1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
10 JH/02 The smtp transport option "multi_domain" is now expanded.
12 JH/03 The smtp transport now requests PRDR by default, if the server offers
15 JH/04 Certificate name checking on server certificates, when exim is a client,
16 is now done by default. The transport option tls_verify_cert_hostname
17 can be used to disable this per-host. The build option
18 EXPERIMENTAL_CERTNAMES is withdrawn.
20 JH/05 The value of the tls_verify_certificates smtp transport and main options
21 default to the word "system" to access the system default CA bundle.
22 For GnuTLS, only version 3.0.20 or later.
24 JH/06 Verification of the server certificate for a TLS connection is now tried
25 (but not required) by default. The verification status is now logged by
26 default, for both outbound TLS and client-certificate supplying inbound
29 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
32 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
33 Status Notification (bounce) messages are now MIME format per RFC 3464.
34 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
35 under the control of the dsn_advertise_hosts option, and routers may
36 have a dsn_lasthop option.
38 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
39 default, modifiable by a malware= option. The list separator for
40 the options can now be changed in the usual way. Bug 68.
42 JH/10 The smtp_receive_timeout main option is now expanded before use.
44 JH/11 The incoming_interface log option now also enables logging of the
45 local interface on delivery outgoing connections.
47 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
48 if the interface and destination host and port all match.
50 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
53 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
54 Patch from Andrew Lewis.
56 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
57 now supports optional time-restrictions, weighting, and priority
58 modifiers per server. Patch originally by <rommer@active.by>.
60 JH/16 The spamd_address main option now supports a mixed list of local
61 and remote servers. Remote servers can be IPv6 addresses, and
64 JH/17 Bug 68: The spamd_address main option now supports an optional
65 timeout value per server.
67 JH/18 Bug 1581: Router and transport options headers_add/remove can
68 now have the list separator specified.
70 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
73 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
76 JH/21 Support for the A6 type of dns record is withdrawn.
78 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
79 rather than the verbs used.
81 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
82 from 255 to 1024 chars.
84 JH/24 Verification callouts now attempt to use TLS by default.
86 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
87 are generic router options now. The defaults didn't change.
89 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
90 Original patch from Alexander Shikoff, worked over by JH.
92 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
93 files if ZCAT_COMMAND is not executable.
95 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
97 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
99 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
100 Normally benign, it bites when the pair was led to by a CNAME;
101 modern usage is to not canoicalize the domain to a CNAME target
102 (and we were inconsistent anyway for A-only vs AAAA+A).
104 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
106 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
107 when evaluating $sender_host_dnssec.
109 JH/31 Check the HELO verification lookup for DNSSEC, adding new
110 $sender_helo_dnssec variable.
112 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
114 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
119 TL/01 When running the test suite, the README says that variables such as
120 no_msglog_check are global and can be placed anywhere in a specific
121 test's script, however it was observed that placement needed to be near
122 the beginning for it to behave that way. Changed the runtest perl
123 script to read through the entire script once to detect and set these
124 variables, reset to the beginning of the script, and then run through
125 the script parsing/test process like normal.
127 TL/02 The BSD's have an arc4random API. One of the functions to induce
128 adding randomness was arc4random_stir(), but it has been removed in
129 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
130 function when detected.
132 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
133 cause callback expansion.
135 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
136 syntax errors in an expansion can be treated as a string instead of
137 logging or causing an error, due to the internal use of bool_lax
138 instead of bool when processing it.
140 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
141 server certificates when making smtp deliveries.
143 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
145 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
147 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
149 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
150 Merged patch from Sebastian Wiedenroth.
152 JH/05 Fix results-pipe from transport process. Several recipients, combined
153 with certificate use, exposed issues where response data items split
154 over buffer boundaries were not parsed properly. This eventually
155 resulted in duplicates being sent. This issue only became common enough
156 to notice due to the introduction of conection certificate information,
157 the item size being so much larger. Found and fixed by Wolfgang Breyha.
159 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
160 size buffer was used, resulting in syntax errors when an expansion
163 JH/07 Add support for directories of certificates when compiled with a GnuTLS
164 version 3.3.6 or later.
166 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
167 is EXPERIMENTAL_EVENT, the main-configuration and transport options
168 both become "event_action", the variables become $event_name, $event_data
169 and $event_defer_errno. There is a new variable $verify_mode, usable in
170 routers, transports and related events. The tls:cert event is now also
171 raised for inbound connections, if the main configuration event_action
174 TL/06 In test suite, disable OCSP for old versions of openssl which contained
175 early OCSP support, but no stapling (appears to be less than 1.0.0).
177 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
178 server certificate names available under the smtp transport option
179 "tls_verify_cert_hostname" now do not permit multi-component wildcard
182 JH/10 Time-related extraction expansions from certificates now use the main
183 option "timezone" setting for output formatting, and are consistent
184 between OpenSSL and GnuTLS compilations. Bug 1541.
186 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
187 encoded parameter in the incoming message. Bug 1558.
189 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
190 include certificate info, eximon was claiming there were spoolfile
193 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
195 JH/14 Log delivery-related information more consistently, using the sequence
196 "H=<name> [<ip>]" wherever possible.
198 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
199 are problematic for Debian distribution, omit them from the release
202 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
204 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
207 JH/17 Fix a null-indirection in certextract expansions when a nondefault
208 output list separator was used.
213 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
214 checkers that were complaining about end of non-void function with no
217 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
218 This was a regression intruduced in 4.83 by another bugfix.
220 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
222 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
223 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
229 TF/01 Correctly close the server side of TLS when forking for delivery.
231 When a message was received over SMTP with TLS, Exim failed to clear up
232 the incoming connection properly after forking off the child process to
233 deliver the message. In some situations the subsequent outgoing
234 delivery connection happened to have the same fd number as the incoming
235 connection previously had. Exim would try to use TLS and fail, logging
236 a "Bad file descriptor" error.
238 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
239 utilities have not been installed.
241 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
242 temporary space as the ACL may create new global variables.
244 TL/01 LDAP support uses per connection or global context settings, depending
245 upon the detected version of the libraries at build time.
247 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
248 to extract and use the src ip:port in logging and expansions as if it
249 were a direct connection from the outside internet. PPv2 support was
250 updated based on HAProxy spec change in May 2014.
252 JH/02 Add ${listextract {number}{list}{success}{fail}}.
254 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
255 Properly escape header and check for NULL return.
257 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
260 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
262 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
263 characters in header names, implemented as a verify condition.
264 Contributed by Michael Fischer v. Mollard.
266 TL/05 Rename SPF condition results err_perm and err_temp to standardized
267 results permerror and temperror. Previous values are deprecated but
268 still accepted. In a future release, err_perm and err_temp will be
269 completely removed, which will be a backward incompatibility if the
270 ACL tests for either of these two old results. Patch contributed by
271 user bes-internal on the mailing list.
273 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
275 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
276 selectors, in both main and reject logs.
278 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
281 JH/07 Add malware type "sock" for talking to simple daemon.
283 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
285 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
286 routers/transports under cutthrough routing.
288 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
289 numbers. Touch up "bool" conditional to keep the same definition.
291 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
293 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
295 JH/12 Expand items in router/transport headers_add or headers_remove lists
296 individually rather than the list as a whole. Bug 1452.
298 Required for reasonable handling of multiple headers_ options when
299 they may be empty; requires that headers_remove items with embedded
300 colons must have them doubled (or the list-separator changed).
302 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
303 view the policy declared in the DMARC record. Currently, $dmarc_status
304 is a combined value of both the record presence and the result of the
307 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
309 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
310 dnslookup router and the smtp transport (applying to the forward
313 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
314 of ldap servers used for a specific lookup. Patch provided by Heiko
317 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
318 New variable $lookup_dnssec_authenticated for observability.
320 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
321 Patch submitted by Lars Timman.
323 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
325 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
326 Requires trusted mode and valid format message id, aborts otherwise.
327 Patch contributed by Heiko Schlichting.
329 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
330 certextract with support for various fields. Bug 1358.
332 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
333 is requested by default, modifiable by smtp transport option
336 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
337 operate on certificate variables to give certificate fingerprints
338 Also new ${sha256:cert_variable}.
340 JH/23 The PRDR feature is moved from being Experimental into the mainline.
342 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
343 Christian Aistleitner.
345 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
347 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
348 file. Patch from Wolfgang Breyha.
350 JH/25 Expand the coverage of the delivery $host and $host_address to
351 client authenticators run in verify callout. Bug 1476.
353 JH/26 Port service names are now accepted for tls_on_connect_ports, to
354 align with daemon_smtp_ports. Bug 72.
356 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
357 support and error reporting did not work properly.
359 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
360 and is readable. Patch from Andrew Colin Kissa.
362 TL/14 Enhance documentation of ${run expansion and how it parses the
363 commandline after expansion, particularly in the case when an
364 unquoted variable expansion results in an empty value.
366 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
368 PP/02 Fix internal collision of T_APL on systems which support RFC3123
369 by renaming away from it. Addresses GH issue 15, reported by
372 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
374 TL/15 SECURITY: prevent double expansion in math comparison functions
375 (can expand unsanitized data). Not remotely exploitable.
382 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
384 PP/02 Make -n do something, by making it not do something.
385 When combined with -bP, the name of an option is not output.
387 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
390 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
391 $sender_host_name and config options to manage this, and basic check
394 PP/05 DSCP support for outbound connections and control modifier for inbound.
396 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
397 (Only plugin which currently uses this is kerberos4, which nobody should
398 be using, but we should make it available and other future plugins might
399 conceivably use it, even though it would break NAT; stuff *should* be
400 using channel bindings instead).
402 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
403 name; added for Sendmail compatibility; requires admin caller.
404 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
405 just ignore it); requires trusted caller.
406 Also parse but ignore: -Ac -Am -X<logfile>
409 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
411 TL/02 Add +smtp_confirmation as a default logging option.
413 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
414 Patch by Magnus Holmgren from 2007-02-20.
416 TL/04 Bugzilla 1281 - Spec typo.
417 Bugzilla 1283 - Spec typo.
418 Bugzilla 1290 - Spec grammar fixes.
420 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
422 TL/06 Add Experimental DMARC support using libopendmarc libraries.
424 TL/07 Fix an out of order global option causing a segfault. Reported to dev
425 mailing list by by Dmitry Isaikin.
427 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
429 JH/02 Support "G" suffix to numbers in ${if comparisons.
431 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
433 NM/01 Bugzilla 1197 - Spec typo
434 Bugzilla 1196 - Spec examples corrections
436 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
438 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
439 gnutls_enable_pkcs11, but renamed to more accurately indicate its
442 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
443 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
445 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
446 "acl {{name}{arg}...}", and optional args on acl condition
449 JH/05 Permit multiple router/transport headers_add/remove lines.
451 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
453 JH/07 Avoid using a waiting database for a single-message-only transport.
454 Performance patch from Paul Fisher. Bugzilla 1262.
456 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
459 JH/09 Add $headers_added variable, with content from use of ACL modifier
460 add_header (but not yet added to the message). Bugzilla 199.
462 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
463 Pulled from Bugzilla 817 by Wolfgang Breyha.
465 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
467 (nb: this is the same fix as in Exim 4.80.1)
469 JH/11 Add A= logging on delivery lines, and a client_set_id option on
472 JH/12 Add optional authenticated_sender logging to A= and a log_selector
475 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
477 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
478 advertise SMTP AUTH mechanism to us, instead of a generic
479 protocol violation error. Also, make Exim more robust to bad
480 data from the Dovecot auth socket.
482 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
484 When a queue runner is handling a message, Exim first routes the
485 recipient addresses, during which it prunes them based on the retry
486 hints database. After that it attempts to deliver the message to
487 any remaining recipients. It then updates the hints database using
490 So if a recipient address works intermittently, it can get repeatedly
491 deferred at routing time. The retry hints record remains fresh so the
492 address never reaches the final cutoff time.
494 This is a fairly common occurrence when a user is bumping up against
495 their storage quota. Exim had some logic in its local delivery code
496 to deal with this. However it did not apply to per-recipient defers
497 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
499 This change adds a proper retry rule check during routing so that the
500 final cutoff time is checked against the message's age. We only do
501 this check if there is an address retry record and there is not a
502 domain retry record; this implies that previous attempts to handle
503 the address had the retry_use_local_parts option turned on. We use
504 this as an approximation for the destination being like a local
505 delivery, as in LMTP.
507 I suspect this new check makes the old local delivery cutoff check
508 redundant, but I have not verified this so I left the code in place.
510 TF/02 Correct gecos expansion when From: is a prefix of the username.
512 Test 0254 submits a message to Exim with the header
516 When I ran the test suite under the user fanf2, Exim expanded
517 the header to contain my full name, whereas it should have added
518 a Resent-Sender: header. It erroneously treats any prefix of the
519 username as equal to the username.
521 This change corrects that bug.
523 GF/01 DCC debug and logging tidyup
524 Error conditions log to paniclog rather than rejectlog.
525 Debug lines prefixed by "DCC: " to remove any ambiguity.
527 TF/03 Avoid unnecessary rebuilds of lookup-related code.
529 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
530 Bug spotted by Jeremy Harris; was flawed since initial commit.
531 Would have resulted in OCSP responses post-SNI triggering an Exim
532 NULL dereference and crash.
534 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
536 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
537 Bug detection, analysis and fix by Samuel Thibault.
538 Bugzilla 1331, Debian bug #698092.
540 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
542 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
543 Server implementation by Todd Lyons, client by JH.
544 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
545 config variable "prdr_enable" controls whether the server
546 advertises the facility. If the client requests PRDR a new
547 acl_data_smtp_prdr ACL is called once for each recipient, after
548 the body content is received and before the acl_smtp_data ACL.
549 The client is controlled by bolth of: a hosts_try_prdr option
550 on the smtp transport, and the server advertisement.
551 Default client logging of deliveries and rejections involving
552 PRDR are flagged with the string "PRDR".
554 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
555 fclose(). Diagnosis by Todd Lyons.
557 PP/17 Update configure.default to handle IPv6 localhost better.
558 Patch by Alain Williams (plus minor tweaks).
561 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
562 This is now consistent with GnuTLS, and is now documented: the
563 previous undocumented portable approach to treating the option as
564 unset was to force an expansion failure. That still works, and
565 an empty string is now equivalent.
567 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
568 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
569 not performing validation itself.
571 PP/20 Added force_command boolean option to pipe transport.
572 Patch from Nick Koston, of cPanel Inc.
574 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
577 TF/04 Added udpsend ACL modifer and hexquote expansion operator
579 PP/21 Fix eximon continuous updating with timestamped log-files.
580 Broken in a format-string cleanup in 4.80, missed when I repaired the
581 other false fix of the same issue.
582 Report and fix from Heiko Schlichting.
585 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
586 Report from Prashanth Katuri.
588 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
589 It's SecureTransport, so affects any MacOS clients which use the
590 system-integrated TLS libraries, including email clients.
592 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
593 using a MIME ACL for non-SMTP local injection.
594 Report and assistance in diagnosis by Warren Baker.
596 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
598 JH/16 Fix comparisons for 64b. Bugzilla 1385.
600 TL/09 Add expansion variable $authenticated_fail_id to keep track of
601 last id that failed so it may be referenced in subsequent ACL's.
603 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
606 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
607 ldap library initialization, allowing self-signed CA's to be
608 used. Also properly sets require_cert option later in code by
609 using NULL (global ldap config) instead of ldap handle (per
610 session). Bug diagnosis and testing by alxgomz.
612 TL/12 Enhanced documentation in the ratelimit.pl script provided in
613 the src/util/ subdirectory.
615 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
616 renamed to Transport Post Delivery Action by Jeremy Harris, as
619 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
620 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
621 redis_servers = needs to be configured which will be used by the redis
622 lookup. Patch from Warren Baker, of The Packet Hub.
624 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
626 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
627 hostname or reverse DNS when processing a host list. Used suggestions
628 from multiple comments on this bug.
630 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
632 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
633 Missed a few lines, added it to make the runtest require no keyboard
636 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
637 contains upper case chars. Make router use caseful_local_part.
639 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
640 support when GnuTLS has been built with p11-kit.
646 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
648 This, or similar/improved, will also be change PP/11 of 4.82.
654 PP/01 Handle short writes when writing local log-files.
655 In practice, only affects FreeBSD (8 onwards).
656 Bugzilla 1053, with thanks to Dmitry Isaikin.
658 NM/01 Bugzilla 949 - Documentation tweak
660 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
663 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
665 PP/02 Implemented gsasl authenticator.
667 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
669 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
670 `pkg-config foo` for cflags/libs.
672 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
673 with rest of GSASL and with heimdal_gssapi.
675 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
676 `pkg-config foo` for cflags/libs for the TLS implementation.
678 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
679 properties get this fed in as external SSF. A number of robustness
680 and debugging improvements to the cyrus_sasl authenticator.
682 PP/08 cyrus_sasl server now expands the server_realm option.
684 PP/09 Bugzilla 1214 - Log authentication information in reject log.
685 Patch by Jeremy Harris.
687 PP/10 Added dbmjz lookup type.
689 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
691 PP/12 MAIL args handles TAB as well as SP, for better interop with
692 non-compliant senders.
693 Analysis and variant patch by Todd Lyons.
695 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
696 Bug report from Lars Müller <lars@samba.org> (via SUSE),
697 Patch from Dirk Mueller <dmueller@suse.com>
699 PP/13 tls_peerdn now print-escaped for spool files.
700 Observed some $tls_peerdn in wild which contained \n, which resulted
701 in spool file corruption.
703 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
704 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
705 or write after TLS renegotiation, which otherwise led to messages
708 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
709 as a tracking header (ie: a signed header comes before the signature).
710 Patch from Wolfgang Breyha.
712 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
713 comma-sep list; embedded commas doubled.
715 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
717 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
719 Report and patch from Dmitry Banschikov.
721 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
722 Removed SSL_clear() after SSL_new() which led to protocol negotiation
723 failures. We appear to now support TLS1.1+ with Exim.
725 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
726 lets Exim select keys and certificates based upon TLS SNI from client.
727 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
728 before an outbound SMTP session. New log_selector, +tls_sni.
730 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
731 NULL dereference. Report and patch from Alun Jones.
733 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
734 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
735 Not seeing resolver debug output on NetBSD, but suspect this is a
736 resolver implementation change.
738 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
739 Left warnings. Added "eximon gdb" invocation mode.
741 PP/21 Defaulting "accept_8bitmime" to true, not false.
743 PP/22 Added -bw for inetd wait mode support.
745 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
746 locate the relevant includes and libraries. Made this the default.
748 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
749 Bugzilla 1246, report and most of solution from Tomasz Kusy.
751 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
752 This may cause build issues on older platforms.
754 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
755 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
756 gnutls_require_mac & gnutls_require_protocols (no longer supported).
757 Added SNI support via GnuTLS too.
758 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
760 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
762 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
763 Applied second patch from Janne, implementing suggestion to default
764 multiple-strings-in-record handling to match SPF spec.
766 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
768 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
769 read-only, out of scope).
770 Patch from Wolfgang Breyha, report from Stuart Northfield.
772 PP/29 Fix three issues highlighted by clang analyser static analysis.
773 Only crash-plausible issue would require the Cambridge-specific
774 iplookup router and a misconfiguration.
775 Report from Marcin Mirosław.
777 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
779 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
780 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
781 As part of this, removing so much warning spew let me fix some minor
782 real issues in debug logging.
784 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
785 assignment on my part. Fixed.
787 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
788 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
789 Janne Snabb (who went above and beyond: thank you).
791 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
792 string otherwise requires a connection and a bunch more work and it's
793 relatively easy to get wrong. Should also expose TLS library linkage
796 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
797 64-bit ${eval} (JH/03).
799 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
800 GNU libc to support some of the 64-bit stuff, should not lead to
801 conflicts. Defined before os.h is pulled in, so if a given platform
802 needs to override this, it can.
804 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
805 protection layer was required, which is not implemented.
806 Bugzilla 1254, patch from Wolfgang Breyha.
808 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
809 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
810 tls_dhparam take prime identifiers. Also unbreak combination of
811 OpenSSL+DH_params+TLSSNI.
813 PP/39 Disable SSLv2 by default in OpenSSL support.
819 PP/01 Solaris build fix for Oracle's LDAP libraries.
820 Bugzilla 1109, patch from Stephen Usher.
822 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
824 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
827 TF/02 Fix a couple more cases where we did not log the error message
828 when unlink() failed. See also change 4.74-TF/03.
830 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
831 lock up or crash if it happened to be inside a call to libc when it
832 got a SIGUSR1 from exiwhat.
834 The SIGUSR1 handler appends the current process status to the process
835 log which is later printed by exiwhat. It used to use the general
836 purpose logging code to do this, but several functions it calls are
837 not safe for signals.
839 The new output code in the SIGUSR1 handler is specific to the process
840 log, and simple enough that it's easy to inspect for signal safety.
841 Removing some special cases also simplifies the general logging code.
842 Removing the spurious timestamps from the process log simplifies
845 TF/04 Improved ratelimit ACL condition.
847 The /noupdate option has been deprecated in favour of /readonly which
848 has clearer semantics. The /leaky, /strict, and /readonly update modes
849 are mutually exclusive. The update mode is no longer included in the
850 database key; it just determines when the database is updated. (This
851 means that when you upgrde Exim will forget old rate measurements.)
853 Exim now checks that the per_* options are used with an update mode that
854 makes sense for the current ACL. For example, when Exim is processing a
855 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
856 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
857 must specify per_mail/readonly. If you omit the update mode it defaults to
858 /leaky where that makes sense (as before) or /readonly where required.
860 The /noupdate option is now undocumented but still supported for
861 backwards compatibility. It is equivalent to /readonly except that in
862 ACLs where /readonly is required you may specify /leaky/noupdate or
863 /strict/noupdate which are treated the same as /readonly.
865 A useful new feature is the /count= option. This is a generalization
866 of the per_byte option, so that you can measure the throughput of other
867 aggregate values. For example, the per_byte option is now equivalent
868 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
870 The per_rcpt option has been generalized using the /count= mechanism
871 (though it's more complicated than the per_byte equivalence). When it is
872 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
873 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
874 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
875 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
876 Note that using per_rcpt with a non-readonly update mode in more than
877 one ACL will cause the recipients to be double-counted. (The per_mail
878 and per_byte options don't have this problem.)
880 The handling of very low rates has changed slightly. If the computed rate
881 is less than the event's count (usually one) then this event is the first
882 after a long gap. In this case the rate is set to the same as this event's
883 count, so that the first message of a spam run is counted properly.
885 The major new feature is a mechanism for counting the rate of unique
886 events. The new per_addr option counts the number of different
887 recipients that someone has sent messages to in the last time period. It
888 behaves like per_rcpt if all the recipient addresses are different, but
889 duplicate recipient addresses do not increase the measured rate. Like
890 the /count= option this is a general mechanism, so the per_addr option
891 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
892 example, measure the rate that a client uses different sender addresses
893 with the options per_mail/unique=$sender_address. There are further
894 details in the main documentation.
896 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
898 TF/06 Removed a few PCRE remnants.
900 TF/07 Automatically extract Exim's version number from tags in the git
901 repository when doing development or release builds.
903 PP/02 Raise smtp_cmd_buffer_size to 16kB.
904 Bugzilla 879. Patch from Paul Fisher.
906 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
907 Heavily based on revision 40f9a89a from Simon Arlott's tree.
910 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
912 PP/05 Variable $av_failed, true if the AV scanner deferred.
913 Bugzilla 1078. Patch from John Horne.
915 PP/06 Stop make process more reliably on build failure.
916 Bugzilla 1087. Patch from Heiko Schlittermann.
918 PP/07 Make maildir_use_size_file an _expandable_ boolean.
919 Bugzilla 1089. Patch from Heiko Schlittermann.
921 PP/08 Handle ${run} returning more data than OS pipe buffer size.
922 Bugzilla 1131. Patch from Holger Weiß.
924 PP/09 Handle IPv6 addresses with SPF.
925 Bugzilla 860. Patch from Wolfgang Breyha.
927 PP/10 GnuTLS: support TLS 1.2 & 1.1.
929 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
932 PP/11 match_* no longer expand right-hand-side by default.
933 New compile-time build option, EXPAND_LISTMATCH_RHS.
934 New expansion conditions, "inlist", "inlisti".
936 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
938 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
940 PP/14 fix log_write() format string regression from TF/03.
941 Bugzilla 1152. Patch from Dmitry Isaikin.
947 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
949 PP/02 Harmonised TLS library version reporting; only show if debugging.
950 Layout now matches that introduced for other libraries in 4.74 PP/03.
952 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
954 PP/04 New "dns_use_edns0" global option.
956 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
959 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
960 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
962 TK/01 Updated PolarSSL code to 0.14.2.
963 Bugzilla 1097. Patch from Andreas Metzler.
965 PP/07 Catch divide-by-zero in ${eval:...}.
968 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
971 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
972 format-string attack -- SECURITY: remote arbitrary code execution.
974 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
975 time unintentionally subject to list matching rules, letting the header
976 cause arbitrary Exim lookups (of items which can occur in lists, *not*
977 arbitrary string expansion). This allowed for information disclosure.
979 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
980 INT_MIN/-1 -- value coerced to INT_MAX.
986 NM/01 Workround for PCRE version dependency in version reporting
989 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
990 This fixes portability to compilers other than gcc, notably
991 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
993 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
994 makefiles for portability to HP-UX and POSIX correctness.
996 PP/01 Permit LOOKUP_foo enabling on the make command-line.
997 Also via indented variable definition in the Makefile.
998 (Debugging by Oliver Heesakkers).
1000 PP/02 Restore caching of spamd results with expanded spamd_address.
1001 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1003 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1004 Improves build reliability. Fix from: Frank Elsner
1006 NM/02 Fix wide character breakage in the rfc2047 coding
1007 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1009 NM/03 Allow underscore in dnslist lookups
1010 Fixes bug 1026. Patch from Graeme Fowler
1012 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1013 Code patches from Adam Ciarcinski of NetBSD.
1015 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1018 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1019 is logged, to avoid truncation. Patch from John Horne.
1021 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1022 Patch from Jakob Hirsch.
1024 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1025 SQL string expansion failure details.
1026 Patch from Andrey Oktyabrski.
1028 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1029 Patch from Simon Arlott.
1031 PP/09 New lookups functionality failed to compile on old gcc which rejects
1032 extern declarations in function scope.
1033 Patch from Oliver Fleischmann
1035 PP/10 Use sig_atomic_t for flags set from signal handlers.
1036 Check getgroups() return and improve debugging.
1037 Fixed developed for diagnosis in bug 927 (which turned out to be
1040 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1041 Patch from Mark Zealey.
1043 PP/12 Bugzilla 1056: Improved spamd server selection.
1044 Patch from Mark Zealey.
1046 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1047 Based on patch from Heiko Schlittermann.
1049 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1050 Patch from Uwe Doering, sign-off by Michael Haardt.
1052 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1053 variable declaration deep within a block. Bug and patch from
1056 PP/15 lookups-Makefile IRIX compatibilty coercion.
1058 PP/16 Make DISABLE_DKIM build knob functional.
1060 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1061 Patch by Simon Arlott
1063 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1064 variable argument macros. Our copy now differs from upstream.
1070 TF/01 Failure to get a lock on a hints database can have serious
1071 consequences so log it to the panic log.
1073 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1074 controlled using the smtp_confirmation log selector.
1076 TF/03 Include the error message when we fail to unlink a spool file.
1078 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1079 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1080 for maintaining out-of-tree patches for some time.
1082 PP/01 Bugzilla 139: Documentation and portability issues.
1083 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1084 Handle per-OS dynamic-module compilation flags.
1086 PP/02 Let /dev/null have normal permissions.
1087 The 4.73 fixes were a little too stringent and complained about the
1088 permissions on /dev/null. Exempt it from some checks.
1089 Reported by Andreas M. Kirchwitz.
1091 PP/03 Report version information for many libraries, including
1092 Exim version information for dynamically loaded libraries. Created
1093 version.h, now support a version extension string for distributors
1094 who patch heavily. Dynamic module ABI change.
1096 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1097 privilege escalation vulnerability whereby the Exim run-time user
1098 can cause root to append content of the attacker's choosing to
1101 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1104 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1105 If dropping privileges for untrusted macros, we disabled normal logging
1106 on the basis that it would fail; for the Exim run-time user, this is not
1107 the case, and it resulted in successful deliveries going unlogged.
1108 Fixed. Reported by Andreas Metzler.
1114 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1115 only prepend for the Resent-* case. Fixes regression introduced in
1116 Exim 4.70 by NM/22 for Bugzilla 607.
1118 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1119 increasing numbers of administrators be bitten by this.
1121 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1123 PP/03 Bugzilla 994: added openssl_options main configuration option.
1125 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1127 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1129 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1131 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1132 without a peer certificate, leading to a segfault because of an
1133 assumption that peers always have certificates. Be a little more
1134 paranoid. Problem reported by Martin Tscholak.
1136 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1137 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1138 NB: ClamAV planning to remove STREAM in "middle of 2010".
1139 CL also introduces -bmalware, various -d+acl logging additions and
1140 more caution in buffer sizes.
1142 PP/09 Implemented reverse_ip expansion operator.
1144 PP/10 Bugzilla 937: provide a "debug" ACL control.
1146 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1148 PP/12 Bugzilla 973: Implement --version.
1150 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1152 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1154 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1156 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1157 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1158 ignore trailing whitespace.
1160 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1162 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1165 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1166 Notification from Dr Andrew Aitchison.
1168 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1169 ExtendedDetectionInfo response format.
1170 Notification from John Horne.
1172 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1175 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1176 XSL and documented dependency on system catalogs, with examples of how
1179 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1182 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1183 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1184 configuration file which is writeable by the Exim user or group.
1186 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1187 of configuration files to cover files specified with the -C option if
1188 they are going to be used with root privileges, not just the default
1191 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1192 option (effectively making it always true).
1194 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1195 files to be used while preserving root privileges.
1197 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1198 that rogue child processes cannot use them.
1200 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1201 run-time user, instead of root.
1203 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1204 Exim run-time user without dropping privileges.
1206 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1207 result string, instead of calling string_vformat() twice with the same
1210 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1211 for other users. Others should always drop root privileges if they use
1212 -C on the command line, even for a whitelisted configure file.
1214 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1216 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1222 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1223 $data_path, and $header_path variables; fixed documentation bugs and
1226 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1227 exipick to access non-standard spools, including the "frozen" queue
1230 NM/01 Bugzilla 965: Support mysql stored procedures.
1231 Patch from Alain Williams
1233 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1235 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1236 Patch from Andreas Metzler
1238 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1239 Patch from Kirill Miazine
1241 NM/05 Bugzilla 671: Added umask to procmail example.
1243 JJ/03 installed exipick 20100323.0, fixing doc bug
1245 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1246 directory. Notification and patch from Dan Rosenberg.
1248 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1250 TK/02 Improve log output when DKIM signing operation fails.
1252 MH/01 Treat the transport option dkim_domain as a colon separated
1253 list, not as a single string, and sign the message with each element,
1254 omitting multiple occurences of the same signer.
1256 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1257 Bugzilla 985, 986. Patch by Simon Arlott
1259 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1260 Patch by Simon Arlott
1262 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1263 MBX locking. Notification from Dan Rosenberg.
1269 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1271 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1273 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1275 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1277 NM/04 Bugzilla 915: Flag broken perl installation during build.
1283 TK/01 Added patch by Johannes Berg that expands the main option
1284 "spamd_address" if it starts with a dollar sign.
1286 TK/02 Write list of recipients to X-Envelope-Sender header when building
1287 the mbox-format spool file for content scanning (suggested by Jakob
1290 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1291 (http://www.dcc-servers.net/) support via dccifd. Activated by
1292 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1294 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1295 by Mark Daniel Reidel <mr@df.eu>.
1297 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1298 When building exim an external PCRE library is now needed -
1299 PCRE is a system library on the majority of modern systems.
1300 See entry on PCRE_LIBS in EDITME file.
1302 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1303 conversation. Added nologin parameter to request.
1304 Patch contributed by Kirill Miazine.
1306 TF/01 Do not log submission mode rewrites if they do not change the address.
1308 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1310 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1311 log files in place. Contributed by Roberto Lima.
1313 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1315 TF/03 Bugzilla 615: When checking the local_parts router precondition
1316 after a local_part_suffix or local_part_prefix option, Exim now
1317 does not use the address's named list lookup cache, since this
1318 contains cached lookups for the whole local part.
1320 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1321 Robert Millan. Documentation is in experimental-spec.txt.
1323 TF/04 Bugzilla 668: Fix parallel build (make -j).
1325 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1327 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1328 Patch provided by Jan Srzednicki.
1330 TF/05 Leading white space used to be stripped from $spam_report which
1331 wrecked the formatting. Now it is preserved.
1333 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1334 that they are available at delivery time.
1336 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1338 TF/08 TLS error reporting now respects the incoming_interface and
1339 incoming_port log selectors.
1341 TF/09 Produce a more useful error message if an SMTP transport's hosts
1342 setting expands to an empty string.
1344 NM/06 Bugzilla 744: EXPN did not work under TLS.
1345 Patch provided by Phil Pennock.
1347 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1348 Patch provided by Richard Godbee.
1350 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1351 acl_smtp_notquit, added index entry.
1353 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1354 Patch provided by Eugene Bujak.
1356 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1357 accept(). Patch provided by Maxim Dounin.
1359 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1360 Patch provided by Phil Pennock.
1362 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1364 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1365 Patch provided by Brad "anomie" Jorsch.
1367 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1368 Patch provided by Dean Brooks.
1370 TK/05 Add native DKIM support (does not depend on external libraries).
1372 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1373 Patch provided by Graeme Fowler.
1375 NM/16 Bugzilla 851: Documentation example syntax fix.
1377 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1379 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1382 NM/19 Bugzilla 745: TLS version reporting.
1383 Patch provided by Phil Pennock.
1385 NM/20 Bugzilla 167: bool: condition support.
1386 Patch provided by Phil Pennock.
1388 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1389 clients. Patch provided by Phil Pennock.
1391 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1392 Patch provided by Brad "anomie" Jorsch.
1394 NM/23 Bugzilla 687: Fix misparses in eximstats.
1395 Patch provided by Heiko Schlittermann.
1397 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1398 Patch provided by Heiko Schlittermann.
1400 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1401 plus update to original patch.
1403 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1405 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1406 Patch provided by David Brownlee.
1408 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1410 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1412 NM/30 Bugzilla 888: TLS documentation bugfixes.
1414 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1416 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1417 Unlike the original bugzilla I have changed all shell scripts in src tree.
1419 NM/33 Bugzilla 898: Transport filter timeout fix.
1420 Patch by Todd Rinaldo.
1422 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1423 Patch by Serge Demonchaux.
1425 NM/35 Bugzilla 39: Base64 decode bug fixes.
1426 Patch by Jakob Hirsch.
1428 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1430 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1432 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1434 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1440 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1441 ALT-N's libdkim that I have put here:
1442 http://duncanthrax.net/exim-experimental/
1444 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1445 were called 'true' and it seems that C99 defines that as a reserved
1446 keyword to be used with 'bool' variable types. That means you could
1447 not include C99-style headers which use bools without triggering
1448 build errors in sieve.c.
1450 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1451 as mailq or other aliases. Changed the --help handling significantly
1452 to do whats expected. exim_usage() emits usage/help information.
1454 SC/01 Added the -bylocaldomain option to eximstats.
1456 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1458 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1460 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1466 PH/01 Another patch from the Sieve maintainer.
1468 PH/02 When an IPv6 address is converted to a string for single-key lookup
1469 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1470 dots are used instead of colons so that keys in lsearch files need not
1471 contain colons. This was done some time before quoting was made available
1472 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1473 (notated using the quote facility) so as to distinguish them from IPv4
1474 keys. This meant that lookups for IP addresses in host lists did not work
1475 for iplsearch lookups.
1477 This has been fixed by arranging for IPv6 addresses to be expressed with
1478 colons if the lookup type is iplsearch. This is not incompatible, because
1479 previously such lookups could never work.
1481 The situation is now rather anomolous, since one *can* have colons in
1482 ordinary lsearch keys. However, making the change in all cases is
1483 incompatible and would probably break a number of configurations.
1485 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1488 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1489 conversion specification without a maximum field width, thereby enabling
1490 a rogue spamd server to cause a buffer overflow. While nobody in their
1491 right mind would setup Exim to query an untrusted spamd server, an
1492 attacker that gains access to a server running spamd could potentially
1493 exploit this vulnerability to run arbitrary code as the Exim user.
1495 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1496 $primary_hostname instead of what libspf2 thinks the hosts name is.
1498 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1499 a directory entry by the name of the lookup key. Previously, if a
1500 symlink pointed to a non-existing file or a file in a directory that
1501 Exim lacked permissions to read, a lookup for a key matching that
1502 symlink would fail. Now it is enough that a matching directory entry
1503 exists, symlink or not. (Bugzilla 503.)
1505 PH/03 The body_linecount and body_zerocount variables are now exported in the
1508 PH/04 Added the $dnslist_matched variable.
1510 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1511 This means they are set thereafter only if the connection becomes
1514 PH/06 Added the client_condition to authenticators so that some can be skipped
1515 by clients under certain conditions.
1517 PH/07 The error message for a badly-placed control=no_multiline_responses left
1518 "_responses" off the end of the name.
1520 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1522 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1523 (without spaces) instead of just copying the configuration text.
1525 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1527 PH/11 Added $max_received_linelength.
1529 PH/12 Added +ignore_defer and +include_defer to host lists.
1531 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1532 way in which PCRE > 7.0 is built.
1534 PH/14 Implemented queue_only_load_latch.
1536 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1537 MAIL command. The effect was to mangle the value on 64-bit systems.
1539 PH/16 Another patch from the Sieve maintainer.
1541 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1543 PH/18 If a system quota error occurred while trying to create the file for
1544 a maildir delivery, the message "Mailbox is full" was not appended to the
1545 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1546 only to a quota excession during the actual writing of the file.
1548 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1549 characters?) which causes problems in log lines. The DN values are now
1550 passed through string_printing() before being added to log lines.
1552 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1553 and InterBase are left for another time.)
1555 PH/21 Added message_body_newlines option.
1557 PH/22 Guard against possible overflow in moan_check_errorcopy().
1559 PH/23 POSIX allows open() to be a macro; guard against that.
1561 PH/24 If the recipient of an error message contained an @ in the local part
1562 (suitably quoted, of course), incorrect values were put in $domain and
1563 $local_part during the evaluation of errors_copy.
1569 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1570 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1573 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1574 issue a MAIL command.
1576 PH/02 In an ACL statement such as
1578 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1580 if a client was not listed at all, or was listed with a value other than
1581 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1582 the condition was not true (as it should be), so access was not denied.
1583 The bug was that the ! inversion was incorrectly passed on to the second
1584 item. This has been fixed.
1586 PH/03 Added additional dnslists conditions == and =& which are different from
1587 = and & when the dns lookup returns more than one IP address.
1589 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1590 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1592 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1593 FSYNC, which compiles an option called disable_fsync that allows for
1594 bypassing fsync(). The documentation is heavily laced with warnings.
1596 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1598 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1599 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1600 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1601 including adding "make clean"; (3) Added -fPIC when compiling the test
1602 dynamically loaded module, to get rid of a warning.
1604 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1605 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1606 = 0s. The bug is otherwise harmless.
1608 PH/07 There was a bug in the dovecot authenticator such that the value of
1609 $auth1 could be overwritten, and so not correctly preserved, after a
1610 successful authentication. This usually meant that the value preserved by
1611 the server_setid option was incorrect.
1613 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1615 PH/09 Installed PCRE release 7.0.
1617 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1618 run for batched SMTP input. It is now run at the start of every message
1619 in the batch. While fixing this I discovered that the process information
1620 (output by running exiwhat) was not always getting set for -bs and -bS
1621 input. This is fixed, and it now also says "batched" for BSMTP.
1623 PH/11 Added control=no_pipelining.
1625 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1626 patch, slightly modified), and move the expansion of helo_data till after
1627 the connection is made in the smtp transport (so it can use these
1630 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1632 PH/14 Added log_selector = +pid.
1634 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1636 PH/16 Add ${if forany and ${if forall.
1638 PH/17 Added dsn_from option to vary the From: line in DSNs.
1640 PH/18 Flush SMTP output before performing a callout, unless control =
1641 no_callout_flush is set.
1643 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1644 was true (the default) a successful delivery failed to delete the retry
1645 item, thus causing premature timeout of the address. The bug is now
1648 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1650 PH/21 Long custom messages for fakedefer and fakereject are now split up
1651 into multiline reponses in the same way that messages for "deny" and
1652 other ACL rejections are.
1654 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1655 with slight modification.
1657 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1658 draft, changing the syntax and factoring some duplicate code".
1660 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1661 for deliveries of the second and subsequent messages over the same SMTP
1664 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1665 ${reduce, with only minor "tidies".
1667 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1669 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1670 expansion side effects.
1672 PH/27 When a message times out after an over-quota error from an Exim-imposed
1673 quota, the bounce message says "mailbox is full". This message was not
1674 being given when it was a system quota that was exceeded. It now should
1677 MH/03 Made $recipients available in local_scan(). local_scan() already has
1678 better access to the recipient list through recipients_list[], but
1679 $recipients can be useful in postmaster-provided expansion strings.
1681 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1682 in the case of a MAIL command with additional options following the
1683 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1684 were accidentally chopped off.
1686 PH/29 SMTP synchronization checks are implemented when a command is read -
1687 there is a check that no more input is waiting when there shouldn't be
1688 any. However, for some commands, a delay in an ACL can mean that it is
1689 some time before the response is written. In this time, more input might
1690 arrive, invalidly. So now there are extra checks after an ACL has run for
1691 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1692 pipelining has not been advertised.
1694 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1696 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1697 returned in the SMTP response when smtp_return_error_details was false.
1698 This has been fixed.
1700 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1701 instead of the C I/O that was originally supplied, because problems were
1702 reported on Solaris.
1704 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1705 Exim which did not show up earlier: it was assuming that a call to
1706 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1707 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1708 was a macro that became an assignment, so it seemed to work. This has
1709 changed to a proper function call with a void return, hence the compile
1710 error. Exim's code has been fixed.
1712 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1715 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1717 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1719 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1720 because it turns out to be too big or there is a timeout) there is no
1721 "Completed" line in the log. When some messages of this type were
1722 selected by exigrep, they were listed as "not completed". Others were
1723 picked up by some special patterns. I have improved the selection
1724 criteria to be more general.
1726 PH/38 The host_find_failed option in the manualroute router can now be set
1727 to "ignore", to completely ignore a host whose IP address cannot be
1728 found. If all hosts are ignored, the behaviour is controlled by the new
1729 host_all_ignored option.
1731 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1732 homing or because of multiple MX records with /mx) generated more than
1733 one IP address, and the following item turned out to be the local host,
1734 all the secondary addresses of the first item were incorrectly removed
1735 from the list, along with the local host and any following hosts (which
1736 is what is supposed to happen).
1738 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1739 whoever called Exim into the -H file. In the case of the daemon it was
1740 behaving confusingly. When first started, it used values for whoever
1741 started the daemon, but after a SIGHUP it used the Exim user (because it
1742 calls itself on a restart). I have changed the code so that it now always
1745 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1746 message are rejected with the same error (e.g. no authentication or bad
1747 sender address), and a DATA command is nevertheless sent (as can happen
1748 with PIPELINING or a stupid MUA), the error message that was given to the
1749 RCPT commands is included in the rejection of the DATA command. This is
1750 intended to be helpful for MUAs that show only the final error to their
1753 PH/42 Another patch from the Sieve maintainer.
1755 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1756 Eximstats - Fixed some broken HTML links and added missing column headers
1758 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1759 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1761 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1763 PH/43 Yet another patch from the Sieve maintainer.
1765 PH/44 I found a way to check for a TCP/IP connection going away before sending
1766 the response to the final '.' that terminates a message, but only in the
1767 case where the client has not sent further data following the '.'
1768 (unfortunately, this is allowed). However, in many cases there won't be
1769 any further data because there won't be any more messages to send. A call
1770 to select() can be used: if it shows that the input is "ready", there is
1771 either input waiting, or the socket has been closed. An attempt to read
1772 the next input character can distinguish the two cases. Previously, Exim
1773 would have sent an OK response which the client would never have see.
1774 This could lead to message repetition. This fix should cure that, at
1775 least in a lot of common cases.
1777 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1778 advertised in response to EHLO.
1784 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1785 fixed by 4.65/MH/01 (is this a record?) are fixed:
1787 (i) An empty string was always treated as zero by the numeric comparison
1788 operators. This behaviour has been restored.
1790 (ii) It is documented that the numeric comparison operators always treat
1791 their arguments as decimal numbers. This was broken in that numbers
1792 starting with 0 were being interpreted as octal.
1794 While fixing these problems I realized that there was another issue that
1795 hadn't been noticed. Values of message_size_limit (both the global option
1796 and the transport option) were treated as octal if they started with 0.
1797 The documentation was vague. These values are now always treated as
1798 decimal, and I will make that clear in the documentation.
1804 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1805 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1808 MH/01 Don't check that the operands of numeric comparison operators are
1809 integers when their expansion is in "skipping" mode (fixes bug
1810 introduced by 4.64-PH/07).
1812 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1813 child addresses, Exim now panics and dies. Previously, because the count
1814 is held in a short int, deliveries were likely to be lost. As such a
1815 large number of recipients for a single message is ridiculous
1816 (performance will be very, very poor), I have chosen to impose a limit
1817 rather than extend the field.
1823 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1824 leftover -K file (the existence of which was triggered by #402).
1825 While we were at it, introduced process PID as part of the -K
1826 filename. This should rule out race conditions when creating
1829 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1830 processing considerably. Previous code took too long for large mails,
1831 triggering a timeout which in turn triggers #401.
1833 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1834 in the DK code in transports.c. sendfile() is not really portable,
1835 hence the _LINUX specificness.
1837 TF/01 In the add_headers option to the mail command in an Exim filter,
1838 there was a bug that Exim would claim a syntax error in any
1839 header after the first one which had an odd number of characters
1842 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1843 callout verification, Exim cached a "reject" for the entire domain. This
1844 is correct for most verifications, but it is not correct for a recipient
1845 verification with use_sender or use_postmaster set, because in that case
1846 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1847 case of MAIL FROM:<> rejection from other early rejections (e.g.
1848 rejection of HELO). When verifying a recipient using a non-null MAIL
1849 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1850 Whatever the result of the callout, the value of the domain cache is
1851 left unchanged (for any other kind of callout, getting as far as trying
1852 RCPT means that the domain itself is ok).
1854 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1857 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1858 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1859 session, the daemon ignores ECONNECT errors and logs others; it now
1860 ignores EPIPE as well.
1862 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1863 (quoted-printable decoding).
1865 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1866 later the small subsequent patch to fix an introduced bug.
1868 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1870 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1872 PH/08 An error is now given if message_size_limit is specified negative.
1874 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1875 to be given (somewhat) arbitrary names.
1877 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1880 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1881 miscellaneous code fixes
1883 PH/10 Added the log_reject_target ACL modifier to specify where to log
1886 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1887 hostname. This is wrong, because it relates to the incoming message (and
1888 probably the interface on which it is arriving) and not to the outgoing
1889 callout (which could be using a different interface). This has been
1890 changed to use the value of the helo_data option from the smtp transport
1891 instead - this is what is used when a message is actually being sent. If
1892 there is no remote transport (possible with a router that sets up host
1893 addresses), $smtp_active_hostname is used.
1895 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1896 tweaks were necessary in order to get it to work (see also 21 below):
1897 (a) The code assumed that strncpy() returns a negative number on buffer
1898 overflow, which isn't the case. Replaced with Exim's string_format()
1900 (b) There were several signed/unsigned issues. I just did the minimum
1901 hacking in of casts. There is scope for a larger refactoring.
1902 (c) The code used strcasecmp() which is not a standard C function.
1903 Replaced with Exim's strcmpic() function.
1904 (d) The code set only $1; it now sets $auth1 as well.
1905 (e) A simple test gave the error "authentication client didn't specify
1906 service in request". It would seem that Dovecot has changed its
1907 interface. Fortunately there's a specification; I followed it and
1908 changed what the client sends and it appears to be working now.
1910 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1913 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1914 address A is aliased to B and C, where B exists and C does not. Without
1915 -v the output is "A verified" because verification stops after a
1916 successful redirection if more than one address is generated. However,
1917 with -v the child addresses are also verified. Exim was outputting "A
1918 failed to verify" and then showing the successful verification for C,
1919 with its parentage. It now outputs "B failed to verify", showing B's
1920 parentage before showing the successful verification of C.
1922 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1923 look up a TXT record in a specific list after matching in a combined
1926 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1927 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1928 they consult the DNS. I had assumed they would set it the way they
1929 wanted; and indeed my experiments on Linux seem to show that in some
1930 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1931 To be on the safe side, however, I have now made the interface to
1932 host_find_byname() similar to host_find_bydns(), with an argument
1933 containing the DNS resolver options. The host_find_byname() function now
1934 sets these options at its start, just as host_find_bydns() does. The smtp
1935 transport options dns_qualify_single and dns_search_parents are passed to
1936 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1937 of host_find_byname() use the default settings of RES_DEFNAMES
1938 (qualify_single) but not RES_DNSRCH (search_parents).
1940 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1941 spool_read_header() do less string testing, by means of a preliminary
1942 switch on the second character of optional "-foo" lines. (This is
1943 overdue, caused by the large number of possibilities that now exist.
1944 Originally there were few.) While I was there, I also converted the
1945 str(n)cmp tests so they don't re-test the leading "-" and the first
1946 character, in the hope this might squeeze out yet more improvement.
1948 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1949 flag allowing group syntax was set by the header_syntax check but not
1950 turned off, possible causing trouble later; (2) The flag was not being
1951 set at all for the header_verify test, causing "group"-style headers to
1952 be rejected. I have now set it in this case, and also caused header_
1953 verify to ignore an empty address taken from a group. While doing this, I
1954 came across some other cases where the code for allowing group syntax
1955 while scanning a header line wasn't quite right (mostly, not resetting
1956 the flag correctly in the right place). These bugs could have caused
1957 trouble for malformed header lines. I hope it is now all correct.
1959 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1960 with the "reply" argument non-NULL. The code, however (which originally
1961 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1962 but it didn't always do it. This confused somebody who was copying the
1963 code for some other use. I have removed all the tests.
1965 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1966 feature that was used to support insecure browsers during the U.S. crypto
1967 embargo. It requires special client support, and Exim is probably the
1968 only MTA that supported it -- and would never use it because real RSA is
1969 always available. This code has been removed, because it had the bad
1970 effect of slowing Exim down by computing (never used) parameters for the
1971 RSA_EXPORT functionality.
1973 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1974 authenticator to fail if there's a tab character in the incoming data
1975 (there should never be unless someone is messing about, as it's supposed
1976 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1977 if the connection is using TLS or if the remote IP is the same as the
1978 local IP, and the "valid-client-cert option" if a client certificate has
1981 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1982 authenticators. This can be used for authorization after authentication
1983 succeeds. (In the case of plaintext, it servers for both authentication
1986 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1987 if any retry times were supplied.
1989 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1990 connection, where there is no client IP address to check. In this
1991 situation, the verify now always succeeds.
1993 PH/25 Applied John Jetmore's -Mset patch.
1995 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1997 PH/27 In a string expansion for a processed (not raw) header when multiple
1998 headers of the same name were present, leading whitespace was being
1999 removed from all of them, but trailing whitespace was being removed only
2000 from the last one. Now trailing whitespace is removed from each header
2001 before concatenation. Completely empty headers in a concatenation (as
2002 before) are ignored.
2004 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2005 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2007 PH/29 [Removed. This was a change that I later backed out, and forgot to
2008 correct the ChangeLog entry (that I had efficiently created) before
2009 committing the later change.]
2011 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2012 address errors (4xx response to RCPT) over the same connection as other
2013 messages routed to the same hosts. Such deliveries are always "forced",
2014 so retry times are not inspected. This resulted in far too many retries
2015 for the affected addresses. The effect occurred only when there were more
2016 hosts than the hosts_max_try setting in the smtp transport when it had
2017 the 4xx errors. Those hosts that it had tried were not added to the list
2018 of hosts for which the message was waiting, so if all were tried, there
2019 was no problem. Two fixes have been applied:
2021 (i) If there are any address or message errors in an SMTP delivery, none
2022 of the hosts (tried or untried) are now added to the list of hosts
2023 for which the message is waiting, so the message should not be a
2024 candidate for sending over the same connection that was used for a
2025 successful delivery of some other message. This seems entirely
2026 reasonable: after all the message is NOT "waiting for some host".
2027 This is so "obvious" that I'm not sure why it wasn't done
2028 previously. Hope I haven't missed anything, but it can't do any
2029 harm, as the worst effect is to miss an optimization.
2031 (ii) If, despite (i), such a delivery is accidentally attempted, the
2032 routing retry time is respected, so at least it doesn't keep
2033 hammering the server.
2035 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2036 in ${readsocket because some servers need this prod.
2038 PH/32 Added some extra debug output when updating a wait-xxx database.
2040 PH/33 The hint "could be header name not terminated by colon", which has been
2041 given for certain expansion errors for a long time, was not being given
2042 for the ${if def:h_colon_omitted{... case.
2044 PH/34 The spec says: "With one important exception, whenever a domain list is
2045 being scanned, $domain contains the subject domain." There was at least
2046 one case where this was not true.
2048 PH/35 The error "getsockname() failed: connection reset by peer" was being
2049 written to the panic log as well as the main log, but it isn't really
2050 panic-worthy as it just means the connection died rather early on. I have
2051 removed the panic log writing for the ECONNRESET error when getsockname()
2054 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2055 runs only) independently of the message's sender address. This meant
2056 that, if the 4xx error was in fact related to the sender, a different
2057 message to the same recipient with a different sender could confuse
2058 things. In particualar, this can happen when sending to a greylisting
2059 server, but other circumstances could also provoke similar problems.
2060 I have changed the default so that the retry time for these errors is now
2061 based a combination of the sender and recipient addresses. This change
2062 can be overridden by setting address_retry_include_sender=false in the
2065 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2066 remote server are returned as part of bounce messages. This was not
2067 happening for LMTP over a pipe (the lmtp transport), but now it is the
2068 same for both kinds of LMTP.
2070 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2071 in header lines that were in fact CNAMEs. This is no longer the case.
2073 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2074 and queue runs started by the daemon processed all messages. This has
2075 been fixed so that -R and -S can now usefully be given with -q<time>.
2077 PH/40 Import PCRE release 6.7 (fixes some bugs).
2079 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2081 PH/42 Give an error if -q is specified more than once.
2083 PH/43 Renamed the variables $interface_address and $interface_port as
2084 $received_ip_address and $received_port, to make it clear that these
2085 values apply to message reception, and not to the outgoing interface when
2086 a message is delivered. (The old names remain recognized, of course.)
2088 PH/44 There was no timeout on the connect() call when using a Unix domain
2089 socket in the ${readsocket expansion. There now is.
2091 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2092 be meaningful with "accept".
2094 SC/01 Eximstats V1.43
2095 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2097 SC/02 Eximstats V1.44
2098 Use a glob alias rather than an array ref in the generated
2099 parser. This improves both readability and performance.
2101 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2102 Collect SpamAssassin and rejection statistics.
2103 Don't display local sender or destination tables unless
2104 there is data to show.
2105 Added average volumes into the top table text output.
2107 SC/04 Eximstats V1.46
2108 Collect data on the number of addresses (recipients)
2109 as well as the number of messages.
2111 SC/05 Eximstats V1.47
2112 Added 'Message too big' to the list of mail rejection
2113 reasons (thanks to Marco Gaiarin).
2115 SC/06 Eximstats V1.48
2116 Mainlog lines which have GMT offsets and are too short to
2117 have a flag are now skipped.
2119 SC/07 Eximstats V1.49 (Alain Williams)
2120 Added the -emptyok flag.
2122 SC/08 Eximstats V1.50
2123 Fixes for obtaining the IP address from reject messages.
2125 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2126 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2127 whitesspace changes from 4.64-PH/27
2129 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2132 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2133 are found, allow negative numbers in numeric criteria)
2135 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2137 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2138 to match changes made in 4.64-PH/43
2140 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2142 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2143 "domain" is actually an IP address, and force a failure. This locks out
2144 those revolvers/nameservers that support "A-for-A" lookups, in
2145 contravention of the specifications.
2147 PH/48 When a host name was looked up from an IP address, and the subsequent
2148 forward lookup of the name timed out, the host name was left in
2149 $sender_host_name, contrary to the specification.
2151 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2152 restricted to single-key lookups, Exim was not diagnosing an error if
2153 * or *@ was used with a query-style lookup.
2155 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2157 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2158 long ago, but noone interested enough thought of it. Let's just say that
2159 the "1.1" means that there are some new functions that weren't there at
2160 some point in the past.
2162 PH/51 Error processing for expansion failure of helo_data from an smtp
2163 transport during callout processing was broken.
2165 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2166 tested/used via the -bh/-bhc/-bs options.
2168 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2169 bug, fixed in subsequent PCRE releases).
2171 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2172 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2174 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2180 SC/01 Use a glob alias rather than an array ref in eximstats generated
2181 parser. This improves both readability and performance.
2183 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2184 Don't display local sender or destination tables in eximstats unless
2185 there is data to show.
2186 Added average volumes into the eximstats top table text output.
2188 SC/03 Collect data on the number of addresses (recipients) as well
2189 as the number of messages in eximstats.
2191 TF/01 Correct an error in the documentation for the redirect router. Exim
2192 does (usually) call initgroups() when daemonizing.
2194 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2195 with consistent privilege compared to when running as a daemon.
2197 TF/03 Note in the spec that $authenticated_id is not set for local
2198 submissions from trusted users.
2200 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2201 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2203 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2204 by adding some example configuration directives to the default
2205 configuration file. A little bit of work is required to uncomment the
2206 directives and define how usernames and passwords are checked, but
2207 there is now a framework to start from.
2209 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2210 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2211 without this. I don't know how relevant this is to other LDAP libraries.
2213 PH/02 Add the verb name to the "unknown ACL verb" error.
2215 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2217 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2219 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2220 directory not expanded when it should be if an expanded home directory
2221 was set for the address (which is overridden by the transport).
2223 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2226 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2227 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2228 because it is too late at that time, and has no effect.
2230 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2231 security issue with \' (bugzilla #107). I could not use the
2232 PQescapeStringConn() function, because it needs a PGconn value as one of
2235 PH/08 When testing addresses using -bt, indicate those final addresses that
2236 are duplicates that would not cause an additional delivery. At least one
2237 person was confused, thinking that -bt output corresponded to deliveries.
2238 (Suppressing duplicates isn't a good idea as you lose the information
2239 about possibly different redirections that led to the duplicates.)
2241 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2242 systems where poll() doesn't work, in particular OS X.
2244 PH/10 Added more information to debugging output for retry time not reached.
2246 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2247 operations in malware.c.
2249 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2252 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2253 syslog_duplication set false, log lines that would normally be written
2254 both the the main log and to the reject log were not written to syslog at
2257 PH/14 In the default configuration, change the use of "message" in ACL warn
2258 statements to "add_header".
2260 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2261 not followed by a command (e.g. "seen endif").
2263 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2264 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2267 PH/17 Added extra conditions to the default value of delay_warning_condition
2271 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2272 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2273 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2276 The Auto-Submitted: and various List- headers are standardised, whereas I
2277 don't think Precedence: ever was.
2279 PH/18 Refactored debugging code in route_finduser() to show more information,
2280 in particular, the error code if getpwnam() issues one.
2282 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2283 This is apparently needed in addition to the PH/07 change above to avoid
2284 any possible encoding problems.
2286 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2287 but not after initializing Perl.
2289 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2290 output them only if debugging. By default they are written stderr,
2291 apparently, which is not desirable.
2293 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2296 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2299 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2301 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2302 authenticated or an ident call has been made. Suppress the default
2303 values for $authenticated_id and $authenticated_sender (but permit -oMai
2304 and -oMas) when testing with -bh.
2306 PH/24 Re-jigged the order of the tests in the default configuration so that the
2307 tests for valid domains and recipients precede the DNS black list and CSA
2308 tests, on the grounds that those ones are more expensive.
2310 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2311 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2312 This bug exists in every version of Exim that I still have, right back to
2315 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2316 However, an attempt to turn on case-sensitivity in a regex key by
2317 including (?-i) didn't work because the subject string was already
2318 lowercased, and the effects were non-intuitive. It turns out that a
2319 one-line patch can be used to allow (?-i) to work as expected.
2325 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2326 other effects) broke the use of negated acl sub-conditions.
2328 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2331 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2332 "Deny" causes Exim to reject the incoming connection with a 554 error.
2333 Unfortunately, if there is a major crisis, such as a disk failure,
2334 tcp-wrappers gives "deny", whereas what one would like would be some
2335 kind of temporary error. A kludge has been added to help with this.
2336 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2337 554 error is used if errno is still zero or contains ENOENT (which occurs
2338 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2341 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2343 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2344 errors. Otherwise a message that provokes a temporary error (when other
2345 messages do not) can cause a whole host to time out.
2347 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2348 addresses were routed directly to files or pipes from a redirect router.
2349 File deliveries just didn't batch; pipe deliveries might have suffered
2352 PH/06 A failure to get a lock for a hints database would erroneously always say
2353 "Failed to get write lock", even when it was really a read lock.
2355 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2356 of 0600. This has been changed to use the value of the lockfile_mode
2357 option (which defaults to 0600).
2359 PH/08 Applied small patch from the Sieve maintainer.
2361 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2362 folder from quota calculations, a direct delivery into this folder messed
2363 up the contents of the maildirsize file. This was because the regex was
2364 used only to exclude .Trash (or whatever) when the size of the mailbox
2365 was calculated. There was no check that a delivery was happening into an
2366 excluded directory. This bug has been fixed by ignoring all quota
2367 processing for deliveries into excluded directories.
2369 PH/10 Added the maildirfolder_create_regex option to appendfile.
2375 PH/01 The code for finding all the local interface addresses on a FreeBSD
2376 system running IPv6 was broken. This may well have applied to all BSD
2377 systems, as well as to others that have similar system calls. The broken
2378 code found IPv4 interfaces correctly, but gave incorrect values for the
2379 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2380 that it would not match correctly against @[] and not recognize the IPv6
2383 PH/02 The ipliteral router was not recognizing addresses of the form user@
2384 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2386 PH/03 Added disable_ipv6.
2388 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2389 decoded form, because it is most often used to construct To: headers
2390 lines in autoreplies, and the decoded form may well be syntactically
2391 invalid. However, $reply_address has leading white space removed, and all
2392 newlines turned into spaces so that the autoreply transport does not
2395 PH/05 If group was specified without a user on a router, and no group or user
2396 was specified on a transport, the group from the router was ignored.
2398 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2399 for visible compile-time settings that can be used to change these
2400 numbers, for those that want even more. Backwards compatibility with old
2401 spool files has been maintained. However, going back to a previous Exim
2402 release will lost any variables that are in spool files.
2404 PH/07 Two small changes when running in the test harness: increase delay when
2405 passing a TCP/IP connection to a new process, in case the original
2406 process has to generate a bounce, and remove special handling of
2407 127.0.0.2 (sic), which is no longer necessary.
2409 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2410 be the same on different OS.
2412 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2415 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2416 whether --show-vars was specified or not
2418 JJ/02 exipick: Added support for new ACL variable spool format introduced
2421 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2422 syntactically invalid From: or Reply-to: line, and a filter used this to
2423 generate an autoreply, and therefore failed to obtain an address for the
2424 autoreply, Exim could try to deliver to a non-existent relative file
2425 name, causing unrelated and misleading errors. What now happens is that
2426 it logs this as a hard delivery error, but does not attempt to create a
2429 PH/11 The exinext utility has a -C option for testing purposes, but although
2430 the given file was scanned by exinext itself; it wasn't being passed on
2431 when Exim was called.
2433 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2434 an end-of-file indication when reading a command response.
2436 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2437 compiled. In many other places in Exim, IPv6 addresses are always
2438 recognized, so I have changed this. It also means that IPv4 domain
2439 literals of the form [IPV4:n.n.n.n] are now always recognized.
2441 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2442 used if the router is not running as root, for example, when verifying at
2443 ACL time, or when using -bh. The debugging output from this situation was
2444 non-existent - all you got was a failure to exec. I have made two
2447 (a) Failures to set uid/gid, the current directory, or a process leader
2448 in a subprocess such as that created by queryprogram now generate
2449 suitable debugging ouput when -d is set.
2451 (b) The queryprogram router detects when it is not running as root,
2452 outputs suitable debugging information if -d is set, and then runs
2453 the subprocess without attempting to change uid/gid.
2455 PH/15 Minor change to Makefile for building test_host (undocumented testing
2458 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2459 additional section of a DNS packet that returns MX or SRV records.
2460 Instead, it always explicitly searches for A/AAAA records. This avoids
2461 major problems that occur when a DNS server includes only records of one
2462 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2463 fixed another bug: if SRV records were looked up and the corresponding
2464 address records were *not* found in the additional section, the port
2465 values from the SRV records were lost.
2467 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2468 using the correct key (the original address) when searching the retry
2469 rules in order to find which one to use for generating the retry hint.
2471 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2472 adding the default one. Similarly, if it contains a Reply-To: header, the
2473 errors_reply_to option, if set, is not used.
2475 PH/19 When calculating a retry time, Exim used to measure the "time since
2476 failure" by looking at the "first failed" field in the retry record. Now
2477 it does not use this if it is later than than the arrival time of the
2478 message. Instead it uses the arrival time. This makes for better
2479 behaviour in cases where some deliveries succeed, thus re-setting the
2480 "first failed" field. An example is a quota failure for a huge message
2481 when small messages continue to be delivered. Without this change, the
2482 "time since failure" will always be short, possible causing more frequent
2483 delivery attempts for the huge message than are intended.
2484 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2486 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2487 $1, $2, $3) because the numerical variables can be reset during some
2488 expansion items (e.g. "match"), thereby losing the authentication data.
2490 PH/21 Make -bV show the size of off_t variables so that the test suite can
2491 decide whether to run tests for quotas > 2G.
2493 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2494 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2495 filecount value is greater than 2G or if a quota value is greater than 2G
2496 on a system where the size of off_t is not greater than 4, a panic error
2499 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2500 never match. The debug and -bh output now contains an explicit error
2501 message indicating a malformed IPv4 address or mask.
2503 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2504 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2505 PH/23 above applies.
2507 PH/25 Do not write to syslog when running in the test harness. The only
2508 occasion when this arises is a failure to open the main or panic logs
2509 (for which there is an explicit test).
2511 PH/26 Added the /no_tell option to "control=freeze".
2513 PH/27 If a host name lookup failed very early in a connection, for example, if
2514 the IP address matched host_lookup and the reverse lookup yielded a name
2515 that did not have a forward lookup, an error message of the form "no IP
2516 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2517 could be logged. Now it outputs the IP address instead of "NULL".
2519 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2520 allows the sender and the authenticated sender to be set when
2521 submitting a message from within Exim. Since child_open_exim() is
2522 documented for local_scan(), the new function should be too.
2524 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2525 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2526 results in an empty string is now treated as unset.
2528 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2530 PH/31 Added sender_verify_fail logging option.
2532 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2533 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2534 tidied the source and removed it altogether.
2536 PH/33 When a queue run was abandoned because the load average was too high, a
2537 log line was always written; now it is written only if the queue_run log
2538 selector is set. In addition, the log line for abandonment now contains
2539 information about the queue run such as the pid. This is always present
2540 in "start" and "stop" lines but was omitted from the "abandon" line.
2542 PH/34 Omit spaces between a header name and the colon in the error message that
2543 is given when verify = headers_syntax fails (if there are lots of them,
2544 the message gets confusing).
2546 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2547 names, as there are now some PTR records that contain slashes. This check
2548 is only to protect against broken name servers that fall over on strange
2549 characters, so the fact that it applies to all lookups doesn't matter.
2551 PH/36 Now that the new test suite is complete, we can remove some of the
2552 special code in Exim that was needed for the old test suite. For example,
2553 sorting DNS records because real resolvers return them in an arbitrary
2554 order. The new test suite's fake resolver always returns records in the
2557 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2558 bounces) except when queue_only is set, to avoid logging races between
2559 the different processes.
2561 PH/38 Panic-die if .include specifies a non-absolute path.
2563 PH/39 A tweak to the "H" retry rule from its user.
2565 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2566 a label. They prevented compilation on older perls.
2568 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2569 a warning to be raised on newish perls.
2571 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2572 on queue. Changes to match documented behaviour of showing count of
2573 messages matching specified criteria.
2575 PH/40 Changed the default ident timeout from 30s to 5s.
2577 PH/41 Added support for the use of login_cap features, on those BSD systems
2578 that have them, for controlling the resources used by pipe deliveries.
2580 PH/42 The content-scanning code uses fopen() to create files in which to put
2581 message data. Previously it was not paying any attention to the mode of
2582 the files. Exim runs with umask(0) because the rest of the code creates
2583 files with open(), and sets the required mode explicitly. Thus, these
2584 files were ending up world-writeable. This was not a big issue, because,
2585 being within the spool directory, they were not world-accessible. I have
2586 created a function called modefopen, which takes an additional mode
2587 argument. It sets umask(777), creates the file, chmods it to the required
2588 mode, then resets the umask. All the relevant calls to fopen() in the
2589 content scanning code have been changed to use this function.
2591 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2592 to 24 hours. This avoids potential overflow problems when processing G
2593 and H retry rules. I suspect nobody ever tinkers with this value.
2595 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2597 PH/45 When the plaintext authenticator is running as a client, the server's
2598 challenges are checked to ensure they are valid base64 strings. By
2599 default, the authentication attempt is cancelled if an invalid string is
2600 received. Setting client_ignore_invalid_base64 true ignores these errors.
2601 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2602 they are received. Thus, the responses can be made to depend on the
2603 challenges. If an invalid string is ignored, an empty string is placed in
2606 PH/46 Messages that are created by the autoreply transport now contains a
2607 References: header, in accordance with RFCs 2822 and 3834.
2609 PH/47 Added authenticated_sender_force to the smtp transport.
2611 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2613 PH/49 Installed latest patch from the Sieve maintainer.
2615 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2616 was also set, the appendfile transport was unnecessarily scanning a
2617 directory of message files (e.g. for maildir delivery) to find the count
2618 of files (along with the size), even though it did not need this
2619 information. It now does the scan only if it needs to find either the
2620 size of the count of files.
2622 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2624 PH/52 Two bugs concerned with error handling when the smtp transport is
2627 (i) Exim was not creating retry information for temporary errors given
2628 for individual recipients after the DATA command when the smtp transport
2629 was used in LMTP mode. This meant that they could be retried too
2630 frequently, and not timed out correctly.
2632 (ii) Exim was setting the flag that allows error details to be returned
2633 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2634 recipients that were returned after the DATA command.
2636 PH/53 This is related to PH/52, but is more general: for any failing address,
2637 when detailed error information was permitted to be returned to the
2638 sender, but the error was temporary, then after the final timeout, only
2639 "retry timeout exceeded" was returned. Now it returns the full error as
2640 well as "retry timeout exceeded".
2642 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2643 do this, and (what is worse) MTAs that accept it.
2645 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2646 will now be deprecated.
2648 PH/56 New os.c-cygwin from the Cygwin maintainer.
2650 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2651 formats (previously only available in exim formats via -bpr, -bpru,
2652 and -bpra. Now also available in native and exiqgrep formats)
2654 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2655 with very large, slow to parse queues
2657 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2659 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2661 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2662 responses to authentication challenges, though it was showing the
2663 challenges; (ii) I've removed the CR characters from the debug output for
2666 PH/58 Allow for the insertion of a newline as well as a space when a string
2667 is turned into more than one encoded-word during RFC 2047 encoding. The
2668 Sieve code now uses this.
2670 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2671 data_4xx, lost_connection, tls_required.
2673 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2674 message was being sent as an SMTP response.
2676 PH/61 Add -l and -k options to exicyclog.
2678 PH/62 When verifying, if an address was redirected to one new address, so that
2679 verification continued, and the new address failed or deferred after
2680 having set something in $address_data, the value of $address_data was not
2681 passed back to the ACL. This was different to the case when no
2682 redirection occurred. The value is now passed back in both cases.
2684 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2685 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2686 use login_cap.h, so on its own it isn't the distinguishing feature. The
2687 new name refers directly to the setclassresources() function.
2689 PH/65 Added configuration files for NetBSD3.
2691 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2693 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2694 is preferred over IPv4.
2696 PH/68 The bounce_return_message and bounce_return_body options were not being
2697 honoured for bounces generated during the reception of non-SMTP messages.
2698 In particular, this applied to messages rejected by the ACL. This bug has
2699 been fixed. However, if bounce_return_message is true and bounce_return_
2700 body is false, the headers that are returned for a non-SMTP message
2701 include only those that have been read before the error was detected.
2702 (In the case of an ACL rejection, they have all been read.)
2704 PH/69 The HTML version of the specification is now built in a directory called
2705 spec_html instead of spec.html, because the latter looks like a path with
2706 a MIME-type, and this confuses some software.
2708 PH/70 Catch two compiler warnings in sieve.c.
2710 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2711 function verify_get_ident() calls ip_connect() to connect a socket, but
2712 if the "connect()" function timed out, ip_connect() used to close the
2713 socket. However, verify_get_ident() also closes the socket later, and in
2714 between Exim writes to the log, which may get opened at this point. When
2715 the socket was closed in ip_connect(), the log could get the same file
2716 descriptor number as the socket. This naturally causes chaos. The fix is
2717 not to close the socket in ip_connect(); the socket should be closed by
2718 the function that creates it. There was only one place in the code where
2719 this was missing, in the iplookup router, which I don't think anybody now
2720 uses, but I've fixed it anyway.
2722 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2723 well as to direct DNS lookups. Otherwise the handling of names in host
2724 lists is inconsistent and therefore confusing.
2730 PH/01 Two changes to the default runtime configuration:
2732 (1) Move the checks for relay_from_hosts and authenticated clients from
2733 after to before the (commented out) DNS black list checks.
2735 (2) Add control=submission to the relay_from_hosts and authenticated
2736 clients checks, on the grounds that messages accepted by these
2737 statements are most likely to be submissions.
2739 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2741 (1) Generate an error if the third argument for the ${prvs expansion is
2744 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2747 (3) Reset the variables that are obtained from the first argument of
2748 ${prvscheck and used in the second argument before leaving the code,
2749 because their memory is reclaimed, so using them afterwards may do
2752 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2753 one (it's much easier than Tom thought :-).
2755 (5) Because of (4), we can now allow for the use of $prvscheck_result
2756 inside the third argument.
2758 PH/03 For some reason, the default setting of PATH when running a command from
2759 a pipe transport was just "/usr/bin". I have changed it to
2762 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2763 anything to be listed in the output from -bV.
2765 PH/05 When a filter generated an autoreply, the entire To: header line was
2766 quoted in the delivery log line, like this:
2768 => >A.N.Other <ano@some.domain> <original@ddress> ...
2770 This has been changed so that it extracts the operative address. There
2771 may be more than one such address. If so, they are comma-separated, like
2774 => >ano@some.domain,ona@other.domain <original@ddress> ...
2776 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2777 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2778 not being looked up in the rDNS to get a host name, Exim was showing the
2779 IP address twice in Received: lines, even though the IP addresses were
2780 identical. For example:
2782 Received: from [1.2.3.4] (helo=[1.2.3.4])
2784 However, if the real host name was known, it was omitting the HELO data
2785 if it matched the actual IP address. This has been tidied up so that it
2786 doesn't show the same IP address twice.
2788 PH/07 When both +timestamp and +memory debugging was on, the value given by
2789 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2790 called by the debug printing, thereby overwriting the timestamp buffer.
2791 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2793 PH/08 When the original message was included in an autoreply transport, it
2794 always said "this is a copy of the message, including all the headers",
2795 even if body_only or headers_only was set. It now gives an appropriate
2798 PH/09 Applied a patch from the Sieve maintainer which:
2800 o fixes some comments
2801 o adds the (disabled) notify extension core
2802 o adds some debug output for the result of if/elsif tests
2803 o points to the current vacation draft in the documentation
2804 and documents the missing references header update
2808 o fixes a bug in processing the envelope test (when testing
2809 multiple envelope elements, the last element determinted the
2812 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2813 Electronic Mail") by including:
2815 Auto-submitted: auto-generated
2817 in the messages that it generates (bounce messages and others, such as
2818 warnings). In the case of bounce messages for non-SMTP mesages, there was
2819 also a typo: it was using "Auto_submitted" (underscore instead of
2820 hyphen). Since every message generated by Exim is necessarily in response
2821 to another message, thes have all been changed to:
2823 Auto-Submitted: auto-replied
2825 in accordance with these statements in the RFC:
2827 The auto-replied keyword:
2829 - SHOULD be used on messages sent in direct response to another
2830 message by an automatic process,
2832 - MUST NOT be used on manually-generated messages,
2834 - MAY be used on Delivery Status Notifications (DSNs) and Message
2835 Disposition Notifications (MDNs),
2837 - MUST NOT be used on messages generated by automatic or periodic
2838 processes, except for messages which are automatic responses to
2841 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2842 to the default Received: header definition.
2844 PH/12 Added log selector acl_warn_skipped (default on).
2846 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2847 variables because (a) they are in the wrong storage pool and (b) even if
2848 they were copied, it wouldn't work properly because of the caching.
2850 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2851 checking when decoding. Apparently there are clients that generate
2852 overlong encoded strings. Why am I not surprised?
2854 PH/15 If the first argument of "${if match_address" was not empty, but did not
2855 contain an "@" character, Exim crashed. Now it writes a panic log message
2856 and treats the condition as false.
2858 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2860 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2861 extension "envelope-auth". The code is finished and in agreement with
2862 other implementations, but there is no documentation so far and in fact,
2863 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2864 not changing the active code.
2866 Print executed "if" and "elsif" statements when debugging is used. This
2867 helps a great deal to understand what a filter does.
2869 Document more things not specified clearly in RFC3028. I had all this
2870 sorted out, when out of a sudden new issues came to my mind. Oops."
2872 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2875 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2876 canonical form (as documented). However, after a host name lookup from
2877 the IP address, check_host() was doing a simple string comparison with
2878 addresses acquired from the DNS when checking that the found name did
2879 have the original IP as one of its addresses. Since any found IPv6
2880 addresses are likely to be in abbreviated form, the comparison could
2881 fail. Luckily, there already exists a function for doing the comparison
2882 by converting both addresses to binary, so now that is used instead of
2883 the text comparison.
2885 PH/20 There was another similar case to PH/19, when a complete host name was
2886 given in a host list; looking up its IP address could give an abbreviated
2887 form, whereas the current host's name might or might not be abbreviated.
2888 The same fix has been applied.
2894 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2895 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2898 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2899 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2901 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2903 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2904 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2905 Also, for "Auto-Submitted", treat anything other than "no" as
2906 non-personal, in accordance with RFC 3834. (Previously it treated
2907 anything starting "auto-" as non-personal.)
2909 TF/01 The control=submission/name=... option had a problem with syntax
2910 errors if the name included a slash character. The /name= option
2911 now slurps the rest of the string, so it can include any characters
2912 but it must come last in the list of options (after /sender_retain
2915 PH/05 Some modifications to the interface to the fake nameserver for the new
2923 TK/01 Added the "success_on_redirect" address verification option. See
2924 NewStuff for rationale and an example.
2926 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2928 PH/02 Patch to exigrep to allow it to work on syslog lines.
2930 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2931 fread() to skip over the body file's header line, because in Cygwin the
2932 header line is locked and is inaccessible.
2934 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2935 co-exist for some time) to make it clear that it is the Exim ID that is
2936 referenced, not the Message-ID: header line.
2938 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2939 string_format() function, because snprintf() does not exist on all
2942 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2943 ${stat: expansion item.
2945 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2946 protocol synchronization error", to keep the pedants happy.
2948 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2949 well as for IRIX systems, when gcc is being used. See the host.c source
2952 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2954 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2957 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2958 transport and to the smtp transport in LMTP mode.
2960 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2962 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2963 run a filter in a subprocess. This could lead to confusion in subsequent
2964 lookups in the parent process. There should also be a search_tidyup() at
2965 the end of the subprocess.
2967 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2968 only if the host matched helo_try_verify_hosts, which caused the
2969 verification to occur when the EHLO/HELO command was issued. The ACL just
2970 tested the remembered result. Now, if a previous verification attempt has
2971 not happened, "verify = helo" does it there and then.
2973 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2975 TK/03 Fix log output including CR from clamd.
2977 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2978 exist provoked a memory error which could cause a segfault.
2980 PH/15 Installed PCRE 6.2
2982 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2984 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2985 of the problem. Specifically, suggested +O2 rather than +O1 for the
2988 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2990 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2991 record, the port was not correctly logged when the outgoing_port log
2992 selector was set (it logged the transort's default port).
2994 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2995 fallback_hosts, and "hosts" in the smtp transport.
2997 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2998 host errors such as "Connection refused".
3000 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3001 authentication with radiusclient 0.4.9:
3003 - Error returned from rc_read_config was caught wrongly
3004 - Username/password not passed on to radius server due to wrong length.
3006 The presumption is that some radiusclient API changes for 4.51/PH/17
3007 were not taken care of correctly. The code is still untested by me (my
3008 Linux distribution still has 0.3.2 of radiusclient), but it was
3009 contributed by a Radius user.
3011 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3012 expanding the "port" option of the smtp transport.
3014 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3015 while reading a MIME header. Thanks to Tom Hughes for a patch.
3017 PH/24 Include config.h inside local_scan.h so that configuration settings are
3020 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3021 that in an ACL for RCPT (for example), you can examine exactly what was
3024 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3025 commands, but it was not correctly comparing the address with the actual
3026 client host address. Thus, it would show the EHLO address in Received:
3027 header lines when this was not necessary.
3029 PH/27 Added the % operator to ${eval:}.
3031 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3032 it should be ignoring failures (because with -C, for example, it has lost
3033 privilege). It wasn't ignoring creation failures other than "already
3036 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3039 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3040 because an input line was too long, either on its own, or by virtue of
3041 too many continuations, the temporary file was not being removed, and the
3042 return code was incorrect.
3044 PH/31 Missing "BOOL" in function definition in filtertest.c.
3046 PH/32 Applied Sieve patches from the maintainer.
3048 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3050 PH/33 Added "verify = not_blind".
3052 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3053 Local/Makefile (with some defaults set). These are used in built scripts
3054 such as exicyclog, but they have never been used in the exim_install
3055 script (though there are many overriding facilities there). I have
3056 arranged that the exim_install script now takes note of these two
3059 PH/35 Installed configuration files for Dragonfly.
3061 PH/36 When a locally submitted message by a trusted user did not contain a
3062 From: header, and the sender address was obtained from -f or from an SMTP
3063 MAIL command, and the trusted user did not use -F to supply a sender
3064 name, $originator_name was incorrectly used when constructing a From:
3065 header. Furthermore, $originator_name was used for submission mode
3066 messages from external hosts without From: headers in a similar way,
3067 which is clearly wrong.
3069 PH/37 Added control=suppress_local_fixups.
3071 PH/38 When log_selector = +received_sender was set, and the addition of the
3072 sender made the log line's construction buffer exactly full, or one byte
3073 less than full, an overflow happened when the terminating "\n" was
3076 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3077 when the result of a list match is failure because a DNS lookup failed.
3079 PH/40 RM_COMMAND is now used in the building process.
3081 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3082 the "build-* directories that it finds.
3084 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3085 address in a domain literal was a prefix of an interface address.
3087 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3088 when verifying a sender address, unless rewrite_headers is false.
3090 PH/44 Wrote a long comment about why errors_to addresses are verified as
3091 recipients, not senders.
3093 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3094 the ratelimit ACL was added.
3096 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3098 PH/46 Added extra information about PostgreSQL errors to the error string.
3100 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3101 avoiding the need to install special zones in a real server. This is
3102 backwards compatible; if it can't find the fake resolver, it drops back.
3103 Thus, both old and new test suites can be run.
3105 TF/02 Added util/ratelimit.pl
3107 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3108 clock is set back in time.
3110 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3111 Candler <B.Candler@pobox.com>.
3113 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3114 OK for addresses that are the result of redirections.
3116 PH/48 A number of further additions for the benefit of the new test suite,
3117 including a fake gethostbyname() that interfaces to the fake DNS resolver
3120 TF/06 The fix for widen_domains has also been applied to qualify_single and
3121 search_parents which are the other dnslookup options that can cause
3124 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3127 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3129 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3130 service" if it encountered an MX record with an empty target hostname.
3131 The message is now "an MX or SRV record indicated no SMTP service".
3133 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3134 if the DNS of the sending site is misconfigured. This is quite a
3135 common situation. This change restores the behaviour of treating a
3136 helo verification defer as a failure.
3138 PH/51 If self=fail was set on a router, the bounce message did not include the
3139 actual error message.
3145 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3147 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3148 command itself ended in error, the underlying message about the transport
3149 filter timeout was being overwritten with the pipe command error. Now the
3150 underlying error message should be appended to the second error message.
3152 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3154 PH/02 Reduce the amount of output that "make" produces by default. Full output
3155 can still be requested.
3157 PH/03 The warning log line about a condition test deferring for a "warn" verb
3158 was being output only once per connection, rather than after each
3159 occurrence (because it was using the same function as for successful
3160 "warn" verbs). This seems wrong, so I have changed it.
3162 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3163 it should not have, which might have caused a crash in the right
3164 circumstances, but probably never did.
3166 PH/04 Installed a modified version of Tony Finch's patch to make submission
3167 mode fix the return path as well as the Sender: header line, and to
3168 add a /name= option so that you can make the user's friendly name appear
3171 TF/03 Added the control = fakedefer ACL modifier.
3173 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3174 Mark Lowes for thorough testing.
3176 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3178 TK/03 Merged latest SRS patch from Miles Wilton.
3180 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3181 with the definition in sysexits.h (which is #included earlier).
3182 Fortunately, Exim does not actually use EX_OK. The code used to try to
3183 preserve the sysexits.h value, by assumimg that macro definitions were
3184 scanned for macro replacements. I have been disabused of this notion,
3185 so now the code just undefines EX_OK before #including unistd.h.
3187 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3188 in the smtp transport. When a block could not be written in a single
3189 write() function, the timeout was being re-applied to each part-write.
3190 This seems wrong - if the receiver was accepting one byte at a time it
3191 would take for ever. The timeout is now adjusted when this happens. It
3192 doesn't have to be particularly precise.
3194 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3195 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3197 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3198 without a domain if the check to <postmaster@domain> fails.
3200 SC/01 Eximstats: added -xls and the ability to specify output files
3201 (patch written by Frank Heydlauf).
3203 SC/02 Eximstats: use FileHandles for outputing results.
3205 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3207 SC/04 Eximstats: fixed display of large numbers with -nvr option
3209 SC/05 Eximstats: fixed merging of reports with empty tables.
3211 SC/06 Eximstats: added the -include_original_destination flag
3213 SC/07 Eximstats: removed tabs and trailing whitespace.
3215 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3217 TK/06 MBOX spool code: Add real "From " MBOX separator line
3218 so the .eml file is really in mbox format (even though
3219 most programs do not really care). Patch from Alex Miller.
3221 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3222 The latter is generated from $received_to and is only set if the
3223 message has one envelope recipient. SA can use these headers,
3224 obviously out-of-the-box. Patch from Alex Miller.
3226 PH/08 The ${def test on a variable was returning false if the variable's
3227 value was "0", contrary to what the specification has always said!
3228 The result should be true unless the variable is empty.
3230 PH/09 The syntax error of a character other than { following "${if
3231 def:variable_name" (after optional whitespace) was not being diagnosed.
3232 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3233 accidental colon was present, for example, could give incorrect results.
3235 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3236 result is used (not including appendfile, where other changes are about
3239 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3240 This involved changing a lot of size variables from int to off_t. It
3241 should work with maildirs and everything.
3243 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3244 spamd dying while we are connected to it.
3246 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3249 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3250 function for BDB 4.3.
3252 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3254 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3255 the sources for PCRE have been re-arranged and more files are now
3258 PH/15 The code I had for printing potentially long long variables in PH/11
3259 above was not the best (it lost precision). The length of off_t variables
3260 is now inspected at build time, and an appropriate printing format (%ld
3261 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3262 to be "long long int" or "long int". This is needed for the internal
3263 formatting function string_vformat().
3265 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3266 the configuration file to be ":syslog", then the script "guesses" where
3267 the logs files are, rather than using the compiled in default. In our
3268 case the guess is not the same as the compiled default, so the script
3269 suddenly stopped working when I started to use syslog. The patch checks
3270 to see if log_file_path is "". If so, it attempts to read it from exim
3271 with no configuration file to get the compiled in version, before it
3272 falls back to the previous guessing code."
3274 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3275 implementing BATV in an Exim configuration. See NewStuff for the gory
3278 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3279 Makefile that are specific to HP-UX.
3281 PH/18 If the "use_postmaster" option was set for a recipient callout together
3282 with the "random" option, the postmaster address was used as the MAIL
3283 FROM address for the random test, but not for the subsequent recipient
3284 test. It is now used for both.
3286 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3287 patch removes a few documentation additions to RFC 3028, because the
3288 latest draft now contains them. It adds the new en;ascii-case comparator
3289 and a new error check for 8bit text in MIME parts. Comparator and
3290 require names are now matched exactly. I enabled the subaddress
3291 extension, but it is not well tested yet (read: it works for me)."
3293 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3294 rework some of the code of TK/09 above to avoid the hardwired use of
3295 "%lld" and "long long". Replaced the call to snprintf() with a call to
3298 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3299 records point to non-existent hosts", "retry timeout exceeded", and
3300 "retry time not reached for any host after a long failure period".
3302 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3303 experimental DomainKeys support:
3305 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3306 (2) On an error such as an illegally used "control", the wrong name for
3307 the control was given.
3309 These problems did NOT occur unless DomainKeys support was compiled.
3311 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3313 PH/24 Added ${if match_ip condition.
3315 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3316 databases so that it will be absolutely obvious if a crash occurs in the
3317 DB library. This is a regular occurrence (often caused by mis-matched
3320 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3321 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3322 picked up on a user's system that detects such things. There doesn't seem
3323 to be a gcc warning option for this - only an attribute that has to be
3324 put on the function's prototype. It seems that in Fedora Core 4 they have
3325 set this on a number of new functions. No doubt there will be more in due
3328 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3329 specify a transport. However, if an address that was verified by such a
3330 router was the subject of a callout, Exim crashed because it tried to
3331 read the rcpt_include_affixes from the non-existent transport. Now it
3332 just assumes that the setting of that option is false. This bug was
3333 introduced by 4.51/PH/31.
3335 PH/28 Changed -d+all to exclude +memory, because that information is very
3336 rarely of interest, but it makes the output a lot bigger. People tend to
3337 do -d+all out of habit.
3339 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3340 code in os-type was giving problems when libc.so lives in lib64, like on
3343 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3344 aren't the modern standard, and it seems that some systems' include files
3345 don't always have them. Exim was already checking for some of the newer
3346 ones like T_AAAA, and defining it itself. I've added checks for all the
3347 record types that Exim uses.
3349 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3350 not automatically generating a new one, as it is supposed to. This
3351 prevented TLS from working. If the file did exist, but contained invalid
3352 data, a new version was generated, as expected. It was only the case of a
3353 non-existent file that was broken.
3355 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3356 with a change in libdomainkeys > 0.64.
3358 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3359 from DNS. If the selector record carries the flag, it now has
3360 precedence over the domain-wide flag.
3362 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3364 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3365 the use of a port name that isn't defined in /etc/services) occurred, the
3366 message was deferred as in a normal delivery, and thus remained on the
3367 spool, instead of being failed because of the mua_wrapper setting. This
3368 is now fixed, and I tidied up some of the mua_wrapper messages at the
3371 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3372 the messages in a hash of arrays rather than using individual hashes.
3373 This is a bit cleaner and results in dramatic memory savings, albeit
3374 at a slight CPU cost.
3376 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3377 as requested by Marc Sherman.
3379 SC/10 Eximstats: added histograms for user specified patterns as requested
3382 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3384 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3385 fopen() in the content-scanning modules that did not already have it.
3391 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3392 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3394 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3396 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3398 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3399 to test Sieve filters that use "vacation".
3401 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3402 that changes the way the GnuTLS parameters are stored in the cache file.
3403 The new format can be generated externally. For backward compatibility,
3404 if the data in the cache doesn't make sense, Exim assumes it has read an
3405 old-format file, and it generates new data and writes a new file. This
3406 means that you can't go back to an older release without removing the
3409 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3410 work if there are any delivery delays because "one_time" forces the
3411 parent to be marked "delivered", so its unseen clone is never tried
3412 again. For this reason, Exim now forbids the simultaneous setting of
3415 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3416 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3417 entry if you want to know the details. The fix, however, neglected to
3418 consider the case when local delivery batching is involved. The test for
3419 "previously delivered" was not happening when checking to see if an
3420 address could be batched with a previous (undelivered) one; under
3421 certain circumstances this could lead to multiple deliveries to the same
3424 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3425 in its include files, and this causes problems building Exim.
3427 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3428 header_syntax) but Exim was just ignoring anything given after a slash.
3429 In particular, this caused confusion with an attempt to use "verify =
3430 reverse_host_lookup/defer_ok". An error is now given when options are
3431 supplied for verify items that do not have them. (Maybe reverse_host_
3432 lookup should have a defer_ok option, but that's a different point.)
3434 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3435 defined by RFC 821) to 2048, because there were problems with some AUTH
3436 commands, and RFC 1869 says the size should be increased for extended
3437 SMTP commands that take arguments.
3439 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3442 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3443 "unknown" error; now it says that the functionality isn't in the binary.
3445 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3446 an address' error message when a string expansion fails (syntax or
3447 whatever). Otherwise the password may appear in the log. Following change
3448 PH/42 below, there is no longer a chance of it appearing in a bounce
3451 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3453 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3454 of its addresses was ever tried. (Bugzilla bug #2.)
3456 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3457 the result incorrectly in the debug output. (It correctly added a newline
3458 to what was transported.)
3460 TF/01 Added $received_time.
3462 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3463 commented out examples of how to interface to a virus scanner and to
3464 SpamAssassin. Also added commented examples of av_scanner and
3465 spamd_address settings.
3467 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3468 and controls are allowed in which ACLs. There were a couple of minor
3469 errors. Some of the entries in the conditions table (which is a table of
3470 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3471 negation of where the condition IS allowed.
3473 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3475 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3476 header file does not have a version number, so I've had to invent a new
3477 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3478 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3479 radiusclient), but it was contributed by a Radius user.
3481 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3482 files or directories, for OpenSSL.
3484 PH/19 When an Exim process that is running as root has to create an Exim log
3485 file, it does so in a subprocess that runs as exim:exim so as to get the
3486 ownership right at creation (otherwise, other Exim processes might see
3487 the file with the wrong ownership). There was no test for failure of this
3488 fork() call, which would lead to the process getting stuck as it waited
3489 for a non-existent subprocess. Forks do occasionally fail when resources
3490 run out. I reviewed all the other calls to fork(); they all seem to check
3493 PH/20 When checking for unexpected SMTP input at connect time (before writing
3494 the banner), Exim was not dealing correctly with a non-positive return
3495 from the read() function. If the client had disconnected by this time,
3496 the result was a log entry for a synchronization error with an empty
3497 string after "input=" when read() returned zero. If read() returned -1
3498 (an event I could not check), uninitialized data bytes were printed.
3499 There were reports of junk text (parts of files, etc) appearing after
3502 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3504 PH/22 Added support for macro redefinition, and (re)definition in between
3505 driver and ACL definitions.
3507 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3508 forgetting to use the resulting value; it was using the unexpanded value.
3510 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3511 hadn't been configured. The fix is from Juergen Kreileder, who
3512 understands it better than I do:
3514 "Here's what I see happening with three configured cyrus_sasl
3515 authenticators configured (plain, login, cram-md5):
3517 On startup auth_cyrus_sasl_init() gets called for each of these.
3518 This means three calls to sasl_listmech() without a specified mech_list.
3519 => SASL tests which mechs of all available mechs actually work
3520 => three warnings about OTP not working
3521 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3523 With the patch, sasl_listmech() also gets called three times. But now
3524 SASL's mech_list option is set to the server_mech specified in the the
3525 authenticator. Or in other words, the answer from sasl_listmech()
3526 gets limited to just the mech you're testing for (which is different
3528 => the return list contains just 'plain' or 'login', 'cram-md5' or
3529 nothing depending on the value of ob->server_mech.
3531 I've just tested the patch: Authentication still works fine,
3532 unavailable mechs specified in the exim configuration are still
3533 caught, and the auth.log warnings about OTP are gone."
3535 PH/25 When debugging is enabled, the contents of the command line are added
3536 to the debugging output, even when log_selector=+arguments is not
3539 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3540 answer is "GNU", and only if the return is "GNU/something" is the answer
3543 PH/27 $acl_verify_message is now set immediately after the failure of a
3544 verification in an ACL, and so is available in subsequent modifiers. In
3545 particular, the message can be preserved by coding like this:
3547 warn !verify = sender
3548 set acl_m0 = $acl_verify_message
3550 Previously, $acl_verify_message was set only while expanding "message"
3551 and "log_message" when a very denied access.
3553 PH/28 Modified OS/os.c-Linux with
3555 -#ifndef OS_LOAD_AVERAGE
3556 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3558 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3559 nomenclature these days.)
3561 PH/29 Installed patch from the Sieve maintainer that adds the options
3562 sieve_useraddress and sieve_subaddress to the redirect router.
3564 PH/30 In these circumstances:
3565 . Two addresses routed to the same list of hosts;
3566 . First host does not offer TLS;
3567 . First host accepts first address;
3568 . First host gives temporary error to second address;
3569 . Second host offers TLS and a TLS session is established;
3570 . Second host accepts second address.
3571 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3572 and peerdn, if requested) that were in fact used only for the second
3575 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3576 attention to any local part prefix or suffix that was matched by the
3577 router that accepted the address. It now behaves in the same way as it
3578 does for delivery: the affixes are removed from the local part unless
3579 rcpt_include_affixes is set on the transport.
3581 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3582 timeout during the DATA phase of an incoming message.
3584 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3585 applied a patch sanctioned by the Sieve maintainer.
3587 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3588 the uid or gid is negative. A case of a negative gid caused this to be
3589 noticed. The fix allows for either to be negative.
3591 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3592 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3595 PH/36 The change PH/12 above was broken. Fixed it.
3597 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3598 the grounds that routing the same address twice would always produce the
3599 same answer. This might have been true once, but it is certainly no
3600 longer true now. Routing a child address may depend on the previous
3601 routing that produced that child. Some complicated redirection strategies
3602 went wrong when messages had multiple recipients, and made Exim's
3603 behaviour dependent on the order in which the addresses were given.
3605 I have moved the duplicate checking until after the routing is complete.
3606 Exim scans the addresses that are assigned to local and remote
3607 transports, and removes any duplicates. This means that more work will be
3608 done, as duplicates will always all be routed, but duplicates are
3609 presumably rare, so I don't expect this is of any significance.
3611 For deliveries to pipes, files, and autoreplies, the duplicate checking
3612 still happens during the routing process, since they are not going to be
3615 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3616 It corrects a timeout issue with spamd. This is Ian's comment: "The
3617 background is that sometimes spamd either never reads data from a
3618 connection it has accepted, or it never writes response data. The exiscan
3619 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3620 blindly assumes that writes won't block so it may never time out."
3622 PH/39 Allow G after quota size as well as K and M.
3624 PH/40 The value set for $authenticated_id in an authenticator may not contain
3625 binary zeroes or newlines because the value is written to log lines and
3626 to spool files. There was no check on this. Now the value is run through
3627 the string_printing() function so that such characters are converted to
3628 printable escape sequences.
3630 PH/41 $message_linecount is a new variable that contains the total number of
3631 lines in the message. Compare $body_linecount, which is the count for the
3634 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3635 bounce and delay warning messages, except in certain special cases, which
3638 (a) An SMTP error message from a remote host;
3639 (b) A message specified in a :fail: redirection;
3640 (c) A message specified in a "fail" command in a system filter;
3641 (d) A message specified in a FAIL return from the queryprogram router;
3642 (e) A message specified by the cannot_route_message router option.
3644 In these cases only, Exim does include the error details in bounce and
3645 warning messages. There are also a few cases where bland messages such
3646 as "unrouteable address" or "local delivery error" are given.
3648 PH/43 $value is now also set for the "else" part of a ${run expansion.
3650 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3651 being worked on, but at least Exim now implements the latest version to
3654 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3655 process to complete was treated as a delivery failure, a timeout while
3656 writing the message to the pipe was logged, but erroneously treated as a
3657 successful delivery. Such timeouts include transport filter timeouts. For
3658 consistency with the overall process timeout, these timeouts are now
3659 treated as errors, giving rise to delivery failures by default. However,
3660 there is now a new Boolean option for the pipe transport called
3661 timeout_defer, which, if set TRUE, converts the failures into defers for
3662 both kinds of timeout. A transport filter timeout is now identified in
3665 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3666 systems where "make" and "gmake" are different, calling "gmake" at top
3667 level broke things. I've arranged for the value of $(MAKE) to be passed
3668 from the Makefile to this script so that it can call the same version of
3672 A note about Exim versions 4.44 and 4.50
3673 ----------------------------------------
3675 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3676 changes of various kinds. As a consequence, a big documentation update was
3677 needed. This delayed the release for rather longer than seemed good, especially
3678 in the light of a couple of (minor) security issues. Therefore, the changes
3679 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3680 release. So 4.44 and 4.50 are in effect two different branches that both start
3683 I have left the 4.50 change log unchanged; it contains all the changes since
3684 4.43. The change log for 4.44 is below; many of its items are identical to
3685 those for 4.50. This seems to be the most sensible way to preserve the
3686 historical information.
3692 1. Minor wording change to the doc/README.SIEVE file.
3694 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3695 computation of the current number of files was incorrect.
3697 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3698 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3701 4. Give more explanation in the error message when the command for a transport
3702 filter fails to execute.
3704 5. There are several places where Exim runs a non-Exim command in a
3705 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3706 was being done only for the command run by the queryprogram router. It is
3707 now done for all such subprocesses. The other cases are: ${run, transport
3708 filters, and the commands run by the lmtp and pipe transports.
3710 6. Added CONFIGURE_GROUP build-time option.
3712 7. Some older OS have a limit of 256 on the maximum number of file
3713 descriptors. Exim was using setrlimit() to set 1000 as a large value
3714 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3715 systems. I've change it so that if it can't get 1000, it tries for 256.
3717 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3718 was an oversight, and furthermore, ever since the addition of extra
3719 controls (e.g. 4.43/32), the checks on when to allow different forms of
3720 "control" were broken. There should now be diagnostics for all cases when a
3721 control that does not make sense is encountered.
3723 9. Added the /retain_sender option to "control=submission".
3725 10. $recipients is now available in the predata ACL (oversight).
3727 11. Tidy the search cache before the fork to do a delivery from a message
3728 received from the command line. Otherwise the child will trigger a lookup
3729 failure and thereby defer the delivery if it tries to use (for example) a
3730 cached ldap connection that the parent has called unbind on.
3732 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3733 of $address_data from the recipient verification was clobbered by the
3734 sender verification.
3736 13. The value of address_data from a sender verification is now available in
3737 $sender_address_data in subsequent conditions in the ACL statement.
3739 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3741 15. Added a new option "connect=<time>" to callout options, to set a different
3744 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3745 was its contents. (It was OK if the option was not defined at all.)
3747 17. A "Completed" log line is now written for messages that are removed from
3748 the spool by the -Mrm option.
3750 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3751 information about exactly what failed.
3753 19. Added -dd to debug only the daemon process.
3755 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3756 handles timeouts, both on the server side and network timeouts. Renamed the
3757 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3759 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3760 It is now set to "smtps".
3762 22. $host_address is now set to the target address during the checking of
3763 ignore_target_hosts.
3765 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3766 being passed; this would have caused $sender_host_name to have been used if
3767 matching the list had actually called for a host name (not very likely,
3768 since this list is usually IP addresses). A host name is now passed as
3771 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3772 code that specifies a non-restarting handler (typically sigaction() in
3773 modern systems) in an attempt to fix a rare and obscure crash bug.
3775 25. Narrowed the window for a race in the daemon that could cause it to ignore
3776 SIGCHLD signals. This is not a major problem, because they are used only to
3777 wake it up if nothing else does.
3779 26. A malformed maildirsize file could cause Exim to calculate negative values
3780 for the mailbox size or file count. Odd effects could occur as a result.
3781 The maildirsize information is now recalculated if the size or filecount
3784 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3785 support for a long time. Removed HAVE_SYS_VFS_H.
3787 28. Installed the latest version of exipick from John Jetmore.
3789 29. In an address list, if the pattern was not a regular expression, an empty
3790 subject address (from a bounce message) matched only if the pattern was an
3791 empty string. Non-empty patterns were not even tested. This was the wrong
3792 because it is perfectly reasonable to use an empty address as part of a
3793 database query. An empty address is now tested by patterns that are
3794 lookups. However, all the other forms of pattern expect the subject to
3795 contain a local part and a domain, and therefore, for them, an empty
3796 address still always fails if the pattern is not itself empty.
3798 30. Exim went into a mad DNS loop when attempting to do a callout where the
3799 host was specified on an smtp transport, and looking it up yielded more
3800 than one IP address.
3802 31. Re-factored the code for checking spool and log partition space into a
3803 function that finds that data and another that does the check. The former
3804 is then used to implement four new variables: $spool_space, $log_space,
3805 $spool_inodes, and $log_inodes.
3807 32. The RFC2047 encoding function was originally intended for short strings
3808 such as real names; it was not keeping to the 75-character limit for
3809 encoded words that the RFC imposes. It now respects the limit, and
3810 generates multiple encoded words if necessary. To be on the safe side, I
3811 have increased the buffer size for the ${rfc2047: expansion operator from
3814 33. It is now permitted to omit both strings after an "if" condition; if the
3815 condition is true, the result is "true". As before, when the second string
3816 is omitted, a false condition yields an empty string. This makes it less
3817 cumbersome to write custom ACL and router conditions.
3819 34. Failure to deliver a bounce message always caused it to be frozen, even if
3820 there was an errors_to setting on the router. The errors_to setting is now
3823 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3824 canonical form (fully expanded) before being placed in
3825 $sender_host_address.
3827 36. The table in the code that translates DNS record types into text (T_A to
3828 "A" for instance) was missing entries for NS and CNAME. It is just possible
3829 that this could have caused confusion if both these types were looked up
3830 for the same domain, because the text type is used as part of Exim's
3831 per-process caching. But the chance of anyone hitting this buglet seems
3834 37. The dnsdb lookup has been extended in a number of ways.
3836 (1) There is a new type, "zns", which walks up the domain tree until it
3837 finds some nameserver records. It should be used with care.
3839 (2) There is a new type, "mxh", which is like "mx" except that it returns
3840 just the host names, not the priorities.
3842 (3) It is now possible to give a list of domains (or IP addresses) to be
3843 looked up. The behaviour when one of the lookups defers can be
3844 controlled by a keyword.
3846 (4) It is now possible to specify the separator character for use when
3847 multiple records are returned.
3849 38. The dnslists ACL condition has been extended: it is now possible to supply
3850 a list of IP addresses and/or domains to be looked up in a particular DNS
3853 39. Added log_selector=+queue_time_overall.
3855 40. When running the queue in the test harness, wait just a tad after forking a
3856 delivery process, to get repeatability of debugging output.
3858 41. Include certificate and key file names in error message when GnuTLS fails
3859 to set them up, because the GnuTLS error message doesn't include the name
3860 of the failing file when there is a problem reading it.
3862 42. Allow both -bf and -bF in the same test run.
3864 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3866 44. The "Exiscan patch" is now merged into the mainline Exim source.
3868 45. Sometimes the final signoff response after QUIT could fail to get
3869 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3870 < 0 before doing a fflush(). This bug looks as though it goes back to the
3871 introduction of TLS in release 3.20, but "sometimes" must have been rare
3872 because the tests only now provoked it.
3874 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3875 (this can affect the format of dates).
3877 47. exim_tidydb, when checking for the continued existence of a message for
3878 which it has found a message-specific retry record, was not finding
3879 messages that were in split spool directories. Consequently, it was
3880 deleting retry records that should have stayed in existence.
3882 48. Steve fixed some bugs in eximstats.
3884 49. The SPA authentication driver was not abandoning authentication and moving
3885 on to the next authenticator when an expansion was forced to fail,
3886 contradicting the general specification for all authenticators. Instead it
3887 was generating a temporary error. It now behaves as specified.
3889 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3890 (the order specifies the preference for clients). The order is now AES256,
3891 AES128, 3DES, ARCFOUR128.
3893 51. Small patch to Sieve code - explicitly set From: when generating an
3896 52. Exim crashed if a remote delivery caused a very long error message to be
3897 recorded - for instance if somebody sent an entire SpamAssassin report back
3898 as a large number of 550 error lines. This bug was coincidentally fixed by
3899 increasing the size of one of Exim's internal buffers (big_buffer) that
3900 happened as part of the Exiscan merge. However, to be on the safe side, I
3901 have made the code more robust (and fixed the comments that describe what
3904 53. Now that there can be additional text after "Completed" in log lines (if
3905 the queue_time_overall log selector is set), a one-byte patch to exigrep
3906 was needed to allow it to recognize "Completed" as not the last thing in
3909 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3910 patch that reportedly fixes this has been added. I am not expert enough to
3911 create a test for it. This is what the patch creator wrote:
3913 "I found a little strange behaviour of ldap code when working with
3914 Windows 2003 AD Domain, where users was placed in more than one
3915 Organization Units. When I tried to give exim partial DN, the exit code
3916 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3917 But simultaneously result of request was absolutely normal ldap result,
3918 so I produce this patch..."
3920 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3921 REFERENCE, so I have modified the code to exclude the patch when that macro
3924 55. Some experimental protocols are using DNS PTR records for new purposes. The
3925 keys for these records are domain names, not reversed IP addresses. The
3926 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3927 leaves it alone. Component reversal etc. now happens only for IP addresses.
3930 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3932 57. Double the size of the debug message buffer (to 2048) so that more of very
3933 long debug lines gets shown.
3935 58. The exicyclog utility now does better if the number of log files to keep
3936 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3938 59. Two changes related to the smtp_active_hostname option:
3940 (1) $smtp_active_hostname is now available as a variable.
3941 (2) The default for smtp_banner uses $smtp_active_hostname instead
3942 of $primary_hostname.
3944 60. The host_aton() function is supposed to be passed a string that is known
3945 to be a valid IP address. However, in the case of IPv6 addresses, it was
3946 not checking this. This is a hostage to fortune. Exim now panics and dies
3947 if the condition is not met. A case was found where this could be provoked
3948 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3949 components; fortuitously, this particular loophole had already been fixed
3950 by change 4.50/55 above.
3952 If there are any other similar loopholes, the new check in host_aton()
3953 itself should stop them being exploited. The report I received stated that
3954 data on the command line could provoke the exploit when Exim was running as
3955 exim, but did not say which command line option was involved. All I could
3956 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3957 running as the user.
3960 61. There was a buffer overflow vulnerability in the SPA authentication code
3961 (which came originally from the Samba project). I have added a test to the
3962 spa_base64_to_bits() function which I hope fixes it.
3965 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3966 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3968 63. The daemon start-up calls getloadavg() while still root for those OS that
3969 need the first call to be done as root, but it missed one case: when
3970 deliver_queue_load_max is set with deliver_drop_privilege. This is
3971 necessary for the benefit of the queue runner, because there is no re-exec
3972 when deliver_drop_privilege is set.
3974 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3975 This has been fixed.
3977 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3978 was in use, was not putting the data itself into the right store pool;
3979 consequently, it could be overwritten for a subsequent message in the same
3980 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3983 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3985 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3986 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3987 the function were treating the return as a boolean value, which happened to
3988 work because 0=false and not-0=true, but is not correct code.
3990 68. The host_aton() function was not handling scoped IPv6 addresses (those
3991 with, for example, "%eth0" on the end) correctly.
3993 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3994 negated items (that is, ~something) in unsigned ints. Some compilers
3995 apparently mutter when there is no cast.
3997 70. If an address verification called from an ACL failed, and did not produce a
3998 user-specific message (i.e. there was only a "system" message), nothing was
3999 put in $acl_verify_message. In this situation, it now puts the system
4002 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4003 session; change 4.31/43 added the unwanted input to the log line - except
4004 that it did not do this in the start of session case. It now does.
4006 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4007 This is wrong and can cause the other end to generate a synchronization
4008 error if it is another Exim or anything else that does the synchronization
4009 check. A QUIT command is no longer sent after a timeout.
4011 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4012 during host lookups.
4014 74. The defer_ok option of callout verification was not working if it was used
4015 when verifying addresses in header lines, that is, for this case:
4017 verify = header_sender/callout=defer_ok
4019 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4020 those file descriptors could be used for SMTP connections. If anything
4021 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4022 could be sent to the SMTP client, causing chaos. The daemon now opens
4023 stdin, stdout, and stderr to /dev/null when it puts itself into the
4026 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4027 log by default. The user can override this with suitable Perl magic.
4029 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4030 the log message when discard triggers, was not working for the DATA ACL or
4031 for the non-SMTP ACL.
4033 78. Error message wording change in sieve.c.
4035 79. If smtp_accept_max_per_host was set, the number of connections could be
4036 restricted to fewer than expected, because the daemon was trying to set up
4037 a new connection before checking whether the processes handling previous
4038 connections had finished. The check for completed processes is now done
4039 earlier. On busy systems, this bug wouldn't be noticed because something
4040 else would have woken the daemon, and it would have reaped the completed
4043 80. If a message was submitted locally by a user whose login name contained one
4044 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4045 It caused a spool format error. I have fixed the spool reading code. A
4046 related problem was that the "from" clause in the Received: line became
4047 illegal because of the space(s). It is now covered by ${quote_local_part.
4049 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4052 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4054 83. Patch from Sieve maintainer to fix unterminated string problem in
4055 "vacation" handling.
4057 84. Some minor changes to the Linux configuration files to help with other
4058 OS variants using glibc.
4060 85. One more patch for Sieve to update vacation handling to latest spec.
4063 ----------------------------------------------------
4064 See the note above about the 4.44 and 4.50 releases.
4065 ----------------------------------------------------
4071 1. Change 4.43/35 introduced a bug that caused file counts to be
4072 incorrectly computed when quota_filecount was set in an appendfile
4075 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4076 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4079 3. Give more explanation in the error message when the command for a transport
4080 filter fails to execute.
4082 4. There are several places where Exim runs a non-Exim command in a
4083 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4084 was being done only for the command run by the queryprogram router. It is
4085 now done for all such subprocesses. The other cases are: ${run, transport
4086 filters, and the commands run by the lmtp and pipe transports.
4088 5. Some older OS have a limit of 256 on the maximum number of file
4089 descriptors. Exim was using setrlimit() to set 1000 as a large value
4090 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4091 systems. I've change it so that if it can't get 1000, it tries for 256.
4093 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4094 was an oversight, and furthermore, ever since the addition of extra
4095 controls (e.g. 4.43/32), the checks on when to allow different forms of
4096 "control" were broken. There should now be diagnostics for all cases when a
4097 control that does not make sense is encountered.
4099 7. $recipients is now available in the predata ACL (oversight).
4101 8. Tidy the search cache before the fork to do a delivery from a message
4102 received from the command line. Otherwise the child will trigger a lookup
4103 failure and thereby defer the delivery if it tries to use (for example) a
4104 cached ldap connection that the parent has called unbind on.
4106 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4107 of $address_data from the recipient verification was clobbered by the
4108 sender verification.
4110 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4111 was its contents. (It was OK if the option was not defined at all.)
4113 11. A "Completed" log line is now written for messages that are removed from
4114 the spool by the -Mrm option.
4116 12. $host_address is now set to the target address during the checking of
4117 ignore_target_hosts.
4119 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4120 being passed; this would have caused $sender_host_name to have been used if
4121 matching the list had actually called for a host name (not very likely,
4122 since this list is usually IP addresses). A host name is now passed as
4125 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4126 code that specifies a non-restarting handler (typically sigaction() in
4127 modern systems) in an attempt to fix a rare and obscure crash bug.
4129 15. Narrowed the window for a race in the daemon that could cause it to ignore
4130 SIGCHLD signals. This is not a major problem, because they are used only to
4131 wake it up if nothing else does.
4133 16. A malformed maildirsize file could cause Exim to calculate negative values
4134 for the mailbox size or file count. Odd effects could occur as a result.
4135 The maildirsize information is now recalculated if the size or filecount
4138 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4139 support for a long time. Removed HAVE_SYS_VFS_H.
4141 18. Updated exipick to current release from John Jetmore.
4143 19. Allow an empty sender to be matched against a lookup in an address list.
4144 Previously the only cases considered were a regular expression, or an
4147 20. Exim went into a mad DNS lookup loop when doing a callout where the
4148 host was specified on the transport, if the DNS lookup yielded more than
4151 21. The RFC2047 encoding function was originally intended for short strings
4152 such as real names; it was not keeping to the 75-character limit for
4153 encoded words that the RFC imposes. It now respects the limit, and
4154 generates multiple encoded words if necessary. To be on the safe side, I
4155 have increased the buffer size for the ${rfc2047: expansion operator from
4158 22. Failure to deliver a bounce message always caused it to be frozen, even if
4159 there was an errors_to setting on the router. The errors_to setting is now
4162 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4163 canonical form (fully expanded) before being placed in
4164 $sender_host_address.
4166 24. Updated eximstats to version 1.33
4168 25. Include certificate and key file names in error message when GnuTLS fails
4169 to set them up, because the GnuTLS error message doesn't include the name
4170 of the failing file when there is a problem reading it.
4172 26. Expand error message when OpenSSL has problems setting up cert/key files.
4175 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4176 (this can affect the format of dates).
4178 28. exim_tidydb, when checking for the continued existence of a message for
4179 which it has found a message-specific retry record, was not finding
4180 messages that were in split spool directories. Consequently, it was
4181 deleting retry records that should have stayed in existence.
4183 29. eximstats updated to version 1.35
4184 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4185 1.35 - bugfix such that pie charts by volume are generated correctly
4187 30. The SPA authentication driver was not abandoning authentication and moving
4188 on to the next authenticator when an expansion was forced to fail,
4189 contradicting the general specification for all authenticators. Instead it
4190 was generating a temporary error. It now behaves as specified.
4192 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4193 (the order specifies the preference for clients). The order is now AES256,
4194 AES128, 3DES, ARCFOUR128.
4196 31. Small patch to Sieve code - explicitly set From: when generating an
4199 32. Exim crashed if a remote delivery caused a very long error message to be
4200 recorded - for instance if somebody sent an entire SpamAssassin report back
4201 as a large number of 550 error lines. This bug was coincidentally fixed by
4202 increasing the size of one of Exim's internal buffers (big_buffer) that
4203 happened as part of the Exiscan merge. However, to be on the safe side, I
4204 have made the code more robust (and fixed the comments that describe what
4207 33. Some experimental protocols are using DNS PTR records for new purposes. The
4208 keys for these records are domain names, not reversed IP addresses. The
4209 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4210 leaves it alone. Component reversal etc. now happens only for IP addresses.
4213 34. The host_aton() function is supposed to be passed a string that is known
4214 to be a valid IP address. However, in the case of IPv6 addresses, it was
4215 not checking this. This is a hostage to fortune. Exim now panics and dies
4216 if the condition is not met. A case was found where this could be provoked
4217 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4218 components; fortuitously, this particular loophole had already been fixed
4219 by change 4.50/55 or 4.44/33 above.
4221 If there are any other similar loopholes, the new check in host_aton()
4222 itself should stop them being exploited. The report I received stated that
4223 data on the command line could provoke the exploit when Exim was running as
4224 exim, but did not say which command line option was involved. All I could
4225 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4226 running as the user.
4229 35. There was a buffer overflow vulnerability in the SPA authentication code
4230 (which came originally from the Samba project). I have added a test to the
4231 spa_base64_to_bits() function which I hope fixes it.
4234 36. The daemon start-up calls getloadavg() while still root for those OS that
4235 need the first call to be done as root, but it missed one case: when
4236 deliver_queue_load_max is set with deliver_drop_privilege. This is
4237 necessary for the benefit of the queue runner, because there is no re-exec
4238 when deliver_drop_privilege is set.
4240 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4241 was in use, was not putting the data itself into the right store pool;
4242 consequently, it could be overwritten for a subsequent message in the same
4243 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4246 38. Sometimes the final signoff response after QUIT could fail to get
4247 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4248 < 0 before doing a fflush(). This bug looks as though it goes back to the
4249 introduction of TLS in release 3.20, but "sometimes" must have been rare
4250 because the tests only now provoked it.
4256 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4257 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4258 Now it returns an integer. A number of calls were still expecting a T/F
4259 return. Fortuitously, in all cases, the tests worked in OK situations,
4260 which is the norm. However, things would have gone wrong on any write
4261 failures on the smtp file descriptor. This function is used when sending
4262 messages over SMTP and also when doing verify callouts.
4264 2. When Exim is called to do synchronous delivery of a locally submitted
4265 message (the -odf or -odi options), it no longer closes stderr before doing
4268 3. Implemented the mua_wrapper option.
4270 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4272 5. Implemented the functions header_remove(), header_testname(),
4273 header_add_at_position(), and receive_remove_recipient(), and exported them
4276 6. If an ACL "warn" statement specified the addition of headers, Exim already
4277 inserted X-ACL-Warn: at the start if there was no header name. However, it
4278 was not making this test for the second and subsequent header lines if
4279 there were newlines in the string. This meant that an invalid header could
4280 be inserted if Exim was badly configured.
4282 7. Allow an ACL "warn" statement to add header lines at the start or after all
4283 the Received: headers, as well as at the end.
4285 8. Added the rcpt_4xx retry error code.
4287 9. Added postmaster_mailfrom=xxx to callout verification option.
4289 10. Added mailfrom=xxxx to the callout verification option, for verify=
4292 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4293 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4295 12. Inserted some casts to stop certain compilers warning when using pointer
4296 differences as field lengths or precisions in printf-type calls (mostly
4297 affecting debugging statements).
4299 13. Added optional readline() support for -be (dynamically loaded).
4301 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4302 same clock tick as a message's arrival, so that its received time was the
4303 same as the "first fail" time on the retry record, and that message
4304 remained on the queue past the ultimate address timeout, every queue runner
4305 would try a delivery (because it was past the ultimate address timeout) but
4306 after another failure, the ultimate address timeout, which should have then
4307 bounced the address, did not kick in. This was a "< instead of <=" error;
4308 in most cases the first failure would have been in the next clock tick
4309 after the received time, and all would be well.
4311 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4312 being recognized when the domain list was tested by the match_domain
4313 condition in an expansion string.
4315 16. Added the ${str2b64: operator.
4317 17. Exim was always calling setrlimit() to set a large limit for the number of
4318 processes, without checking whether the existing limit was already
4319 adequate. (It did check for the limit on file descriptors.) Furthermore,
4320 errors from getrlimit() and setrlimit() were being ignored. Now they are
4321 logged to the main and panic logs, but Exim does carry on, to try to do its
4322 job under whatever limits there are.
4324 18. Imported PCRE 5.0.
4326 19. Trivial typo in log message " temporarily refused connection" (the leading
4329 20. If the log selector return_path_on_delivery was set and an address was
4330 redirected to /dev/null, the delivery process crashed because it assumed
4331 that a return path would always be set for a "successful" delivery. In this
4332 case, the whole delivery is bypassed as an optimization, and therefore no
4335 21. Internal re-arrangement: the function for sending a challenge and reading
4336 a response while authentication was assuming a zero-terminated challenge
4337 string. It's now changed to take a pointer and a length, to allow for
4338 binary data in such strings.
4340 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4342 23. Exim was not respecting finduser_retries when seeking the login of the
4343 uid under which it was called; it was always trying 10 times. (The default
4344 setting of finduser_retries is zero.) Also, it was sleeping after the final
4345 failure, which is pointless.
4347 24. Implemented tls_on_connect_ports.
4349 25. Implemented acl_smtp_predata.
4351 26. If the domain in control=submission is set empty, Exim assumes that the
4352 authenticated id is a complete email address when it generates From: or
4353 Sender: header lines.
4355 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4356 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4357 chown and chgrp in /bin and hostname in /usr/bin.
4359 28. Exim was keeping the "process log" file open after each use, just as it
4360 does for the main log. This opens the possibility of it remaining open for
4361 long periods when the USR1 signal hits a daemon. Occasional processlog
4362 errors were reported, that could have been caused by this. Anyway, it seems
4363 much more sensible not to leave this file open at all, so that is what now
4366 29. The long-running daemon process does not normally write to the log once it
4367 has entered its main loop, and it closes the log before doing so. This is
4368 so that log files can straightforwardly be renamed and moved. However,
4369 there are a couple of unusual error situations where the daemon does write
4370 log entries, and I had neglected to close the log afterwards.
4372 30. The text of an SMTP error response that was received during a remote
4373 delivery was being truncated at 512 bytes. This is too short for some of
4374 the long messages that one sometimes sees. I've increased the limit to
4377 31. It is now possible to make retry rules that apply only when a message has a
4378 specific sender, in particular, an empty sender.
4380 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4381 it possible to be selective about when SMTP synchronization is enforced.
4383 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4385 32. Implemented hosts_connection_nolog.
4387 33. Added an ACL for QUIT.
4389 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4392 35. Added mailbox_size and mailbox_filecount to appendfile.
4394 36. Added control = no_multiline_responses to ACLs.
4396 37. There was a bug in the logic of the code that waits for the clock to tick
4397 in the case where the clock went backwards by a substantial amount such
4398 that the microsecond fraction of "now" was more than the microsecond
4399 fraction of "then" (but the whole seconds number was less).
4401 38. Added support for the libradius Radius client library this is found on
4402 FreeBSD (previously only the radiusclient library was supported).
4408 1. When certain lookups returned multiple values in the form name=value, the
4409 quoting of the values was not always being done properly. Specifically:
4410 (a) If the value started with a double quote, but contained no whitespace,
4412 (b) If the value contained whitespace other than a space character (i.e.
4413 tabs or newlines or carriage returns) it was not quoted.
4414 This fix has been applied to the mysql and pgsql lookups by writing a
4415 separate quoting function and calling it from the lookup code. The fix
4416 should probably also be applied to nisplus, ibase and oracle lookups, but
4417 since I cannot test any of those, I have not disturbed their existing code.
4419 2. A hit in the callout cache for a specific address caused a log line with no
4420 reason for rejecting RCPT. Now it says "Previous (cached) callout
4421 verification failure".
4423 3. There was an off-by-one bug in the queryprogram router. An over-long
4424 return line was truncated at 256 instead of 255 characters, thereby
4425 overflowing its buffer with the terminating zero. As well as fixing this, I
4426 have increased the buffer size to 1024 (and made a note to document this).
4428 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4429 when Exim is waiting for an SMTP response from a remote server, Exim
4430 restarts its select() call on the socket, thereby resetting its timeout.
4431 This is not a problem when such interrupts are rare. Somebody set up a cron
4432 job to run exiwhat every 2 minutes, which is less than the normal select()
4433 timeout (5 or 10 minutes). This meant that the select() timeout never
4434 kicked in because it was always reset. I have fixed this by comparing the
4435 time when an interrupt arrives with the time at the start of the first call
4436 to select(). If more time than the timeout has elapsed, the interrupt is
4437 treated as a timeout.
4439 5. Some internal re-factoring in preparation for the addition of Sieve
4440 extensions (by MH). In particular, the "personal" test is moved to a
4441 separate function, and given an option for scanning Cc: and Bcc: (which is
4442 not set for Exim filters).
4444 6. When Exim created an email address using the login of the caller as the
4445 local part (e.g. when creating a From: or Sender: header line), it was not
4446 quoting the local part when it contained special characters such as @.
4448 7. Installed new OpenBSD configuration files.
4450 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4451 try to make them clearer.
4453 9. Callout options, other than the timeout value, were being ignored when
4454 verifying sender addresses in header lines. For example, when using
4456 verify = header_sender/callout=no_cache
4458 the cache was (incorrectly) being used.
4460 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4461 only the Cygwin environment.
4463 11. When return_path_on_delivery was set as a log selector, if different remote
4464 addresses in the same message used different return paths and parallel
4465 remote delivery occurred, the wrong values would sometimes be logged.
4466 (Whenever a remote delivery process finished, the return path value from
4467 the most recently started remote delivery process was logged.)
4469 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4470 lines when AUTH and/or TLS are in use. This is the "received protocol"
4471 field. Exim used to use "asmtp" for authenticated SMTP, without any
4472 indication (in the protocol name) for TLS use. Now it follows the RFC and
4473 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4474 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4475 names appear in log lines as well as in Received: header lines.
4477 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4478 extensions, and comparison tests, and to fix some bugs.
4480 14. Changes to the "personal" filter test:
4482 (1) The test was buggy in that it was just doing the equivalent of
4483 "contains" tests on header lines. For example, if a user's address was
4484 anne@some.where, the "personal" test would incorrectly be true for
4486 To: susanne@some.where
4488 This test is now done by extracting each address from the header in turn,
4489 and checking the entire address. Other tests that are part of "personal"
4490 are now done using regular expressions (for example, to check local parts
4491 of addresses in From: header lines).
4493 (2) The list of non-personal local parts in From: addresses has been
4494 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4495 taken from the Sieve specification recommendations.
4497 (3) If the message contains any header line starting with "List-" it is
4498 treated as non-personal.
4500 (4) The test for "circular" in the Subject: header line has been removed
4501 because it now seems ill-conceived.
4503 15. Minor typos in src/EDITME comments corrected.
4505 16. Installed latest exipick from John Jetmore.
4507 17. If headers_add on a router specified a text string that was too long for
4508 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4509 of string_sprintf() is now avoided.
4511 18. $message_body_size was not set (it was always zero) when running the DATA
4512 ACL and the local_scan() function.
4514 19. For the "mail" command in an Exim filter, no default was being set for
4515 the once_repeat time, causing a random time value to be used if "once" was
4516 specified. (If the value happened to be <= 0, no repeat happened.) The
4517 default is now 0s, meaning "never repeat". The "vacation" command was OK
4518 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4519 (I found it when inspecting the code).
4521 20. There is now an overall timeout for performing a callout verification. It
4522 defaults to 4 times the callout timeout, which applies to individual SMTP
4523 commands during the callout. The overall timeout applies when there is more
4524 than one host that can be tried. The timeout is checked before trying the
4525 next host. This prevents very long delays if there are a large number of
4526 hosts and all are timing out (e.g. when the network connections are timing
4527 out). The value of the overall timeout can be changed by specifying an
4528 additional sub-option for "callout", called "maxwait". For example:
4530 verify = sender/callout=5s,maxwait=20s
4532 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4533 to the end before writing, but this should make it even safer).
4535 22. Exim was forgetting that it had advertised PIPELINING for the second and
4536 subsequent messages on an SMTP connection. It was also not resetting its
4537 memory on STARTTLS and an internal HELO.
4539 23. When Exim logs an SMTP synchronization error within a session, it now
4540 records whether PIPELINING has been advertised or not.
4542 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4543 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4544 rather than long int.
4546 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4548 26. Added the never_mail option to autoreply.
4554 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4555 crash if the getsockname() call failed; this can happen if a connection is
4556 closed very soon after it is established. The problem was simply in the
4557 order in which certain operations were done, causing Exim to try to write
4558 to the SMTP stream before it had set up the file descriptor. The bug has
4559 been fixed by making things happen in the correct order.
4565 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4566 before the connection was closed, thus losing the rejection response.
4568 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4569 some early Solaris releases, but causes trouble in current releases where
4570 socklen_t is defined.
4572 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4575 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4578 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4579 that releasing the top of it at the end releases what was used for sub-
4580 expansions (unless the block got too big). However, discard this block if
4581 the first thing is a variable or header, so that we can use its block when
4582 it is dynamic (useful for very large $message_headers, for example).
4584 6. Lookups now cache *every* query, not just the most recent. A new, separate
4585 store pool is used for this. It can be recovered when all lookup caches are
4586 flushed. Lookups now release memory at the end of their result strings.
4587 This has involved some general refactoring of the lookup sources.
4589 7. Some code has been added to the store_xxx() functions to reduce the amount
4590 of flapping under certain conditions.
4592 8. log_incoming_interface used to affect only the <= reception log lines. Now
4593 it causes the local interface and port to be added to several more SMTP log
4594 lines, for example "SMTP connection from", and rejection lines.
4596 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4598 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4600 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4601 could be overwritten at the end of the current message (or the start of a
4602 new message if it was set in a HELO ACL). The value is now preserved for
4603 the duration of the SMTP connection.
4605 12. If a transport had a headers_rewrite setting, and a matching header line
4606 contained an unqualified address, that address was qualified, even if it
4607 did not match any rewriting rules. The underlying bug was that the values
4608 of the flags that permit the existence of unqualified sender and recipient
4609 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4610 non-local messages, and by -bnq for local messages) were not being
4611 preserved with the message after it was received.
4613 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4614 "next input=" as part of the text comprising the host identity instead of
4615 the correct text. The code was using the same buffer for two different
4616 strings. However, depending on which order the printing function evaluated
4617 its arguments, the bug did not always show up. Under Linux, for example, my
4618 test suite worked just fine.
4620 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4621 doesn't work with some older versions of Perl. It has been changed to "my",
4622 which in any case is probably the better facility to use.
4624 15. A really picky compiler found some instances of statements for creating
4625 error messages that either had too many or two few arguments for the format
4628 16. The size of the buffer for calls to the DNS resolver has been increased
4629 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4630 for addresses that have a lot of PTR records. This alleviates a problem; it
4631 does not fully solve it.
4633 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4634 buffer now truncates the list and logs the incident, which is the same
4635 action as happens when Exim is looking up a host name and its aliases.
4636 Previously in this situation something unpredictable would happen;
4637 sometimes it was "internal error: store_reset failed".
4639 18. If a server dropped the connection unexpectedly when an Exim client was
4640 using GnuTLS and trying to read a response, the client delivery process
4641 crashed while trying to generate an error log message.
4643 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4644 string, for example:
4646 warn message = H1: something\nH2: something
4648 the text was added as a single header line from Exim's point of view
4649 though it ended up OK in the delivered message. However, searching for the
4650 second and subsequent header lines using $h_h2: did not work. This has been
4651 fixed. Similarly, if a system filter added multiple headers in this way,
4652 the routers could not see them.
4654 20. Expanded the error message when iplsearch is called with an invalid key to
4655 suggest using net-iplsearch in a host list.
4657 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4658 ACLs are no longer actually imposed (and a message to that effect is
4661 22. If a "gecos" field in a passwd entry contained escaped characters, in
4662 particular, if it contained a \" sequence, Exim got it wrong when building
4663 a From: or a Sender: header from that name. A second bug also caused
4664 incorrect handling when an unquoted " was present following a character
4665 that needed quoting.
4667 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4668 was not being matched caselessly.
4670 24. Arranged for all hyphens in the exim.8 source to be escaped with
4673 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4674 properly. Recipient callout cache records were still being keyed to include
4675 the sender, even when use_sender was set false. This led to far more
4676 callouts that were necessary. The sender is no longer included in the key
4677 when use_sender is false.
4679 26. Added "control = submission" modifier to ACLs.
4681 27. Added the ${base62d: operator to decode base 62 numbers.
4683 28. dnsdb lookups can now access SRV records.
4685 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4686 the configuration file.
4688 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4689 (-v) mode. This makes the output for a verbose queue run more intelligible.
4691 31. Added a use_postmaster feature to recipient callouts.
4693 32. Added the $body_zerocount variable, containing the number of binary zero
4694 bytes in the message body.
4696 33. The time of last modification of the "new" subdirectory is now used as the
4697 "mailbox time last read" when there is a quota error for a maildir
4700 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4702 35. Added +ignore_unknown as a special item in host lists.
4704 36. Code for decoding IPv6 addresses in host lists is now included, even if
4705 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4706 address was recognized as an IP address, but was then not correctly decoded
4707 into binary, causing unexpected and incorrect effects when compared with
4714 1. Very minor rewording of debugging text in manualroute to say "list of
4715 hosts" instead of "hostlist".
4717 2. If verify=header_syntax was set, and a header line with an unqualified
4718 address (no domain) and a large number of spaces between the end of the
4719 name and the colon was received, the reception process suffered a buffer
4720 overflow, and (when I tested it) crashed. This was caused by some obsolete
4721 code that should have been removed. The fix is to remove it!
4723 3. When running in the test harness, delay a bit after writing a bounce
4724 message to get a bit more predictability in the log output.
4726 4. Added a call to search_tidyup() just before forking a reception process. In
4727 theory, someone could use a lookup in the expansion of smtp_accept_max_
4728 per_host which, without the tidyup, could leave open a database connection.
4730 5. Added the variables $recipient_data and $sender_data which get set from a
4731 lookup success in an ACL "recipients" or "senders" condition, or a router
4732 "senders" option, similar to $domain_data and $local_part_data.
4734 6. Moved the writing of debug_print from before to after the "senders" test
4737 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4738 problems for message scanning, either using a data ACL, or using
4739 local_scan() because the Received: header was not generated till after they
4740 were called (in order to set the time as the time of reception completion).
4741 I have revised the way this works. The header is now generated after the
4742 body is received, but before the ACL or local_scan() are called. After they
4743 are run, the timestamp in the header is updated.
4749 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4750 before starting a queue runner without re-exec. This happened only when
4751 deliver_drop_privilege was set or when the Exim user was set to root. The
4752 effect of the bug was that timeouts during subsequent deliveries caused
4753 crashes instead of being properly handled. The handler is now left at its
4754 default (and expected) setting.
4756 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4757 message, again when deliver_drop_privilege is set or Exim is run as root.
4758 The bug described in (1) was not present in this case, but the tidying up
4759 of the other signals was missing. I have made the two cases consistent.
4761 3. The ignore_target_hosts setting on a manualroute router was being ignored
4762 for hosts that were looked up using the /MX notation.
4764 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4767 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4768 operated on the sender address. After changing the $sender_address to <>
4769 for the sender address verify, Exim was re-instated it as the original
4770 (before rewriting) address, but remembering that it had rewritten it, so it
4771 wasn't rewriting it again. This bug also had the effect of breaking the
4772 sender address verification caching when the sender address was rewritten.
4774 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4775 This has been changed so that if the ip literal address matches
4776 ignore_target_hosts, the router declines.
4778 7. Added expansion conditions match_domain, match_address, and match_local_
4779 part (NOT match_host).
4781 8. The placeholder for the Received: header didn't have a length field set.
4783 9. Added code to Exim itself and to exim_lock to test for a specific race
4784 condition that could lead to file corruption when using MBX delivery. The
4785 issue is with the lockfile that is created in /tmp. If this file is removed
4786 after a process has opened it but before that process has acquired a lock,
4787 there is the potential for a second process to recreate the file and also
4788 acquire a lock. This could lead to two Exim processes writing to the file
4789 at the same time. The added code performs the same test as UW imapd; it
4790 checks after acquiring the lock that its file descriptor still refers to
4791 the same named file.
4793 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4794 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4797 11. Added the smtp_active_hostname option. If used, this will typically be made
4798 to depend on the incoming interface address. Because $interface_address is
4799 not set up until the daemon has forked a reception process, error responses
4800 that can happen earlier (such as "too many connections") no longer contain
4803 12. If an expansion in a condition on a "warn" statement fails because a lookup
4804 defers, the "warn" statement is abandoned, and the next ACL statement is
4805 processed. Previously this caused the whole ACL to be aborted.
4807 13. Added the iplsearch lookup type.
4809 14. Added ident_timeout as a log selector.
4811 15. Added tls_certificate_verified as a log selector.
4813 16. Added a global option tls_require_ciphers (compare the smtp transport
4814 option of the same name). This controls incoming TLS connections.
4816 17. I finally figured out how to make tls_require_ciphers do a similar thing
4817 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4818 before starting the TLS session.
4820 18. Tabs are now shown as \t in -bP output.
4822 19. If the log selector return_path_on_delivery was set, Exim crashed when
4823 bouncing a message because it had too many Received: header lines.
4825 20. If two routers both had headers_remove settings, and the first one included
4826 a superfluous trailing colon, the final name in the first list and the
4827 first name in the second list were incorrectly joined into one item (with a
4828 colon in the middle).
4834 1. Added -C and -D options to the exinext utility, mainly to make it easier
4835 to include in the automated testing, but these could be helpful when
4836 multiple configurations are in use.
4838 2. The exinext utility was not formatting the output nicely when there was
4839 an alternate port involved in the retry record key, nor when there was a
4840 message id as well (for retries that were specific to a specific message
4841 and a specific host). It was also confused by IPv6 addresses, because of
4842 the additional colons they contain. I have fixed the IPv4 problem, and
4843 patched it up to do a reasonable job for IPv6.
4845 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4846 delivery, the log line now contains "pipelined" if PIPELINING was used.
4848 4. An SMTP transport process used to panic and die if the bind() call to set
4849 an explicit outgoing interface failed. This has been changed; it is now
4850 treated in the same way as a connect() failure.
4852 5. A reference to $sender_host_name in the part of a conditional expansion
4853 that was being skipped was still causing a DNS lookup. This no longer
4856 6. The def: expansion condition was not recognizing references to header lines
4857 that used bh_ and bheader_.
4859 7. Added the _cache feature to named lists.
4861 8. The code for checking quota_filecount in the appendfile transport was
4862 allowing one more file than it should have been.
4864 9. For compatibility with Sendmail, the command line option
4872 and sets the incoming protocol and host name (for trusted callers). The
4873 host name and its colon can be omitted when only the protocol is to be set.
4874 Note the Exim already has two private options, -pd and -ps, that refer to
4875 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4876 "s", but I don't think that's a major issue.
4878 10. A number of refactoring changes to the code, none of which should affect
4881 (a) The number of logging options was getting close to filling up the
4882 32-bit word that was used as a bit map. I have split them into two classes:
4883 those that are passed in the argument to log_write(), and those that are
4884 only ever tested independently outside of that function. These are now in
4885 separate 32-bit words, so there is plenty of room for expansion again.
4886 There is no change in the user interface or the logging behaviour.
4888 (b) When building, for example, log lines, the code previously used a
4889 macro that called string_cat() twice, in order to add two strings. This is
4890 not really sufficiently general. Furthermore, there was one instance where
4891 it was actually wrong because one of the argument was used twice, and in
4892 one call a function was used. (As it happened, calling the function twice
4893 did not affect the overall behaviour.) The macro has been replaced by a
4894 function that can join an arbitrary number of extra strings onto a growing
4897 (c) The code for expansion conditions now uses a table and a binary chop
4898 instead of a serial search (which was left over from when there were very
4899 few conditions). Also, it now recognizes conditions like "pam" even when
4900 the relevant support is not compiled in: a suitably worded error message is
4901 given if an attempt is made to use such a condition.
4903 11. Added ${time_interval:xxxxx}.
4905 12. A bug was causing one of the ddress fields not to be passed back correctly
4906 from remote delivery subprocesses. The field in question was not being
4907 subsequently used, so this caused to problems in practice.
4909 13. Added new log selectors queue_time and deliver_time.
4911 14. Might have fixed a bug in maildirsizefile handling that threw up
4912 "unexpected character" debug warnings, and recalculated the data
4913 unnecessarily. In any case, I expanded the warning message to give more
4916 15. Added the message "Restricted characters in address" to the statements in
4917 the default ACL that block characters like @ and % in local parts.
4919 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4920 Three changes have been made:
4922 (a) There was a serious bug; a negative response to MAIL caused the whole
4923 recipient domain to be cached as invalid, thereby blocking all messages
4924 to all local parts at the same domain, from all senders. This bug has
4925 been fixed. The domain is no longer cached after a negative response to
4926 MAIL if the sender used is not empty.
4928 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4931 (c) A new callout option, "use_sender" has been added for people who want
4932 the modified behaviour.
4938 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4941 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4942 indeed breaks things for older releases.
4944 3. Added additional logging to the case where there is a problem reading data
4945 from a filter that is running in a subprocess using a pipe, in order to
4946 try to track down a specific problem.
4948 4. Testing facility fudge: when running in the test harness and attempting
4949 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4950 getting "No route to host". Convert this to a timeout.
4952 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4955 6. Some OS don't have socklen_t but use size_t instead. This affects the
4956 fifth argument of getsockopt() amongst other things. This is now
4957 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4958 can be set for individual OS. I have set it for SunOS5, OSF1, and
4959 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4960 some earlier ones do not.
4962 7. Change 4.30/15 was not doing the test caselessly.
4964 8. The standard form for an IPv6 address literal was being rejected by address
4965 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4966 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4967 this, as well as the form without the "IPv6" on the front (but only when
4968 address literals are enabled, of course).
4970 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4972 10. Exim crashed if a message with an empty sender address specified by -f
4973 encountered a router with an errors_to setting. This could be provoked only
4974 by a command such as
4978 where an empty string was supplied; "<>" did not hit this bug.
4980 11. Installed PCRE release 4.5.
4982 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4983 remained set. It is now erased.
4985 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4986 times from message ids (which are base 36 rather than the normal 62).
4988 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4989 were being counted as actual protocol errors, and logged if the log
4990 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4991 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4992 and DATA following a set of rejected RCPTs do not count as protocol errors.
4993 In other words, Exim assumes they were pipelined, though this may not
4994 actually be the case. Of course, in all cases the client gets an
4995 appropriate error code.
4997 15. If a lookup fails in an ACL condition, a message about the failure may
4998 be available; it is used if testing the ACL cannot continue, because most
4999 such messages specify what the cause of the deferral is. However, some
5000 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5001 that caused an old message to be retained and used if a later statement
5002 caused a defer, replacing the real cause of the deferral.
5004 16. If an IP address had so many PTR records that the DNS lookup buffer
5005 was not large enough to hold them, Exim could crash while trying to process
5006 the truncated data. It now detects and logs this case.
5008 17. Further to 4.21/58, another change has been made: if (and only if) the
5009 first line of a message (the first header line) ends with CRLF, a bare LF
5010 in a subsequent header line has a space inserted after it, so as not to
5011 terminate the header.
5013 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5014 unnecessarily, used atoi() instead of strtol(), and didn't check the
5015 termination when getting file sizes from file names by regex.
5017 19. Completely re-implemented the support for maildirsize files, in the light
5018 of a number of problems with the previous contributed implementation
5019 (4.30/29). In particular:
5021 . If the quota is zero, the maildirsize file is maintained, but no quota is
5024 . If the maildir directory does not exist, it is created before any attempt
5025 to write a maildirsize file.
5027 . The quota value in the file is just a cache; if the quota is changed in
5028 the transport, the new value overrides.
5030 . A regular expression is available for excluding directories from the
5033 20. The autoreply transport checks the characters in options that define the
5034 message's headers; it allows continued headers, but it was checking with
5035 isspace() after an embedded newline instead of explicitly looking for a
5038 21. If all the "regular" hosts to which an address was routed had passed their
5039 expiry times, and had not reached their retry times, the address was
5040 bounced, even if fallback hosts were defined. Now Exim should go on to try
5043 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5044 equivalent code in the SMTP transport. Some hosts send humungous responses
5045 to HELO/EHLO, more than 1024 it seems.
5047 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5048 gives compiler warnings in some environments. I've now done it "properly",
5051 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5052 (because of problems with the built-in one) was declared to return uschar *
5053 instead of char *, causing compiler failure.
5055 25. Fixed a file descriptor leak when processing alias/forward files.
5057 26. Fixed a minor format string issue in dbfn.c.
5059 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5061 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5062 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5065 29. When a sender address is verified, it is cached, to save repeating the test
5066 when there is more than one recipient in a message. However, when the
5067 verification involves a callout, it is possible for different callout
5068 options to be set for different recipients. It is too complicated to keep
5069 track of this in the cache, so now Exim always runs a verification when a
5070 callout is required, relying on the callout cache for the optimization.
5071 The overhead is duplication of the address routing, but this should not be
5074 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5075 to be verified with callout=postmaster, and the main callout worked but the
5076 postmaster check failed, the verification correctly failed. However, if a
5077 subsequent RCPT command asked for sender verification *without* the
5078 postmaster check, incorrect caching caused this verification also to fail,
5081 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5082 it was not caching the DNS options (qualify_single, search_parents) that
5083 were used when the lookup failed. A subsequent lookup with different
5084 options therefore always gave the same answer, though there were cases
5085 where it should not have. (Example: a "domains = !$mx_any" option on a
5086 dnslookup router: the "domains" option is always processed without any
5087 widening, but the router might have qualify_single set.) Now Exim uses the
5088 cached value only when the same options are set.
5090 32. Added John Jetmore's "exipick" utility to the distribution.
5092 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5093 than a timeout (e.g. a certificate is required, and is not provided), an
5094 Exim server now closes the connection immediately. Previously it waited for
5095 the client to close - but if the client is SSL, it seems that they each
5096 wait for each other, leading to a delay before one of them times out.
5098 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5099 maintained 0.8.x compatibility because I don't think many are using it, and
5100 it is clearly obsolete.
5102 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5105 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5106 client certificate was expired. A simple patch fixes this, though I don't
5107 understand the full logic of why the verify callback is called multiple
5110 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5111 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5112 which causes problems with some clients (such as the Certicom SSL Plus
5113 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5114 disables the coutermeasure allowing Eudora to connect."
5116 38. Exim was not checking that a write() to a log file succeeded. This could
5117 lead to Bad Things if a log got too big, in particular if it hit a file
5118 size limit. Exim now panics and dies if it cannot write to a log file, just
5119 as it does if it cannot open a log file.
5121 39. Modified OS/Makefile-Linux so that it now contains
5123 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5125 The two -D definitions ensure that Exim is compiled with large file
5126 support, which makes it possible to handle log files that are bigger than
5129 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5130 instance) a domain was checked against a named list that involved a lookup,
5131 causing $domain_data to be set, then another domain was checked against the
5132 same list, then the first domain was re-checked, the value of $domain_data
5133 after the final check could be wrong. In particular, if the second check
5134 failed, it could be set empty. This bug probably also applied to
5137 41. The strip_trailing_dot option was not being applied to the address given
5138 with the -f command-line option.
5140 42. The code for reading a message's header from the spool was incrementing
5141 $received_count, but never initializing it. This meant that the value was
5142 incorrect (doubled) while delivering a message in the same process in which
5143 it was received. In the most common configuration of Exim, this never
5144 happens - a fresh exec is done - but it can happen when
5145 deliver_drop_privilege is set.
5147 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5148 it now includes up to 150 characters of the unexpected data in the log
5151 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5152 and building data strings. The size of both of these buffers was 10 000
5153 bytes - far larger than anybody would *ever* want, thought I. Needless to
5154 say, somebody hit the limit. I have increased the maximum line length to
5155 20 000 and the maximum data length of concatenated lines to 100 000. I have
5156 also fixed two bugs, because there was no checking on these buffers. Tsk,
5157 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5158 buffer is too small.
5160 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5161 lsearch lookups. Now it does.
5163 46. When parsing a route_list item in a manualroute router, a fixed-length
5164 buffer was used for the list of hosts. I made this 1024 bytes long,
5165 thinking that nobody would ever have a list of hosts that long. Wrong.
5166 Somebody had a whole pile of complicated expansion conditions, and the
5167 string was silently truncated, leading to an expansion error. It turns out
5168 that it is easier to change to an unlimited length (owing to other changes
5169 that have happened since this code was originally written) than to build
5170 structure for giving a limitation error. The length of the item that
5171 expands into the list of hosts is now unlimited.
5173 47. The lsearch lookup could not handle data where the length of text line was
5174 more than 4095 characters. Such lines were truncated, leading to shortened
5175 data being returned. It should now handle lines of any length.
5177 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5178 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5181 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5182 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5183 no longer happens in comment lines. A list of replacements is now placed
5184 at the head of all of the source files, except those whose only change is
5185 to replace PERL_COMMAND in the very first #! line.
5187 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5188 messages on the queue, with a bottom-up merge sort, using code contributed
5189 by Michael Haardt. This should make operations like -bp somewhat faster on
5190 large queues. It won't affect queue runners, except when queue_run_in_order
5193 51. Installed eximstats 1.31 in the distribution.
5195 52. Added support for SRV lookups to the dnslookup router.
5197 53. If an ACL referred to $message_body or $message_body_end, the value was not
5198 reset for any messages that followed in the same SMTP session.
5200 54. The store-handling optimization for building very long strings was not
5201 differentiating between the different store pools. I don't think this
5202 actually made any difference in practice, but I've tidied it.
5204 55. While running the routers to verify a sender address, $sender_address
5205 was still set to the sender address. This is wrong, because when routing to
5206 send a bounce to the sender, it would be empty. Therefore, I have changed
5207 it so that, while verifying a sender address, $sender_address is set to <>.
5208 (There is no change to what happens when verifying a recipient address.)
5210 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5211 target A or AAAA records (if not already returned) without resetting the
5212 qualify_single or search_parents options of the DNS resolver. These are
5213 inappropriate in this case because the targets of MX and SRV records must
5214 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5215 target that, when qualified, matched something in the local domain. These
5216 two options are now turned off when doing these lookups.
5218 57. It seems that at least some releases of Reiserfs (which does not have the
5219 concept of a fixed number of inodes) returns zero and not -1 for the
5220 number of available inodes. This interacted badly with check_spool_inodes,
5221 which assumed that -1 was the "no such thing" setting. What I have done is
5222 to check that the total number of inodes is greater than zero before doing
5223 the test of how many are available.
5225 58. When a "warn" ACL statement has a log_message modifier, the message is
5226 remembered, and not repeated. This is to avoid a lot of repetition when a
5227 message has many recipients that cause the same warning to be written.
5228 Howewer, Exim was preserving the list of already written lines for an
5229 entire SMTP session, which doesn't seem right. The memory is now reset if a
5230 new message is started.
5232 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5233 debugging output unless log_rewrite was also set.
5235 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5236 of (int)(handle) when we know that handle contains (void *)(-1).
5238 61. The Exim daemon panic-logs an error return when it closes the incoming
5239 connection. However "connection reset by peer" seems to be common, and
5240 isn't really an error worthy of noting specially, so that particular error
5243 62. When Exim is trying to find all the local interfaces, it used to panic and
5244 die if the ioctl to get the interface flags failed. However, it seems that
5245 on at least one OS (Solaris 9) it is possible to have an interface that is
5246 included in the list of interfaces, but for which you get a failure error
5247 for this call. This happens when the interface is not "plumbed" into a
5248 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5249 failure of the "get flags" call assumes that the interface is down.
5251 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5252 makes life easier for people who are doing arithmetic on fields extracted
5253 from dates, where you often get leading zeros that should not be
5254 interpreted as octal.
5256 64. Added qualify_domain to the redirect router, to override the global
5259 65. If a pathologically long header line contained very many addresses (the
5260 report of this problem mentioned 10 000) and each of them was rewritten,
5261 Exim could use up a very large amount of memory. (It kept on making new
5262 copies of the header line as it rewrote, and never released the old ones.)
5263 At the expense of a bit more processing, the header rewriting function has
5264 been changed so that it no longer eats memory in this way.
5266 66. The generation of the Received: header has been moved from the time that a
5267 message starts to be received, to the time that it finishes. The timestamp
5268 in the Received: header should now be very close to that of the <= log
5269 line. There are two side-effects of this change:
5271 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5272 logged header lines no longer include the local Received: line, because
5273 it has not yet been created. The same applies to a copy of the message
5274 that is returned to a non-SMTP sender when a message is rejected.
5276 (b) When a filter file is tested using -bf, no additional Received: header
5277 is added to the test message. After some thought, I decided that this
5280 This change does not affect the value of $received_for. It is still set
5281 after address rewriting, but before local_scan() is called.
5283 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5285 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5286 gave an unhelpful panic error message, and a defer error. I have managed to
5287 change this behaviour so that it now rejects any supplied certificate,
5288 which seems right, as the list of acceptable certificates is empty.
5290 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5291 gave an unhelpful defer error. I have not managed to make this reject any
5292 supplied certificates, but the error message it gives is "no certificate
5293 supplied", which is not helpful.
5295 70. exigrep's output now also includes lines that are not associated with any
5296 message, but which match the given pattern. Implemented by a patch from
5297 Martin Sluka, which also tidied up the Perl a bit.
5299 71. Recipient callout verification, like sender verification, was using <> in
5300 the MAIL FROM command. This isn't really the right thing, since the actual
5301 sender may affect whether the remote host accepts the recipient or not. I
5302 have changed it to use the actual sender in the callout; this means that
5303 the cache record is now keyed on a recipient/sender pair, not just the
5304 recipient address. There doesn't seem to be a real danger of callout loops,
5305 since a callout by the remote host to check the sender would use <>.
5306 [SEE ABOVE: changed after hitting problems.]
5308 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5309 temporary errors. However, in the case of such a code being given after
5310 the end of a data transmission (i.e. after ".") Exim was failing to write
5311 a retry record for the message. (Yes, there was some broken host that was
5312 actually sending 8xx at this point.)
5314 73. An unknown lookup type in a host list could cause Exim to panic-die when
5315 the list was checked. (An example that provoked this was putting <; in the
5316 middle of a list instead of at the start.) If this happened during a DATA
5317 ACL check, a -D file could be left lying around. This kind of configuration
5318 error no longer causes Exim to die; instead it causes a defer errror. The
5319 incident is still logged to the main and panic logs.
5321 74. Buglet left over from Exim 3 conversion. The message "too many messages
5322 in one connection" was written to the rejectlog but not the mainlog, except
5323 when address rewriting (yes!) was being logged.
5325 75. Added write_rejectlog option.
5327 76. When a system filter was run not as root (that is, when system_filter_user
5328 was set), the values of the $n variables were not being returned to the
5329 main process; thus, they were not subsequently available in the $sn
5332 77. Added +return_path_on_delivery log selector.
5334 78. A connection timeout was being treated differently from recipients deferred
5335 when testing hosts_max_try with a message that was older than the host's
5336 retry timeout. (The host should not be counted, thus allowing all hosts to
5337 be tried at least once before bouncing.) This may have been the cause of an
5338 occasionally reported bug whereby a message would remain on the queue
5339 longer than the retry timeout, but would be bounced if a delivery was
5340 forced. I say "may" because I never totally pinned down the problem;
5341 setting up timeout/retry tests is difficult. See also the next item.
5343 79. The ultimate address timeout was not being applied to errors that involved
5344 a combination of host plus message (for example, a timeout on a MAIL
5345 command). When an address resolved to a number of possible hosts, and they
5346 were not all tried for each delivery (e.g. because of hosts_max_try), a
5347 message could remain on the queue longer than the retry timeout.
5349 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5352 81. Fixed an obscure SMTP outgoing bug which required at least the following
5353 conditions: (a) there was another message waiting for the same server;
5354 (b) the server returned 5xx to all RCPT commands in the first message so
5355 that the message was not completed; (c) the server dropped the connection
5356 or gave a negative response to the RSET that Exim sends to abort the
5357 transaction. The observed case was a dropped connection after DATA that had
5358 been sent in pipelining mode. That is, the server had advertised PIPELINING
5359 but was not implementing it correctly. The effect of the bug was incorrect
5360 behaviour, such as trying another host, and this could lead to a crash.
5366 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5367 and daemon.c were passed as pointers to ints; they should have been
5368 pointers to socklen_t variables (which are typically unsigned ints).
5370 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5373 3. Fixed a really odd bug that affected only the testing scheme; patching a
5374 certain fixed string in the binary changed the value of another string that
5375 happened to be identical to the end of the original first string.
5377 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5378 name", it returns that address as the IP address. On some operating
5379 systems (e.g. Solaris), it also passes back the IP address string as the
5380 "host name". However, on others (e.g. Linux), it passes back an empty
5381 string. Exim wasn't checking for this, and was changing the host name to an
5382 empty string, assuming it had been canonicized.
5384 5. Although rare, it is permitted to have more than one PTR record for a given
5385 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5386 all the names associated with an address, because they do in Solaris.
5387 However, it seems that they do not in Linux for data that comes from the
5388 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5389 I found this out when I moved to a new Linux workstation and tried to run
5390 the Exim test suite.
5392 To get round this problem I have changed the code so that it now does its
5393 own call to the DNS to look up PTR records when searching for a host name.
5394 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5395 addresses that are only in /etc/hosts are still found.
5397 This behaviour is, however, controlled by an option called host_lookup_
5398 order, which defaults to "bydns:byaddr". If people want to use the other
5399 order, or indeed, just use one or the other means of lookup, they can
5400 specify it in this variable.
5402 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5403 some operating systems, this comes back from gethostbyaddr() as an empty
5404 string, and this is what Exim used to test for. However, it seems that in
5405 other systems, "." is yielded. Exim now tests for this case too.
5407 7. The values of check_spool_space and check_log_space are now held internally
5408 as a number of kilobytes instead of an absolute number of bytes. If a
5409 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5410 kilobyte. This means that much larger values can be stored.
5412 8. Exim monitor: an attempt to get the action menu when not actually pointing
5413 at a message produces an empty menu entitled "No message selected". This
5414 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5415 no entries in it ("Shell widget menu has zero width and/or height"). So I
5416 have added a single, blank menu entry in this case.
5418 9. Added ${quote_local_part.
5420 10. MIME decoding is now applied to the contents of Subject: header lines when
5423 11. Now that a reference to $sender_host_address automatically causes a reverse
5424 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5425 host lookup before query-style lookups in lists that might use this
5426 variable. This has therefore been abolished, and the "net-" prefix is no
5427 longer necessary for query-style lookups.
5429 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5430 have been a typo for LFLAGS, so it has been changed.
5432 13. The install script calls Exim with "-C /dev/null" in order to find the
5433 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5434 to be output. Howeve, since Exim outputs its version number before the
5435 error, it didn't break the script. It just looked ugly. I fixed this by
5436 always allowing "-C /dev/null" if the caller is root.
5438 14. Ignore overlarge ACL variable number when reading spool file - insurance
5439 against a later release with more variables having written the file.
5441 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5442 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5443 this, as well as the form without the "IPv6" on the front.
5445 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5446 OS/Makefile-Darwin file.
5448 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5449 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5451 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5452 variable. However, some OS use ints for inodes. I've added cast to long int
5453 to get rid of the compiler warning.
5455 19. I had forgotten to lock out "/../" in configuration file names when
5456 ALT_CONFIG_PREFIX was set.
5458 20. Routers used for verification do not need to specify transports. However,
5459 if such a router generated a host list, and callout was configured, Exim
5460 crashed, because it could not find a port number from the (non-existent)
5461 transport. It now assumes port 25 in this circumstance.
5463 21. Added the -t option to exigrep.
5465 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5466 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5467 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5468 them is set, LOOKUP_LSEARCH is forced.
5470 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5472 24. Added sender and host information to the "rejected by local_scan()" log
5473 line; previously there was no indication of these.
5475 25. Added .include_if_exists.
5477 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5478 Linux. It turns out that not all file systems support this. Apparently some
5479 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5480 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5481 not supported on the file descriptor, is now ignored when Exim is trying to
5482 sync a directory. This applies only to Linux.
5484 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5486 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5487 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5490 29. Added support for maildirsize files from supplied patch (modified a bit).
5492 30. The use of :fail: followed by an empty string could lead Exim to respond to
5493 sender verification failures with (e.g.):
5495 550 Verification failed for <xxx>
5496 550 Sender verify failed
5498 where the first response line was missing the '-' that indicates it is not
5499 the final line of the response.
5501 31. The loop for finding the name of the user that called Exim had a hardwired
5502 limit of 10; it now uses the value of finduser_retries, which is used for
5503 all other user lookups.
5505 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5508 33. Exim was neglecting to zero errno before one call of strtol() when
5509 expanding a string and expecting an integer value. On some systems this
5510 resulted in spurious "integer overflow" errors. Also, it was casting the
5511 result into an int without checking.
5513 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5514 did not work. The code looks as if it has *never* worked, though it appears
5515 to have been documented since at least releast 1.62. I have made it work.
5517 35. The "timeout_DNS" error in retry rules, also documented since at least
5518 1.62, also never worked. As it isn't clear exactly what this means, and
5519 clearly it isn't a major issue, I have abolished the feature by treating it
5520 as "timeout", and writing a warning to the main and panic logs.
5522 36. The display of retry rules for -brt wasn't always showing the error code
5525 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5526 timeout_connect_A, timeout_connect_MX.
5528 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5529 to the empty sender.
5531 39. The daemon was not analysing the content of -oX till after it had closed
5532 stderr and disconnected from the controlling terminal. This meant that any
5533 syntax errors were only noted on the panic log, and the return code from
5534 the command was 0. By re-arranging the code a little, I've made the
5535 decoding happen first, so such errors now appear on stderr, and the return
5536 code is 1. However, the actual setting up of the sockets still happens in
5537 the disconnected process, so errors there are still only recorded on the
5540 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5541 connections (as happens on some IP stacks) was logged at start up time as
5542 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5543 it from "IPv6 and IPv4", which means that two separate sockets are being
5546 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5547 says whether AF_INET or AF_INET6 was passed as an argument.
5549 42. Exiwhat output was messed up when time zones were included in log
5552 43. Exiwhat now gives more information about the daemon's listening ports,
5553 and whether -tls-on-connect was used.
5555 44. The "port" option of the smtp transport is now expanded.
5557 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5558 silently ignored. Now an error message is written to the main and panic
5561 46. There's a new ACL modifier called "logwrite" which writes to a log file
5562 as soon as it is encountered.
5564 47. Added $local_user_uid and $local_user_gid at routing time.
5566 48. Exim crashed when trying to verify a sender address that was being
5569 49. Exim was recognizing only a space character after ".include". It now also
5570 recognizes a tab character.
5572 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5573 extracting the relevant information from the specification. The man page no
5574 longer contains scrambled data for the -d option, and I've added a section
5575 at the front about calling Exim under different names.
5577 51. Added "extra_headers" argument to the "mail" command in filter files.
5579 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5582 53. Installed eximstats 1.29.
5584 54. Added transport_filter_timeout as a generic transport option.
5586 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5587 Cc: header lines. This was required by RFC 822, but it not required by RFC
5590 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5591 incoming messages that did not have them. Now it does so only if the
5592 message originates locally, that is, if there is no associated remote host
5593 address. When Resent- header lines are present, this applies to the Resent-
5594 lines rather than the non-Resent- lines.
5596 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5597 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5599 58. Messages for configuration errors now include the name of the main
5600 configuration file - useful now that there may be more than one file in a
5601 list (.included file names were always shown).
5603 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5604 for those rare installations that do not start the daemon as root or run it
5605 setuid root. I've cut out the call to initgroups() if the daemon is not
5608 60. The Exim user and group can now be bound into the binary as text strings
5609 that are looked up at the start of Exim's processing.
5611 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5613 62. Added $mailstore_basename variable.
5615 63. Installed patch to sieve.c from Michael Haardt.
5617 64. When Exim failed to open the panic log after failing to open the main log,
5618 the original message it was trying to log was written to stderr and debug
5619 output, but if they were not available (the usual case in production), it
5620 was lost. Now it is written to syslog before the two lines that record the
5621 failures to open the logs.
5623 65. Users' Exim filters run in subprocesses under the user's uid. It is
5624 possible for a "deliver" command or an alias in a "personal" command to
5625 provoke an address rewrite. If logging of address rewriting is configured,
5626 this fails because the process is not running as root or exim. There may be
5627 a better way of dealing with this, but for the moment (because 4.30 needs
5628 to be released), I have disabled address rewrite logging when running a
5629 filter in a non-root, non-exim process.
5635 1. The buildconfig auxiliary program wasn't quoting the value set for
5636 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5637 not defined. This bug was masked in 4.22 by the effect that was fixed in
5640 2. Some messages that were rejected after a message id was allocated were
5641 shown as "incomplete" by exigrep. It no longer does this for messages that
5642 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5644 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5645 have allow_domain_literals set, the ID did not get logged in the <= line.
5646 Domain literals are now always recognized in Message-ID: header lines.
5648 4. The first argument for a ${extract expansion item is the key name or field
5649 number. Leading and trailing spaces in this item were not being ignored,
5650 causing some misleading effects.
5652 5. When deliver_drop_privilege was set, single queue runner processes started
5653 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5654 same command in the process it spins off) were not dropping privilege.
5656 6. When the daemon running as "exim" started a queue runner, it always
5657 re-executed Exim in the spun-off process. This is a waste of effort when
5658 deliver_drop_privilege is set. The new process now just calls the
5659 queue-runner function directly.
5665 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5668 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5669 ignored. Though the use of -oP was forcing the writing of a pid file, it
5670 was always written to the default place.
5672 3. If the message "no IP address found for host xxxx" is generated during
5673 incoming verification, it is now followed by identification of the incoming
5674 connection (so you can more easily find what provoked it).
5676 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5678 5. Added some features to "harden" Exim a bit more against certain attacks:
5680 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5681 be put in Local/Makefile. This is like the never_users runtime option,
5682 but it cannot be overridden. The default setting is "root".
5684 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5685 prefix string with which any file named in a -C command line option
5688 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5689 is retained for -C and -D only if the caller of Exim is root. Without
5690 it, the exim user may also use -C and -D and retain privilege.
5692 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5693 command line option is disabled.
5695 6. Macro names set by the -D option must start with an upper case letter, just
5696 like macro names defined in the configuration file.
5698 7. Added "dereference=" facility to LDAP.
5700 8. Two instances of the typo "uknown" in the source files are fixed.
5702 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5703 the Configure-Makefile script screwed up while processing it.
5705 10. Incorporated PCRE 4.4.
5707 11. The SMTP synchronization check was not operating right at the start of an
5708 SMTP session. For example, it could not catch a HELO sent before the client
5709 waited for the greeting. There is now a check for outstanding input at the
5710 point when the greeting is written. Because of the duplex, asynchronous
5711 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5712 way, but not yet received, when the check is performed.
5714 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5715 on TCP/IP sockets, because this apparently causes some broken clients to
5718 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5719 unchanged) from the Cygwin maintainer.
5721 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5722 was supported instead of testing for mbx. Effectively a typo.
5724 15. The spa authenticator server code was not checking that the input it
5725 received was valid base64.
5727 16. The debug output line for the "set" modifier in ACLs was not showing the
5728 name of the variable that was being set.
5730 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5732 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5733 lookup on its own. Something else was needed to trigger the lookup. For
5734 example, a match in host_lookup or the need for a host name in a host list.
5735 Now, if $sender_host_name is referenced and the host name has not yet been
5736 looked up, a lookup is performed. If the lookup fails, the variable remains
5737 empty, and $host_lookup_failed is set to "1".
5739 19. Added "eqi" as a case-independent comparison operator.
5741 20. The saslauthd authentication condition could segfault if neither service
5742 nor realm was specified.
5744 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5745 error message that was logged was misleading, and incoming SMTP
5746 connections were dropped. The message is now more accurate, and temporary
5747 errors are given to SMTP connections.
5749 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5750 (except RSET) with a 421 error, until QUIT is received. However, it was
5751 failing to send a response to QUIT.
5753 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5754 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5755 ACL is now run after the helo_try_verify_hosts code.
5757 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5758 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5759 case-independent, so other case variants are also recognized). Apparently
5760 some systems use these upper case variants.
5762 25. If more than two messages were waiting for the same host, and a transport
5763 filter was specified for the transport, Exim sent two messages over the
5764 same TCP/IP connection, and then failed with "socket operation on non-
5765 socket" when it tried to send the third.
5767 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5769 27. The extern definition of crypt16() in expand.c was not being excluded when
5770 the OS had its own crypt16() function.
5772 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5773 as a preferred synonym for return_size_limit, both as an option and as an
5776 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5778 30. Changed the default configuration ACL to relax the local part checking rule
5779 for addresses that are not in any local domains. For these addresses,
5780 slashes and pipe symbols are allowed within local parts, but the sequence
5781 /../ is explicitly forbidden.
5783 31. SPA server authentication was not clearing the challenge buffer before
5786 32. log_message in a "warn" ACL statement was writing to the reject log as
5787 well as to the main log, which contradicts the documentation and doesn't
5788 seem right (because no rejection is happening). So I have stopped it.
5790 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5791 However, I am unable to do any testing of this.
5793 34. Fixed an infelicity in the appendfile transport. When checking directories
5794 for a mailbox, to see if any needed to be created, it was accidentally
5795 using path names with one or more superfluous leading slashes; tracing
5796 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5798 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5799 contents are added to the log line that is written for every discarded
5800 recipient. (Previously a log_message setting was ignored.)
5802 36. The ${quote: operator now quotes the string if it is empty.
5804 37. The install script runs exim in order to find its version number. If for
5805 some reason other than non-existence or emptiness, which it checks, it
5806 could not run './exim', it was installing it with an empty version number,
5807 i.e. as "exim-". This error state is now caught, and the installation is
5810 38. An argument was missing from the function that creates an error message
5811 when Exim fails to connect to the socket for saslauthd authentication.
5812 This could cause Exim to crash, or give a corrupted message.
5814 39. Added isip, isip4, and isip6 to ${if conditions.
5816 40. The ACL variables $acl_xx are now saved with the message, and can be
5817 accessed later in routers, transports, and filters.
5819 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5820 strings in the file are not string-expanded.
5822 42. If a MAIL command specified a SIZE value that was too large to fit into an
5823 int variable, the check against message_size_limit failed. Such values are
5824 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5825 day this will have to be increased, but I don't think I want to be around
5826 when emails are that large.
5833 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5834 iconv() is not standard in FreeBSD.
5836 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5837 IPv6 enabled. The observed symptom was a segmentation fault on return from
5838 the function os_common_find_running_interfaces() in src/os.c.
5840 3. In the check_special_case() function in daemon.c I had used "errno" as an
5841 argument name, which causes warnings on some systems. This was basically a
5842 typo, since it was named "eno" in the comments!
5844 4. The code that waits for the clock to tick (at a resolution of some fraction
5845 of a second) so as to ensure message-id uniqueness was always waiting for
5846 at least one whole tick, when it could have waited for less. [This is
5847 almost certainly not relevant at current processor speeds, where it is
5848 unlikely to ever wait at all. But we try to future-proof.]
5850 5. The function that sleeps for a time interval that includes fractions of a
5851 second contained a race. It did not block SIGALRM between setting the
5852 timer, and suspending (a couple of lines later). If the interval was short
5853 and the sigsuspend() was delayed until after it had expired, the suspension
5854 never ended. On busy systems this could lead to processes getting stuck for
5857 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5858 process, before it forks any delivery processes. The open lookup caching
5859 mechanism meant that the open file or database connection was passed into
5860 the delivery process. The problem was that delivery processes always tidy
5861 up cached lookup data. This could cause a problem for the next delivery
5862 process started by the queue runner, because the external queue runner
5863 process does not know about the closure. So the next delivery process
5864 still has data in the lookup cache. In the case of a file lookup, there was
5865 no problem because closing a file descriptor in a subprocess doesn't affect
5866 the parent. However, if the lookup was caching a connection to a database,
5867 the connection was closed, and the second delivery process was likely to
5868 see errors such as "PGSQL: query failed: server closed the connection
5869 unexpectedly". The problem has been fixed by closing all cached lookups
5870 in a queue runner before running a delivery process.
5872 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5873 seem to have the "const" qualifier which it has on other OS. I've
5876 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5877 *of the same type* (client or server) with the same public name that an
5878 error should be diagnosed.
5880 9. When Exim looked up a host name for an IP address, but failed to find the
5881 original IP address when looking up the host name (a safety check), it
5882 output the message "<ip address> does not match any IP for NULL", which was
5883 confusing, to say the least. The bug was that the host name should have
5884 appeared instead of "NULL".
5886 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5887 user that is built into the binary, and the -C or -D options is used, root
5888 privilege is dropped before the configuration file is read. In addition,
5889 logging is switched to stderr instead of the normal log files. If the
5890 configuration then re-defines the Exim user, the unprivileged environment
5891 is probably not what is expected, so Exim logs a panic warning message (but
5894 However, if deliver_drop_privilege is set, the unprivileged state may well
5895 be exactly what is intended, so the warning has been cut out in that case,
5896 and Exim is allowed to try to write to its normal log files.
5902 1. smtp_return_error_details was not giving details for temporary sender
5903 or receiver verification errors.
5905 2. Diagnose a configuration error if two authenticators have the same public
5908 3. Exim used not to create the message log file for a message until the first
5909 delivery attempt. This could be confusing when incoming messages were held
5910 for policy or load reasons. The message log file is now created at the time
5911 the message is received, and an initial "Received" line is written to it.
5913 4. The automatically generated man page for command line options had a minor
5914 bug that caused no ill effects; however, a more serious problem was that
5915 the procedure for building the man page automatically didn't always
5916 operate. Consequently, release 4.20 contains an out-of-date version. This
5917 shouldn't happen again.
5919 5. When building Exim with embedded Perl support, the script that builds the
5920 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5921 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5923 6. The freeze_tell option was not being used for messages that were frozen on
5924 arrival, either by an ACL or by local_scan().
5926 7. Added the smtp_incomplete_transaction log selector.
5928 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5929 was accepting AUTH without a new EHLO.
5931 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5932 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5935 10. Logging of TCP/IP connections (when configured) now happens in the main
5936 daemon process instead of the child process, so that the TCP/IP connection
5937 count is more accurate (but it can never be perfect).
5939 11. The use of "drop" in a nested ACL was not being handled correctly in the
5940 outer ACL. Now, if condition failure induced by the nested "drop" causes
5941 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5942 or "require"), the connection is dropped.
5944 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5945 that yield "discard" can now be used with an "accept" or a "discard" verb,
5946 but an error is generated for any others (because I can't see a useful way
5947 to define what should happen).
5949 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5950 are now processed in the same way as lines in the Exim configuration file.
5951 In particular, continuation lines are supported.
5953 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5955 15. Added -ti meaning -t -i.
5957 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5958 domains, and warn by logging if others are found.
5960 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5961 returned from GIFCONF when Exim is trying to find the list of interfaces on
5962 a host. The code in os.c has been modified to copy each ifreq to an aligned
5963 structure in all cases.
5965 Also, in some cases, the returned ifreq's were being copied to a 'struct
5966 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5967 means the last couple of bytes of an IPv6 address could be chopped if the
5968 ifreq contained only a normal sockaddr (14 bytes storage).
5970 18. Named domain lists were not supported in the hosts_treat_as_local option.
5971 An entry such as +xxxx was not recognized, and was treated as a literal
5974 19. Ensure that header lines added by a DATA ACL are included in the reject log
5975 if the ACL subsequently rejects the message.
5977 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5978 MD5 (which is deprecated).
5980 21. When testing a filter file using -bf, Exim was writing a message when it
5981 took the sender from a "From " line in the message, but it was not doing so
5982 when it took $return_path from a Return-Path: header line. It now does.
5984 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5985 with a valid header line field name (a series of printing characters
5986 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5988 23. Changed "disc" in the source to "disk" to conform to the documentation and
5989 the book and for uniformity.
5991 24. Ignore Sendmail's -Ooption=value command line item.
5993 25. When execve() failed while trying to run a command in a pipe transport,
5994 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5995 could be confused with a return value of 69 from the command itself. This
5996 has been changed to 127, the value the shell returns if it is asked to run
5997 a non-existent command. The wording for the related log line suggests a
5998 non-existent command as the problem.
6000 26. If received_header_text expands to an empty string, do not add a Received:
6001 header line to the message. (Well, it adds a token one on the spool, but
6002 marks it "old" so that it doesn't get used or transmitted.)
6004 27. Installed eximstats 1.28 (addition of -nt option).
6006 28. There was no check for failure on the call to getsockname() in the daemon
6007 code. This can fail if there is a shortage of resources on the system, with
6008 ENOMEM, for example. A temporary error is now given on failure.
6010 29. Contrary to the C standard, it seems that in some environments, the
6011 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6012 program. Exim now does this explicitly; it affects the formatting of
6013 timestamps using strftime().
6015 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6016 complaints. I've now initialized all the variables, to avoid this.
6018 32. Header lines added by a system filter were not being "seen" during
6019 transport-time rewrites.
6021 33. The info_callback() function passed to OpenSSL is set up with type void
6022 (*)(SSL *, int, int), as described somewhere. However, when calling the
6023 function (actually a macro) that sets it up, the type void(*)() is
6024 expected. I've put in a cast to prevent warnings from picky compilers.
6026 34. If a DNS black list lookup found a CNAME record, but there were no A
6027 records associated with the domain it pointed at, Exim crashed.
6029 35. If a DNS black list lookup returned more than one A record, Exim ignored
6030 all but the first. It now scans all returned addresses if a particular IP
6031 value is being sought. In this situation, the contents of the
6032 $dnslist_value variable are a list of all the addresses, separated by a
6035 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6036 to accept a host name and all its aliases if the forward lookup for any of
6037 them yielded the IP address of the incoming connection. Now it accepts only
6038 those names whose forward lookup yields the correct IP address. Any other
6039 names are discarded. This closes a loophole whereby a rogue DNS
6040 administrator could create reverse DNS records to break through a
6041 wildcarded host restriction in an ACL.
6043 37. If a user filter or a system filter that ran in a subprocess used any of
6044 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6045 the wrong values were passed to the pipe command ($thisaddress had the
6046 value of $0, $0 had the value of $1, etc). This bug was introduced by
6047 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6049 38. Improved the line breaking for long SMTP error messages from ACLs.
6050 Previously, if there was no break point between 40 and 75 characters, Exim
6051 left the rest of the message alone. Two changes have been made: (a) I've
6052 reduced the minimum length to 35 characters; (b) if it can't find a break
6053 point between 35 and 75 characters, it looks ahead and uses the first one
6054 that it finds. This may give the occasional overlong line, but at least the
6055 remaining text gets split now.
6057 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6058 file descriptors might be low, and that setting 1000 would always raise it.
6059 It turns out that in some environments, the limit is already over 1000 and
6060 that lowering it causes trouble. So now Exim takes care not to decrease it.
6062 40. When delivering a message, the value of $return_path is set to $sender_
6063 address at the start of routing (routers may change the value). By an
6064 oversight, this default was not being set up when an address was tested by
6065 -bt or -bv, which affected the outcome if any router or filter referred to
6068 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6069 log only when "message" or "log_message" are set. However, if one of the
6070 conditions was an address verification, or a call to a nested ACL, the
6071 messages generated by the underlying test were being passed through. This
6072 no longer happens. The underlying message is available in $acl_verify_
6073 message for both "message" and "log_message" expansions, so it can be
6074 passed through if needed.
6076 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6077 new expansion $bh_ to give the encoded byte string without charset
6078 translation. Translation happens only if iconv() is available; HAVE_ICONV
6079 indicates this at build time. HEADERS_CHARSET gives the charset to
6080 translate to; headers_charset can change it in the configuration, and
6081 "headers charset" can change it in an individual filter file.
6083 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6084 that creates RFC 2047 encoded "words" labels them as that charset instead
6085 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6086 expansion operator; (ii) when Exim creates a From: line for a local
6087 message; (iii) when a header line is rewritten to include a "phrase" part.
6089 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6090 buggy, causing it to skip the first lines of messages whose message ID
6091 ended in 'D'. This would not have bitten before Exim release 4.14, because
6092 message IDs were unlikely to end in 'D' before then. The effect was to have
6093 incorrect size information for certain domains.
6095 45. #include "config.h" was missing at the start of the crypt16.c module. This
6096 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6099 46. If there was a timeout during a "random" callout check, Exim treated it as
6100 a failure of the random address, and carried on sending RSET and the real
6101 address. If the delay was just some slowness somewhere, the response to the
6102 original RCPT would be taken as a response to RSET and so on, causing
6103 mayhem of various kinds.
6105 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6106 when I implemented it. It didn't allow for the fact that some option values
6107 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6108 the right test for positive values.
6110 48. Domain names in DNS records are case-independent. Exim always looks them up
6111 in lower case. Some resolvers return domain names in exactly the case they
6112 appear in the zone file, that is, they may contain uppercase letters. Not
6113 all resolvers do this - some return always lower case. Exim was treating a
6114 change of case by a resolver as a change of domain, similar to a widening
6115 of a domain abbreviation. This triggered its re-routing code and so it was
6116 trying to route what was effectively the same domain again. This normally
6117 caused routing to fail (because the router wouldn't handle the domain
6118 twice). Now Exim checks for this case specially, and just changes the
6119 casing of the domain that it ultimately uses when it transmits the message
6122 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6125 50. If a filter generated a file delivery with a non-absolute name (possible if
6126 no home directory exists for the router), the forbid_file option was not
6129 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6130 the existing equality matching.
6132 52. Exim was using ints instead of ino_t variables in some places where it was
6133 dealing with inode numbers.
6135 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6136 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6137 called TMPDIR, and if it finds it is different, it changes its value.
6139 54. The smtp_printf() function is now made available to local_scan() so
6140 additional output lines can be written before returning. There is also an
6141 smtp_fflush() function to enable the detection of a dropped connection.
6142 The variables smtp_input and smtp_batched_input are exported to
6145 55. Changed the default runtime configuration: the message "Unknown user"
6146 has been removed from the ACL, and instead placed on the localuser router,
6147 using the cannot_route_message feature. This means that any verification
6148 failures that generate their own messages won't get overridden. Similarly,
6149 the "Unrouteable address" message that was in the ACL for unverifiable
6150 relay addresses has also been removed.
6152 56. Added hosts_avoid_esmtp to the smtp transport.
6154 57. The exicyclog script was not checking for the esoteric option
6155 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6156 will work only if exicyclog is run under the appropriate euid.
6158 58. Following a discussion on the list, the rules by which Exim recognises line
6159 endings on incoming messages have been changed. The -dropcr and drop_cr
6160 options are now no-ops, retained only for backwards compatibility. The
6161 following line terminators are recognized: LF CRLF CR. However, special
6162 processing applies to CR:
6164 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6165 nor a local message in the state where . is a terminator.
6167 (ii) If a bare CR is encountered in a header line, an extra space is added
6168 after the line terminator so as not to end the header. The reasoning
6169 behind this is that bare CRs in header lines are most likely either
6170 to be mistakes, or people trying to play silly games.
6172 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6173 was being incorrectly given as 18 bytes larger than it should have been.
6174 This is a VOB (very old bug).
6176 60. This may never have affected anything current, but just in case it has:
6177 When the local host is found other than at the start of a list of hosts,
6178 the local host, those with the same MX, and any that follow, are discarded.
6179 When the list in question was part of a longer list of hosts, the following
6180 hosts (not currently being processed) were also being discarded. This no
6181 longer happens. I'm not sure if this situation could ever has previously
6184 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6187 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6188 header. This is something that is recommended in a new Internet Draft, and
6189 is something that is documented as being done by Sendmail. There are two
6190 possible values. For messages generated by the autoreply transport, Exim
6193 Auto-Submitted: auto-replied
6195 whereas for all other generated messages (e.g. bounces) it adds
6197 Auto-Submitted: auto-generated
6199 63. The "personal" condition in filters now includes a test for the
6200 Auto-Submitted: header. If it contains the string "auto-" the message it
6201 not considered personal.
6203 64. Added rcpt_include_affixes as a generic transport option.
6205 65. Added queue_only_override (default true).
6207 66. Added the syslog_duplication option.
6209 67. If what should have been the first header line of a message consisted of
6210 a space followed by a colon, Exim was mis-interpreting it as a header line.
6211 It isn't of course - it is syntactically invalid and should therefore be
6212 treated as the start of the message body. The misbehaviour could have
6213 caused a number of strange effects, including loss of data in subsequent
6214 header lines, and spool format errors.
6216 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6217 client host had authenticated. This control can now be exercised by an ACL
6218 for more flexibility.
6220 69. By default, callouts do not happen when testing with -bh. There is now a
6221 variant, -bhc, which does actually run the callout code, including
6222 consulting and updating the callout cache.
6224 70. Added support for saslauthd authentication, courtesy of Alexander
6227 71. If statvfs() failed on the spool or log directories while checking their
6228 size for availability, Exim confusingly gave the error "space shortage".
6229 Furthermore, in debugging mode it crashed with a floating point exception.
6230 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6231 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6232 serious problem, Exim now writes to the main and panic logs when this
6233 happens, with details of the failure. It then refuses to accept the
6234 incoming message, giving the message "spool directory problem" or "log
6235 directory problem" with a 421 code for SMTP messages.
6237 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6238 0, 1, and 2 exist, because some OS complain for execs without them (see
6239 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6240 descriptors. However, the code omitted to check that the open succeeded,
6241 causing mysterious errors if for some reason the permissions on /dev/null
6242 got screwed. Now Exim writes a message to the main and panic logs, and
6243 bombs out if it can't open /dev/null.
6245 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6246 interact so that it is all more flexible. It is supposed to remain
6247 backwards compatible. Also added extra_local_interfaces.
6249 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6250 to bomb out with an assertion failure - to the client this appears as a
6251 connection drop. This problem occurs in the part of the code that was taken
6252 from the Samba project. Fortunately, the assertion is in a very simple
6253 function, so I have fixed this by reproducing the function inline in the
6254 one place where it is called, and arranging for authentication to fail
6255 instead of killing the process with assert().
6257 75. The SPA client code was not working when the server requested OEM rather
6258 than Unicode encoding.
6260 76. Added code to make require_files with a specific uid setting more usable in
6261 the case where statting the file as root fails - usually a non-root-mounted
6262 NFS file system. When this happens and the failure is EACCES, Exim now
6263 forks a subprocess and does the per-uid checking as the relevant uid.
6265 77. Added process_log_path.
6267 78. If log_file_path was not explicitly set, a setting of check_log_space or
6268 check_log_inodes was ignored.
6270 79. If a space check for the spool or log partitions fails, the incident is now
6271 logged. Of course, in the latter case the data may get lost...
6273 80. Added the %p formatting code to string_format() so that it can be used to
6274 print addresses in debug_print(). Adjusted all the address printing in the
6275 debugging in store.c to use %p rather than %d.
6277 81. There was a concern that a line of code in smtp_in.c could overflow a
6278 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6279 initially expressed, the concern was not well-founded, because trailing
6280 spaces are removed early. However, if the trailing spaces were followed by
6281 a NULL, they did not get removed, so the overflow was possible. Two fixes
6284 (a) I re-wrote the offending code in a cleaner fashion.
6285 (b) If an incoming SMTP command contains a NULL character, it is rejected
6288 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6289 runs initgroups(), so that if the Exim user is in any additional groups,
6290 they will be used during message reception.
6296 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob