 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5468 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5471 29. Added support for maildirsize files from supplied patch (modified a bit).
5473 30. The use of :fail: followed by an empty string could lead Exim to respond to
5474 sender verification failures with (e.g.):
5476 550 Verification failed for <xxx>
5477 550 Sender verify failed
5479 where the first response line was missing the '-' that indicates it is not
5480 the final line of the response.
5482 31. The loop for finding the name of the user that called Exim had a hardwired
5483 limit of 10; it now uses the value of finduser_retries, which is used for
5484 all other user lookups.
5486 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5489 33. Exim was neglecting to zero errno before one call of strtol() when
5490 expanding a string and expecting an integer value. On some systems this
5491 resulted in spurious "integer overflow" errors. Also, it was casting the
5492 result into an int without checking.
5494 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5495 did not work. The code looks as if it has *never* worked, though it appears
5496 to have been documented since at least releast 1.62. I have made it work.
5498 35. The "timeout_DNS" error in retry rules, also documented since at least
5499 1.62, also never worked. As it isn't clear exactly what this means, and
5500 clearly it isn't a major issue, I have abolished the feature by treating it
5501 as "timeout", and writing a warning to the main and panic logs.
5503 36. The display of retry rules for -brt wasn't always showing the error code
5506 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5507 timeout_connect_A, timeout_connect_MX.
5509 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5510 to the empty sender.
5512 39. The daemon was not analysing the content of -oX till after it had closed
5513 stderr and disconnected from the controlling terminal. This meant that any
5514 syntax errors were only noted on the panic log, and the return code from
5515 the command was 0. By re-arranging the code a little, I've made the
5516 decoding happen first, so such errors now appear on stderr, and the return
5517 code is 1. However, the actual setting up of the sockets still happens in
5518 the disconnected process, so errors there are still only recorded on the
5521 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5522 connections (as happens on some IP stacks) was logged at start up time as
5523 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5524 it from "IPv6 and IPv4", which means that two separate sockets are being
5527 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5528 says whether AF_INET or AF_INET6 was passed as an argument.
5530 42. Exiwhat output was messed up when time zones were included in log
5533 43. Exiwhat now gives more information about the daemon's listening ports,
5534 and whether -tls-on-connect was used.
5536 44. The "port" option of the smtp transport is now expanded.
5538 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5539 silently ignored. Now an error message is written to the main and panic
5542 46. There's a new ACL modifier called "logwrite" which writes to a log file
5543 as soon as it is encountered.
5545 47. Added $local_user_uid and $local_user_gid at routing time.
5547 48. Exim crashed when trying to verify a sender address that was being
5550 49. Exim was recognizing only a space character after ".include". It now also
5551 recognizes a tab character.
5553 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5554 extracting the relevant information from the specification. The man page no
5555 longer contains scrambled data for the -d option, and I've added a section
5556 at the front about calling Exim under different names.
5558 51. Added "extra_headers" argument to the "mail" command in filter files.
5560 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5563 53. Installed eximstats 1.29.
5565 54. Added transport_filter_timeout as a generic transport option.
5567 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5568 Cc: header lines. This was required by RFC 822, but it not required by RFC
5571 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5572 incoming messages that did not have them. Now it does so only if the
5573 message originates locally, that is, if there is no associated remote host
5574 address. When Resent- header lines are present, this applies to the Resent-
5575 lines rather than the non-Resent- lines.
5577 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5578 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5580 58. Messages for configuration errors now include the name of the main
5581 configuration file - useful now that there may be more than one file in a
5582 list (.included file names were always shown).
5584 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5585 for those rare installations that do not start the daemon as root or run it
5586 setuid root. I've cut out the call to initgroups() if the daemon is not
5589 60. The Exim user and group can now be bound into the binary as text strings
5590 that are looked up at the start of Exim's processing.
5592 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5594 62. Added $mailstore_basename variable.
5596 63. Installed patch to sieve.c from Michael Haardt.
5598 64. When Exim failed to open the panic log after failing to open the main log,
5599 the original message it was trying to log was written to stderr and debug
5600 output, but if they were not available (the usual case in production), it
5601 was lost. Now it is written to syslog before the two lines that record the
5602 failures to open the logs.
5604 65. Users' Exim filters run in subprocesses under the user's uid. It is
5605 possible for a "deliver" command or an alias in a "personal" command to
5606 provoke an address rewrite. If logging of address rewriting is configured,
5607 this fails because the process is not running as root or exim. There may be
5608 a better way of dealing with this, but for the moment (because 4.30 needs
5609 to be released), I have disabled address rewrite logging when running a
5610 filter in a non-root, non-exim process.
5616 1. The buildconfig auxiliary program wasn't quoting the value set for
5617 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5618 not defined. This bug was masked in 4.22 by the effect that was fixed in
5621 2. Some messages that were rejected after a message id was allocated were
5622 shown as "incomplete" by exigrep. It no longer does this for messages that
5623 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5625 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5626 have allow_domain_literals set, the ID did not get logged in the <= line.
5627 Domain literals are now always recognized in Message-ID: header lines.
5629 4. The first argument for a ${extract expansion item is the key name or field
5630 number. Leading and trailing spaces in this item were not being ignored,
5631 causing some misleading effects.
5633 5. When deliver_drop_privilege was set, single queue runner processes started
5634 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5635 same command in the process it spins off) were not dropping privilege.
5637 6. When the daemon running as "exim" started a queue runner, it always
5638 re-executed Exim in the spun-off process. This is a waste of effort when
5639 deliver_drop_privilege is set. The new process now just calls the
5640 queue-runner function directly.
5646 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5649 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5650 ignored. Though the use of -oP was forcing the writing of a pid file, it
5651 was always written to the default place.
5653 3. If the message "no IP address found for host xxxx" is generated during
5654 incoming verification, it is now followed by identification of the incoming
5655 connection (so you can more easily find what provoked it).
5657 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5659 5. Added some features to "harden" Exim a bit more against certain attacks:
5661 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5662 be put in Local/Makefile. This is like the never_users runtime option,
5663 but it cannot be overridden. The default setting is "root".
5665 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5666 prefix string with which any file named in a -C command line option
5669 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5670 is retained for -C and -D only if the caller of Exim is root. Without
5671 it, the exim user may also use -C and -D and retain privilege.
5673 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5674 command line option is disabled.
5676 6. Macro names set by the -D option must start with an upper case letter, just
5677 like macro names defined in the configuration file.
5679 7. Added "dereference=" facility to LDAP.
5681 8. Two instances of the typo "uknown" in the source files are fixed.
5683 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5684 the Configure-Makefile script screwed up while processing it.
5686 10. Incorporated PCRE 4.4.
5688 11. The SMTP synchronization check was not operating right at the start of an
5689 SMTP session. For example, it could not catch a HELO sent before the client
5690 waited for the greeting. There is now a check for outstanding input at the
5691 point when the greeting is written. Because of the duplex, asynchronous
5692 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5693 way, but not yet received, when the check is performed.
5695 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5696 on TCP/IP sockets, because this apparently causes some broken clients to
5699 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5700 unchanged) from the Cygwin maintainer.
5702 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5703 was supported instead of testing for mbx. Effectively a typo.
5705 15. The spa authenticator server code was not checking that the input it
5706 received was valid base64.
5708 16. The debug output line for the "set" modifier in ACLs was not showing the
5709 name of the variable that was being set.
5711 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5713 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5714 lookup on its own. Something else was needed to trigger the lookup. For
5715 example, a match in host_lookup or the need for a host name in a host list.
5716 Now, if $sender_host_name is referenced and the host name has not yet been
5717 looked up, a lookup is performed. If the lookup fails, the variable remains
5718 empty, and $host_lookup_failed is set to "1".
5720 19. Added "eqi" as a case-independent comparison operator.
5722 20. The saslauthd authentication condition could segfault if neither service
5723 nor realm was specified.
5725 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5726 error message that was logged was misleading, and incoming SMTP
5727 connections were dropped. The message is now more accurate, and temporary
5728 errors are given to SMTP connections.
5730 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5731 (except RSET) with a 421 error, until QUIT is received. However, it was
5732 failing to send a response to QUIT.
5734 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5735 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5736 ACL is now run after the helo_try_verify_hosts code.
5738 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5739 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5740 case-independent, so other case variants are also recognized). Apparently
5741 some systems use these upper case variants.
5743 25. If more than two messages were waiting for the same host, and a transport
5744 filter was specified for the transport, Exim sent two messages over the
5745 same TCP/IP connection, and then failed with "socket operation on non-
5746 socket" when it tried to send the third.
5748 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5750 27. The extern definition of crypt16() in expand.c was not being excluded when
5751 the OS had its own crypt16() function.
5753 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5754 as a preferred synonym for return_size_limit, both as an option and as an
5757 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5759 30. Changed the default configuration ACL to relax the local part checking rule
5760 for addresses that are not in any local domains. For these addresses,
5761 slashes and pipe symbols are allowed within local parts, but the sequence
5762 /../ is explicitly forbidden.
5764 31. SPA server authentication was not clearing the challenge buffer before
5767 32. log_message in a "warn" ACL statement was writing to the reject log as
5768 well as to the main log, which contradicts the documentation and doesn't
5769 seem right (because no rejection is happening). So I have stopped it.
5771 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5772 However, I am unable to do any testing of this.
5774 34. Fixed an infelicity in the appendfile transport. When checking directories
5775 for a mailbox, to see if any needed to be created, it was accidentally
5776 using path names with one or more superfluous leading slashes; tracing
5777 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5779 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5780 contents are added to the log line that is written for every discarded
5781 recipient. (Previously a log_message setting was ignored.)
5783 36. The ${quote: operator now quotes the string if it is empty.
5785 37. The install script runs exim in order to find its version number. If for
5786 some reason other than non-existence or emptiness, which it checks, it
5787 could not run './exim', it was installing it with an empty version number,
5788 i.e. as "exim-". This error state is now caught, and the installation is
5791 38. An argument was missing from the function that creates an error message
5792 when Exim fails to connect to the socket for saslauthd authentication.
5793 This could cause Exim to crash, or give a corrupted message.
5795 39. Added isip, isip4, and isip6 to ${if conditions.
5797 40. The ACL variables $acl_xx are now saved with the message, and can be
5798 accessed later in routers, transports, and filters.
5800 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5801 strings in the file are not string-expanded.
5803 42. If a MAIL command specified a SIZE value that was too large to fit into an
5804 int variable, the check against message_size_limit failed. Such values are
5805 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5806 day this will have to be increased, but I don't think I want to be around
5807 when emails are that large.
5814 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5815 iconv() is not standard in FreeBSD.
5817 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5818 IPv6 enabled. The observed symptom was a segmentation fault on return from
5819 the function os_common_find_running_interfaces() in src/os.c.
5821 3. In the check_special_case() function in daemon.c I had used "errno" as an
5822 argument name, which causes warnings on some systems. This was basically a
5823 typo, since it was named "eno" in the comments!
5825 4. The code that waits for the clock to tick (at a resolution of some fraction
5826 of a second) so as to ensure message-id uniqueness was always waiting for
5827 at least one whole tick, when it could have waited for less. [This is
5828 almost certainly not relevant at current processor speeds, where it is
5829 unlikely to ever wait at all. But we try to future-proof.]
5831 5. The function that sleeps for a time interval that includes fractions of a
5832 second contained a race. It did not block SIGALRM between setting the
5833 timer, and suspending (a couple of lines later). If the interval was short
5834 and the sigsuspend() was delayed until after it had expired, the suspension
5835 never ended. On busy systems this could lead to processes getting stuck for
5838 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5839 process, before it forks any delivery processes. The open lookup caching
5840 mechanism meant that the open file or database connection was passed into
5841 the delivery process. The problem was that delivery processes always tidy
5842 up cached lookup data. This could cause a problem for the next delivery
5843 process started by the queue runner, because the external queue runner
5844 process does not know about the closure. So the next delivery process
5845 still has data in the lookup cache. In the case of a file lookup, there was
5846 no problem because closing a file descriptor in a subprocess doesn't affect
5847 the parent. However, if the lookup was caching a connection to a database,
5848 the connection was closed, and the second delivery process was likely to
5849 see errors such as "PGSQL: query failed: server closed the connection
5850 unexpectedly". The problem has been fixed by closing all cached lookups
5851 in a queue runner before running a delivery process.
5853 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5854 seem to have the "const" qualifier which it has on other OS. I've
5857 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5858 *of the same type* (client or server) with the same public name that an
5859 error should be diagnosed.
5861 9. When Exim looked up a host name for an IP address, but failed to find the
5862 original IP address when looking up the host name (a safety check), it
5863 output the message "<ip address> does not match any IP for NULL", which was
5864 confusing, to say the least. The bug was that the host name should have
5865 appeared instead of "NULL".
5867 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5868 user that is built into the binary, and the -C or -D options is used, root
5869 privilege is dropped before the configuration file is read. In addition,
5870 logging is switched to stderr instead of the normal log files. If the
5871 configuration then re-defines the Exim user, the unprivileged environment
5872 is probably not what is expected, so Exim logs a panic warning message (but
5875 However, if deliver_drop_privilege is set, the unprivileged state may well
5876 be exactly what is intended, so the warning has been cut out in that case,
5877 and Exim is allowed to try to write to its normal log files.
5883 1. smtp_return_error_details was not giving details for temporary sender
5884 or receiver verification errors.
5886 2. Diagnose a configuration error if two authenticators have the same public
5889 3. Exim used not to create the message log file for a message until the first
5890 delivery attempt. This could be confusing when incoming messages were held
5891 for policy or load reasons. The message log file is now created at the time
5892 the message is received, and an initial "Received" line is written to it.
5894 4. The automatically generated man page for command line options had a minor
5895 bug that caused no ill effects; however, a more serious problem was that
5896 the procedure for building the man page automatically didn't always
5897 operate. Consequently, release 4.20 contains an out-of-date version. This
5898 shouldn't happen again.
5900 5. When building Exim with embedded Perl support, the script that builds the
5901 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5902 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5904 6. The freeze_tell option was not being used for messages that were frozen on
5905 arrival, either by an ACL or by local_scan().
5907 7. Added the smtp_incomplete_transaction log selector.
5909 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5910 was accepting AUTH without a new EHLO.
5912 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5913 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5916 10. Logging of TCP/IP connections (when configured) now happens in the main
5917 daemon process instead of the child process, so that the TCP/IP connection
5918 count is more accurate (but it can never be perfect).
5920 11. The use of "drop" in a nested ACL was not being handled correctly in the
5921 outer ACL. Now, if condition failure induced by the nested "drop" causes
5922 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5923 or "require"), the connection is dropped.
5925 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5926 that yield "discard" can now be used with an "accept" or a "discard" verb,
5927 but an error is generated for any others (because I can't see a useful way
5928 to define what should happen).
5930 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5931 are now processed in the same way as lines in the Exim configuration file.
5932 In particular, continuation lines are supported.
5934 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5936 15. Added -ti meaning -t -i.
5938 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5939 domains, and warn by logging if others are found.
5941 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5942 returned from GIFCONF when Exim is trying to find the list of interfaces on
5943 a host. The code in os.c has been modified to copy each ifreq to an aligned
5944 structure in all cases.
5946 Also, in some cases, the returned ifreq's were being copied to a 'struct
5947 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5948 means the last couple of bytes of an IPv6 address could be chopped if the
5949 ifreq contained only a normal sockaddr (14 bytes storage).
5951 18. Named domain lists were not supported in the hosts_treat_as_local option.
5952 An entry such as +xxxx was not recognized, and was treated as a literal
5955 19. Ensure that header lines added by a DATA ACL are included in the reject log
5956 if the ACL subsequently rejects the message.
5958 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5959 MD5 (which is deprecated).
5961 21. When testing a filter file using -bf, Exim was writing a message when it
5962 took the sender from a "From " line in the message, but it was not doing so
5963 when it took $return_path from a Return-Path: header line. It now does.
5965 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5966 with a valid header line field name (a series of printing characters
5967 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5969 23. Changed "disc" in the source to "disk" to conform to the documentation and
5970 the book and for uniformity.
5972 24. Ignore Sendmail's -Ooption=value command line item.
5974 25. When execve() failed while trying to run a command in a pipe transport,
5975 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5976 could be confused with a return value of 69 from the command itself. This
5977 has been changed to 127, the value the shell returns if it is asked to run
5978 a non-existent command. The wording for the related log line suggests a
5979 non-existent command as the problem.
5981 26. If received_header_text expands to an empty string, do not add a Received:
5982 header line to the message. (Well, it adds a token one on the spool, but
5983 marks it "old" so that it doesn't get used or transmitted.)
5985 27. Installed eximstats 1.28 (addition of -nt option).
5987 28. There was no check for failure on the call to getsockname() in the daemon
5988 code. This can fail if there is a shortage of resources on the system, with
5989 ENOMEM, for example. A temporary error is now given on failure.
5991 29. Contrary to the C standard, it seems that in some environments, the
5992 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5993 program. Exim now does this explicitly; it affects the formatting of
5994 timestamps using strftime().
5996 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5997 complaints. I've now initialized all the variables, to avoid this.
5999 32. Header lines added by a system filter were not being "seen" during
6000 transport-time rewrites.
6002 33. The info_callback() function passed to OpenSSL is set up with type void
6003 (*)(SSL *, int, int), as described somewhere. However, when calling the
6004 function (actually a macro) that sets it up, the type void(*)() is
6005 expected. I've put in a cast to prevent warnings from picky compilers.
6007 34. If a DNS black list lookup found a CNAME record, but there were no A
6008 records associated with the domain it pointed at, Exim crashed.
6010 35. If a DNS black list lookup returned more than one A record, Exim ignored
6011 all but the first. It now scans all returned addresses if a particular IP
6012 value is being sought. In this situation, the contents of the
6013 $dnslist_value variable are a list of all the addresses, separated by a
6016 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6017 to accept a host name and all its aliases if the forward lookup for any of
6018 them yielded the IP address of the incoming connection. Now it accepts only
6019 those names whose forward lookup yields the correct IP address. Any other
6020 names are discarded. This closes a loophole whereby a rogue DNS
6021 administrator could create reverse DNS records to break through a
6022 wildcarded host restriction in an ACL.
6024 37. If a user filter or a system filter that ran in a subprocess used any of
6025 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6026 the wrong values were passed to the pipe command ($thisaddress had the
6027 value of $0, $0 had the value of $1, etc). This bug was introduced by
6028 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6030 38. Improved the line breaking for long SMTP error messages from ACLs.
6031 Previously, if there was no break point between 40 and 75 characters, Exim
6032 left the rest of the message alone. Two changes have been made: (a) I've
6033 reduced the minimum length to 35 characters; (b) if it can't find a break
6034 point between 35 and 75 characters, it looks ahead and uses the first one
6035 that it finds. This may give the occasional overlong line, but at least the
6036 remaining text gets split now.
6038 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6039 file descriptors might be low, and that setting 1000 would always raise it.
6040 It turns out that in some environments, the limit is already over 1000 and
6041 that lowering it causes trouble. So now Exim takes care not to decrease it.
6043 40. When delivering a message, the value of $return_path is set to $sender_
6044 address at the start of routing (routers may change the value). By an
6045 oversight, this default was not being set up when an address was tested by
6046 -bt or -bv, which affected the outcome if any router or filter referred to
6049 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6050 log only when "message" or "log_message" are set. However, if one of the
6051 conditions was an address verification, or a call to a nested ACL, the
6052 messages generated by the underlying test were being passed through. This
6053 no longer happens. The underlying message is available in $acl_verify_
6054 message for both "message" and "log_message" expansions, so it can be
6055 passed through if needed.
6057 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6058 new expansion $bh_ to give the encoded byte string without charset
6059 translation. Translation happens only if iconv() is available; HAVE_ICONV
6060 indicates this at build time. HEADERS_CHARSET gives the charset to
6061 translate to; headers_charset can change it in the configuration, and
6062 "headers charset" can change it in an individual filter file.
6064 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6065 that creates RFC 2047 encoded "words" labels them as that charset instead
6066 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6067 expansion operator; (ii) when Exim creates a From: line for a local
6068 message; (iii) when a header line is rewritten to include a "phrase" part.
6070 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6071 buggy, causing it to skip the first lines of messages whose message ID
6072 ended in 'D'. This would not have bitten before Exim release 4.14, because
6073 message IDs were unlikely to end in 'D' before then. The effect was to have
6074 incorrect size information for certain domains.
6076 45. #include "config.h" was missing at the start of the crypt16.c module. This
6077 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6080 46. If there was a timeout during a "random" callout check, Exim treated it as
6081 a failure of the random address, and carried on sending RSET and the real
6082 address. If the delay was just some slowness somewhere, the response to the
6083 original RCPT would be taken as a response to RSET and so on, causing
6084 mayhem of various kinds.
6086 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6087 when I implemented it. It didn't allow for the fact that some option values
6088 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6089 the right test for positive values.
6091 48. Domain names in DNS records are case-independent. Exim always looks them up
6092 in lower case. Some resolvers return domain names in exactly the case they
6093 appear in the zone file, that is, they may contain uppercase letters. Not
6094 all resolvers do this - some return always lower case. Exim was treating a
6095 change of case by a resolver as a change of domain, similar to a widening
6096 of a domain abbreviation. This triggered its re-routing code and so it was
6097 trying to route what was effectively the same domain again. This normally
6098 caused routing to fail (because the router wouldn't handle the domain
6099 twice). Now Exim checks for this case specially, and just changes the
6100 casing of the domain that it ultimately uses when it transmits the message
6103 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6106 50. If a filter generated a file delivery with a non-absolute name (possible if
6107 no home directory exists for the router), the forbid_file option was not
6110 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6111 the existing equality matching.
6113 52. Exim was using ints instead of ino_t variables in some places where it was
6114 dealing with inode numbers.
6116 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6117 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6118 called TMPDIR, and if it finds it is different, it changes its value.
6120 54. The smtp_printf() function is now made available to local_scan() so
6121 additional output lines can be written before returning. There is also an
6122 smtp_fflush() function to enable the detection of a dropped connection.
6123 The variables smtp_input and smtp_batched_input are exported to
6126 55. Changed the default runtime configuration: the message "Unknown user"
6127 has been removed from the ACL, and instead placed on the localuser router,
6128 using the cannot_route_message feature. This means that any verification
6129 failures that generate their own messages won't get overridden. Similarly,
6130 the "Unrouteable address" message that was in the ACL for unverifiable
6131 relay addresses has also been removed.
6133 56. Added hosts_avoid_esmtp to the smtp transport.
6135 57. The exicyclog script was not checking for the esoteric option
6136 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6137 will work only if exicyclog is run under the appropriate euid.
6139 58. Following a discussion on the list, the rules by which Exim recognises line
6140 endings on incoming messages have been changed. The -dropcr and drop_cr
6141 options are now no-ops, retained only for backwards compatibility. The
6142 following line terminators are recognized: LF CRLF CR. However, special
6143 processing applies to CR:
6145 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6146 nor a local message in the state where . is a terminator.
6148 (ii) If a bare CR is encountered in a header line, an extra space is added
6149 after the line terminator so as not to end the header. The reasoning
6150 behind this is that bare CRs in header lines are most likely either
6151 to be mistakes, or people trying to play silly games.
6153 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6154 was being incorrectly given as 18 bytes larger than it should have been.
6155 This is a VOB (very old bug).
6157 60. This may never have affected anything current, but just in case it has:
6158 When the local host is found other than at the start of a list of hosts,
6159 the local host, those with the same MX, and any that follow, are discarded.
6160 When the list in question was part of a longer list of hosts, the following
6161 hosts (not currently being processed) were also being discarded. This no
6162 longer happens. I'm not sure if this situation could ever has previously
6165 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6168 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6169 header. This is something that is recommended in a new Internet Draft, and
6170 is something that is documented as being done by Sendmail. There are two
6171 possible values. For messages generated by the autoreply transport, Exim
6174 Auto-Submitted: auto-replied
6176 whereas for all other generated messages (e.g. bounces) it adds
6178 Auto-Submitted: auto-generated
6180 63. The "personal" condition in filters now includes a test for the
6181 Auto-Submitted: header. If it contains the string "auto-" the message it
6182 not considered personal.
6184 64. Added rcpt_include_affixes as a generic transport option.
6186 65. Added queue_only_override (default true).
6188 66. Added the syslog_duplication option.
6190 67. If what should have been the first header line of a message consisted of
6191 a space followed by a colon, Exim was mis-interpreting it as a header line.
6192 It isn't of course - it is syntactically invalid and should therefore be
6193 treated as the start of the message body. The misbehaviour could have
6194 caused a number of strange effects, including loss of data in subsequent
6195 header lines, and spool format errors.
6197 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6198 client host had authenticated. This control can now be exercised by an ACL
6199 for more flexibility.
6201 69. By default, callouts do not happen when testing with -bh. There is now a
6202 variant, -bhc, which does actually run the callout code, including
6203 consulting and updating the callout cache.
6205 70. Added support for saslauthd authentication, courtesy of Alexander
6208 71. If statvfs() failed on the spool or log directories while checking their
6209 size for availability, Exim confusingly gave the error "space shortage".
6210 Furthermore, in debugging mode it crashed with a floating point exception.
6211 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6212 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6213 serious problem, Exim now writes to the main and panic logs when this
6214 happens, with details of the failure. It then refuses to accept the
6215 incoming message, giving the message "spool directory problem" or "log
6216 directory problem" with a 421 code for SMTP messages.
6218 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6219 0, 1, and 2 exist, because some OS complain for execs without them (see
6220 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6221 descriptors. However, the code omitted to check that the open succeeded,
6222 causing mysterious errors if for some reason the permissions on /dev/null
6223 got screwed. Now Exim writes a message to the main and panic logs, and
6224 bombs out if it can't open /dev/null.
6226 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6227 interact so that it is all more flexible. It is supposed to remain
6228 backwards compatible. Also added extra_local_interfaces.
6230 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6231 to bomb out with an assertion failure - to the client this appears as a
6232 connection drop. This problem occurs in the part of the code that was taken
6233 from the Samba project. Fortunately, the assertion is in a very simple
6234 function, so I have fixed this by reproducing the function inline in the
6235 one place where it is called, and arranging for authentication to fail
6236 instead of killing the process with assert().
6238 75. The SPA client code was not working when the server requested OEM rather
6239 than Unicode encoding.
6241 76. Added code to make require_files with a specific uid setting more usable in
6242 the case where statting the file as root fails - usually a non-root-mounted
6243 NFS file system. When this happens and the failure is EACCES, Exim now
6244 forks a subprocess and does the per-uid checking as the relevant uid.
6246 77. Added process_log_path.
6248 78. If log_file_path was not explicitly set, a setting of check_log_space or
6249 check_log_inodes was ignored.
6251 79. If a space check for the spool or log partitions fails, the incident is now
6252 logged. Of course, in the latter case the data may get lost...
6254 80. Added the %p formatting code to string_format() so that it can be used to
6255 print addresses in debug_print(). Adjusted all the address printing in the
6256 debugging in store.c to use %p rather than %d.
6258 81. There was a concern that a line of code in smtp_in.c could overflow a
6259 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6260 initially expressed, the concern was not well-founded, because trailing
6261 spaces are removed early. However, if the trailing spaces were followed by
6262 a NULL, they did not get removed, so the overflow was possible. Two fixes
6265 (a) I re-wrote the offending code in a cleaner fashion.
6266 (b) If an incoming SMTP command contains a NULL character, it is rejected
6269 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6270 runs initgroups(), so that if the Exim user is in any additional groups,
6271 they will be used during message reception.
6277 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob