1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
8 and 3.4.4 - once the server is enabled to respond to an OCSP request
9 it does even when not requested, resulting in a stapling non-aware
10 client dropping the TLS connection.
12 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
13 support variable-length bit vectors. No functional change.
15 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
16 The I= interface field on outgoing lines is now after the H= remote
17 host field, same as incoming lines. There is a separate
18 outgoing_interface log selector which allows you to disable the
21 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
22 If not running log_selector +smtp_connection the mainlog would be held
23 open indefinitely after a "too many connections" event, including to a
24 deleted file after a log rotate. Leave the per net connection logging
25 leaving it open for efficiency as that will be quickly detected by the
26 check on the next write.
28 HS/01 Bug 1671: Fix post transport crash.
29 Processing the wait-<transport> messages could crash the delivery
30 process if the message IDs didn't exist for some reason. When
31 using 'split_spool_directory=yes' the construction of the spool
32 file name failed already, exposing the same netto behaviour.
34 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
35 mime_regex ACL conditions.
40 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
43 JH/02 The smtp transport option "multi_domain" is now expanded.
45 JH/03 The smtp transport now requests PRDR by default, if the server offers
48 JH/04 Certificate name checking on server certificates, when exim is a client,
49 is now done by default. The transport option tls_verify_cert_hostnames
50 can be used to disable this per-host. The build option
51 EXPERIMENTAL_CERTNAMES is withdrawn.
53 JH/05 The value of the tls_verify_certificates smtp transport and main options
54 default to the word "system" to access the system default CA bundle.
55 For GnuTLS, only version 3.0.20 or later.
57 JH/06 Verification of the server certificate for a TLS connection is now tried
58 (but not required) by default. The verification status is now logged by
59 default, for both outbound TLS and client-certificate supplying inbound
62 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
65 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
66 Status Notification (bounce) messages are now MIME format per RFC 3464.
67 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
68 under the control of the dsn_advertise_hosts option, and routers may
69 have a dsn_lasthop option.
71 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
72 default, modifiable by a malware= option. The list separator for
73 the options can now be changed in the usual way. Bug 68.
75 JH/10 The smtp_receive_timeout main option is now expanded before use.
77 JH/11 The incoming_interface log option now also enables logging of the
78 local interface on delivery outgoing connections.
80 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
81 if the interface and destination host and port all match.
83 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
86 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
87 Patch from Andrew Lewis.
89 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
90 now supports optional time-restrictions, weighting, and priority
91 modifiers per server. Patch originally by <rommer@active.by>.
93 JH/16 The spamd_address main option now supports a mixed list of local
94 and remote servers. Remote servers can be IPv6 addresses, and
97 JH/17 Bug 68: The spamd_address main option now supports an optional
98 timeout value per server.
100 JH/18 Bug 1581: Router and transport options headers_add/remove can
101 now have the list separator specified.
103 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
106 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
109 JH/21 Support for the A6 type of dns record is withdrawn.
111 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
112 rather than the verbs used.
114 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
115 from 255 to 1024 chars.
117 JH/24 Verification callouts now attempt to use TLS by default.
119 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
120 are generic router options now. The defaults didn't change.
122 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
123 Original patch from Alexander Shikoff, worked over by JH.
125 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
126 files if ZCAT_COMMAND is not executable.
128 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
130 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
132 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
133 Normally benign, it bites when the pair was led to by a CNAME;
134 modern usage is to not canoicalize the domain to a CNAME target
135 (and we were inconsistent anyway for A-only vs AAAA+A).
137 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
139 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
140 when evaluating $sender_host_dnssec.
142 JH/31 Check the HELO verification lookup for DNSSEC, adding new
143 $sender_helo_dnssec variable.
145 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
147 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
149 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
151 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
152 documented as working, but never had. Support all but $spam_report.
154 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
155 added for tls authenticator.
160 TL/01 When running the test suite, the README says that variables such as
161 no_msglog_check are global and can be placed anywhere in a specific
162 test's script, however it was observed that placement needed to be near
163 the beginning for it to behave that way. Changed the runtest perl
164 script to read through the entire script once to detect and set these
165 variables, reset to the beginning of the script, and then run through
166 the script parsing/test process like normal.
168 TL/02 The BSD's have an arc4random API. One of the functions to induce
169 adding randomness was arc4random_stir(), but it has been removed in
170 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
171 function when detected.
173 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
174 cause callback expansion.
176 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
177 syntax errors in an expansion can be treated as a string instead of
178 logging or causing an error, due to the internal use of bool_lax
179 instead of bool when processing it.
181 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
182 server certificates when making smtp deliveries.
184 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
186 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
188 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
190 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
191 Merged patch from Sebastian Wiedenroth.
193 JH/05 Fix results-pipe from transport process. Several recipients, combined
194 with certificate use, exposed issues where response data items split
195 over buffer boundaries were not parsed properly. This eventually
196 resulted in duplicates being sent. This issue only became common enough
197 to notice due to the introduction of conection certificate information,
198 the item size being so much larger. Found and fixed by Wolfgang Breyha.
200 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
201 size buffer was used, resulting in syntax errors when an expansion
204 JH/07 Add support for directories of certificates when compiled with a GnuTLS
205 version 3.3.6 or later.
207 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
208 is EXPERIMENTAL_EVENT, the main-configuration and transport options
209 both become "event_action", the variables become $event_name, $event_data
210 and $event_defer_errno. There is a new variable $verify_mode, usable in
211 routers, transports and related events. The tls:cert event is now also
212 raised for inbound connections, if the main configuration event_action
215 TL/06 In test suite, disable OCSP for old versions of openssl which contained
216 early OCSP support, but no stapling (appears to be less than 1.0.0).
218 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
219 server certificate names available under the smtp transport option
220 "tls_verify_cert_hostname" now do not permit multi-component wildcard
223 JH/10 Time-related extraction expansions from certificates now use the main
224 option "timezone" setting for output formatting, and are consistent
225 between OpenSSL and GnuTLS compilations. Bug 1541.
227 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
228 encoded parameter in the incoming message. Bug 1558.
230 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
231 include certificate info, eximon was claiming there were spoolfile
234 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
236 JH/14 Log delivery-related information more consistently, using the sequence
237 "H=<name> [<ip>]" wherever possible.
239 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
240 are problematic for Debian distribution, omit them from the release
243 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
245 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
248 JH/17 Fix a null-indirection in certextract expansions when a nondefault
249 output list separator was used.
254 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
255 checkers that were complaining about end of non-void function with no
258 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
259 This was a regression intruduced in 4.83 by another bugfix.
261 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
263 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
264 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
270 TF/01 Correctly close the server side of TLS when forking for delivery.
272 When a message was received over SMTP with TLS, Exim failed to clear up
273 the incoming connection properly after forking off the child process to
274 deliver the message. In some situations the subsequent outgoing
275 delivery connection happened to have the same fd number as the incoming
276 connection previously had. Exim would try to use TLS and fail, logging
277 a "Bad file descriptor" error.
279 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
280 utilities have not been installed.
282 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
283 temporary space as the ACL may create new global variables.
285 TL/01 LDAP support uses per connection or global context settings, depending
286 upon the detected version of the libraries at build time.
288 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
289 to extract and use the src ip:port in logging and expansions as if it
290 were a direct connection from the outside internet. PPv2 support was
291 updated based on HAProxy spec change in May 2014.
293 JH/02 Add ${listextract {number}{list}{success}{fail}}.
295 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
296 Properly escape header and check for NULL return.
298 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
301 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
303 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
304 characters in header names, implemented as a verify condition.
305 Contributed by Michael Fischer v. Mollard.
307 TL/05 Rename SPF condition results err_perm and err_temp to standardized
308 results permerror and temperror. Previous values are deprecated but
309 still accepted. In a future release, err_perm and err_temp will be
310 completely removed, which will be a backward incompatibility if the
311 ACL tests for either of these two old results. Patch contributed by
312 user bes-internal on the mailing list.
314 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
316 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
317 selectors, in both main and reject logs.
319 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
322 JH/07 Add malware type "sock" for talking to simple daemon.
324 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
326 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
327 routers/transports under cutthrough routing.
329 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
330 numbers. Touch up "bool" conditional to keep the same definition.
332 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
334 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
336 JH/12 Expand items in router/transport headers_add or headers_remove lists
337 individually rather than the list as a whole. Bug 1452.
339 Required for reasonable handling of multiple headers_ options when
340 they may be empty; requires that headers_remove items with embedded
341 colons must have them doubled (or the list-separator changed).
343 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
344 view the policy declared in the DMARC record. Currently, $dmarc_status
345 is a combined value of both the record presence and the result of the
348 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
350 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
351 dnslookup router and the smtp transport (applying to the forward
354 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
355 of ldap servers used for a specific lookup. Patch provided by Heiko
358 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
359 New variable $lookup_dnssec_authenticated for observability.
361 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
362 Patch submitted by Lars Timman.
364 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
366 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
367 Requires trusted mode and valid format message id, aborts otherwise.
368 Patch contributed by Heiko Schlichting.
370 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
371 certextract with support for various fields. Bug 1358.
373 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
374 is requested by default, modifiable by smtp transport option
377 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
378 operate on certificate variables to give certificate fingerprints
379 Also new ${sha256:cert_variable}.
381 JH/23 The PRDR feature is moved from being Experimental into the mainline.
383 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
384 Christian Aistleitner.
386 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
388 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
389 file. Patch from Wolfgang Breyha.
391 JH/25 Expand the coverage of the delivery $host and $host_address to
392 client authenticators run in verify callout. Bug 1476.
394 JH/26 Port service names are now accepted for tls_on_connect_ports, to
395 align with daemon_smtp_ports. Bug 72.
397 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
398 support and error reporting did not work properly.
400 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
401 and is readable. Patch from Andrew Colin Kissa.
403 TL/14 Enhance documentation of ${run expansion and how it parses the
404 commandline after expansion, particularly in the case when an
405 unquoted variable expansion results in an empty value.
407 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
409 PP/02 Fix internal collision of T_APL on systems which support RFC3123
410 by renaming away from it. Addresses GH issue 15, reported by
413 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
415 TL/15 SECURITY: prevent double expansion in math comparison functions
416 (can expand unsanitized data). Not remotely exploitable.
423 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
425 PP/02 Make -n do something, by making it not do something.
426 When combined with -bP, the name of an option is not output.
428 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
431 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
432 $sender_host_name and config options to manage this, and basic check
435 PP/05 DSCP support for outbound connections and control modifier for inbound.
437 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
438 (Only plugin which currently uses this is kerberos4, which nobody should
439 be using, but we should make it available and other future plugins might
440 conceivably use it, even though it would break NAT; stuff *should* be
441 using channel bindings instead).
443 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
444 name; added for Sendmail compatibility; requires admin caller.
445 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
446 just ignore it); requires trusted caller.
447 Also parse but ignore: -Ac -Am -X<logfile>
450 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
452 TL/02 Add +smtp_confirmation as a default logging option.
454 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
455 Patch by Magnus Holmgren from 2007-02-20.
457 TL/04 Bugzilla 1281 - Spec typo.
458 Bugzilla 1283 - Spec typo.
459 Bugzilla 1290 - Spec grammar fixes.
461 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
463 TL/06 Add Experimental DMARC support using libopendmarc libraries.
465 TL/07 Fix an out of order global option causing a segfault. Reported to dev
466 mailing list by by Dmitry Isaikin.
468 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
470 JH/02 Support "G" suffix to numbers in ${if comparisons.
472 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
474 NM/01 Bugzilla 1197 - Spec typo
475 Bugzilla 1196 - Spec examples corrections
477 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
479 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
480 gnutls_enable_pkcs11, but renamed to more accurately indicate its
483 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
484 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
486 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
487 "acl {{name}{arg}...}", and optional args on acl condition
490 JH/05 Permit multiple router/transport headers_add/remove lines.
492 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
494 JH/07 Avoid using a waiting database for a single-message-only transport.
495 Performance patch from Paul Fisher. Bugzilla 1262.
497 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
500 JH/09 Add $headers_added variable, with content from use of ACL modifier
501 add_header (but not yet added to the message). Bugzilla 199.
503 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
504 Pulled from Bugzilla 817 by Wolfgang Breyha.
506 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
508 (nb: this is the same fix as in Exim 4.80.1)
510 JH/11 Add A= logging on delivery lines, and a client_set_id option on
513 JH/12 Add optional authenticated_sender logging to A= and a log_selector
516 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
518 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
519 advertise SMTP AUTH mechanism to us, instead of a generic
520 protocol violation error. Also, make Exim more robust to bad
521 data from the Dovecot auth socket.
523 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
525 When a queue runner is handling a message, Exim first routes the
526 recipient addresses, during which it prunes them based on the retry
527 hints database. After that it attempts to deliver the message to
528 any remaining recipients. It then updates the hints database using
531 So if a recipient address works intermittently, it can get repeatedly
532 deferred at routing time. The retry hints record remains fresh so the
533 address never reaches the final cutoff time.
535 This is a fairly common occurrence when a user is bumping up against
536 their storage quota. Exim had some logic in its local delivery code
537 to deal with this. However it did not apply to per-recipient defers
538 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
540 This change adds a proper retry rule check during routing so that the
541 final cutoff time is checked against the message's age. We only do
542 this check if there is an address retry record and there is not a
543 domain retry record; this implies that previous attempts to handle
544 the address had the retry_use_local_parts option turned on. We use
545 this as an approximation for the destination being like a local
546 delivery, as in LMTP.
548 I suspect this new check makes the old local delivery cutoff check
549 redundant, but I have not verified this so I left the code in place.
551 TF/02 Correct gecos expansion when From: is a prefix of the username.
553 Test 0254 submits a message to Exim with the header
557 When I ran the test suite under the user fanf2, Exim expanded
558 the header to contain my full name, whereas it should have added
559 a Resent-Sender: header. It erroneously treats any prefix of the
560 username as equal to the username.
562 This change corrects that bug.
564 GF/01 DCC debug and logging tidyup
565 Error conditions log to paniclog rather than rejectlog.
566 Debug lines prefixed by "DCC: " to remove any ambiguity.
568 TF/03 Avoid unnecessary rebuilds of lookup-related code.
570 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
571 Bug spotted by Jeremy Harris; was flawed since initial commit.
572 Would have resulted in OCSP responses post-SNI triggering an Exim
573 NULL dereference and crash.
575 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
577 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
578 Bug detection, analysis and fix by Samuel Thibault.
579 Bugzilla 1331, Debian bug #698092.
581 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
583 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
584 Server implementation by Todd Lyons, client by JH.
585 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
586 config variable "prdr_enable" controls whether the server
587 advertises the facility. If the client requests PRDR a new
588 acl_data_smtp_prdr ACL is called once for each recipient, after
589 the body content is received and before the acl_smtp_data ACL.
590 The client is controlled by bolth of: a hosts_try_prdr option
591 on the smtp transport, and the server advertisement.
592 Default client logging of deliveries and rejections involving
593 PRDR are flagged with the string "PRDR".
595 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
596 fclose(). Diagnosis by Todd Lyons.
598 PP/17 Update configure.default to handle IPv6 localhost better.
599 Patch by Alain Williams (plus minor tweaks).
602 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
603 This is now consistent with GnuTLS, and is now documented: the
604 previous undocumented portable approach to treating the option as
605 unset was to force an expansion failure. That still works, and
606 an empty string is now equivalent.
608 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
609 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
610 not performing validation itself.
612 PP/20 Added force_command boolean option to pipe transport.
613 Patch from Nick Koston, of cPanel Inc.
615 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
618 TF/04 Added udpsend ACL modifer and hexquote expansion operator
620 PP/21 Fix eximon continuous updating with timestamped log-files.
621 Broken in a format-string cleanup in 4.80, missed when I repaired the
622 other false fix of the same issue.
623 Report and fix from Heiko Schlichting.
626 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
627 Report from Prashanth Katuri.
629 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
630 It's SecureTransport, so affects any MacOS clients which use the
631 system-integrated TLS libraries, including email clients.
633 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
634 using a MIME ACL for non-SMTP local injection.
635 Report and assistance in diagnosis by Warren Baker.
637 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
639 JH/16 Fix comparisons for 64b. Bugzilla 1385.
641 TL/09 Add expansion variable $authenticated_fail_id to keep track of
642 last id that failed so it may be referenced in subsequent ACL's.
644 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
647 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
648 ldap library initialization, allowing self-signed CA's to be
649 used. Also properly sets require_cert option later in code by
650 using NULL (global ldap config) instead of ldap handle (per
651 session). Bug diagnosis and testing by alxgomz.
653 TL/12 Enhanced documentation in the ratelimit.pl script provided in
654 the src/util/ subdirectory.
656 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
657 renamed to Transport Post Delivery Action by Jeremy Harris, as
660 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
661 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
662 redis_servers = needs to be configured which will be used by the redis
663 lookup. Patch from Warren Baker, of The Packet Hub.
665 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
667 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
668 hostname or reverse DNS when processing a host list. Used suggestions
669 from multiple comments on this bug.
671 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
673 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
674 Missed a few lines, added it to make the runtest require no keyboard
677 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
678 contains upper case chars. Make router use caseful_local_part.
680 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
681 support when GnuTLS has been built with p11-kit.
687 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
689 This, or similar/improved, will also be change PP/11 of 4.82.
695 PP/01 Handle short writes when writing local log-files.
696 In practice, only affects FreeBSD (8 onwards).
697 Bugzilla 1053, with thanks to Dmitry Isaikin.
699 NM/01 Bugzilla 949 - Documentation tweak
701 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
704 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
706 PP/02 Implemented gsasl authenticator.
708 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
710 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
711 `pkg-config foo` for cflags/libs.
713 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
714 with rest of GSASL and with heimdal_gssapi.
716 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
717 `pkg-config foo` for cflags/libs for the TLS implementation.
719 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
720 properties get this fed in as external SSF. A number of robustness
721 and debugging improvements to the cyrus_sasl authenticator.
723 PP/08 cyrus_sasl server now expands the server_realm option.
725 PP/09 Bugzilla 1214 - Log authentication information in reject log.
726 Patch by Jeremy Harris.
728 PP/10 Added dbmjz lookup type.
730 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
732 PP/12 MAIL args handles TAB as well as SP, for better interop with
733 non-compliant senders.
734 Analysis and variant patch by Todd Lyons.
736 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
737 Bug report from Lars Müller <lars@samba.org> (via SUSE),
738 Patch from Dirk Mueller <dmueller@suse.com>
740 PP/13 tls_peerdn now print-escaped for spool files.
741 Observed some $tls_peerdn in wild which contained \n, which resulted
742 in spool file corruption.
744 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
745 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
746 or write after TLS renegotiation, which otherwise led to messages
749 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
750 as a tracking header (ie: a signed header comes before the signature).
751 Patch from Wolfgang Breyha.
753 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
754 comma-sep list; embedded commas doubled.
756 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
758 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
760 Report and patch from Dmitry Banschikov.
762 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
763 Removed SSL_clear() after SSL_new() which led to protocol negotiation
764 failures. We appear to now support TLS1.1+ with Exim.
766 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
767 lets Exim select keys and certificates based upon TLS SNI from client.
768 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
769 before an outbound SMTP session. New log_selector, +tls_sni.
771 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
772 NULL dereference. Report and patch from Alun Jones.
774 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
775 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
776 Not seeing resolver debug output on NetBSD, but suspect this is a
777 resolver implementation change.
779 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
780 Left warnings. Added "eximon gdb" invocation mode.
782 PP/21 Defaulting "accept_8bitmime" to true, not false.
784 PP/22 Added -bw for inetd wait mode support.
786 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
787 locate the relevant includes and libraries. Made this the default.
789 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
790 Bugzilla 1246, report and most of solution from Tomasz Kusy.
792 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
793 This may cause build issues on older platforms.
795 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
796 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
797 gnutls_require_mac & gnutls_require_protocols (no longer supported).
798 Added SNI support via GnuTLS too.
799 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
801 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
803 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
804 Applied second patch from Janne, implementing suggestion to default
805 multiple-strings-in-record handling to match SPF spec.
807 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
809 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
810 read-only, out of scope).
811 Patch from Wolfgang Breyha, report from Stuart Northfield.
813 PP/29 Fix three issues highlighted by clang analyser static analysis.
814 Only crash-plausible issue would require the Cambridge-specific
815 iplookup router and a misconfiguration.
816 Report from Marcin Mirosław.
818 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
820 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
821 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
822 As part of this, removing so much warning spew let me fix some minor
823 real issues in debug logging.
825 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
826 assignment on my part. Fixed.
828 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
829 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
830 Janne Snabb (who went above and beyond: thank you).
832 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
833 string otherwise requires a connection and a bunch more work and it's
834 relatively easy to get wrong. Should also expose TLS library linkage
837 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
838 64-bit ${eval} (JH/03).
840 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
841 GNU libc to support some of the 64-bit stuff, should not lead to
842 conflicts. Defined before os.h is pulled in, so if a given platform
843 needs to override this, it can.
845 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
846 protection layer was required, which is not implemented.
847 Bugzilla 1254, patch from Wolfgang Breyha.
849 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
850 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
851 tls_dhparam take prime identifiers. Also unbreak combination of
852 OpenSSL+DH_params+TLSSNI.
854 PP/39 Disable SSLv2 by default in OpenSSL support.
860 PP/01 Solaris build fix for Oracle's LDAP libraries.
861 Bugzilla 1109, patch from Stephen Usher.
863 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
865 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
868 TF/02 Fix a couple more cases where we did not log the error message
869 when unlink() failed. See also change 4.74-TF/03.
871 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
872 lock up or crash if it happened to be inside a call to libc when it
873 got a SIGUSR1 from exiwhat.
875 The SIGUSR1 handler appends the current process status to the process
876 log which is later printed by exiwhat. It used to use the general
877 purpose logging code to do this, but several functions it calls are
878 not safe for signals.
880 The new output code in the SIGUSR1 handler is specific to the process
881 log, and simple enough that it's easy to inspect for signal safety.
882 Removing some special cases also simplifies the general logging code.
883 Removing the spurious timestamps from the process log simplifies
886 TF/04 Improved ratelimit ACL condition.
888 The /noupdate option has been deprecated in favour of /readonly which
889 has clearer semantics. The /leaky, /strict, and /readonly update modes
890 are mutually exclusive. The update mode is no longer included in the
891 database key; it just determines when the database is updated. (This
892 means that when you upgrde Exim will forget old rate measurements.)
894 Exim now checks that the per_* options are used with an update mode that
895 makes sense for the current ACL. For example, when Exim is processing a
896 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
897 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
898 must specify per_mail/readonly. If you omit the update mode it defaults to
899 /leaky where that makes sense (as before) or /readonly where required.
901 The /noupdate option is now undocumented but still supported for
902 backwards compatibility. It is equivalent to /readonly except that in
903 ACLs where /readonly is required you may specify /leaky/noupdate or
904 /strict/noupdate which are treated the same as /readonly.
906 A useful new feature is the /count= option. This is a generalization
907 of the per_byte option, so that you can measure the throughput of other
908 aggregate values. For example, the per_byte option is now equivalent
909 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
911 The per_rcpt option has been generalized using the /count= mechanism
912 (though it's more complicated than the per_byte equivalence). When it is
913 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
914 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
915 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
916 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
917 Note that using per_rcpt with a non-readonly update mode in more than
918 one ACL will cause the recipients to be double-counted. (The per_mail
919 and per_byte options don't have this problem.)
921 The handling of very low rates has changed slightly. If the computed rate
922 is less than the event's count (usually one) then this event is the first
923 after a long gap. In this case the rate is set to the same as this event's
924 count, so that the first message of a spam run is counted properly.
926 The major new feature is a mechanism for counting the rate of unique
927 events. The new per_addr option counts the number of different
928 recipients that someone has sent messages to in the last time period. It
929 behaves like per_rcpt if all the recipient addresses are different, but
930 duplicate recipient addresses do not increase the measured rate. Like
931 the /count= option this is a general mechanism, so the per_addr option
932 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
933 example, measure the rate that a client uses different sender addresses
934 with the options per_mail/unique=$sender_address. There are further
935 details in the main documentation.
937 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
939 TF/06 Removed a few PCRE remnants.
941 TF/07 Automatically extract Exim's version number from tags in the git
942 repository when doing development or release builds.
944 PP/02 Raise smtp_cmd_buffer_size to 16kB.
945 Bugzilla 879. Patch from Paul Fisher.
947 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
948 Heavily based on revision 40f9a89a from Simon Arlott's tree.
951 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
953 PP/05 Variable $av_failed, true if the AV scanner deferred.
954 Bugzilla 1078. Patch from John Horne.
956 PP/06 Stop make process more reliably on build failure.
957 Bugzilla 1087. Patch from Heiko Schlittermann.
959 PP/07 Make maildir_use_size_file an _expandable_ boolean.
960 Bugzilla 1089. Patch from Heiko Schlittermann.
962 PP/08 Handle ${run} returning more data than OS pipe buffer size.
963 Bugzilla 1131. Patch from Holger Weiß.
965 PP/09 Handle IPv6 addresses with SPF.
966 Bugzilla 860. Patch from Wolfgang Breyha.
968 PP/10 GnuTLS: support TLS 1.2 & 1.1.
970 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
973 PP/11 match_* no longer expand right-hand-side by default.
974 New compile-time build option, EXPAND_LISTMATCH_RHS.
975 New expansion conditions, "inlist", "inlisti".
977 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
979 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
981 PP/14 fix log_write() format string regression from TF/03.
982 Bugzilla 1152. Patch from Dmitry Isaikin.
988 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
990 PP/02 Harmonised TLS library version reporting; only show if debugging.
991 Layout now matches that introduced for other libraries in 4.74 PP/03.
993 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
995 PP/04 New "dns_use_edns0" global option.
997 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1000 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1001 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1003 TK/01 Updated PolarSSL code to 0.14.2.
1004 Bugzilla 1097. Patch from Andreas Metzler.
1006 PP/07 Catch divide-by-zero in ${eval:...}.
1007 Fixes bugzilla 1102.
1009 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1012 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1013 format-string attack -- SECURITY: remote arbitrary code execution.
1015 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1016 time unintentionally subject to list matching rules, letting the header
1017 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1018 arbitrary string expansion). This allowed for information disclosure.
1020 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1021 INT_MIN/-1 -- value coerced to INT_MAX.
1027 NM/01 Workround for PCRE version dependency in version reporting
1030 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1031 This fixes portability to compilers other than gcc, notably
1032 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1034 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1035 makefiles for portability to HP-UX and POSIX correctness.
1037 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1038 Also via indented variable definition in the Makefile.
1039 (Debugging by Oliver Heesakkers).
1041 PP/02 Restore caching of spamd results with expanded spamd_address.
1042 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1044 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1045 Improves build reliability. Fix from: Frank Elsner
1047 NM/02 Fix wide character breakage in the rfc2047 coding
1048 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1050 NM/03 Allow underscore in dnslist lookups
1051 Fixes bug 1026. Patch from Graeme Fowler
1053 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1054 Code patches from Adam Ciarcinski of NetBSD.
1056 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1059 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1060 is logged, to avoid truncation. Patch from John Horne.
1062 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1063 Patch from Jakob Hirsch.
1065 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1066 SQL string expansion failure details.
1067 Patch from Andrey Oktyabrski.
1069 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1070 Patch from Simon Arlott.
1072 PP/09 New lookups functionality failed to compile on old gcc which rejects
1073 extern declarations in function scope.
1074 Patch from Oliver Fleischmann
1076 PP/10 Use sig_atomic_t for flags set from signal handlers.
1077 Check getgroups() return and improve debugging.
1078 Fixed developed for diagnosis in bug 927 (which turned out to be
1081 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1082 Patch from Mark Zealey.
1084 PP/12 Bugzilla 1056: Improved spamd server selection.
1085 Patch from Mark Zealey.
1087 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1088 Based on patch from Heiko Schlittermann.
1090 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1091 Patch from Uwe Doering, sign-off by Michael Haardt.
1093 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1094 variable declaration deep within a block. Bug and patch from
1097 PP/15 lookups-Makefile IRIX compatibilty coercion.
1099 PP/16 Make DISABLE_DKIM build knob functional.
1101 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1102 Patch by Simon Arlott
1104 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1105 variable argument macros. Our copy now differs from upstream.
1111 TF/01 Failure to get a lock on a hints database can have serious
1112 consequences so log it to the panic log.
1114 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1115 controlled using the smtp_confirmation log selector.
1117 TF/03 Include the error message when we fail to unlink a spool file.
1119 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1120 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1121 for maintaining out-of-tree patches for some time.
1123 PP/01 Bugzilla 139: Documentation and portability issues.
1124 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1125 Handle per-OS dynamic-module compilation flags.
1127 PP/02 Let /dev/null have normal permissions.
1128 The 4.73 fixes were a little too stringent and complained about the
1129 permissions on /dev/null. Exempt it from some checks.
1130 Reported by Andreas M. Kirchwitz.
1132 PP/03 Report version information for many libraries, including
1133 Exim version information for dynamically loaded libraries. Created
1134 version.h, now support a version extension string for distributors
1135 who patch heavily. Dynamic module ABI change.
1137 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1138 privilege escalation vulnerability whereby the Exim run-time user
1139 can cause root to append content of the attacker's choosing to
1142 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1145 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1146 If dropping privileges for untrusted macros, we disabled normal logging
1147 on the basis that it would fail; for the Exim run-time user, this is not
1148 the case, and it resulted in successful deliveries going unlogged.
1149 Fixed. Reported by Andreas Metzler.
1155 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1156 only prepend for the Resent-* case. Fixes regression introduced in
1157 Exim 4.70 by NM/22 for Bugzilla 607.
1159 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1160 increasing numbers of administrators be bitten by this.
1162 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1164 PP/03 Bugzilla 994: added openssl_options main configuration option.
1166 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1168 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1170 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1172 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1173 without a peer certificate, leading to a segfault because of an
1174 assumption that peers always have certificates. Be a little more
1175 paranoid. Problem reported by Martin Tscholak.
1177 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1178 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1179 NB: ClamAV planning to remove STREAM in "middle of 2010".
1180 CL also introduces -bmalware, various -d+acl logging additions and
1181 more caution in buffer sizes.
1183 PP/09 Implemented reverse_ip expansion operator.
1185 PP/10 Bugzilla 937: provide a "debug" ACL control.
1187 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1189 PP/12 Bugzilla 973: Implement --version.
1191 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1193 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1195 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1197 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1198 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1199 ignore trailing whitespace.
1201 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1203 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1206 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1207 Notification from Dr Andrew Aitchison.
1209 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1210 ExtendedDetectionInfo response format.
1211 Notification from John Horne.
1213 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1216 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1217 XSL and documented dependency on system catalogs, with examples of how
1220 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1223 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1224 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1225 configuration file which is writeable by the Exim user or group.
1227 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1228 of configuration files to cover files specified with the -C option if
1229 they are going to be used with root privileges, not just the default
1232 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1233 option (effectively making it always true).
1235 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1236 files to be used while preserving root privileges.
1238 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1239 that rogue child processes cannot use them.
1241 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1242 run-time user, instead of root.
1244 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1245 Exim run-time user without dropping privileges.
1247 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1248 result string, instead of calling string_vformat() twice with the same
1251 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1252 for other users. Others should always drop root privileges if they use
1253 -C on the command line, even for a whitelisted configure file.
1255 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1257 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1263 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1264 $data_path, and $header_path variables; fixed documentation bugs and
1267 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1268 exipick to access non-standard spools, including the "frozen" queue
1271 NM/01 Bugzilla 965: Support mysql stored procedures.
1272 Patch from Alain Williams
1274 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1276 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1277 Patch from Andreas Metzler
1279 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1280 Patch from Kirill Miazine
1282 NM/05 Bugzilla 671: Added umask to procmail example.
1284 JJ/03 installed exipick 20100323.0, fixing doc bug
1286 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1287 directory. Notification and patch from Dan Rosenberg.
1289 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1291 TK/02 Improve log output when DKIM signing operation fails.
1293 MH/01 Treat the transport option dkim_domain as a colon separated
1294 list, not as a single string, and sign the message with each element,
1295 omitting multiple occurences of the same signer.
1297 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1298 Bugzilla 985, 986. Patch by Simon Arlott
1300 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1301 Patch by Simon Arlott
1303 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1304 MBX locking. Notification from Dan Rosenberg.
1310 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1312 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1314 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1316 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1318 NM/04 Bugzilla 915: Flag broken perl installation during build.
1324 TK/01 Added patch by Johannes Berg that expands the main option
1325 "spamd_address" if it starts with a dollar sign.
1327 TK/02 Write list of recipients to X-Envelope-Sender header when building
1328 the mbox-format spool file for content scanning (suggested by Jakob
1331 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1332 (http://www.dcc-servers.net/) support via dccifd. Activated by
1333 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1335 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1336 by Mark Daniel Reidel <mr@df.eu>.
1338 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1339 When building exim an external PCRE library is now needed -
1340 PCRE is a system library on the majority of modern systems.
1341 See entry on PCRE_LIBS in EDITME file.
1343 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1344 conversation. Added nologin parameter to request.
1345 Patch contributed by Kirill Miazine.
1347 TF/01 Do not log submission mode rewrites if they do not change the address.
1349 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1351 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1352 log files in place. Contributed by Roberto Lima.
1354 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1356 TF/03 Bugzilla 615: When checking the local_parts router precondition
1357 after a local_part_suffix or local_part_prefix option, Exim now
1358 does not use the address's named list lookup cache, since this
1359 contains cached lookups for the whole local part.
1361 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1362 Robert Millan. Documentation is in experimental-spec.txt.
1364 TF/04 Bugzilla 668: Fix parallel build (make -j).
1366 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1368 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1369 Patch provided by Jan Srzednicki.
1371 TF/05 Leading white space used to be stripped from $spam_report which
1372 wrecked the formatting. Now it is preserved.
1374 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1375 that they are available at delivery time.
1377 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1379 TF/08 TLS error reporting now respects the incoming_interface and
1380 incoming_port log selectors.
1382 TF/09 Produce a more useful error message if an SMTP transport's hosts
1383 setting expands to an empty string.
1385 NM/06 Bugzilla 744: EXPN did not work under TLS.
1386 Patch provided by Phil Pennock.
1388 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1389 Patch provided by Richard Godbee.
1391 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1392 acl_smtp_notquit, added index entry.
1394 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1395 Patch provided by Eugene Bujak.
1397 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1398 accept(). Patch provided by Maxim Dounin.
1400 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1401 Patch provided by Phil Pennock.
1403 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1405 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1406 Patch provided by Brad "anomie" Jorsch.
1408 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1409 Patch provided by Dean Brooks.
1411 TK/05 Add native DKIM support (does not depend on external libraries).
1413 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1414 Patch provided by Graeme Fowler.
1416 NM/16 Bugzilla 851: Documentation example syntax fix.
1418 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1420 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1423 NM/19 Bugzilla 745: TLS version reporting.
1424 Patch provided by Phil Pennock.
1426 NM/20 Bugzilla 167: bool: condition support.
1427 Patch provided by Phil Pennock.
1429 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1430 clients. Patch provided by Phil Pennock.
1432 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1433 Patch provided by Brad "anomie" Jorsch.
1435 NM/23 Bugzilla 687: Fix misparses in eximstats.
1436 Patch provided by Heiko Schlittermann.
1438 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1439 Patch provided by Heiko Schlittermann.
1441 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1442 plus update to original patch.
1444 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1446 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1447 Patch provided by David Brownlee.
1449 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1451 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1453 NM/30 Bugzilla 888: TLS documentation bugfixes.
1455 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1457 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1458 Unlike the original bugzilla I have changed all shell scripts in src tree.
1460 NM/33 Bugzilla 898: Transport filter timeout fix.
1461 Patch by Todd Rinaldo.
1463 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1464 Patch by Serge Demonchaux.
1466 NM/35 Bugzilla 39: Base64 decode bug fixes.
1467 Patch by Jakob Hirsch.
1469 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1471 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1473 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1475 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1481 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1482 ALT-N's libdkim that I have put here:
1483 http://duncanthrax.net/exim-experimental/
1485 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1486 were called 'true' and it seems that C99 defines that as a reserved
1487 keyword to be used with 'bool' variable types. That means you could
1488 not include C99-style headers which use bools without triggering
1489 build errors in sieve.c.
1491 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1492 as mailq or other aliases. Changed the --help handling significantly
1493 to do whats expected. exim_usage() emits usage/help information.
1495 SC/01 Added the -bylocaldomain option to eximstats.
1497 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1499 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1501 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1507 PH/01 Another patch from the Sieve maintainer.
1509 PH/02 When an IPv6 address is converted to a string for single-key lookup
1510 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1511 dots are used instead of colons so that keys in lsearch files need not
1512 contain colons. This was done some time before quoting was made available
1513 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1514 (notated using the quote facility) so as to distinguish them from IPv4
1515 keys. This meant that lookups for IP addresses in host lists did not work
1516 for iplsearch lookups.
1518 This has been fixed by arranging for IPv6 addresses to be expressed with
1519 colons if the lookup type is iplsearch. This is not incompatible, because
1520 previously such lookups could never work.
1522 The situation is now rather anomolous, since one *can* have colons in
1523 ordinary lsearch keys. However, making the change in all cases is
1524 incompatible and would probably break a number of configurations.
1526 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1529 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1530 conversion specification without a maximum field width, thereby enabling
1531 a rogue spamd server to cause a buffer overflow. While nobody in their
1532 right mind would setup Exim to query an untrusted spamd server, an
1533 attacker that gains access to a server running spamd could potentially
1534 exploit this vulnerability to run arbitrary code as the Exim user.
1536 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1537 $primary_hostname instead of what libspf2 thinks the hosts name is.
1539 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1540 a directory entry by the name of the lookup key. Previously, if a
1541 symlink pointed to a non-existing file or a file in a directory that
1542 Exim lacked permissions to read, a lookup for a key matching that
1543 symlink would fail. Now it is enough that a matching directory entry
1544 exists, symlink or not. (Bugzilla 503.)
1546 PH/03 The body_linecount and body_zerocount variables are now exported in the
1549 PH/04 Added the $dnslist_matched variable.
1551 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1552 This means they are set thereafter only if the connection becomes
1555 PH/06 Added the client_condition to authenticators so that some can be skipped
1556 by clients under certain conditions.
1558 PH/07 The error message for a badly-placed control=no_multiline_responses left
1559 "_responses" off the end of the name.
1561 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1563 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1564 (without spaces) instead of just copying the configuration text.
1566 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1568 PH/11 Added $max_received_linelength.
1570 PH/12 Added +ignore_defer and +include_defer to host lists.
1572 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1573 way in which PCRE > 7.0 is built.
1575 PH/14 Implemented queue_only_load_latch.
1577 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1578 MAIL command. The effect was to mangle the value on 64-bit systems.
1580 PH/16 Another patch from the Sieve maintainer.
1582 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1584 PH/18 If a system quota error occurred while trying to create the file for
1585 a maildir delivery, the message "Mailbox is full" was not appended to the
1586 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1587 only to a quota excession during the actual writing of the file.
1589 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1590 characters?) which causes problems in log lines. The DN values are now
1591 passed through string_printing() before being added to log lines.
1593 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1594 and InterBase are left for another time.)
1596 PH/21 Added message_body_newlines option.
1598 PH/22 Guard against possible overflow in moan_check_errorcopy().
1600 PH/23 POSIX allows open() to be a macro; guard against that.
1602 PH/24 If the recipient of an error message contained an @ in the local part
1603 (suitably quoted, of course), incorrect values were put in $domain and
1604 $local_part during the evaluation of errors_copy.
1610 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1611 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1614 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1615 issue a MAIL command.
1617 PH/02 In an ACL statement such as
1619 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1621 if a client was not listed at all, or was listed with a value other than
1622 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1623 the condition was not true (as it should be), so access was not denied.
1624 The bug was that the ! inversion was incorrectly passed on to the second
1625 item. This has been fixed.
1627 PH/03 Added additional dnslists conditions == and =& which are different from
1628 = and & when the dns lookup returns more than one IP address.
1630 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1631 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1633 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1634 FSYNC, which compiles an option called disable_fsync that allows for
1635 bypassing fsync(). The documentation is heavily laced with warnings.
1637 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1639 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1640 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1641 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1642 including adding "make clean"; (3) Added -fPIC when compiling the test
1643 dynamically loaded module, to get rid of a warning.
1645 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1646 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1647 = 0s. The bug is otherwise harmless.
1649 PH/07 There was a bug in the dovecot authenticator such that the value of
1650 $auth1 could be overwritten, and so not correctly preserved, after a
1651 successful authentication. This usually meant that the value preserved by
1652 the server_setid option was incorrect.
1654 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1656 PH/09 Installed PCRE release 7.0.
1658 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1659 run for batched SMTP input. It is now run at the start of every message
1660 in the batch. While fixing this I discovered that the process information
1661 (output by running exiwhat) was not always getting set for -bs and -bS
1662 input. This is fixed, and it now also says "batched" for BSMTP.
1664 PH/11 Added control=no_pipelining.
1666 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1667 patch, slightly modified), and move the expansion of helo_data till after
1668 the connection is made in the smtp transport (so it can use these
1671 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1673 PH/14 Added log_selector = +pid.
1675 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1677 PH/16 Add ${if forany and ${if forall.
1679 PH/17 Added dsn_from option to vary the From: line in DSNs.
1681 PH/18 Flush SMTP output before performing a callout, unless control =
1682 no_callout_flush is set.
1684 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1685 was true (the default) a successful delivery failed to delete the retry
1686 item, thus causing premature timeout of the address. The bug is now
1689 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1691 PH/21 Long custom messages for fakedefer and fakereject are now split up
1692 into multiline reponses in the same way that messages for "deny" and
1693 other ACL rejections are.
1695 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1696 with slight modification.
1698 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1699 draft, changing the syntax and factoring some duplicate code".
1701 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1702 for deliveries of the second and subsequent messages over the same SMTP
1705 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1706 ${reduce, with only minor "tidies".
1708 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1710 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1711 expansion side effects.
1713 PH/27 When a message times out after an over-quota error from an Exim-imposed
1714 quota, the bounce message says "mailbox is full". This message was not
1715 being given when it was a system quota that was exceeded. It now should
1718 MH/03 Made $recipients available in local_scan(). local_scan() already has
1719 better access to the recipient list through recipients_list[], but
1720 $recipients can be useful in postmaster-provided expansion strings.
1722 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1723 in the case of a MAIL command with additional options following the
1724 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1725 were accidentally chopped off.
1727 PH/29 SMTP synchronization checks are implemented when a command is read -
1728 there is a check that no more input is waiting when there shouldn't be
1729 any. However, for some commands, a delay in an ACL can mean that it is
1730 some time before the response is written. In this time, more input might
1731 arrive, invalidly. So now there are extra checks after an ACL has run for
1732 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1733 pipelining has not been advertised.
1735 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1737 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1738 returned in the SMTP response when smtp_return_error_details was false.
1739 This has been fixed.
1741 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1742 instead of the C I/O that was originally supplied, because problems were
1743 reported on Solaris.
1745 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1746 Exim which did not show up earlier: it was assuming that a call to
1747 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1748 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1749 was a macro that became an assignment, so it seemed to work. This has
1750 changed to a proper function call with a void return, hence the compile
1751 error. Exim's code has been fixed.
1753 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1756 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1758 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1760 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1761 because it turns out to be too big or there is a timeout) there is no
1762 "Completed" line in the log. When some messages of this type were
1763 selected by exigrep, they were listed as "not completed". Others were
1764 picked up by some special patterns. I have improved the selection
1765 criteria to be more general.
1767 PH/38 The host_find_failed option in the manualroute router can now be set
1768 to "ignore", to completely ignore a host whose IP address cannot be
1769 found. If all hosts are ignored, the behaviour is controlled by the new
1770 host_all_ignored option.
1772 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1773 homing or because of multiple MX records with /mx) generated more than
1774 one IP address, and the following item turned out to be the local host,
1775 all the secondary addresses of the first item were incorrectly removed
1776 from the list, along with the local host and any following hosts (which
1777 is what is supposed to happen).
1779 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1780 whoever called Exim into the -H file. In the case of the daemon it was
1781 behaving confusingly. When first started, it used values for whoever
1782 started the daemon, but after a SIGHUP it used the Exim user (because it
1783 calls itself on a restart). I have changed the code so that it now always
1786 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1787 message are rejected with the same error (e.g. no authentication or bad
1788 sender address), and a DATA command is nevertheless sent (as can happen
1789 with PIPELINING or a stupid MUA), the error message that was given to the
1790 RCPT commands is included in the rejection of the DATA command. This is
1791 intended to be helpful for MUAs that show only the final error to their
1794 PH/42 Another patch from the Sieve maintainer.
1796 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1797 Eximstats - Fixed some broken HTML links and added missing column headers
1799 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1800 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1802 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1804 PH/43 Yet another patch from the Sieve maintainer.
1806 PH/44 I found a way to check for a TCP/IP connection going away before sending
1807 the response to the final '.' that terminates a message, but only in the
1808 case where the client has not sent further data following the '.'
1809 (unfortunately, this is allowed). However, in many cases there won't be
1810 any further data because there won't be any more messages to send. A call
1811 to select() can be used: if it shows that the input is "ready", there is
1812 either input waiting, or the socket has been closed. An attempt to read
1813 the next input character can distinguish the two cases. Previously, Exim
1814 would have sent an OK response which the client would never have see.
1815 This could lead to message repetition. This fix should cure that, at
1816 least in a lot of common cases.
1818 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1819 advertised in response to EHLO.
1825 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1826 fixed by 4.65/MH/01 (is this a record?) are fixed:
1828 (i) An empty string was always treated as zero by the numeric comparison
1829 operators. This behaviour has been restored.
1831 (ii) It is documented that the numeric comparison operators always treat
1832 their arguments as decimal numbers. This was broken in that numbers
1833 starting with 0 were being interpreted as octal.
1835 While fixing these problems I realized that there was another issue that
1836 hadn't been noticed. Values of message_size_limit (both the global option
1837 and the transport option) were treated as octal if they started with 0.
1838 The documentation was vague. These values are now always treated as
1839 decimal, and I will make that clear in the documentation.
1845 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1846 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1849 MH/01 Don't check that the operands of numeric comparison operators are
1850 integers when their expansion is in "skipping" mode (fixes bug
1851 introduced by 4.64-PH/07).
1853 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1854 child addresses, Exim now panics and dies. Previously, because the count
1855 is held in a short int, deliveries were likely to be lost. As such a
1856 large number of recipients for a single message is ridiculous
1857 (performance will be very, very poor), I have chosen to impose a limit
1858 rather than extend the field.
1864 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1865 leftover -K file (the existence of which was triggered by #402).
1866 While we were at it, introduced process PID as part of the -K
1867 filename. This should rule out race conditions when creating
1870 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1871 processing considerably. Previous code took too long for large mails,
1872 triggering a timeout which in turn triggers #401.
1874 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1875 in the DK code in transports.c. sendfile() is not really portable,
1876 hence the _LINUX specificness.
1878 TF/01 In the add_headers option to the mail command in an Exim filter,
1879 there was a bug that Exim would claim a syntax error in any
1880 header after the first one which had an odd number of characters
1883 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1884 callout verification, Exim cached a "reject" for the entire domain. This
1885 is correct for most verifications, but it is not correct for a recipient
1886 verification with use_sender or use_postmaster set, because in that case
1887 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1888 case of MAIL FROM:<> rejection from other early rejections (e.g.
1889 rejection of HELO). When verifying a recipient using a non-null MAIL
1890 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1891 Whatever the result of the callout, the value of the domain cache is
1892 left unchanged (for any other kind of callout, getting as far as trying
1893 RCPT means that the domain itself is ok).
1895 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1898 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1899 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1900 session, the daemon ignores ECONNECT errors and logs others; it now
1901 ignores EPIPE as well.
1903 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1904 (quoted-printable decoding).
1906 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1907 later the small subsequent patch to fix an introduced bug.
1909 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1911 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1913 PH/08 An error is now given if message_size_limit is specified negative.
1915 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1916 to be given (somewhat) arbitrary names.
1918 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1921 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1922 miscellaneous code fixes
1924 PH/10 Added the log_reject_target ACL modifier to specify where to log
1927 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1928 hostname. This is wrong, because it relates to the incoming message (and
1929 probably the interface on which it is arriving) and not to the outgoing
1930 callout (which could be using a different interface). This has been
1931 changed to use the value of the helo_data option from the smtp transport
1932 instead - this is what is used when a message is actually being sent. If
1933 there is no remote transport (possible with a router that sets up host
1934 addresses), $smtp_active_hostname is used.
1936 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1937 tweaks were necessary in order to get it to work (see also 21 below):
1938 (a) The code assumed that strncpy() returns a negative number on buffer
1939 overflow, which isn't the case. Replaced with Exim's string_format()
1941 (b) There were several signed/unsigned issues. I just did the minimum
1942 hacking in of casts. There is scope for a larger refactoring.
1943 (c) The code used strcasecmp() which is not a standard C function.
1944 Replaced with Exim's strcmpic() function.
1945 (d) The code set only $1; it now sets $auth1 as well.
1946 (e) A simple test gave the error "authentication client didn't specify
1947 service in request". It would seem that Dovecot has changed its
1948 interface. Fortunately there's a specification; I followed it and
1949 changed what the client sends and it appears to be working now.
1951 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1954 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1955 address A is aliased to B and C, where B exists and C does not. Without
1956 -v the output is "A verified" because verification stops after a
1957 successful redirection if more than one address is generated. However,
1958 with -v the child addresses are also verified. Exim was outputting "A
1959 failed to verify" and then showing the successful verification for C,
1960 with its parentage. It now outputs "B failed to verify", showing B's
1961 parentage before showing the successful verification of C.
1963 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1964 look up a TXT record in a specific list after matching in a combined
1967 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1968 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1969 they consult the DNS. I had assumed they would set it the way they
1970 wanted; and indeed my experiments on Linux seem to show that in some
1971 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1972 To be on the safe side, however, I have now made the interface to
1973 host_find_byname() similar to host_find_bydns(), with an argument
1974 containing the DNS resolver options. The host_find_byname() function now
1975 sets these options at its start, just as host_find_bydns() does. The smtp
1976 transport options dns_qualify_single and dns_search_parents are passed to
1977 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1978 of host_find_byname() use the default settings of RES_DEFNAMES
1979 (qualify_single) but not RES_DNSRCH (search_parents).
1981 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1982 spool_read_header() do less string testing, by means of a preliminary
1983 switch on the second character of optional "-foo" lines. (This is
1984 overdue, caused by the large number of possibilities that now exist.
1985 Originally there were few.) While I was there, I also converted the
1986 str(n)cmp tests so they don't re-test the leading "-" and the first
1987 character, in the hope this might squeeze out yet more improvement.
1989 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1990 flag allowing group syntax was set by the header_syntax check but not
1991 turned off, possible causing trouble later; (2) The flag was not being
1992 set at all for the header_verify test, causing "group"-style headers to
1993 be rejected. I have now set it in this case, and also caused header_
1994 verify to ignore an empty address taken from a group. While doing this, I
1995 came across some other cases where the code for allowing group syntax
1996 while scanning a header line wasn't quite right (mostly, not resetting
1997 the flag correctly in the right place). These bugs could have caused
1998 trouble for malformed header lines. I hope it is now all correct.
2000 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2001 with the "reply" argument non-NULL. The code, however (which originally
2002 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2003 but it didn't always do it. This confused somebody who was copying the
2004 code for some other use. I have removed all the tests.
2006 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2007 feature that was used to support insecure browsers during the U.S. crypto
2008 embargo. It requires special client support, and Exim is probably the
2009 only MTA that supported it -- and would never use it because real RSA is
2010 always available. This code has been removed, because it had the bad
2011 effect of slowing Exim down by computing (never used) parameters for the
2012 RSA_EXPORT functionality.
2014 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2015 authenticator to fail if there's a tab character in the incoming data
2016 (there should never be unless someone is messing about, as it's supposed
2017 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2018 if the connection is using TLS or if the remote IP is the same as the
2019 local IP, and the "valid-client-cert option" if a client certificate has
2022 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2023 authenticators. This can be used for authorization after authentication
2024 succeeds. (In the case of plaintext, it servers for both authentication
2027 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2028 if any retry times were supplied.
2030 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2031 connection, where there is no client IP address to check. In this
2032 situation, the verify now always succeeds.
2034 PH/25 Applied John Jetmore's -Mset patch.
2036 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2038 PH/27 In a string expansion for a processed (not raw) header when multiple
2039 headers of the same name were present, leading whitespace was being
2040 removed from all of them, but trailing whitespace was being removed only
2041 from the last one. Now trailing whitespace is removed from each header
2042 before concatenation. Completely empty headers in a concatenation (as
2043 before) are ignored.
2045 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2046 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2048 PH/29 [Removed. This was a change that I later backed out, and forgot to
2049 correct the ChangeLog entry (that I had efficiently created) before
2050 committing the later change.]
2052 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2053 address errors (4xx response to RCPT) over the same connection as other
2054 messages routed to the same hosts. Such deliveries are always "forced",
2055 so retry times are not inspected. This resulted in far too many retries
2056 for the affected addresses. The effect occurred only when there were more
2057 hosts than the hosts_max_try setting in the smtp transport when it had
2058 the 4xx errors. Those hosts that it had tried were not added to the list
2059 of hosts for which the message was waiting, so if all were tried, there
2060 was no problem. Two fixes have been applied:
2062 (i) If there are any address or message errors in an SMTP delivery, none
2063 of the hosts (tried or untried) are now added to the list of hosts
2064 for which the message is waiting, so the message should not be a
2065 candidate for sending over the same connection that was used for a
2066 successful delivery of some other message. This seems entirely
2067 reasonable: after all the message is NOT "waiting for some host".
2068 This is so "obvious" that I'm not sure why it wasn't done
2069 previously. Hope I haven't missed anything, but it can't do any
2070 harm, as the worst effect is to miss an optimization.
2072 (ii) If, despite (i), such a delivery is accidentally attempted, the
2073 routing retry time is respected, so at least it doesn't keep
2074 hammering the server.
2076 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2077 in ${readsocket because some servers need this prod.
2079 PH/32 Added some extra debug output when updating a wait-xxx database.
2081 PH/33 The hint "could be header name not terminated by colon", which has been
2082 given for certain expansion errors for a long time, was not being given
2083 for the ${if def:h_colon_omitted{... case.
2085 PH/34 The spec says: "With one important exception, whenever a domain list is
2086 being scanned, $domain contains the subject domain." There was at least
2087 one case where this was not true.
2089 PH/35 The error "getsockname() failed: connection reset by peer" was being
2090 written to the panic log as well as the main log, but it isn't really
2091 panic-worthy as it just means the connection died rather early on. I have
2092 removed the panic log writing for the ECONNRESET error when getsockname()
2095 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2096 runs only) independently of the message's sender address. This meant
2097 that, if the 4xx error was in fact related to the sender, a different
2098 message to the same recipient with a different sender could confuse
2099 things. In particualar, this can happen when sending to a greylisting
2100 server, but other circumstances could also provoke similar problems.
2101 I have changed the default so that the retry time for these errors is now
2102 based a combination of the sender and recipient addresses. This change
2103 can be overridden by setting address_retry_include_sender=false in the
2106 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2107 remote server are returned as part of bounce messages. This was not
2108 happening for LMTP over a pipe (the lmtp transport), but now it is the
2109 same for both kinds of LMTP.
2111 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2112 in header lines that were in fact CNAMEs. This is no longer the case.
2114 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2115 and queue runs started by the daemon processed all messages. This has
2116 been fixed so that -R and -S can now usefully be given with -q<time>.
2118 PH/40 Import PCRE release 6.7 (fixes some bugs).
2120 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2122 PH/42 Give an error if -q is specified more than once.
2124 PH/43 Renamed the variables $interface_address and $interface_port as
2125 $received_ip_address and $received_port, to make it clear that these
2126 values apply to message reception, and not to the outgoing interface when
2127 a message is delivered. (The old names remain recognized, of course.)
2129 PH/44 There was no timeout on the connect() call when using a Unix domain
2130 socket in the ${readsocket expansion. There now is.
2132 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2133 be meaningful with "accept".
2135 SC/01 Eximstats V1.43
2136 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2138 SC/02 Eximstats V1.44
2139 Use a glob alias rather than an array ref in the generated
2140 parser. This improves both readability and performance.
2142 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2143 Collect SpamAssassin and rejection statistics.
2144 Don't display local sender or destination tables unless
2145 there is data to show.
2146 Added average volumes into the top table text output.
2148 SC/04 Eximstats V1.46
2149 Collect data on the number of addresses (recipients)
2150 as well as the number of messages.
2152 SC/05 Eximstats V1.47
2153 Added 'Message too big' to the list of mail rejection
2154 reasons (thanks to Marco Gaiarin).
2156 SC/06 Eximstats V1.48
2157 Mainlog lines which have GMT offsets and are too short to
2158 have a flag are now skipped.
2160 SC/07 Eximstats V1.49 (Alain Williams)
2161 Added the -emptyok flag.
2163 SC/08 Eximstats V1.50
2164 Fixes for obtaining the IP address from reject messages.
2166 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2167 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2168 whitesspace changes from 4.64-PH/27
2170 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2173 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2174 are found, allow negative numbers in numeric criteria)
2176 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2178 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2179 to match changes made in 4.64-PH/43
2181 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2183 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2184 "domain" is actually an IP address, and force a failure. This locks out
2185 those revolvers/nameservers that support "A-for-A" lookups, in
2186 contravention of the specifications.
2188 PH/48 When a host name was looked up from an IP address, and the subsequent
2189 forward lookup of the name timed out, the host name was left in
2190 $sender_host_name, contrary to the specification.
2192 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2193 restricted to single-key lookups, Exim was not diagnosing an error if
2194 * or *@ was used with a query-style lookup.
2196 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2198 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2199 long ago, but noone interested enough thought of it. Let's just say that
2200 the "1.1" means that there are some new functions that weren't there at
2201 some point in the past.
2203 PH/51 Error processing for expansion failure of helo_data from an smtp
2204 transport during callout processing was broken.
2206 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2207 tested/used via the -bh/-bhc/-bs options.
2209 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2210 bug, fixed in subsequent PCRE releases).
2212 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2213 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2215 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2221 SC/01 Use a glob alias rather than an array ref in eximstats generated
2222 parser. This improves both readability and performance.
2224 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2225 Don't display local sender or destination tables in eximstats unless
2226 there is data to show.
2227 Added average volumes into the eximstats top table text output.
2229 SC/03 Collect data on the number of addresses (recipients) as well
2230 as the number of messages in eximstats.
2232 TF/01 Correct an error in the documentation for the redirect router. Exim
2233 does (usually) call initgroups() when daemonizing.
2235 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2236 with consistent privilege compared to when running as a daemon.
2238 TF/03 Note in the spec that $authenticated_id is not set for local
2239 submissions from trusted users.
2241 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2242 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2244 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2245 by adding some example configuration directives to the default
2246 configuration file. A little bit of work is required to uncomment the
2247 directives and define how usernames and passwords are checked, but
2248 there is now a framework to start from.
2250 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2251 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2252 without this. I don't know how relevant this is to other LDAP libraries.
2254 PH/02 Add the verb name to the "unknown ACL verb" error.
2256 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2258 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2260 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2261 directory not expanded when it should be if an expanded home directory
2262 was set for the address (which is overridden by the transport).
2264 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2267 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2268 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2269 because it is too late at that time, and has no effect.
2271 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2272 security issue with \' (bugzilla #107). I could not use the
2273 PQescapeStringConn() function, because it needs a PGconn value as one of
2276 PH/08 When testing addresses using -bt, indicate those final addresses that
2277 are duplicates that would not cause an additional delivery. At least one
2278 person was confused, thinking that -bt output corresponded to deliveries.
2279 (Suppressing duplicates isn't a good idea as you lose the information
2280 about possibly different redirections that led to the duplicates.)
2282 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2283 systems where poll() doesn't work, in particular OS X.
2285 PH/10 Added more information to debugging output for retry time not reached.
2287 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2288 operations in malware.c.
2290 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2293 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2294 syslog_duplication set false, log lines that would normally be written
2295 both the the main log and to the reject log were not written to syslog at
2298 PH/14 In the default configuration, change the use of "message" in ACL warn
2299 statements to "add_header".
2301 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2302 not followed by a command (e.g. "seen endif").
2304 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2305 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2308 PH/17 Added extra conditions to the default value of delay_warning_condition
2312 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2313 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2314 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2317 The Auto-Submitted: and various List- headers are standardised, whereas I
2318 don't think Precedence: ever was.
2320 PH/18 Refactored debugging code in route_finduser() to show more information,
2321 in particular, the error code if getpwnam() issues one.
2323 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2324 This is apparently needed in addition to the PH/07 change above to avoid
2325 any possible encoding problems.
2327 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2328 but not after initializing Perl.
2330 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2331 output them only if debugging. By default they are written stderr,
2332 apparently, which is not desirable.
2334 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2337 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2340 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2342 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2343 authenticated or an ident call has been made. Suppress the default
2344 values for $authenticated_id and $authenticated_sender (but permit -oMai
2345 and -oMas) when testing with -bh.
2347 PH/24 Re-jigged the order of the tests in the default configuration so that the
2348 tests for valid domains and recipients precede the DNS black list and CSA
2349 tests, on the grounds that those ones are more expensive.
2351 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2352 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2353 This bug exists in every version of Exim that I still have, right back to
2356 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2357 However, an attempt to turn on case-sensitivity in a regex key by
2358 including (?-i) didn't work because the subject string was already
2359 lowercased, and the effects were non-intuitive. It turns out that a
2360 one-line patch can be used to allow (?-i) to work as expected.
2366 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2367 other effects) broke the use of negated acl sub-conditions.
2369 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2372 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2373 "Deny" causes Exim to reject the incoming connection with a 554 error.
2374 Unfortunately, if there is a major crisis, such as a disk failure,
2375 tcp-wrappers gives "deny", whereas what one would like would be some
2376 kind of temporary error. A kludge has been added to help with this.
2377 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2378 554 error is used if errno is still zero or contains ENOENT (which occurs
2379 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2382 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2384 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2385 errors. Otherwise a message that provokes a temporary error (when other
2386 messages do not) can cause a whole host to time out.
2388 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2389 addresses were routed directly to files or pipes from a redirect router.
2390 File deliveries just didn't batch; pipe deliveries might have suffered
2393 PH/06 A failure to get a lock for a hints database would erroneously always say
2394 "Failed to get write lock", even when it was really a read lock.
2396 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2397 of 0600. This has been changed to use the value of the lockfile_mode
2398 option (which defaults to 0600).
2400 PH/08 Applied small patch from the Sieve maintainer.
2402 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2403 folder from quota calculations, a direct delivery into this folder messed
2404 up the contents of the maildirsize file. This was because the regex was
2405 used only to exclude .Trash (or whatever) when the size of the mailbox
2406 was calculated. There was no check that a delivery was happening into an
2407 excluded directory. This bug has been fixed by ignoring all quota
2408 processing for deliveries into excluded directories.
2410 PH/10 Added the maildirfolder_create_regex option to appendfile.
2416 PH/01 The code for finding all the local interface addresses on a FreeBSD
2417 system running IPv6 was broken. This may well have applied to all BSD
2418 systems, as well as to others that have similar system calls. The broken
2419 code found IPv4 interfaces correctly, but gave incorrect values for the
2420 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2421 that it would not match correctly against @[] and not recognize the IPv6
2424 PH/02 The ipliteral router was not recognizing addresses of the form user@
2425 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2427 PH/03 Added disable_ipv6.
2429 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2430 decoded form, because it is most often used to construct To: headers
2431 lines in autoreplies, and the decoded form may well be syntactically
2432 invalid. However, $reply_address has leading white space removed, and all
2433 newlines turned into spaces so that the autoreply transport does not
2436 PH/05 If group was specified without a user on a router, and no group or user
2437 was specified on a transport, the group from the router was ignored.
2439 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2440 for visible compile-time settings that can be used to change these
2441 numbers, for those that want even more. Backwards compatibility with old
2442 spool files has been maintained. However, going back to a previous Exim
2443 release will lost any variables that are in spool files.
2445 PH/07 Two small changes when running in the test harness: increase delay when
2446 passing a TCP/IP connection to a new process, in case the original
2447 process has to generate a bounce, and remove special handling of
2448 127.0.0.2 (sic), which is no longer necessary.
2450 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2451 be the same on different OS.
2453 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2456 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2457 whether --show-vars was specified or not
2459 JJ/02 exipick: Added support for new ACL variable spool format introduced
2462 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2463 syntactically invalid From: or Reply-to: line, and a filter used this to
2464 generate an autoreply, and therefore failed to obtain an address for the
2465 autoreply, Exim could try to deliver to a non-existent relative file
2466 name, causing unrelated and misleading errors. What now happens is that
2467 it logs this as a hard delivery error, but does not attempt to create a
2470 PH/11 The exinext utility has a -C option for testing purposes, but although
2471 the given file was scanned by exinext itself; it wasn't being passed on
2472 when Exim was called.
2474 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2475 an end-of-file indication when reading a command response.
2477 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2478 compiled. In many other places in Exim, IPv6 addresses are always
2479 recognized, so I have changed this. It also means that IPv4 domain
2480 literals of the form [IPV4:n.n.n.n] are now always recognized.
2482 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2483 used if the router is not running as root, for example, when verifying at
2484 ACL time, or when using -bh. The debugging output from this situation was
2485 non-existent - all you got was a failure to exec. I have made two
2488 (a) Failures to set uid/gid, the current directory, or a process leader
2489 in a subprocess such as that created by queryprogram now generate
2490 suitable debugging ouput when -d is set.
2492 (b) The queryprogram router detects when it is not running as root,
2493 outputs suitable debugging information if -d is set, and then runs
2494 the subprocess without attempting to change uid/gid.
2496 PH/15 Minor change to Makefile for building test_host (undocumented testing
2499 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2500 additional section of a DNS packet that returns MX or SRV records.
2501 Instead, it always explicitly searches for A/AAAA records. This avoids
2502 major problems that occur when a DNS server includes only records of one
2503 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2504 fixed another bug: if SRV records were looked up and the corresponding
2505 address records were *not* found in the additional section, the port
2506 values from the SRV records were lost.
2508 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2509 using the correct key (the original address) when searching the retry
2510 rules in order to find which one to use for generating the retry hint.
2512 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2513 adding the default one. Similarly, if it contains a Reply-To: header, the
2514 errors_reply_to option, if set, is not used.
2516 PH/19 When calculating a retry time, Exim used to measure the "time since
2517 failure" by looking at the "first failed" field in the retry record. Now
2518 it does not use this if it is later than than the arrival time of the
2519 message. Instead it uses the arrival time. This makes for better
2520 behaviour in cases where some deliveries succeed, thus re-setting the
2521 "first failed" field. An example is a quota failure for a huge message
2522 when small messages continue to be delivered. Without this change, the
2523 "time since failure" will always be short, possible causing more frequent
2524 delivery attempts for the huge message than are intended.
2525 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2527 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2528 $1, $2, $3) because the numerical variables can be reset during some
2529 expansion items (e.g. "match"), thereby losing the authentication data.
2531 PH/21 Make -bV show the size of off_t variables so that the test suite can
2532 decide whether to run tests for quotas > 2G.
2534 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2535 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2536 filecount value is greater than 2G or if a quota value is greater than 2G
2537 on a system where the size of off_t is not greater than 4, a panic error
2540 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2541 never match. The debug and -bh output now contains an explicit error
2542 message indicating a malformed IPv4 address or mask.
2544 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2545 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2546 PH/23 above applies.
2548 PH/25 Do not write to syslog when running in the test harness. The only
2549 occasion when this arises is a failure to open the main or panic logs
2550 (for which there is an explicit test).
2552 PH/26 Added the /no_tell option to "control=freeze".
2554 PH/27 If a host name lookup failed very early in a connection, for example, if
2555 the IP address matched host_lookup and the reverse lookup yielded a name
2556 that did not have a forward lookup, an error message of the form "no IP
2557 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2558 could be logged. Now it outputs the IP address instead of "NULL".
2560 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2561 allows the sender and the authenticated sender to be set when
2562 submitting a message from within Exim. Since child_open_exim() is
2563 documented for local_scan(), the new function should be too.
2565 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2566 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2567 results in an empty string is now treated as unset.
2569 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2571 PH/31 Added sender_verify_fail logging option.
2573 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2574 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2575 tidied the source and removed it altogether.
2577 PH/33 When a queue run was abandoned because the load average was too high, a
2578 log line was always written; now it is written only if the queue_run log
2579 selector is set. In addition, the log line for abandonment now contains
2580 information about the queue run such as the pid. This is always present
2581 in "start" and "stop" lines but was omitted from the "abandon" line.
2583 PH/34 Omit spaces between a header name and the colon in the error message that
2584 is given when verify = headers_syntax fails (if there are lots of them,
2585 the message gets confusing).
2587 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2588 names, as there are now some PTR records that contain slashes. This check
2589 is only to protect against broken name servers that fall over on strange
2590 characters, so the fact that it applies to all lookups doesn't matter.
2592 PH/36 Now that the new test suite is complete, we can remove some of the
2593 special code in Exim that was needed for the old test suite. For example,
2594 sorting DNS records because real resolvers return them in an arbitrary
2595 order. The new test suite's fake resolver always returns records in the
2598 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2599 bounces) except when queue_only is set, to avoid logging races between
2600 the different processes.
2602 PH/38 Panic-die if .include specifies a non-absolute path.
2604 PH/39 A tweak to the "H" retry rule from its user.
2606 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2607 a label. They prevented compilation on older perls.
2609 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2610 a warning to be raised on newish perls.
2612 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2613 on queue. Changes to match documented behaviour of showing count of
2614 messages matching specified criteria.
2616 PH/40 Changed the default ident timeout from 30s to 5s.
2618 PH/41 Added support for the use of login_cap features, on those BSD systems
2619 that have them, for controlling the resources used by pipe deliveries.
2621 PH/42 The content-scanning code uses fopen() to create files in which to put
2622 message data. Previously it was not paying any attention to the mode of
2623 the files. Exim runs with umask(0) because the rest of the code creates
2624 files with open(), and sets the required mode explicitly. Thus, these
2625 files were ending up world-writeable. This was not a big issue, because,
2626 being within the spool directory, they were not world-accessible. I have
2627 created a function called modefopen, which takes an additional mode
2628 argument. It sets umask(777), creates the file, chmods it to the required
2629 mode, then resets the umask. All the relevant calls to fopen() in the
2630 content scanning code have been changed to use this function.
2632 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2633 to 24 hours. This avoids potential overflow problems when processing G
2634 and H retry rules. I suspect nobody ever tinkers with this value.
2636 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2638 PH/45 When the plaintext authenticator is running as a client, the server's
2639 challenges are checked to ensure they are valid base64 strings. By
2640 default, the authentication attempt is cancelled if an invalid string is
2641 received. Setting client_ignore_invalid_base64 true ignores these errors.
2642 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2643 they are received. Thus, the responses can be made to depend on the
2644 challenges. If an invalid string is ignored, an empty string is placed in
2647 PH/46 Messages that are created by the autoreply transport now contains a
2648 References: header, in accordance with RFCs 2822 and 3834.
2650 PH/47 Added authenticated_sender_force to the smtp transport.
2652 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2654 PH/49 Installed latest patch from the Sieve maintainer.
2656 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2657 was also set, the appendfile transport was unnecessarily scanning a
2658 directory of message files (e.g. for maildir delivery) to find the count
2659 of files (along with the size), even though it did not need this
2660 information. It now does the scan only if it needs to find either the
2661 size of the count of files.
2663 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2665 PH/52 Two bugs concerned with error handling when the smtp transport is
2668 (i) Exim was not creating retry information for temporary errors given
2669 for individual recipients after the DATA command when the smtp transport
2670 was used in LMTP mode. This meant that they could be retried too
2671 frequently, and not timed out correctly.
2673 (ii) Exim was setting the flag that allows error details to be returned
2674 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2675 recipients that were returned after the DATA command.
2677 PH/53 This is related to PH/52, but is more general: for any failing address,
2678 when detailed error information was permitted to be returned to the
2679 sender, but the error was temporary, then after the final timeout, only
2680 "retry timeout exceeded" was returned. Now it returns the full error as
2681 well as "retry timeout exceeded".
2683 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2684 do this, and (what is worse) MTAs that accept it.
2686 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2687 will now be deprecated.
2689 PH/56 New os.c-cygwin from the Cygwin maintainer.
2691 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2692 formats (previously only available in exim formats via -bpr, -bpru,
2693 and -bpra. Now also available in native and exiqgrep formats)
2695 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2696 with very large, slow to parse queues
2698 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2700 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2702 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2703 responses to authentication challenges, though it was showing the
2704 challenges; (ii) I've removed the CR characters from the debug output for
2707 PH/58 Allow for the insertion of a newline as well as a space when a string
2708 is turned into more than one encoded-word during RFC 2047 encoding. The
2709 Sieve code now uses this.
2711 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2712 data_4xx, lost_connection, tls_required.
2714 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2715 message was being sent as an SMTP response.
2717 PH/61 Add -l and -k options to exicyclog.
2719 PH/62 When verifying, if an address was redirected to one new address, so that
2720 verification continued, and the new address failed or deferred after
2721 having set something in $address_data, the value of $address_data was not
2722 passed back to the ACL. This was different to the case when no
2723 redirection occurred. The value is now passed back in both cases.
2725 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2726 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2727 use login_cap.h, so on its own it isn't the distinguishing feature. The
2728 new name refers directly to the setclassresources() function.
2730 PH/65 Added configuration files for NetBSD3.
2732 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2734 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2735 is preferred over IPv4.
2737 PH/68 The bounce_return_message and bounce_return_body options were not being
2738 honoured for bounces generated during the reception of non-SMTP messages.
2739 In particular, this applied to messages rejected by the ACL. This bug has
2740 been fixed. However, if bounce_return_message is true and bounce_return_
2741 body is false, the headers that are returned for a non-SMTP message
2742 include only those that have been read before the error was detected.
2743 (In the case of an ACL rejection, they have all been read.)
2745 PH/69 The HTML version of the specification is now built in a directory called
2746 spec_html instead of spec.html, because the latter looks like a path with
2747 a MIME-type, and this confuses some software.
2749 PH/70 Catch two compiler warnings in sieve.c.
2751 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2752 function verify_get_ident() calls ip_connect() to connect a socket, but
2753 if the "connect()" function timed out, ip_connect() used to close the
2754 socket. However, verify_get_ident() also closes the socket later, and in
2755 between Exim writes to the log, which may get opened at this point. When
2756 the socket was closed in ip_connect(), the log could get the same file
2757 descriptor number as the socket. This naturally causes chaos. The fix is
2758 not to close the socket in ip_connect(); the socket should be closed by
2759 the function that creates it. There was only one place in the code where
2760 this was missing, in the iplookup router, which I don't think anybody now
2761 uses, but I've fixed it anyway.
2763 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2764 well as to direct DNS lookups. Otherwise the handling of names in host
2765 lists is inconsistent and therefore confusing.
2771 PH/01 Two changes to the default runtime configuration:
2773 (1) Move the checks for relay_from_hosts and authenticated clients from
2774 after to before the (commented out) DNS black list checks.
2776 (2) Add control=submission to the relay_from_hosts and authenticated
2777 clients checks, on the grounds that messages accepted by these
2778 statements are most likely to be submissions.
2780 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2782 (1) Generate an error if the third argument for the ${prvs expansion is
2785 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2788 (3) Reset the variables that are obtained from the first argument of
2789 ${prvscheck and used in the second argument before leaving the code,
2790 because their memory is reclaimed, so using them afterwards may do
2793 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2794 one (it's much easier than Tom thought :-).
2796 (5) Because of (4), we can now allow for the use of $prvscheck_result
2797 inside the third argument.
2799 PH/03 For some reason, the default setting of PATH when running a command from
2800 a pipe transport was just "/usr/bin". I have changed it to
2803 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2804 anything to be listed in the output from -bV.
2806 PH/05 When a filter generated an autoreply, the entire To: header line was
2807 quoted in the delivery log line, like this:
2809 => >A.N.Other <ano@some.domain> <original@ddress> ...
2811 This has been changed so that it extracts the operative address. There
2812 may be more than one such address. If so, they are comma-separated, like
2815 => >ano@some.domain,ona@other.domain <original@ddress> ...
2817 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2818 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2819 not being looked up in the rDNS to get a host name, Exim was showing the
2820 IP address twice in Received: lines, even though the IP addresses were
2821 identical. For example:
2823 Received: from [1.2.3.4] (helo=[1.2.3.4])
2825 However, if the real host name was known, it was omitting the HELO data
2826 if it matched the actual IP address. This has been tidied up so that it
2827 doesn't show the same IP address twice.
2829 PH/07 When both +timestamp and +memory debugging was on, the value given by
2830 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2831 called by the debug printing, thereby overwriting the timestamp buffer.
2832 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2834 PH/08 When the original message was included in an autoreply transport, it
2835 always said "this is a copy of the message, including all the headers",
2836 even if body_only or headers_only was set. It now gives an appropriate
2839 PH/09 Applied a patch from the Sieve maintainer which:
2841 o fixes some comments
2842 o adds the (disabled) notify extension core
2843 o adds some debug output for the result of if/elsif tests
2844 o points to the current vacation draft in the documentation
2845 and documents the missing references header update
2849 o fixes a bug in processing the envelope test (when testing
2850 multiple envelope elements, the last element determinted the
2853 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2854 Electronic Mail") by including:
2856 Auto-submitted: auto-generated
2858 in the messages that it generates (bounce messages and others, such as
2859 warnings). In the case of bounce messages for non-SMTP mesages, there was
2860 also a typo: it was using "Auto_submitted" (underscore instead of
2861 hyphen). Since every message generated by Exim is necessarily in response
2862 to another message, thes have all been changed to:
2864 Auto-Submitted: auto-replied
2866 in accordance with these statements in the RFC:
2868 The auto-replied keyword:
2870 - SHOULD be used on messages sent in direct response to another
2871 message by an automatic process,
2873 - MUST NOT be used on manually-generated messages,
2875 - MAY be used on Delivery Status Notifications (DSNs) and Message
2876 Disposition Notifications (MDNs),
2878 - MUST NOT be used on messages generated by automatic or periodic
2879 processes, except for messages which are automatic responses to
2882 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2883 to the default Received: header definition.
2885 PH/12 Added log selector acl_warn_skipped (default on).
2887 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2888 variables because (a) they are in the wrong storage pool and (b) even if
2889 they were copied, it wouldn't work properly because of the caching.
2891 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2892 checking when decoding. Apparently there are clients that generate
2893 overlong encoded strings. Why am I not surprised?
2895 PH/15 If the first argument of "${if match_address" was not empty, but did not
2896 contain an "@" character, Exim crashed. Now it writes a panic log message
2897 and treats the condition as false.
2899 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2901 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2902 extension "envelope-auth". The code is finished and in agreement with
2903 other implementations, but there is no documentation so far and in fact,
2904 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2905 not changing the active code.
2907 Print executed "if" and "elsif" statements when debugging is used. This
2908 helps a great deal to understand what a filter does.
2910 Document more things not specified clearly in RFC3028. I had all this
2911 sorted out, when out of a sudden new issues came to my mind. Oops."
2913 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2916 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2917 canonical form (as documented). However, after a host name lookup from
2918 the IP address, check_host() was doing a simple string comparison with
2919 addresses acquired from the DNS when checking that the found name did
2920 have the original IP as one of its addresses. Since any found IPv6
2921 addresses are likely to be in abbreviated form, the comparison could
2922 fail. Luckily, there already exists a function for doing the comparison
2923 by converting both addresses to binary, so now that is used instead of
2924 the text comparison.
2926 PH/20 There was another similar case to PH/19, when a complete host name was
2927 given in a host list; looking up its IP address could give an abbreviated
2928 form, whereas the current host's name might or might not be abbreviated.
2929 The same fix has been applied.
2935 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2936 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2939 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2940 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2942 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2944 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2945 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2946 Also, for "Auto-Submitted", treat anything other than "no" as
2947 non-personal, in accordance with RFC 3834. (Previously it treated
2948 anything starting "auto-" as non-personal.)
2950 TF/01 The control=submission/name=... option had a problem with syntax
2951 errors if the name included a slash character. The /name= option
2952 now slurps the rest of the string, so it can include any characters
2953 but it must come last in the list of options (after /sender_retain
2956 PH/05 Some modifications to the interface to the fake nameserver for the new
2964 TK/01 Added the "success_on_redirect" address verification option. See
2965 NewStuff for rationale and an example.
2967 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2969 PH/02 Patch to exigrep to allow it to work on syslog lines.
2971 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2972 fread() to skip over the body file's header line, because in Cygwin the
2973 header line is locked and is inaccessible.
2975 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2976 co-exist for some time) to make it clear that it is the Exim ID that is
2977 referenced, not the Message-ID: header line.
2979 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2980 string_format() function, because snprintf() does not exist on all
2983 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2984 ${stat: expansion item.
2986 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2987 protocol synchronization error", to keep the pedants happy.
2989 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2990 well as for IRIX systems, when gcc is being used. See the host.c source
2993 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2995 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2998 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2999 transport and to the smtp transport in LMTP mode.
3001 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3003 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3004 run a filter in a subprocess. This could lead to confusion in subsequent
3005 lookups in the parent process. There should also be a search_tidyup() at
3006 the end of the subprocess.
3008 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3009 only if the host matched helo_try_verify_hosts, which caused the
3010 verification to occur when the EHLO/HELO command was issued. The ACL just
3011 tested the remembered result. Now, if a previous verification attempt has
3012 not happened, "verify = helo" does it there and then.
3014 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3016 TK/03 Fix log output including CR from clamd.
3018 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3019 exist provoked a memory error which could cause a segfault.
3021 PH/15 Installed PCRE 6.2
3023 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3025 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3026 of the problem. Specifically, suggested +O2 rather than +O1 for the
3029 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3031 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3032 record, the port was not correctly logged when the outgoing_port log
3033 selector was set (it logged the transort's default port).
3035 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3036 fallback_hosts, and "hosts" in the smtp transport.
3038 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3039 host errors such as "Connection refused".
3041 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3042 authentication with radiusclient 0.4.9:
3044 - Error returned from rc_read_config was caught wrongly
3045 - Username/password not passed on to radius server due to wrong length.
3047 The presumption is that some radiusclient API changes for 4.51/PH/17
3048 were not taken care of correctly. The code is still untested by me (my
3049 Linux distribution still has 0.3.2 of radiusclient), but it was
3050 contributed by a Radius user.
3052 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3053 expanding the "port" option of the smtp transport.
3055 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3056 while reading a MIME header. Thanks to Tom Hughes for a patch.
3058 PH/24 Include config.h inside local_scan.h so that configuration settings are
3061 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3062 that in an ACL for RCPT (for example), you can examine exactly what was
3065 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3066 commands, but it was not correctly comparing the address with the actual
3067 client host address. Thus, it would show the EHLO address in Received:
3068 header lines when this was not necessary.
3070 PH/27 Added the % operator to ${eval:}.
3072 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3073 it should be ignoring failures (because with -C, for example, it has lost
3074 privilege). It wasn't ignoring creation failures other than "already
3077 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3080 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3081 because an input line was too long, either on its own, or by virtue of
3082 too many continuations, the temporary file was not being removed, and the
3083 return code was incorrect.
3085 PH/31 Missing "BOOL" in function definition in filtertest.c.
3087 PH/32 Applied Sieve patches from the maintainer.
3089 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3091 PH/33 Added "verify = not_blind".
3093 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3094 Local/Makefile (with some defaults set). These are used in built scripts
3095 such as exicyclog, but they have never been used in the exim_install
3096 script (though there are many overriding facilities there). I have
3097 arranged that the exim_install script now takes note of these two
3100 PH/35 Installed configuration files for Dragonfly.
3102 PH/36 When a locally submitted message by a trusted user did not contain a
3103 From: header, and the sender address was obtained from -f or from an SMTP
3104 MAIL command, and the trusted user did not use -F to supply a sender
3105 name, $originator_name was incorrectly used when constructing a From:
3106 header. Furthermore, $originator_name was used for submission mode
3107 messages from external hosts without From: headers in a similar way,
3108 which is clearly wrong.
3110 PH/37 Added control=suppress_local_fixups.
3112 PH/38 When log_selector = +received_sender was set, and the addition of the
3113 sender made the log line's construction buffer exactly full, or one byte
3114 less than full, an overflow happened when the terminating "\n" was
3117 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3118 when the result of a list match is failure because a DNS lookup failed.
3120 PH/40 RM_COMMAND is now used in the building process.
3122 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3123 the "build-* directories that it finds.
3125 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3126 address in a domain literal was a prefix of an interface address.
3128 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3129 when verifying a sender address, unless rewrite_headers is false.
3131 PH/44 Wrote a long comment about why errors_to addresses are verified as
3132 recipients, not senders.
3134 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3135 the ratelimit ACL was added.
3137 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3139 PH/46 Added extra information about PostgreSQL errors to the error string.
3141 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3142 avoiding the need to install special zones in a real server. This is
3143 backwards compatible; if it can't find the fake resolver, it drops back.
3144 Thus, both old and new test suites can be run.
3146 TF/02 Added util/ratelimit.pl
3148 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3149 clock is set back in time.
3151 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3152 Candler <B.Candler@pobox.com>.
3154 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3155 OK for addresses that are the result of redirections.
3157 PH/48 A number of further additions for the benefit of the new test suite,
3158 including a fake gethostbyname() that interfaces to the fake DNS resolver
3161 TF/06 The fix for widen_domains has also been applied to qualify_single and
3162 search_parents which are the other dnslookup options that can cause
3165 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3168 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3170 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3171 service" if it encountered an MX record with an empty target hostname.
3172 The message is now "an MX or SRV record indicated no SMTP service".
3174 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3175 if the DNS of the sending site is misconfigured. This is quite a
3176 common situation. This change restores the behaviour of treating a
3177 helo verification defer as a failure.
3179 PH/51 If self=fail was set on a router, the bounce message did not include the
3180 actual error message.
3186 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3188 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3189 command itself ended in error, the underlying message about the transport
3190 filter timeout was being overwritten with the pipe command error. Now the
3191 underlying error message should be appended to the second error message.
3193 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3195 PH/02 Reduce the amount of output that "make" produces by default. Full output
3196 can still be requested.
3198 PH/03 The warning log line about a condition test deferring for a "warn" verb
3199 was being output only once per connection, rather than after each
3200 occurrence (because it was using the same function as for successful
3201 "warn" verbs). This seems wrong, so I have changed it.
3203 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3204 it should not have, which might have caused a crash in the right
3205 circumstances, but probably never did.
3207 PH/04 Installed a modified version of Tony Finch's patch to make submission
3208 mode fix the return path as well as the Sender: header line, and to
3209 add a /name= option so that you can make the user's friendly name appear
3212 TF/03 Added the control = fakedefer ACL modifier.
3214 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3215 Mark Lowes for thorough testing.
3217 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3219 TK/03 Merged latest SRS patch from Miles Wilton.
3221 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3222 with the definition in sysexits.h (which is #included earlier).
3223 Fortunately, Exim does not actually use EX_OK. The code used to try to
3224 preserve the sysexits.h value, by assumimg that macro definitions were
3225 scanned for macro replacements. I have been disabused of this notion,
3226 so now the code just undefines EX_OK before #including unistd.h.
3228 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3229 in the smtp transport. When a block could not be written in a single
3230 write() function, the timeout was being re-applied to each part-write.
3231 This seems wrong - if the receiver was accepting one byte at a time it
3232 would take for ever. The timeout is now adjusted when this happens. It
3233 doesn't have to be particularly precise.
3235 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3236 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3238 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3239 without a domain if the check to <postmaster@domain> fails.
3241 SC/01 Eximstats: added -xls and the ability to specify output files
3242 (patch written by Frank Heydlauf).
3244 SC/02 Eximstats: use FileHandles for outputing results.
3246 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3248 SC/04 Eximstats: fixed display of large numbers with -nvr option
3250 SC/05 Eximstats: fixed merging of reports with empty tables.
3252 SC/06 Eximstats: added the -include_original_destination flag
3254 SC/07 Eximstats: removed tabs and trailing whitespace.
3256 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3258 TK/06 MBOX spool code: Add real "From " MBOX separator line
3259 so the .eml file is really in mbox format (even though
3260 most programs do not really care). Patch from Alex Miller.
3262 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3263 The latter is generated from $received_to and is only set if the
3264 message has one envelope recipient. SA can use these headers,
3265 obviously out-of-the-box. Patch from Alex Miller.
3267 PH/08 The ${def test on a variable was returning false if the variable's
3268 value was "0", contrary to what the specification has always said!
3269 The result should be true unless the variable is empty.
3271 PH/09 The syntax error of a character other than { following "${if
3272 def:variable_name" (after optional whitespace) was not being diagnosed.
3273 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3274 accidental colon was present, for example, could give incorrect results.
3276 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3277 result is used (not including appendfile, where other changes are about
3280 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3281 This involved changing a lot of size variables from int to off_t. It
3282 should work with maildirs and everything.
3284 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3285 spamd dying while we are connected to it.
3287 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3290 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3291 function for BDB 4.3.
3293 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3295 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3296 the sources for PCRE have been re-arranged and more files are now
3299 PH/15 The code I had for printing potentially long long variables in PH/11
3300 above was not the best (it lost precision). The length of off_t variables
3301 is now inspected at build time, and an appropriate printing format (%ld
3302 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3303 to be "long long int" or "long int". This is needed for the internal
3304 formatting function string_vformat().
3306 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3307 the configuration file to be ":syslog", then the script "guesses" where
3308 the logs files are, rather than using the compiled in default. In our
3309 case the guess is not the same as the compiled default, so the script
3310 suddenly stopped working when I started to use syslog. The patch checks
3311 to see if log_file_path is "". If so, it attempts to read it from exim
3312 with no configuration file to get the compiled in version, before it
3313 falls back to the previous guessing code."
3315 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3316 implementing BATV in an Exim configuration. See NewStuff for the gory
3319 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3320 Makefile that are specific to HP-UX.
3322 PH/18 If the "use_postmaster" option was set for a recipient callout together
3323 with the "random" option, the postmaster address was used as the MAIL
3324 FROM address for the random test, but not for the subsequent recipient
3325 test. It is now used for both.
3327 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3328 patch removes a few documentation additions to RFC 3028, because the
3329 latest draft now contains them. It adds the new en;ascii-case comparator
3330 and a new error check for 8bit text in MIME parts. Comparator and
3331 require names are now matched exactly. I enabled the subaddress
3332 extension, but it is not well tested yet (read: it works for me)."
3334 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3335 rework some of the code of TK/09 above to avoid the hardwired use of
3336 "%lld" and "long long". Replaced the call to snprintf() with a call to
3339 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3340 records point to non-existent hosts", "retry timeout exceeded", and
3341 "retry time not reached for any host after a long failure period".
3343 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3344 experimental DomainKeys support:
3346 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3347 (2) On an error such as an illegally used "control", the wrong name for
3348 the control was given.
3350 These problems did NOT occur unless DomainKeys support was compiled.
3352 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3354 PH/24 Added ${if match_ip condition.
3356 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3357 databases so that it will be absolutely obvious if a crash occurs in the
3358 DB library. This is a regular occurrence (often caused by mis-matched
3361 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3362 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3363 picked up on a user's system that detects such things. There doesn't seem
3364 to be a gcc warning option for this - only an attribute that has to be
3365 put on the function's prototype. It seems that in Fedora Core 4 they have
3366 set this on a number of new functions. No doubt there will be more in due
3369 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3370 specify a transport. However, if an address that was verified by such a
3371 router was the subject of a callout, Exim crashed because it tried to
3372 read the rcpt_include_affixes from the non-existent transport. Now it
3373 just assumes that the setting of that option is false. This bug was
3374 introduced by 4.51/PH/31.
3376 PH/28 Changed -d+all to exclude +memory, because that information is very
3377 rarely of interest, but it makes the output a lot bigger. People tend to
3378 do -d+all out of habit.
3380 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3381 code in os-type was giving problems when libc.so lives in lib64, like on
3384 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3385 aren't the modern standard, and it seems that some systems' include files
3386 don't always have them. Exim was already checking for some of the newer
3387 ones like T_AAAA, and defining it itself. I've added checks for all the
3388 record types that Exim uses.
3390 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3391 not automatically generating a new one, as it is supposed to. This
3392 prevented TLS from working. If the file did exist, but contained invalid
3393 data, a new version was generated, as expected. It was only the case of a
3394 non-existent file that was broken.
3396 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3397 with a change in libdomainkeys > 0.64.
3399 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3400 from DNS. If the selector record carries the flag, it now has
3401 precedence over the domain-wide flag.
3403 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3405 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3406 the use of a port name that isn't defined in /etc/services) occurred, the
3407 message was deferred as in a normal delivery, and thus remained on the
3408 spool, instead of being failed because of the mua_wrapper setting. This
3409 is now fixed, and I tidied up some of the mua_wrapper messages at the
3412 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3413 the messages in a hash of arrays rather than using individual hashes.
3414 This is a bit cleaner and results in dramatic memory savings, albeit
3415 at a slight CPU cost.
3417 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3418 as requested by Marc Sherman.
3420 SC/10 Eximstats: added histograms for user specified patterns as requested
3423 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3425 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3426 fopen() in the content-scanning modules that did not already have it.
3432 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3433 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3435 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3437 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3439 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3440 to test Sieve filters that use "vacation".
3442 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3443 that changes the way the GnuTLS parameters are stored in the cache file.
3444 The new format can be generated externally. For backward compatibility,
3445 if the data in the cache doesn't make sense, Exim assumes it has read an
3446 old-format file, and it generates new data and writes a new file. This
3447 means that you can't go back to an older release without removing the
3450 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3451 work if there are any delivery delays because "one_time" forces the
3452 parent to be marked "delivered", so its unseen clone is never tried
3453 again. For this reason, Exim now forbids the simultaneous setting of
3456 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3457 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3458 entry if you want to know the details. The fix, however, neglected to
3459 consider the case when local delivery batching is involved. The test for
3460 "previously delivered" was not happening when checking to see if an
3461 address could be batched with a previous (undelivered) one; under
3462 certain circumstances this could lead to multiple deliveries to the same
3465 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3466 in its include files, and this causes problems building Exim.
3468 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3469 header_syntax) but Exim was just ignoring anything given after a slash.
3470 In particular, this caused confusion with an attempt to use "verify =
3471 reverse_host_lookup/defer_ok". An error is now given when options are
3472 supplied for verify items that do not have them. (Maybe reverse_host_
3473 lookup should have a defer_ok option, but that's a different point.)
3475 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3476 defined by RFC 821) to 2048, because there were problems with some AUTH
3477 commands, and RFC 1869 says the size should be increased for extended
3478 SMTP commands that take arguments.
3480 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3483 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3484 "unknown" error; now it says that the functionality isn't in the binary.
3486 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3487 an address' error message when a string expansion fails (syntax or
3488 whatever). Otherwise the password may appear in the log. Following change
3489 PH/42 below, there is no longer a chance of it appearing in a bounce
3492 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3494 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3495 of its addresses was ever tried. (Bugzilla bug #2.)
3497 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3498 the result incorrectly in the debug output. (It correctly added a newline
3499 to what was transported.)
3501 TF/01 Added $received_time.
3503 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3504 commented out examples of how to interface to a virus scanner and to
3505 SpamAssassin. Also added commented examples of av_scanner and
3506 spamd_address settings.
3508 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3509 and controls are allowed in which ACLs. There were a couple of minor
3510 errors. Some of the entries in the conditions table (which is a table of
3511 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3512 negation of where the condition IS allowed.
3514 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3516 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3517 header file does not have a version number, so I've had to invent a new
3518 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3519 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3520 radiusclient), but it was contributed by a Radius user.
3522 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3523 files or directories, for OpenSSL.
3525 PH/19 When an Exim process that is running as root has to create an Exim log
3526 file, it does so in a subprocess that runs as exim:exim so as to get the
3527 ownership right at creation (otherwise, other Exim processes might see
3528 the file with the wrong ownership). There was no test for failure of this
3529 fork() call, which would lead to the process getting stuck as it waited
3530 for a non-existent subprocess. Forks do occasionally fail when resources
3531 run out. I reviewed all the other calls to fork(); they all seem to check
3534 PH/20 When checking for unexpected SMTP input at connect time (before writing
3535 the banner), Exim was not dealing correctly with a non-positive return
3536 from the read() function. If the client had disconnected by this time,
3537 the result was a log entry for a synchronization error with an empty
3538 string after "input=" when read() returned zero. If read() returned -1
3539 (an event I could not check), uninitialized data bytes were printed.
3540 There were reports of junk text (parts of files, etc) appearing after
3543 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3545 PH/22 Added support for macro redefinition, and (re)definition in between
3546 driver and ACL definitions.
3548 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3549 forgetting to use the resulting value; it was using the unexpanded value.
3551 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3552 hadn't been configured. The fix is from Juergen Kreileder, who
3553 understands it better than I do:
3555 "Here's what I see happening with three configured cyrus_sasl
3556 authenticators configured (plain, login, cram-md5):
3558 On startup auth_cyrus_sasl_init() gets called for each of these.
3559 This means three calls to sasl_listmech() without a specified mech_list.
3560 => SASL tests which mechs of all available mechs actually work
3561 => three warnings about OTP not working
3562 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3564 With the patch, sasl_listmech() also gets called three times. But now
3565 SASL's mech_list option is set to the server_mech specified in the the
3566 authenticator. Or in other words, the answer from sasl_listmech()
3567 gets limited to just the mech you're testing for (which is different
3569 => the return list contains just 'plain' or 'login', 'cram-md5' or
3570 nothing depending on the value of ob->server_mech.
3572 I've just tested the patch: Authentication still works fine,
3573 unavailable mechs specified in the exim configuration are still
3574 caught, and the auth.log warnings about OTP are gone."
3576 PH/25 When debugging is enabled, the contents of the command line are added
3577 to the debugging output, even when log_selector=+arguments is not
3580 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3581 answer is "GNU", and only if the return is "GNU/something" is the answer
3584 PH/27 $acl_verify_message is now set immediately after the failure of a
3585 verification in an ACL, and so is available in subsequent modifiers. In
3586 particular, the message can be preserved by coding like this:
3588 warn !verify = sender
3589 set acl_m0 = $acl_verify_message
3591 Previously, $acl_verify_message was set only while expanding "message"
3592 and "log_message" when a very denied access.
3594 PH/28 Modified OS/os.c-Linux with
3596 -#ifndef OS_LOAD_AVERAGE
3597 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3599 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3600 nomenclature these days.)
3602 PH/29 Installed patch from the Sieve maintainer that adds the options
3603 sieve_useraddress and sieve_subaddress to the redirect router.
3605 PH/30 In these circumstances:
3606 . Two addresses routed to the same list of hosts;
3607 . First host does not offer TLS;
3608 . First host accepts first address;
3609 . First host gives temporary error to second address;
3610 . Second host offers TLS and a TLS session is established;
3611 . Second host accepts second address.
3612 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3613 and peerdn, if requested) that were in fact used only for the second
3616 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3617 attention to any local part prefix or suffix that was matched by the
3618 router that accepted the address. It now behaves in the same way as it
3619 does for delivery: the affixes are removed from the local part unless
3620 rcpt_include_affixes is set on the transport.
3622 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3623 timeout during the DATA phase of an incoming message.
3625 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3626 applied a patch sanctioned by the Sieve maintainer.
3628 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3629 the uid or gid is negative. A case of a negative gid caused this to be
3630 noticed. The fix allows for either to be negative.
3632 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3633 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3636 PH/36 The change PH/12 above was broken. Fixed it.
3638 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3639 the grounds that routing the same address twice would always produce the
3640 same answer. This might have been true once, but it is certainly no
3641 longer true now. Routing a child address may depend on the previous
3642 routing that produced that child. Some complicated redirection strategies
3643 went wrong when messages had multiple recipients, and made Exim's
3644 behaviour dependent on the order in which the addresses were given.
3646 I have moved the duplicate checking until after the routing is complete.
3647 Exim scans the addresses that are assigned to local and remote
3648 transports, and removes any duplicates. This means that more work will be
3649 done, as duplicates will always all be routed, but duplicates are
3650 presumably rare, so I don't expect this is of any significance.
3652 For deliveries to pipes, files, and autoreplies, the duplicate checking
3653 still happens during the routing process, since they are not going to be
3656 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3657 It corrects a timeout issue with spamd. This is Ian's comment: "The
3658 background is that sometimes spamd either never reads data from a
3659 connection it has accepted, or it never writes response data. The exiscan
3660 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3661 blindly assumes that writes won't block so it may never time out."
3663 PH/39 Allow G after quota size as well as K and M.
3665 PH/40 The value set for $authenticated_id in an authenticator may not contain
3666 binary zeroes or newlines because the value is written to log lines and
3667 to spool files. There was no check on this. Now the value is run through
3668 the string_printing() function so that such characters are converted to
3669 printable escape sequences.
3671 PH/41 $message_linecount is a new variable that contains the total number of
3672 lines in the message. Compare $body_linecount, which is the count for the
3675 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3676 bounce and delay warning messages, except in certain special cases, which
3679 (a) An SMTP error message from a remote host;
3680 (b) A message specified in a :fail: redirection;
3681 (c) A message specified in a "fail" command in a system filter;
3682 (d) A message specified in a FAIL return from the queryprogram router;
3683 (e) A message specified by the cannot_route_message router option.
3685 In these cases only, Exim does include the error details in bounce and
3686 warning messages. There are also a few cases where bland messages such
3687 as "unrouteable address" or "local delivery error" are given.
3689 PH/43 $value is now also set for the "else" part of a ${run expansion.
3691 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3692 being worked on, but at least Exim now implements the latest version to
3695 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3696 process to complete was treated as a delivery failure, a timeout while
3697 writing the message to the pipe was logged, but erroneously treated as a
3698 successful delivery. Such timeouts include transport filter timeouts. For
3699 consistency with the overall process timeout, these timeouts are now
3700 treated as errors, giving rise to delivery failures by default. However,
3701 there is now a new Boolean option for the pipe transport called
3702 timeout_defer, which, if set TRUE, converts the failures into defers for
3703 both kinds of timeout. A transport filter timeout is now identified in
3706 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3707 systems where "make" and "gmake" are different, calling "gmake" at top
3708 level broke things. I've arranged for the value of $(MAKE) to be passed
3709 from the Makefile to this script so that it can call the same version of
3713 A note about Exim versions 4.44 and 4.50
3714 ----------------------------------------
3716 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3717 changes of various kinds. As a consequence, a big documentation update was
3718 needed. This delayed the release for rather longer than seemed good, especially
3719 in the light of a couple of (minor) security issues. Therefore, the changes
3720 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3721 release. So 4.44 and 4.50 are in effect two different branches that both start
3724 I have left the 4.50 change log unchanged; it contains all the changes since
3725 4.43. The change log for 4.44 is below; many of its items are identical to
3726 those for 4.50. This seems to be the most sensible way to preserve the
3727 historical information.
3733 1. Minor wording change to the doc/README.SIEVE file.
3735 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3736 computation of the current number of files was incorrect.
3738 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3739 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3742 4. Give more explanation in the error message when the command for a transport
3743 filter fails to execute.
3745 5. There are several places where Exim runs a non-Exim command in a
3746 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3747 was being done only for the command run by the queryprogram router. It is
3748 now done for all such subprocesses. The other cases are: ${run, transport
3749 filters, and the commands run by the lmtp and pipe transports.
3751 6. Added CONFIGURE_GROUP build-time option.
3753 7. Some older OS have a limit of 256 on the maximum number of file
3754 descriptors. Exim was using setrlimit() to set 1000 as a large value
3755 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3756 systems. I've change it so that if it can't get 1000, it tries for 256.
3758 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3759 was an oversight, and furthermore, ever since the addition of extra
3760 controls (e.g. 4.43/32), the checks on when to allow different forms of
3761 "control" were broken. There should now be diagnostics for all cases when a
3762 control that does not make sense is encountered.
3764 9. Added the /retain_sender option to "control=submission".
3766 10. $recipients is now available in the predata ACL (oversight).
3768 11. Tidy the search cache before the fork to do a delivery from a message
3769 received from the command line. Otherwise the child will trigger a lookup
3770 failure and thereby defer the delivery if it tries to use (for example) a
3771 cached ldap connection that the parent has called unbind on.
3773 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3774 of $address_data from the recipient verification was clobbered by the
3775 sender verification.
3777 13. The value of address_data from a sender verification is now available in
3778 $sender_address_data in subsequent conditions in the ACL statement.
3780 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3782 15. Added a new option "connect=<time>" to callout options, to set a different
3785 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3786 was its contents. (It was OK if the option was not defined at all.)
3788 17. A "Completed" log line is now written for messages that are removed from
3789 the spool by the -Mrm option.
3791 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3792 information about exactly what failed.
3794 19. Added -dd to debug only the daemon process.
3796 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3797 handles timeouts, both on the server side and network timeouts. Renamed the
3798 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3800 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3801 It is now set to "smtps".
3803 22. $host_address is now set to the target address during the checking of
3804 ignore_target_hosts.
3806 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3807 being passed; this would have caused $sender_host_name to have been used if
3808 matching the list had actually called for a host name (not very likely,
3809 since this list is usually IP addresses). A host name is now passed as
3812 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3813 code that specifies a non-restarting handler (typically sigaction() in
3814 modern systems) in an attempt to fix a rare and obscure crash bug.
3816 25. Narrowed the window for a race in the daemon that could cause it to ignore
3817 SIGCHLD signals. This is not a major problem, because they are used only to
3818 wake it up if nothing else does.
3820 26. A malformed maildirsize file could cause Exim to calculate negative values
3821 for the mailbox size or file count. Odd effects could occur as a result.
3822 The maildirsize information is now recalculated if the size or filecount
3825 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3826 support for a long time. Removed HAVE_SYS_VFS_H.
3828 28. Installed the latest version of exipick from John Jetmore.
3830 29. In an address list, if the pattern was not a regular expression, an empty
3831 subject address (from a bounce message) matched only if the pattern was an
3832 empty string. Non-empty patterns were not even tested. This was the wrong
3833 because it is perfectly reasonable to use an empty address as part of a
3834 database query. An empty address is now tested by patterns that are
3835 lookups. However, all the other forms of pattern expect the subject to
3836 contain a local part and a domain, and therefore, for them, an empty
3837 address still always fails if the pattern is not itself empty.
3839 30. Exim went into a mad DNS loop when attempting to do a callout where the
3840 host was specified on an smtp transport, and looking it up yielded more
3841 than one IP address.
3843 31. Re-factored the code for checking spool and log partition space into a
3844 function that finds that data and another that does the check. The former
3845 is then used to implement four new variables: $spool_space, $log_space,
3846 $spool_inodes, and $log_inodes.
3848 32. The RFC2047 encoding function was originally intended for short strings
3849 such as real names; it was not keeping to the 75-character limit for
3850 encoded words that the RFC imposes. It now respects the limit, and
3851 generates multiple encoded words if necessary. To be on the safe side, I
3852 have increased the buffer size for the ${rfc2047: expansion operator from
3855 33. It is now permitted to omit both strings after an "if" condition; if the
3856 condition is true, the result is "true". As before, when the second string
3857 is omitted, a false condition yields an empty string. This makes it less
3858 cumbersome to write custom ACL and router conditions.
3860 34. Failure to deliver a bounce message always caused it to be frozen, even if
3861 there was an errors_to setting on the router. The errors_to setting is now
3864 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3865 canonical form (fully expanded) before being placed in
3866 $sender_host_address.
3868 36. The table in the code that translates DNS record types into text (T_A to
3869 "A" for instance) was missing entries for NS and CNAME. It is just possible
3870 that this could have caused confusion if both these types were looked up
3871 for the same domain, because the text type is used as part of Exim's
3872 per-process caching. But the chance of anyone hitting this buglet seems
3875 37. The dnsdb lookup has been extended in a number of ways.
3877 (1) There is a new type, "zns", which walks up the domain tree until it
3878 finds some nameserver records. It should be used with care.
3880 (2) There is a new type, "mxh", which is like "mx" except that it returns
3881 just the host names, not the priorities.
3883 (3) It is now possible to give a list of domains (or IP addresses) to be
3884 looked up. The behaviour when one of the lookups defers can be
3885 controlled by a keyword.
3887 (4) It is now possible to specify the separator character for use when
3888 multiple records are returned.
3890 38. The dnslists ACL condition has been extended: it is now possible to supply
3891 a list of IP addresses and/or domains to be looked up in a particular DNS
3894 39. Added log_selector=+queue_time_overall.
3896 40. When running the queue in the test harness, wait just a tad after forking a
3897 delivery process, to get repeatability of debugging output.
3899 41. Include certificate and key file names in error message when GnuTLS fails
3900 to set them up, because the GnuTLS error message doesn't include the name
3901 of the failing file when there is a problem reading it.
3903 42. Allow both -bf and -bF in the same test run.
3905 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3907 44. The "Exiscan patch" is now merged into the mainline Exim source.
3909 45. Sometimes the final signoff response after QUIT could fail to get
3910 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3911 < 0 before doing a fflush(). This bug looks as though it goes back to the
3912 introduction of TLS in release 3.20, but "sometimes" must have been rare
3913 because the tests only now provoked it.
3915 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3916 (this can affect the format of dates).
3918 47. exim_tidydb, when checking for the continued existence of a message for
3919 which it has found a message-specific retry record, was not finding
3920 messages that were in split spool directories. Consequently, it was
3921 deleting retry records that should have stayed in existence.
3923 48. Steve fixed some bugs in eximstats.
3925 49. The SPA authentication driver was not abandoning authentication and moving
3926 on to the next authenticator when an expansion was forced to fail,
3927 contradicting the general specification for all authenticators. Instead it
3928 was generating a temporary error. It now behaves as specified.
3930 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3931 (the order specifies the preference for clients). The order is now AES256,
3932 AES128, 3DES, ARCFOUR128.
3934 51. Small patch to Sieve code - explicitly set From: when generating an
3937 52. Exim crashed if a remote delivery caused a very long error message to be
3938 recorded - for instance if somebody sent an entire SpamAssassin report back
3939 as a large number of 550 error lines. This bug was coincidentally fixed by
3940 increasing the size of one of Exim's internal buffers (big_buffer) that
3941 happened as part of the Exiscan merge. However, to be on the safe side, I
3942 have made the code more robust (and fixed the comments that describe what
3945 53. Now that there can be additional text after "Completed" in log lines (if
3946 the queue_time_overall log selector is set), a one-byte patch to exigrep
3947 was needed to allow it to recognize "Completed" as not the last thing in
3950 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3951 patch that reportedly fixes this has been added. I am not expert enough to
3952 create a test for it. This is what the patch creator wrote:
3954 "I found a little strange behaviour of ldap code when working with
3955 Windows 2003 AD Domain, where users was placed in more than one
3956 Organization Units. When I tried to give exim partial DN, the exit code
3957 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3958 But simultaneously result of request was absolutely normal ldap result,
3959 so I produce this patch..."
3961 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3962 REFERENCE, so I have modified the code to exclude the patch when that macro
3965 55. Some experimental protocols are using DNS PTR records for new purposes. The
3966 keys for these records are domain names, not reversed IP addresses. The
3967 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3968 leaves it alone. Component reversal etc. now happens only for IP addresses.
3971 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3973 57. Double the size of the debug message buffer (to 2048) so that more of very
3974 long debug lines gets shown.
3976 58. The exicyclog utility now does better if the number of log files to keep
3977 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3979 59. Two changes related to the smtp_active_hostname option:
3981 (1) $smtp_active_hostname is now available as a variable.
3982 (2) The default for smtp_banner uses $smtp_active_hostname instead
3983 of $primary_hostname.
3985 60. The host_aton() function is supposed to be passed a string that is known
3986 to be a valid IP address. However, in the case of IPv6 addresses, it was
3987 not checking this. This is a hostage to fortune. Exim now panics and dies
3988 if the condition is not met. A case was found where this could be provoked
3989 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3990 components; fortuitously, this particular loophole had already been fixed
3991 by change 4.50/55 above.
3993 If there are any other similar loopholes, the new check in host_aton()
3994 itself should stop them being exploited. The report I received stated that
3995 data on the command line could provoke the exploit when Exim was running as
3996 exim, but did not say which command line option was involved. All I could
3997 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3998 running as the user.
4001 61. There was a buffer overflow vulnerability in the SPA authentication code
4002 (which came originally from the Samba project). I have added a test to the
4003 spa_base64_to_bits() function which I hope fixes it.
4006 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4007 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4009 63. The daemon start-up calls getloadavg() while still root for those OS that
4010 need the first call to be done as root, but it missed one case: when
4011 deliver_queue_load_max is set with deliver_drop_privilege. This is
4012 necessary for the benefit of the queue runner, because there is no re-exec
4013 when deliver_drop_privilege is set.
4015 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4016 This has been fixed.
4018 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4019 was in use, was not putting the data itself into the right store pool;
4020 consequently, it could be overwritten for a subsequent message in the same
4021 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4024 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4026 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4027 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4028 the function were treating the return as a boolean value, which happened to
4029 work because 0=false and not-0=true, but is not correct code.
4031 68. The host_aton() function was not handling scoped IPv6 addresses (those
4032 with, for example, "%eth0" on the end) correctly.
4034 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4035 negated items (that is, ~something) in unsigned ints. Some compilers
4036 apparently mutter when there is no cast.
4038 70. If an address verification called from an ACL failed, and did not produce a
4039 user-specific message (i.e. there was only a "system" message), nothing was
4040 put in $acl_verify_message. In this situation, it now puts the system
4043 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4044 session; change 4.31/43 added the unwanted input to the log line - except
4045 that it did not do this in the start of session case. It now does.
4047 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4048 This is wrong and can cause the other end to generate a synchronization
4049 error if it is another Exim or anything else that does the synchronization
4050 check. A QUIT command is no longer sent after a timeout.
4052 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4053 during host lookups.
4055 74. The defer_ok option of callout verification was not working if it was used
4056 when verifying addresses in header lines, that is, for this case:
4058 verify = header_sender/callout=defer_ok
4060 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4061 those file descriptors could be used for SMTP connections. If anything
4062 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4063 could be sent to the SMTP client, causing chaos. The daemon now opens
4064 stdin, stdout, and stderr to /dev/null when it puts itself into the
4067 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4068 log by default. The user can override this with suitable Perl magic.
4070 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4071 the log message when discard triggers, was not working for the DATA ACL or
4072 for the non-SMTP ACL.
4074 78. Error message wording change in sieve.c.
4076 79. If smtp_accept_max_per_host was set, the number of connections could be
4077 restricted to fewer than expected, because the daemon was trying to set up
4078 a new connection before checking whether the processes handling previous
4079 connections had finished. The check for completed processes is now done
4080 earlier. On busy systems, this bug wouldn't be noticed because something
4081 else would have woken the daemon, and it would have reaped the completed
4084 80. If a message was submitted locally by a user whose login name contained one
4085 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4086 It caused a spool format error. I have fixed the spool reading code. A
4087 related problem was that the "from" clause in the Received: line became
4088 illegal because of the space(s). It is now covered by ${quote_local_part.
4090 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4093 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4095 83. Patch from Sieve maintainer to fix unterminated string problem in
4096 "vacation" handling.
4098 84. Some minor changes to the Linux configuration files to help with other
4099 OS variants using glibc.
4101 85. One more patch for Sieve to update vacation handling to latest spec.
4104 ----------------------------------------------------
4105 See the note above about the 4.44 and 4.50 releases.
4106 ----------------------------------------------------
4112 1. Change 4.43/35 introduced a bug that caused file counts to be
4113 incorrectly computed when quota_filecount was set in an appendfile
4116 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4117 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4120 3. Give more explanation in the error message when the command for a transport
4121 filter fails to execute.
4123 4. There are several places where Exim runs a non-Exim command in a
4124 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4125 was being done only for the command run by the queryprogram router. It is
4126 now done for all such subprocesses. The other cases are: ${run, transport
4127 filters, and the commands run by the lmtp and pipe transports.
4129 5. Some older OS have a limit of 256 on the maximum number of file
4130 descriptors. Exim was using setrlimit() to set 1000 as a large value
4131 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4132 systems. I've change it so that if it can't get 1000, it tries for 256.
4134 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4135 was an oversight, and furthermore, ever since the addition of extra
4136 controls (e.g. 4.43/32), the checks on when to allow different forms of
4137 "control" were broken. There should now be diagnostics for all cases when a
4138 control that does not make sense is encountered.
4140 7. $recipients is now available in the predata ACL (oversight).
4142 8. Tidy the search cache before the fork to do a delivery from a message
4143 received from the command line. Otherwise the child will trigger a lookup
4144 failure and thereby defer the delivery if it tries to use (for example) a
4145 cached ldap connection that the parent has called unbind on.
4147 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4148 of $address_data from the recipient verification was clobbered by the
4149 sender verification.
4151 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4152 was its contents. (It was OK if the option was not defined at all.)
4154 11. A "Completed" log line is now written for messages that are removed from
4155 the spool by the -Mrm option.
4157 12. $host_address is now set to the target address during the checking of
4158 ignore_target_hosts.
4160 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4161 being passed; this would have caused $sender_host_name to have been used if
4162 matching the list had actually called for a host name (not very likely,
4163 since this list is usually IP addresses). A host name is now passed as
4166 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4167 code that specifies a non-restarting handler (typically sigaction() in
4168 modern systems) in an attempt to fix a rare and obscure crash bug.
4170 15. Narrowed the window for a race in the daemon that could cause it to ignore
4171 SIGCHLD signals. This is not a major problem, because they are used only to
4172 wake it up if nothing else does.
4174 16. A malformed maildirsize file could cause Exim to calculate negative values
4175 for the mailbox size or file count. Odd effects could occur as a result.
4176 The maildirsize information is now recalculated if the size or filecount
4179 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4180 support for a long time. Removed HAVE_SYS_VFS_H.
4182 18. Updated exipick to current release from John Jetmore.
4184 19. Allow an empty sender to be matched against a lookup in an address list.
4185 Previously the only cases considered were a regular expression, or an
4188 20. Exim went into a mad DNS lookup loop when doing a callout where the
4189 host was specified on the transport, if the DNS lookup yielded more than
4192 21. The RFC2047 encoding function was originally intended for short strings
4193 such as real names; it was not keeping to the 75-character limit for
4194 encoded words that the RFC imposes. It now respects the limit, and
4195 generates multiple encoded words if necessary. To be on the safe side, I
4196 have increased the buffer size for the ${rfc2047: expansion operator from
4199 22. Failure to deliver a bounce message always caused it to be frozen, even if
4200 there was an errors_to setting on the router. The errors_to setting is now
4203 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4204 canonical form (fully expanded) before being placed in
4205 $sender_host_address.
4207 24. Updated eximstats to version 1.33
4209 25. Include certificate and key file names in error message when GnuTLS fails
4210 to set them up, because the GnuTLS error message doesn't include the name
4211 of the failing file when there is a problem reading it.
4213 26. Expand error message when OpenSSL has problems setting up cert/key files.
4216 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4217 (this can affect the format of dates).
4219 28. exim_tidydb, when checking for the continued existence of a message for
4220 which it has found a message-specific retry record, was not finding
4221 messages that were in split spool directories. Consequently, it was
4222 deleting retry records that should have stayed in existence.
4224 29. eximstats updated to version 1.35
4225 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4226 1.35 - bugfix such that pie charts by volume are generated correctly
4228 30. The SPA authentication driver was not abandoning authentication and moving
4229 on to the next authenticator when an expansion was forced to fail,
4230 contradicting the general specification for all authenticators. Instead it
4231 was generating a temporary error. It now behaves as specified.
4233 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4234 (the order specifies the preference for clients). The order is now AES256,
4235 AES128, 3DES, ARCFOUR128.
4237 31. Small patch to Sieve code - explicitly set From: when generating an
4240 32. Exim crashed if a remote delivery caused a very long error message to be
4241 recorded - for instance if somebody sent an entire SpamAssassin report back
4242 as a large number of 550 error lines. This bug was coincidentally fixed by
4243 increasing the size of one of Exim's internal buffers (big_buffer) that
4244 happened as part of the Exiscan merge. However, to be on the safe side, I
4245 have made the code more robust (and fixed the comments that describe what
4248 33. Some experimental protocols are using DNS PTR records for new purposes. The
4249 keys for these records are domain names, not reversed IP addresses. The
4250 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4251 leaves it alone. Component reversal etc. now happens only for IP addresses.
4254 34. The host_aton() function is supposed to be passed a string that is known
4255 to be a valid IP address. However, in the case of IPv6 addresses, it was
4256 not checking this. This is a hostage to fortune. Exim now panics and dies
4257 if the condition is not met. A case was found where this could be provoked
4258 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4259 components; fortuitously, this particular loophole had already been fixed
4260 by change 4.50/55 or 4.44/33 above.
4262 If there are any other similar loopholes, the new check in host_aton()
4263 itself should stop them being exploited. The report I received stated that
4264 data on the command line could provoke the exploit when Exim was running as
4265 exim, but did not say which command line option was involved. All I could
4266 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4267 running as the user.
4270 35. There was a buffer overflow vulnerability in the SPA authentication code
4271 (which came originally from the Samba project). I have added a test to the
4272 spa_base64_to_bits() function which I hope fixes it.
4275 36. The daemon start-up calls getloadavg() while still root for those OS that
4276 need the first call to be done as root, but it missed one case: when
4277 deliver_queue_load_max is set with deliver_drop_privilege. This is
4278 necessary for the benefit of the queue runner, because there is no re-exec
4279 when deliver_drop_privilege is set.
4281 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4282 was in use, was not putting the data itself into the right store pool;
4283 consequently, it could be overwritten for a subsequent message in the same
4284 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4287 38. Sometimes the final signoff response after QUIT could fail to get
4288 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4289 < 0 before doing a fflush(). This bug looks as though it goes back to the
4290 introduction of TLS in release 3.20, but "sometimes" must have been rare
4291 because the tests only now provoked it.
4297 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4298 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4299 Now it returns an integer. A number of calls were still expecting a T/F
4300 return. Fortuitously, in all cases, the tests worked in OK situations,
4301 which is the norm. However, things would have gone wrong on any write
4302 failures on the smtp file descriptor. This function is used when sending
4303 messages over SMTP and also when doing verify callouts.
4305 2. When Exim is called to do synchronous delivery of a locally submitted
4306 message (the -odf or -odi options), it no longer closes stderr before doing
4309 3. Implemented the mua_wrapper option.
4311 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4313 5. Implemented the functions header_remove(), header_testname(),
4314 header_add_at_position(), and receive_remove_recipient(), and exported them
4317 6. If an ACL "warn" statement specified the addition of headers, Exim already
4318 inserted X-ACL-Warn: at the start if there was no header name. However, it
4319 was not making this test for the second and subsequent header lines if
4320 there were newlines in the string. This meant that an invalid header could
4321 be inserted if Exim was badly configured.
4323 7. Allow an ACL "warn" statement to add header lines at the start or after all
4324 the Received: headers, as well as at the end.
4326 8. Added the rcpt_4xx retry error code.
4328 9. Added postmaster_mailfrom=xxx to callout verification option.
4330 10. Added mailfrom=xxxx to the callout verification option, for verify=
4333 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4334 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4336 12. Inserted some casts to stop certain compilers warning when using pointer
4337 differences as field lengths or precisions in printf-type calls (mostly
4338 affecting debugging statements).
4340 13. Added optional readline() support for -be (dynamically loaded).
4342 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4343 same clock tick as a message's arrival, so that its received time was the
4344 same as the "first fail" time on the retry record, and that message
4345 remained on the queue past the ultimate address timeout, every queue runner
4346 would try a delivery (because it was past the ultimate address timeout) but
4347 after another failure, the ultimate address timeout, which should have then
4348 bounced the address, did not kick in. This was a "< instead of <=" error;
4349 in most cases the first failure would have been in the next clock tick
4350 after the received time, and all would be well.
4352 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4353 being recognized when the domain list was tested by the match_domain
4354 condition in an expansion string.
4356 16. Added the ${str2b64: operator.
4358 17. Exim was always calling setrlimit() to set a large limit for the number of
4359 processes, without checking whether the existing limit was already
4360 adequate. (It did check for the limit on file descriptors.) Furthermore,
4361 errors from getrlimit() and setrlimit() were being ignored. Now they are
4362 logged to the main and panic logs, but Exim does carry on, to try to do its
4363 job under whatever limits there are.
4365 18. Imported PCRE 5.0.
4367 19. Trivial typo in log message " temporarily refused connection" (the leading
4370 20. If the log selector return_path_on_delivery was set and an address was
4371 redirected to /dev/null, the delivery process crashed because it assumed
4372 that a return path would always be set for a "successful" delivery. In this
4373 case, the whole delivery is bypassed as an optimization, and therefore no
4376 21. Internal re-arrangement: the function for sending a challenge and reading
4377 a response while authentication was assuming a zero-terminated challenge
4378 string. It's now changed to take a pointer and a length, to allow for
4379 binary data in such strings.
4381 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4383 23. Exim was not respecting finduser_retries when seeking the login of the
4384 uid under which it was called; it was always trying 10 times. (The default
4385 setting of finduser_retries is zero.) Also, it was sleeping after the final
4386 failure, which is pointless.
4388 24. Implemented tls_on_connect_ports.
4390 25. Implemented acl_smtp_predata.
4392 26. If the domain in control=submission is set empty, Exim assumes that the
4393 authenticated id is a complete email address when it generates From: or
4394 Sender: header lines.
4396 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4397 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4398 chown and chgrp in /bin and hostname in /usr/bin.
4400 28. Exim was keeping the "process log" file open after each use, just as it
4401 does for the main log. This opens the possibility of it remaining open for
4402 long periods when the USR1 signal hits a daemon. Occasional processlog
4403 errors were reported, that could have been caused by this. Anyway, it seems
4404 much more sensible not to leave this file open at all, so that is what now
4407 29. The long-running daemon process does not normally write to the log once it
4408 has entered its main loop, and it closes the log before doing so. This is
4409 so that log files can straightforwardly be renamed and moved. However,
4410 there are a couple of unusual error situations where the daemon does write
4411 log entries, and I had neglected to close the log afterwards.
4413 30. The text of an SMTP error response that was received during a remote
4414 delivery was being truncated at 512 bytes. This is too short for some of
4415 the long messages that one sometimes sees. I've increased the limit to
4418 31. It is now possible to make retry rules that apply only when a message has a
4419 specific sender, in particular, an empty sender.
4421 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4422 it possible to be selective about when SMTP synchronization is enforced.
4424 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4426 32. Implemented hosts_connection_nolog.
4428 33. Added an ACL for QUIT.
4430 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4433 35. Added mailbox_size and mailbox_filecount to appendfile.
4435 36. Added control = no_multiline_responses to ACLs.
4437 37. There was a bug in the logic of the code that waits for the clock to tick
4438 in the case where the clock went backwards by a substantial amount such
4439 that the microsecond fraction of "now" was more than the microsecond
4440 fraction of "then" (but the whole seconds number was less).
4442 38. Added support for the libradius Radius client library this is found on
4443 FreeBSD (previously only the radiusclient library was supported).
4449 1. When certain lookups returned multiple values in the form name=value, the
4450 quoting of the values was not always being done properly. Specifically:
4451 (a) If the value started with a double quote, but contained no whitespace,
4453 (b) If the value contained whitespace other than a space character (i.e.
4454 tabs or newlines or carriage returns) it was not quoted.
4455 This fix has been applied to the mysql and pgsql lookups by writing a
4456 separate quoting function and calling it from the lookup code. The fix
4457 should probably also be applied to nisplus, ibase and oracle lookups, but
4458 since I cannot test any of those, I have not disturbed their existing code.
4460 2. A hit in the callout cache for a specific address caused a log line with no
4461 reason for rejecting RCPT. Now it says "Previous (cached) callout
4462 verification failure".
4464 3. There was an off-by-one bug in the queryprogram router. An over-long
4465 return line was truncated at 256 instead of 255 characters, thereby
4466 overflowing its buffer with the terminating zero. As well as fixing this, I
4467 have increased the buffer size to 1024 (and made a note to document this).
4469 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4470 when Exim is waiting for an SMTP response from a remote server, Exim
4471 restarts its select() call on the socket, thereby resetting its timeout.
4472 This is not a problem when such interrupts are rare. Somebody set up a cron
4473 job to run exiwhat every 2 minutes, which is less than the normal select()
4474 timeout (5 or 10 minutes). This meant that the select() timeout never
4475 kicked in because it was always reset. I have fixed this by comparing the
4476 time when an interrupt arrives with the time at the start of the first call
4477 to select(). If more time than the timeout has elapsed, the interrupt is
4478 treated as a timeout.
4480 5. Some internal re-factoring in preparation for the addition of Sieve
4481 extensions (by MH). In particular, the "personal" test is moved to a
4482 separate function, and given an option for scanning Cc: and Bcc: (which is
4483 not set for Exim filters).
4485 6. When Exim created an email address using the login of the caller as the
4486 local part (e.g. when creating a From: or Sender: header line), it was not
4487 quoting the local part when it contained special characters such as @.
4489 7. Installed new OpenBSD configuration files.
4491 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4492 try to make them clearer.
4494 9. Callout options, other than the timeout value, were being ignored when
4495 verifying sender addresses in header lines. For example, when using
4497 verify = header_sender/callout=no_cache
4499 the cache was (incorrectly) being used.
4501 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4502 only the Cygwin environment.
4504 11. When return_path_on_delivery was set as a log selector, if different remote
4505 addresses in the same message used different return paths and parallel
4506 remote delivery occurred, the wrong values would sometimes be logged.
4507 (Whenever a remote delivery process finished, the return path value from
4508 the most recently started remote delivery process was logged.)
4510 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4511 lines when AUTH and/or TLS are in use. This is the "received protocol"
4512 field. Exim used to use "asmtp" for authenticated SMTP, without any
4513 indication (in the protocol name) for TLS use. Now it follows the RFC and
4514 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4515 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4516 names appear in log lines as well as in Received: header lines.
4518 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4519 extensions, and comparison tests, and to fix some bugs.
4521 14. Changes to the "personal" filter test:
4523 (1) The test was buggy in that it was just doing the equivalent of
4524 "contains" tests on header lines. For example, if a user's address was
4525 anne@some.where, the "personal" test would incorrectly be true for
4527 To: susanne@some.where
4529 This test is now done by extracting each address from the header in turn,
4530 and checking the entire address. Other tests that are part of "personal"
4531 are now done using regular expressions (for example, to check local parts
4532 of addresses in From: header lines).
4534 (2) The list of non-personal local parts in From: addresses has been
4535 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4536 taken from the Sieve specification recommendations.
4538 (3) If the message contains any header line starting with "List-" it is
4539 treated as non-personal.
4541 (4) The test for "circular" in the Subject: header line has been removed
4542 because it now seems ill-conceived.
4544 15. Minor typos in src/EDITME comments corrected.
4546 16. Installed latest exipick from John Jetmore.
4548 17. If headers_add on a router specified a text string that was too long for
4549 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4550 of string_sprintf() is now avoided.
4552 18. $message_body_size was not set (it was always zero) when running the DATA
4553 ACL and the local_scan() function.
4555 19. For the "mail" command in an Exim filter, no default was being set for
4556 the once_repeat time, causing a random time value to be used if "once" was
4557 specified. (If the value happened to be <= 0, no repeat happened.) The
4558 default is now 0s, meaning "never repeat". The "vacation" command was OK
4559 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4560 (I found it when inspecting the code).
4562 20. There is now an overall timeout for performing a callout verification. It
4563 defaults to 4 times the callout timeout, which applies to individual SMTP
4564 commands during the callout. The overall timeout applies when there is more
4565 than one host that can be tried. The timeout is checked before trying the
4566 next host. This prevents very long delays if there are a large number of
4567 hosts and all are timing out (e.g. when the network connections are timing
4568 out). The value of the overall timeout can be changed by specifying an
4569 additional sub-option for "callout", called "maxwait". For example:
4571 verify = sender/callout=5s,maxwait=20s
4573 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4574 to the end before writing, but this should make it even safer).
4576 22. Exim was forgetting that it had advertised PIPELINING for the second and
4577 subsequent messages on an SMTP connection. It was also not resetting its
4578 memory on STARTTLS and an internal HELO.
4580 23. When Exim logs an SMTP synchronization error within a session, it now
4581 records whether PIPELINING has been advertised or not.
4583 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4584 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4585 rather than long int.
4587 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4589 26. Added the never_mail option to autoreply.
4595 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4596 crash if the getsockname() call failed; this can happen if a connection is
4597 closed very soon after it is established. The problem was simply in the
4598 order in which certain operations were done, causing Exim to try to write
4599 to the SMTP stream before it had set up the file descriptor. The bug has
4600 been fixed by making things happen in the correct order.
4606 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4607 before the connection was closed, thus losing the rejection response.
4609 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4610 some early Solaris releases, but causes trouble in current releases where
4611 socklen_t is defined.
4613 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4616 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4619 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4620 that releasing the top of it at the end releases what was used for sub-
4621 expansions (unless the block got too big). However, discard this block if
4622 the first thing is a variable or header, so that we can use its block when
4623 it is dynamic (useful for very large $message_headers, for example).
4625 6. Lookups now cache *every* query, not just the most recent. A new, separate
4626 store pool is used for this. It can be recovered when all lookup caches are
4627 flushed. Lookups now release memory at the end of their result strings.
4628 This has involved some general refactoring of the lookup sources.
4630 7. Some code has been added to the store_xxx() functions to reduce the amount
4631 of flapping under certain conditions.
4633 8. log_incoming_interface used to affect only the <= reception log lines. Now
4634 it causes the local interface and port to be added to several more SMTP log
4635 lines, for example "SMTP connection from", and rejection lines.
4637 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4639 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4641 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4642 could be overwritten at the end of the current message (or the start of a
4643 new message if it was set in a HELO ACL). The value is now preserved for
4644 the duration of the SMTP connection.
4646 12. If a transport had a headers_rewrite setting, and a matching header line
4647 contained an unqualified address, that address was qualified, even if it
4648 did not match any rewriting rules. The underlying bug was that the values
4649 of the flags that permit the existence of unqualified sender and recipient
4650 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4651 non-local messages, and by -bnq for local messages) were not being
4652 preserved with the message after it was received.
4654 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4655 "next input=" as part of the text comprising the host identity instead of
4656 the correct text. The code was using the same buffer for two different
4657 strings. However, depending on which order the printing function evaluated
4658 its arguments, the bug did not always show up. Under Linux, for example, my
4659 test suite worked just fine.
4661 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4662 doesn't work with some older versions of Perl. It has been changed to "my",
4663 which in any case is probably the better facility to use.
4665 15. A really picky compiler found some instances of statements for creating
4666 error messages that either had too many or two few arguments for the format
4669 16. The size of the buffer for calls to the DNS resolver has been increased
4670 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4671 for addresses that have a lot of PTR records. This alleviates a problem; it
4672 does not fully solve it.
4674 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4675 buffer now truncates the list and logs the incident, which is the same
4676 action as happens when Exim is looking up a host name and its aliases.
4677 Previously in this situation something unpredictable would happen;
4678 sometimes it was "internal error: store_reset failed".
4680 18. If a server dropped the connection unexpectedly when an Exim client was
4681 using GnuTLS and trying to read a response, the client delivery process
4682 crashed while trying to generate an error log message.
4684 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4685 string, for example:
4687 warn message = H1: something\nH2: something
4689 the text was added as a single header line from Exim's point of view
4690 though it ended up OK in the delivered message. However, searching for the
4691 second and subsequent header lines using $h_h2: did not work. This has been
4692 fixed. Similarly, if a system filter added multiple headers in this way,
4693 the routers could not see them.
4695 20. Expanded the error message when iplsearch is called with an invalid key to
4696 suggest using net-iplsearch in a host list.
4698 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4699 ACLs are no longer actually imposed (and a message to that effect is
4702 22. If a "gecos" field in a passwd entry contained escaped characters, in
4703 particular, if it contained a \" sequence, Exim got it wrong when building
4704 a From: or a Sender: header from that name. A second bug also caused
4705 incorrect handling when an unquoted " was present following a character
4706 that needed quoting.
4708 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4709 was not being matched caselessly.
4711 24. Arranged for all hyphens in the exim.8 source to be escaped with
4714 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4715 properly. Recipient callout cache records were still being keyed to include
4716 the sender, even when use_sender was set false. This led to far more
4717 callouts that were necessary. The sender is no longer included in the key
4718 when use_sender is false.
4720 26. Added "control = submission" modifier to ACLs.
4722 27. Added the ${base62d: operator to decode base 62 numbers.
4724 28. dnsdb lookups can now access SRV records.
4726 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4727 the configuration file.
4729 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4730 (-v) mode. This makes the output for a verbose queue run more intelligible.
4732 31. Added a use_postmaster feature to recipient callouts.
4734 32. Added the $body_zerocount variable, containing the number of binary zero
4735 bytes in the message body.
4737 33. The time of last modification of the "new" subdirectory is now used as the
4738 "mailbox time last read" when there is a quota error for a maildir
4741 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4743 35. Added +ignore_unknown as a special item in host lists.
4745 36. Code for decoding IPv6 addresses in host lists is now included, even if
4746 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4747 address was recognized as an IP address, but was then not correctly decoded
4748 into binary, causing unexpected and incorrect effects when compared with
4755 1. Very minor rewording of debugging text in manualroute to say "list of
4756 hosts" instead of "hostlist".
4758 2. If verify=header_syntax was set, and a header line with an unqualified
4759 address (no domain) and a large number of spaces between the end of the
4760 name and the colon was received, the reception process suffered a buffer
4761 overflow, and (when I tested it) crashed. This was caused by some obsolete
4762 code that should have been removed. The fix is to remove it!
4764 3. When running in the test harness, delay a bit after writing a bounce
4765 message to get a bit more predictability in the log output.
4767 4. Added a call to search_tidyup() just before forking a reception process. In
4768 theory, someone could use a lookup in the expansion of smtp_accept_max_
4769 per_host which, without the tidyup, could leave open a database connection.
4771 5. Added the variables $recipient_data and $sender_data which get set from a
4772 lookup success in an ACL "recipients" or "senders" condition, or a router
4773 "senders" option, similar to $domain_data and $local_part_data.
4775 6. Moved the writing of debug_print from before to after the "senders" test
4778 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4779 problems for message scanning, either using a data ACL, or using
4780 local_scan() because the Received: header was not generated till after they
4781 were called (in order to set the time as the time of reception completion).
4782 I have revised the way this works. The header is now generated after the
4783 body is received, but before the ACL or local_scan() are called. After they
4784 are run, the timestamp in the header is updated.
4790 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4791 before starting a queue runner without re-exec. This happened only when
4792 deliver_drop_privilege was set or when the Exim user was set to root. The
4793 effect of the bug was that timeouts during subsequent deliveries caused
4794 crashes instead of being properly handled. The handler is now left at its
4795 default (and expected) setting.
4797 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4798 message, again when deliver_drop_privilege is set or Exim is run as root.
4799 The bug described in (1) was not present in this case, but the tidying up
4800 of the other signals was missing. I have made the two cases consistent.
4802 3. The ignore_target_hosts setting on a manualroute router was being ignored
4803 for hosts that were looked up using the /MX notation.
4805 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4808 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4809 operated on the sender address. After changing the $sender_address to <>
4810 for the sender address verify, Exim was re-instated it as the original
4811 (before rewriting) address, but remembering that it had rewritten it, so it
4812 wasn't rewriting it again. This bug also had the effect of breaking the
4813 sender address verification caching when the sender address was rewritten.
4815 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4816 This has been changed so that if the ip literal address matches
4817 ignore_target_hosts, the router declines.
4819 7. Added expansion conditions match_domain, match_address, and match_local_
4820 part (NOT match_host).
4822 8. The placeholder for the Received: header didn't have a length field set.
4824 9. Added code to Exim itself and to exim_lock to test for a specific race
4825 condition that could lead to file corruption when using MBX delivery. The
4826 issue is with the lockfile that is created in /tmp. If this file is removed
4827 after a process has opened it but before that process has acquired a lock,
4828 there is the potential for a second process to recreate the file and also
4829 acquire a lock. This could lead to two Exim processes writing to the file
4830 at the same time. The added code performs the same test as UW imapd; it
4831 checks after acquiring the lock that its file descriptor still refers to
4832 the same named file.
4834 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4835 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4838 11. Added the smtp_active_hostname option. If used, this will typically be made
4839 to depend on the incoming interface address. Because $interface_address is
4840 not set up until the daemon has forked a reception process, error responses
4841 that can happen earlier (such as "too many connections") no longer contain
4844 12. If an expansion in a condition on a "warn" statement fails because a lookup
4845 defers, the "warn" statement is abandoned, and the next ACL statement is
4846 processed. Previously this caused the whole ACL to be aborted.
4848 13. Added the iplsearch lookup type.
4850 14. Added ident_timeout as a log selector.
4852 15. Added tls_certificate_verified as a log selector.
4854 16. Added a global option tls_require_ciphers (compare the smtp transport
4855 option of the same name). This controls incoming TLS connections.
4857 17. I finally figured out how to make tls_require_ciphers do a similar thing
4858 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4859 before starting the TLS session.
4861 18. Tabs are now shown as \t in -bP output.
4863 19. If the log selector return_path_on_delivery was set, Exim crashed when
4864 bouncing a message because it had too many Received: header lines.
4866 20. If two routers both had headers_remove settings, and the first one included
4867 a superfluous trailing colon, the final name in the first list and the
4868 first name in the second list were incorrectly joined into one item (with a
4869 colon in the middle).
4875 1. Added -C and -D options to the exinext utility, mainly to make it easier
4876 to include in the automated testing, but these could be helpful when
4877 multiple configurations are in use.
4879 2. The exinext utility was not formatting the output nicely when there was
4880 an alternate port involved in the retry record key, nor when there was a
4881 message id as well (for retries that were specific to a specific message
4882 and a specific host). It was also confused by IPv6 addresses, because of
4883 the additional colons they contain. I have fixed the IPv4 problem, and
4884 patched it up to do a reasonable job for IPv6.
4886 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4887 delivery, the log line now contains "pipelined" if PIPELINING was used.
4889 4. An SMTP transport process used to panic and die if the bind() call to set
4890 an explicit outgoing interface failed. This has been changed; it is now
4891 treated in the same way as a connect() failure.
4893 5. A reference to $sender_host_name in the part of a conditional expansion
4894 that was being skipped was still causing a DNS lookup. This no longer
4897 6. The def: expansion condition was not recognizing references to header lines
4898 that used bh_ and bheader_.
4900 7. Added the _cache feature to named lists.
4902 8. The code for checking quota_filecount in the appendfile transport was
4903 allowing one more file than it should have been.
4905 9. For compatibility with Sendmail, the command line option
4913 and sets the incoming protocol and host name (for trusted callers). The
4914 host name and its colon can be omitted when only the protocol is to be set.
4915 Note the Exim already has two private options, -pd and -ps, that refer to
4916 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4917 "s", but I don't think that's a major issue.
4919 10. A number of refactoring changes to the code, none of which should affect
4922 (a) The number of logging options was getting close to filling up the
4923 32-bit word that was used as a bit map. I have split them into two classes:
4924 those that are passed in the argument to log_write(), and those that are
4925 only ever tested independently outside of that function. These are now in
4926 separate 32-bit words, so there is plenty of room for expansion again.
4927 There is no change in the user interface or the logging behaviour.
4929 (b) When building, for example, log lines, the code previously used a
4930 macro that called string_cat() twice, in order to add two strings. This is
4931 not really sufficiently general. Furthermore, there was one instance where
4932 it was actually wrong because one of the argument was used twice, and in
4933 one call a function was used. (As it happened, calling the function twice
4934 did not affect the overall behaviour.) The macro has been replaced by a
4935 function that can join an arbitrary number of extra strings onto a growing
4938 (c) The code for expansion conditions now uses a table and a binary chop
4939 instead of a serial search (which was left over from when there were very
4940 few conditions). Also, it now recognizes conditions like "pam" even when
4941 the relevant support is not compiled in: a suitably worded error message is
4942 given if an attempt is made to use such a condition.
4944 11. Added ${time_interval:xxxxx}.
4946 12. A bug was causing one of the ddress fields not to be passed back correctly
4947 from remote delivery subprocesses. The field in question was not being
4948 subsequently used, so this caused to problems in practice.
4950 13. Added new log selectors queue_time and deliver_time.
4952 14. Might have fixed a bug in maildirsizefile handling that threw up
4953 "unexpected character" debug warnings, and recalculated the data
4954 unnecessarily. In any case, I expanded the warning message to give more
4957 15. Added the message "Restricted characters in address" to the statements in
4958 the default ACL that block characters like @ and % in local parts.
4960 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4961 Three changes have been made:
4963 (a) There was a serious bug; a negative response to MAIL caused the whole
4964 recipient domain to be cached as invalid, thereby blocking all messages
4965 to all local parts at the same domain, from all senders. This bug has
4966 been fixed. The domain is no longer cached after a negative response to
4967 MAIL if the sender used is not empty.
4969 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4972 (c) A new callout option, "use_sender" has been added for people who want
4973 the modified behaviour.
4979 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4982 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4983 indeed breaks things for older releases.
4985 3. Added additional logging to the case where there is a problem reading data
4986 from a filter that is running in a subprocess using a pipe, in order to
4987 try to track down a specific problem.
4989 4. Testing facility fudge: when running in the test harness and attempting
4990 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4991 getting "No route to host". Convert this to a timeout.
4993 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4996 6. Some OS don't have socklen_t but use size_t instead. This affects the
4997 fifth argument of getsockopt() amongst other things. This is now
4998 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4999 can be set for individual OS. I have set it for SunOS5, OSF1, and
5000 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5001 some earlier ones do not.
5003 7. Change 4.30/15 was not doing the test caselessly.
5005 8. The standard form for an IPv6 address literal was being rejected by address
5006 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5007 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5008 this, as well as the form without the "IPv6" on the front (but only when
5009 address literals are enabled, of course).
5011 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5013 10. Exim crashed if a message with an empty sender address specified by -f
5014 encountered a router with an errors_to setting. This could be provoked only
5015 by a command such as
5019 where an empty string was supplied; "<>" did not hit this bug.
5021 11. Installed PCRE release 4.5.
5023 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5024 remained set. It is now erased.
5026 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5027 times from message ids (which are base 36 rather than the normal 62).
5029 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5030 were being counted as actual protocol errors, and logged if the log
5031 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5032 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5033 and DATA following a set of rejected RCPTs do not count as protocol errors.
5034 In other words, Exim assumes they were pipelined, though this may not
5035 actually be the case. Of course, in all cases the client gets an
5036 appropriate error code.
5038 15. If a lookup fails in an ACL condition, a message about the failure may
5039 be available; it is used if testing the ACL cannot continue, because most
5040 such messages specify what the cause of the deferral is. However, some
5041 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5042 that caused an old message to be retained and used if a later statement
5043 caused a defer, replacing the real cause of the deferral.
5045 16. If an IP address had so many PTR records that the DNS lookup buffer
5046 was not large enough to hold them, Exim could crash while trying to process
5047 the truncated data. It now detects and logs this case.
5049 17. Further to 4.21/58, another change has been made: if (and only if) the
5050 first line of a message (the first header line) ends with CRLF, a bare LF
5051 in a subsequent header line has a space inserted after it, so as not to
5052 terminate the header.
5054 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5055 unnecessarily, used atoi() instead of strtol(), and didn't check the
5056 termination when getting file sizes from file names by regex.
5058 19. Completely re-implemented the support for maildirsize files, in the light
5059 of a number of problems with the previous contributed implementation
5060 (4.30/29). In particular:
5062 . If the quota is zero, the maildirsize file is maintained, but no quota is
5065 . If the maildir directory does not exist, it is created before any attempt
5066 to write a maildirsize file.
5068 . The quota value in the file is just a cache; if the quota is changed in
5069 the transport, the new value overrides.
5071 . A regular expression is available for excluding directories from the
5074 20. The autoreply transport checks the characters in options that define the
5075 message's headers; it allows continued headers, but it was checking with
5076 isspace() after an embedded newline instead of explicitly looking for a
5079 21. If all the "regular" hosts to which an address was routed had passed their
5080 expiry times, and had not reached their retry times, the address was
5081 bounced, even if fallback hosts were defined. Now Exim should go on to try
5084 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5085 equivalent code in the SMTP transport. Some hosts send humungous responses
5086 to HELO/EHLO, more than 1024 it seems.
5088 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5089 gives compiler warnings in some environments. I've now done it "properly",
5092 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5093 (because of problems with the built-in one) was declared to return uschar *
5094 instead of char *, causing compiler failure.
5096 25. Fixed a file descriptor leak when processing alias/forward files.
5098 26. Fixed a minor format string issue in dbfn.c.
5100 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5102 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5103 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5106 29. When a sender address is verified, it is cached, to save repeating the test
5107 when there is more than one recipient in a message. However, when the
5108 verification involves a callout, it is possible for different callout
5109 options to be set for different recipients. It is too complicated to keep
5110 track of this in the cache, so now Exim always runs a verification when a
5111 callout is required, relying on the callout cache for the optimization.
5112 The overhead is duplication of the address routing, but this should not be
5115 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5116 to be verified with callout=postmaster, and the main callout worked but the
5117 postmaster check failed, the verification correctly failed. However, if a
5118 subsequent RCPT command asked for sender verification *without* the
5119 postmaster check, incorrect caching caused this verification also to fail,
5122 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5123 it was not caching the DNS options (qualify_single, search_parents) that
5124 were used when the lookup failed. A subsequent lookup with different
5125 options therefore always gave the same answer, though there were cases
5126 where it should not have. (Example: a "domains = !$mx_any" option on a
5127 dnslookup router: the "domains" option is always processed without any
5128 widening, but the router might have qualify_single set.) Now Exim uses the
5129 cached value only when the same options are set.
5131 32. Added John Jetmore's "exipick" utility to the distribution.
5133 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5134 than a timeout (e.g. a certificate is required, and is not provided), an
5135 Exim server now closes the connection immediately. Previously it waited for
5136 the client to close - but if the client is SSL, it seems that they each
5137 wait for each other, leading to a delay before one of them times out.
5139 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5140 maintained 0.8.x compatibility because I don't think many are using it, and
5141 it is clearly obsolete.
5143 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5146 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5147 client certificate was expired. A simple patch fixes this, though I don't
5148 understand the full logic of why the verify callback is called multiple
5151 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5152 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5153 which causes problems with some clients (such as the Certicom SSL Plus
5154 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5155 disables the coutermeasure allowing Eudora to connect."
5157 38. Exim was not checking that a write() to a log file succeeded. This could
5158 lead to Bad Things if a log got too big, in particular if it hit a file
5159 size limit. Exim now panics and dies if it cannot write to a log file, just
5160 as it does if it cannot open a log file.
5162 39. Modified OS/Makefile-Linux so that it now contains
5164 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5166 The two -D definitions ensure that Exim is compiled with large file
5167 support, which makes it possible to handle log files that are bigger than
5170 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5171 instance) a domain was checked against a named list that involved a lookup,
5172 causing $domain_data to be set, then another domain was checked against the
5173 same list, then the first domain was re-checked, the value of $domain_data
5174 after the final check could be wrong. In particular, if the second check
5175 failed, it could be set empty. This bug probably also applied to
5178 41. The strip_trailing_dot option was not being applied to the address given
5179 with the -f command-line option.
5181 42. The code for reading a message's header from the spool was incrementing
5182 $received_count, but never initializing it. This meant that the value was
5183 incorrect (doubled) while delivering a message in the same process in which
5184 it was received. In the most common configuration of Exim, this never
5185 happens - a fresh exec is done - but it can happen when
5186 deliver_drop_privilege is set.
5188 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5189 it now includes up to 150 characters of the unexpected data in the log
5192 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5193 and building data strings. The size of both of these buffers was 10 000
5194 bytes - far larger than anybody would *ever* want, thought I. Needless to
5195 say, somebody hit the limit. I have increased the maximum line length to
5196 20 000 and the maximum data length of concatenated lines to 100 000. I have
5197 also fixed two bugs, because there was no checking on these buffers. Tsk,
5198 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5199 buffer is too small.
5201 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5202 lsearch lookups. Now it does.
5204 46. When parsing a route_list item in a manualroute router, a fixed-length
5205 buffer was used for the list of hosts. I made this 1024 bytes long,
5206 thinking that nobody would ever have a list of hosts that long. Wrong.
5207 Somebody had a whole pile of complicated expansion conditions, and the
5208 string was silently truncated, leading to an expansion error. It turns out
5209 that it is easier to change to an unlimited length (owing to other changes
5210 that have happened since this code was originally written) than to build
5211 structure for giving a limitation error. The length of the item that
5212 expands into the list of hosts is now unlimited.
5214 47. The lsearch lookup could not handle data where the length of text line was
5215 more than 4095 characters. Such lines were truncated, leading to shortened
5216 data being returned. It should now handle lines of any length.
5218 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5219 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5222 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5223 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5224 no longer happens in comment lines. A list of replacements is now placed
5225 at the head of all of the source files, except those whose only change is
5226 to replace PERL_COMMAND in the very first #! line.
5228 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5229 messages on the queue, with a bottom-up merge sort, using code contributed
5230 by Michael Haardt. This should make operations like -bp somewhat faster on
5231 large queues. It won't affect queue runners, except when queue_run_in_order
5234 51. Installed eximstats 1.31 in the distribution.
5236 52. Added support for SRV lookups to the dnslookup router.
5238 53. If an ACL referred to $message_body or $message_body_end, the value was not
5239 reset for any messages that followed in the same SMTP session.
5241 54. The store-handling optimization for building very long strings was not
5242 differentiating between the different store pools. I don't think this
5243 actually made any difference in practice, but I've tidied it.
5245 55. While running the routers to verify a sender address, $sender_address
5246 was still set to the sender address. This is wrong, because when routing to
5247 send a bounce to the sender, it would be empty. Therefore, I have changed
5248 it so that, while verifying a sender address, $sender_address is set to <>.
5249 (There is no change to what happens when verifying a recipient address.)
5251 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5252 target A or AAAA records (if not already returned) without resetting the
5253 qualify_single or search_parents options of the DNS resolver. These are
5254 inappropriate in this case because the targets of MX and SRV records must
5255 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5256 target that, when qualified, matched something in the local domain. These
5257 two options are now turned off when doing these lookups.
5259 57. It seems that at least some releases of Reiserfs (which does not have the
5260 concept of a fixed number of inodes) returns zero and not -1 for the
5261 number of available inodes. This interacted badly with check_spool_inodes,
5262 which assumed that -1 was the "no such thing" setting. What I have done is
5263 to check that the total number of inodes is greater than zero before doing
5264 the test of how many are available.
5266 58. When a "warn" ACL statement has a log_message modifier, the message is
5267 remembered, and not repeated. This is to avoid a lot of repetition when a
5268 message has many recipients that cause the same warning to be written.
5269 Howewer, Exim was preserving the list of already written lines for an
5270 entire SMTP session, which doesn't seem right. The memory is now reset if a
5271 new message is started.
5273 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5274 debugging output unless log_rewrite was also set.
5276 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5277 of (int)(handle) when we know that handle contains (void *)(-1).
5279 61. The Exim daemon panic-logs an error return when it closes the incoming
5280 connection. However "connection reset by peer" seems to be common, and
5281 isn't really an error worthy of noting specially, so that particular error
5284 62. When Exim is trying to find all the local interfaces, it used to panic and
5285 die if the ioctl to get the interface flags failed. However, it seems that
5286 on at least one OS (Solaris 9) it is possible to have an interface that is
5287 included in the list of interfaces, but for which you get a failure error
5288 for this call. This happens when the interface is not "plumbed" into a
5289 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5290 failure of the "get flags" call assumes that the interface is down.
5292 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5293 makes life easier for people who are doing arithmetic on fields extracted
5294 from dates, where you often get leading zeros that should not be
5295 interpreted as octal.
5297 64. Added qualify_domain to the redirect router, to override the global
5300 65. If a pathologically long header line contained very many addresses (the
5301 report of this problem mentioned 10 000) and each of them was rewritten,
5302 Exim could use up a very large amount of memory. (It kept on making new
5303 copies of the header line as it rewrote, and never released the old ones.)
5304 At the expense of a bit more processing, the header rewriting function has
5305 been changed so that it no longer eats memory in this way.
5307 66. The generation of the Received: header has been moved from the time that a
5308 message starts to be received, to the time that it finishes. The timestamp
5309 in the Received: header should now be very close to that of the <= log
5310 line. There are two side-effects of this change:
5312 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5313 logged header lines no longer include the local Received: line, because
5314 it has not yet been created. The same applies to a copy of the message
5315 that is returned to a non-SMTP sender when a message is rejected.
5317 (b) When a filter file is tested using -bf, no additional Received: header
5318 is added to the test message. After some thought, I decided that this
5321 This change does not affect the value of $received_for. It is still set
5322 after address rewriting, but before local_scan() is called.
5324 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5326 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5327 gave an unhelpful panic error message, and a defer error. I have managed to
5328 change this behaviour so that it now rejects any supplied certificate,
5329 which seems right, as the list of acceptable certificates is empty.
5331 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5332 gave an unhelpful defer error. I have not managed to make this reject any
5333 supplied certificates, but the error message it gives is "no certificate
5334 supplied", which is not helpful.
5336 70. exigrep's output now also includes lines that are not associated with any
5337 message, but which match the given pattern. Implemented by a patch from
5338 Martin Sluka, which also tidied up the Perl a bit.
5340 71. Recipient callout verification, like sender verification, was using <> in
5341 the MAIL FROM command. This isn't really the right thing, since the actual
5342 sender may affect whether the remote host accepts the recipient or not. I
5343 have changed it to use the actual sender in the callout; this means that
5344 the cache record is now keyed on a recipient/sender pair, not just the
5345 recipient address. There doesn't seem to be a real danger of callout loops,
5346 since a callout by the remote host to check the sender would use <>.
5347 [SEE ABOVE: changed after hitting problems.]
5349 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5350 temporary errors. However, in the case of such a code being given after
5351 the end of a data transmission (i.e. after ".") Exim was failing to write
5352 a retry record for the message. (Yes, there was some broken host that was
5353 actually sending 8xx at this point.)
5355 73. An unknown lookup type in a host list could cause Exim to panic-die when
5356 the list was checked. (An example that provoked this was putting <; in the
5357 middle of a list instead of at the start.) If this happened during a DATA
5358 ACL check, a -D file could be left lying around. This kind of configuration
5359 error no longer causes Exim to die; instead it causes a defer errror. The
5360 incident is still logged to the main and panic logs.
5362 74. Buglet left over from Exim 3 conversion. The message "too many messages
5363 in one connection" was written to the rejectlog but not the mainlog, except
5364 when address rewriting (yes!) was being logged.
5366 75. Added write_rejectlog option.
5368 76. When a system filter was run not as root (that is, when system_filter_user
5369 was set), the values of the $n variables were not being returned to the
5370 main process; thus, they were not subsequently available in the $sn
5373 77. Added +return_path_on_delivery log selector.
5375 78. A connection timeout was being treated differently from recipients deferred
5376 when testing hosts_max_try with a message that was older than the host's
5377 retry timeout. (The host should not be counted, thus allowing all hosts to
5378 be tried at least once before bouncing.) This may have been the cause of an
5379 occasionally reported bug whereby a message would remain on the queue
5380 longer than the retry timeout, but would be bounced if a delivery was
5381 forced. I say "may" because I never totally pinned down the problem;
5382 setting up timeout/retry tests is difficult. See also the next item.
5384 79. The ultimate address timeout was not being applied to errors that involved
5385 a combination of host plus message (for example, a timeout on a MAIL
5386 command). When an address resolved to a number of possible hosts, and they
5387 were not all tried for each delivery (e.g. because of hosts_max_try), a
5388 message could remain on the queue longer than the retry timeout.
5390 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5393 81. Fixed an obscure SMTP outgoing bug which required at least the following
5394 conditions: (a) there was another message waiting for the same server;
5395 (b) the server returned 5xx to all RCPT commands in the first message so
5396 that the message was not completed; (c) the server dropped the connection
5397 or gave a negative response to the RSET that Exim sends to abort the
5398 transaction. The observed case was a dropped connection after DATA that had
5399 been sent in pipelining mode. That is, the server had advertised PIPELINING
5400 but was not implementing it correctly. The effect of the bug was incorrect
5401 behaviour, such as trying another host, and this could lead to a crash.
5407 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5408 and daemon.c were passed as pointers to ints; they should have been
5409 pointers to socklen_t variables (which are typically unsigned ints).
5411 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5414 3. Fixed a really odd bug that affected only the testing scheme; patching a
5415 certain fixed string in the binary changed the value of another string that
5416 happened to be identical to the end of the original first string.
5418 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5419 name", it returns that address as the IP address. On some operating
5420 systems (e.g. Solaris), it also passes back the IP address string as the
5421 "host name". However, on others (e.g. Linux), it passes back an empty
5422 string. Exim wasn't checking for this, and was changing the host name to an
5423 empty string, assuming it had been canonicized.
5425 5. Although rare, it is permitted to have more than one PTR record for a given
5426 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5427 all the names associated with an address, because they do in Solaris.
5428 However, it seems that they do not in Linux for data that comes from the
5429 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5430 I found this out when I moved to a new Linux workstation and tried to run
5431 the Exim test suite.
5433 To get round this problem I have changed the code so that it now does its
5434 own call to the DNS to look up PTR records when searching for a host name.
5435 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5436 addresses that are only in /etc/hosts are still found.
5438 This behaviour is, however, controlled by an option called host_lookup_
5439 order, which defaults to "bydns:byaddr". If people want to use the other
5440 order, or indeed, just use one or the other means of lookup, they can
5441 specify it in this variable.
5443 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5444 some operating systems, this comes back from gethostbyaddr() as an empty
5445 string, and this is what Exim used to test for. However, it seems that in
5446 other systems, "." is yielded. Exim now tests for this case too.
5448 7. The values of check_spool_space and check_log_space are now held internally
5449 as a number of kilobytes instead of an absolute number of bytes. If a
5450 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5451 kilobyte. This means that much larger values can be stored.
5453 8. Exim monitor: an attempt to get the action menu when not actually pointing
5454 at a message produces an empty menu entitled "No message selected". This
5455 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5456 no entries in it ("Shell widget menu has zero width and/or height"). So I
5457 have added a single, blank menu entry in this case.
5459 9. Added ${quote_local_part.
5461 10. MIME decoding is now applied to the contents of Subject: header lines when
5464 11. Now that a reference to $sender_host_address automatically causes a reverse
5465 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5466 host lookup before query-style lookups in lists that might use this
5467 variable. This has therefore been abolished, and the "net-" prefix is no
5468 longer necessary for query-style lookups.
5470 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5471 have been a typo for LFLAGS, so it has been changed.
5473 13. The install script calls Exim with "-C /dev/null" in order to find the
5474 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5475 to be output. Howeve, since Exim outputs its version number before the
5476 error, it didn't break the script. It just looked ugly. I fixed this by
5477 always allowing "-C /dev/null" if the caller is root.
5479 14. Ignore overlarge ACL variable number when reading spool file - insurance
5480 against a later release with more variables having written the file.
5482 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5483 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5484 this, as well as the form without the "IPv6" on the front.
5486 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5487 OS/Makefile-Darwin file.
5489 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5490 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5492 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5493 variable. However, some OS use ints for inodes. I've added cast to long int
5494 to get rid of the compiler warning.
5496 19. I had forgotten to lock out "/../" in configuration file names when
5497 ALT_CONFIG_PREFIX was set.
5499 20. Routers used for verification do not need to specify transports. However,
5500 if such a router generated a host list, and callout was configured, Exim
5501 crashed, because it could not find a port number from the (non-existent)
5502 transport. It now assumes port 25 in this circumstance.
5504 21. Added the -t option to exigrep.
5506 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5507 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5508 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5509 them is set, LOOKUP_LSEARCH is forced.
5511 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5513 24. Added sender and host information to the "rejected by local_scan()" log
5514 line; previously there was no indication of these.
5516 25. Added .include_if_exists.
5518 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5519 Linux. It turns out that not all file systems support this. Apparently some
5520 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5521 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5522 not supported on the file descriptor, is now ignored when Exim is trying to
5523 sync a directory. This applies only to Linux.
5525 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5527 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5528 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5531 29. Added support for maildirsize files from supplied patch (modified a bit).
5533 30. The use of :fail: followed by an empty string could lead Exim to respond to
5534 sender verification failures with (e.g.):
5536 550 Verification failed for <xxx>
5537 550 Sender verify failed
5539 where the first response line was missing the '-' that indicates it is not
5540 the final line of the response.
5542 31. The loop for finding the name of the user that called Exim had a hardwired
5543 limit of 10; it now uses the value of finduser_retries, which is used for
5544 all other user lookups.
5546 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5549 33. Exim was neglecting to zero errno before one call of strtol() when
5550 expanding a string and expecting an integer value. On some systems this
5551 resulted in spurious "integer overflow" errors. Also, it was casting the
5552 result into an int without checking.
5554 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5555 did not work. The code looks as if it has *never* worked, though it appears
5556 to have been documented since at least releast 1.62. I have made it work.
5558 35. The "timeout_DNS" error in retry rules, also documented since at least
5559 1.62, also never worked. As it isn't clear exactly what this means, and
5560 clearly it isn't a major issue, I have abolished the feature by treating it
5561 as "timeout", and writing a warning to the main and panic logs.
5563 36. The display of retry rules for -brt wasn't always showing the error code
5566 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5567 timeout_connect_A, timeout_connect_MX.
5569 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5570 to the empty sender.
5572 39. The daemon was not analysing the content of -oX till after it had closed
5573 stderr and disconnected from the controlling terminal. This meant that any
5574 syntax errors were only noted on the panic log, and the return code from
5575 the command was 0. By re-arranging the code a little, I've made the
5576 decoding happen first, so such errors now appear on stderr, and the return
5577 code is 1. However, the actual setting up of the sockets still happens in
5578 the disconnected process, so errors there are still only recorded on the
5581 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5582 connections (as happens on some IP stacks) was logged at start up time as
5583 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5584 it from "IPv6 and IPv4", which means that two separate sockets are being
5587 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5588 says whether AF_INET or AF_INET6 was passed as an argument.
5590 42. Exiwhat output was messed up when time zones were included in log
5593 43. Exiwhat now gives more information about the daemon's listening ports,
5594 and whether -tls-on-connect was used.
5596 44. The "port" option of the smtp transport is now expanded.
5598 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5599 silently ignored. Now an error message is written to the main and panic
5602 46. There's a new ACL modifier called "logwrite" which writes to a log file
5603 as soon as it is encountered.
5605 47. Added $local_user_uid and $local_user_gid at routing time.
5607 48. Exim crashed when trying to verify a sender address that was being
5610 49. Exim was recognizing only a space character after ".include". It now also
5611 recognizes a tab character.
5613 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5614 extracting the relevant information from the specification. The man page no
5615 longer contains scrambled data for the -d option, and I've added a section
5616 at the front about calling Exim under different names.
5618 51. Added "extra_headers" argument to the "mail" command in filter files.
5620 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5623 53. Installed eximstats 1.29.
5625 54. Added transport_filter_timeout as a generic transport option.
5627 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5628 Cc: header lines. This was required by RFC 822, but it not required by RFC
5631 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5632 incoming messages that did not have them. Now it does so only if the
5633 message originates locally, that is, if there is no associated remote host
5634 address. When Resent- header lines are present, this applies to the Resent-
5635 lines rather than the non-Resent- lines.
5637 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5638 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5640 58. Messages for configuration errors now include the name of the main
5641 configuration file - useful now that there may be more than one file in a
5642 list (.included file names were always shown).
5644 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5645 for those rare installations that do not start the daemon as root or run it
5646 setuid root. I've cut out the call to initgroups() if the daemon is not
5649 60. The Exim user and group can now be bound into the binary as text strings
5650 that are looked up at the start of Exim's processing.
5652 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5654 62. Added $mailstore_basename variable.
5656 63. Installed patch to sieve.c from Michael Haardt.
5658 64. When Exim failed to open the panic log after failing to open the main log,
5659 the original message it was trying to log was written to stderr and debug
5660 output, but if they were not available (the usual case in production), it
5661 was lost. Now it is written to syslog before the two lines that record the
5662 failures to open the logs.
5664 65. Users' Exim filters run in subprocesses under the user's uid. It is
5665 possible for a "deliver" command or an alias in a "personal" command to
5666 provoke an address rewrite. If logging of address rewriting is configured,
5667 this fails because the process is not running as root or exim. There may be
5668 a better way of dealing with this, but for the moment (because 4.30 needs
5669 to be released), I have disabled address rewrite logging when running a
5670 filter in a non-root, non-exim process.
5676 1. The buildconfig auxiliary program wasn't quoting the value set for
5677 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5678 not defined. This bug was masked in 4.22 by the effect that was fixed in
5681 2. Some messages that were rejected after a message id was allocated were
5682 shown as "incomplete" by exigrep. It no longer does this for messages that
5683 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5685 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5686 have allow_domain_literals set, the ID did not get logged in the <= line.
5687 Domain literals are now always recognized in Message-ID: header lines.
5689 4. The first argument for a ${extract expansion item is the key name or field
5690 number. Leading and trailing spaces in this item were not being ignored,
5691 causing some misleading effects.
5693 5. When deliver_drop_privilege was set, single queue runner processes started
5694 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5695 same command in the process it spins off) were not dropping privilege.
5697 6. When the daemon running as "exim" started a queue runner, it always
5698 re-executed Exim in the spun-off process. This is a waste of effort when
5699 deliver_drop_privilege is set. The new process now just calls the
5700 queue-runner function directly.
5706 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5709 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5710 ignored. Though the use of -oP was forcing the writing of a pid file, it
5711 was always written to the default place.
5713 3. If the message "no IP address found for host xxxx" is generated during
5714 incoming verification, it is now followed by identification of the incoming
5715 connection (so you can more easily find what provoked it).
5717 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5719 5. Added some features to "harden" Exim a bit more against certain attacks:
5721 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5722 be put in Local/Makefile. This is like the never_users runtime option,
5723 but it cannot be overridden. The default setting is "root".
5725 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5726 prefix string with which any file named in a -C command line option
5729 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5730 is retained for -C and -D only if the caller of Exim is root. Without
5731 it, the exim user may also use -C and -D and retain privilege.
5733 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5734 command line option is disabled.
5736 6. Macro names set by the -D option must start with an upper case letter, just
5737 like macro names defined in the configuration file.
5739 7. Added "dereference=" facility to LDAP.
5741 8. Two instances of the typo "uknown" in the source files are fixed.
5743 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5744 the Configure-Makefile script screwed up while processing it.
5746 10. Incorporated PCRE 4.4.
5748 11. The SMTP synchronization check was not operating right at the start of an
5749 SMTP session. For example, it could not catch a HELO sent before the client
5750 waited for the greeting. There is now a check for outstanding input at the
5751 point when the greeting is written. Because of the duplex, asynchronous
5752 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5753 way, but not yet received, when the check is performed.
5755 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5756 on TCP/IP sockets, because this apparently causes some broken clients to
5759 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5760 unchanged) from the Cygwin maintainer.
5762 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5763 was supported instead of testing for mbx. Effectively a typo.
5765 15. The spa authenticator server code was not checking that the input it
5766 received was valid base64.
5768 16. The debug output line for the "set" modifier in ACLs was not showing the
5769 name of the variable that was being set.
5771 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5773 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5774 lookup on its own. Something else was needed to trigger the lookup. For
5775 example, a match in host_lookup or the need for a host name in a host list.
5776 Now, if $sender_host_name is referenced and the host name has not yet been
5777 looked up, a lookup is performed. If the lookup fails, the variable remains
5778 empty, and $host_lookup_failed is set to "1".
5780 19. Added "eqi" as a case-independent comparison operator.
5782 20. The saslauthd authentication condition could segfault if neither service
5783 nor realm was specified.
5785 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5786 error message that was logged was misleading, and incoming SMTP
5787 connections were dropped. The message is now more accurate, and temporary
5788 errors are given to SMTP connections.
5790 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5791 (except RSET) with a 421 error, until QUIT is received. However, it was
5792 failing to send a response to QUIT.
5794 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5795 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5796 ACL is now run after the helo_try_verify_hosts code.
5798 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5799 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5800 case-independent, so other case variants are also recognized). Apparently
5801 some systems use these upper case variants.
5803 25. If more than two messages were waiting for the same host, and a transport
5804 filter was specified for the transport, Exim sent two messages over the
5805 same TCP/IP connection, and then failed with "socket operation on non-
5806 socket" when it tried to send the third.
5808 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5810 27. The extern definition of crypt16() in expand.c was not being excluded when
5811 the OS had its own crypt16() function.
5813 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5814 as a preferred synonym for return_size_limit, both as an option and as an
5817 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5819 30. Changed the default configuration ACL to relax the local part checking rule
5820 for addresses that are not in any local domains. For these addresses,
5821 slashes and pipe symbols are allowed within local parts, but the sequence
5822 /../ is explicitly forbidden.
5824 31. SPA server authentication was not clearing the challenge buffer before
5827 32. log_message in a "warn" ACL statement was writing to the reject log as
5828 well as to the main log, which contradicts the documentation and doesn't
5829 seem right (because no rejection is happening). So I have stopped it.
5831 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5832 However, I am unable to do any testing of this.
5834 34. Fixed an infelicity in the appendfile transport. When checking directories
5835 for a mailbox, to see if any needed to be created, it was accidentally
5836 using path names with one or more superfluous leading slashes; tracing
5837 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5839 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5840 contents are added to the log line that is written for every discarded
5841 recipient. (Previously a log_message setting was ignored.)
5843 36. The ${quote: operator now quotes the string if it is empty.
5845 37. The install script runs exim in order to find its version number. If for
5846 some reason other than non-existence or emptiness, which it checks, it
5847 could not run './exim', it was installing it with an empty version number,
5848 i.e. as "exim-". This error state is now caught, and the installation is
5851 38. An argument was missing from the function that creates an error message
5852 when Exim fails to connect to the socket for saslauthd authentication.
5853 This could cause Exim to crash, or give a corrupted message.
5855 39. Added isip, isip4, and isip6 to ${if conditions.
5857 40. The ACL variables $acl_xx are now saved with the message, and can be
5858 accessed later in routers, transports, and filters.
5860 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5861 strings in the file are not string-expanded.
5863 42. If a MAIL command specified a SIZE value that was too large to fit into an
5864 int variable, the check against message_size_limit failed. Such values are
5865 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5866 day this will have to be increased, but I don't think I want to be around
5867 when emails are that large.
5874 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5875 iconv() is not standard in FreeBSD.
5877 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5878 IPv6 enabled. The observed symptom was a segmentation fault on return from
5879 the function os_common_find_running_interfaces() in src/os.c.
5881 3. In the check_special_case() function in daemon.c I had used "errno" as an
5882 argument name, which causes warnings on some systems. This was basically a
5883 typo, since it was named "eno" in the comments!
5885 4. The code that waits for the clock to tick (at a resolution of some fraction
5886 of a second) so as to ensure message-id uniqueness was always waiting for
5887 at least one whole tick, when it could have waited for less. [This is
5888 almost certainly not relevant at current processor speeds, where it is
5889 unlikely to ever wait at all. But we try to future-proof.]
5891 5. The function that sleeps for a time interval that includes fractions of a
5892 second contained a race. It did not block SIGALRM between setting the
5893 timer, and suspending (a couple of lines later). If the interval was short
5894 and the sigsuspend() was delayed until after it had expired, the suspension
5895 never ended. On busy systems this could lead to processes getting stuck for
5898 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5899 process, before it forks any delivery processes. The open lookup caching
5900 mechanism meant that the open file or database connection was passed into
5901 the delivery process. The problem was that delivery processes always tidy
5902 up cached lookup data. This could cause a problem for the next delivery
5903 process started by the queue runner, because the external queue runner
5904 process does not know about the closure. So the next delivery process
5905 still has data in the lookup cache. In the case of a file lookup, there was
5906 no problem because closing a file descriptor in a subprocess doesn't affect
5907 the parent. However, if the lookup was caching a connection to a database,
5908 the connection was closed, and the second delivery process was likely to
5909 see errors such as "PGSQL: query failed: server closed the connection
5910 unexpectedly". The problem has been fixed by closing all cached lookups
5911 in a queue runner before running a delivery process.
5913 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5914 seem to have the "const" qualifier which it has on other OS. I've
5917 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5918 *of the same type* (client or server) with the same public name that an
5919 error should be diagnosed.
5921 9. When Exim looked up a host name for an IP address, but failed to find the
5922 original IP address when looking up the host name (a safety check), it
5923 output the message "<ip address> does not match any IP for NULL", which was
5924 confusing, to say the least. The bug was that the host name should have
5925 appeared instead of "NULL".
5927 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5928 user that is built into the binary, and the -C or -D options is used, root
5929 privilege is dropped before the configuration file is read. In addition,
5930 logging is switched to stderr instead of the normal log files. If the
5931 configuration then re-defines the Exim user, the unprivileged environment
5932 is probably not what is expected, so Exim logs a panic warning message (but
5935 However, if deliver_drop_privilege is set, the unprivileged state may well
5936 be exactly what is intended, so the warning has been cut out in that case,
5937 and Exim is allowed to try to write to its normal log files.
5943 1. smtp_return_error_details was not giving details for temporary sender
5944 or receiver verification errors.
5946 2. Diagnose a configuration error if two authenticators have the same public
5949 3. Exim used not to create the message log file for a message until the first
5950 delivery attempt. This could be confusing when incoming messages were held
5951 for policy or load reasons. The message log file is now created at the time
5952 the message is received, and an initial "Received" line is written to it.
5954 4. The automatically generated man page for command line options had a minor
5955 bug that caused no ill effects; however, a more serious problem was that
5956 the procedure for building the man page automatically didn't always
5957 operate. Consequently, release 4.20 contains an out-of-date version. This
5958 shouldn't happen again.
5960 5. When building Exim with embedded Perl support, the script that builds the
5961 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5962 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5964 6. The freeze_tell option was not being used for messages that were frozen on
5965 arrival, either by an ACL or by local_scan().
5967 7. Added the smtp_incomplete_transaction log selector.
5969 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5970 was accepting AUTH without a new EHLO.
5972 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5973 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5976 10. Logging of TCP/IP connections (when configured) now happens in the main
5977 daemon process instead of the child process, so that the TCP/IP connection
5978 count is more accurate (but it can never be perfect).
5980 11. The use of "drop" in a nested ACL was not being handled correctly in the
5981 outer ACL. Now, if condition failure induced by the nested "drop" causes
5982 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5983 or "require"), the connection is dropped.
5985 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5986 that yield "discard" can now be used with an "accept" or a "discard" verb,
5987 but an error is generated for any others (because I can't see a useful way
5988 to define what should happen).
5990 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5991 are now processed in the same way as lines in the Exim configuration file.
5992 In particular, continuation lines are supported.
5994 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5996 15. Added -ti meaning -t -i.
5998 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5999 domains, and warn by logging if others are found.
6001 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6002 returned from GIFCONF when Exim is trying to find the list of interfaces on
6003 a host. The code in os.c has been modified to copy each ifreq to an aligned
6004 structure in all cases.
6006 Also, in some cases, the returned ifreq's were being copied to a 'struct
6007 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6008 means the last couple of bytes of an IPv6 address could be chopped if the
6009 ifreq contained only a normal sockaddr (14 bytes storage).
6011 18. Named domain lists were not supported in the hosts_treat_as_local option.
6012 An entry such as +xxxx was not recognized, and was treated as a literal
6015 19. Ensure that header lines added by a DATA ACL are included in the reject log
6016 if the ACL subsequently rejects the message.
6018 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6019 MD5 (which is deprecated).
6021 21. When testing a filter file using -bf, Exim was writing a message when it
6022 took the sender from a "From " line in the message, but it was not doing so
6023 when it took $return_path from a Return-Path: header line. It now does.
6025 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6026 with a valid header line field name (a series of printing characters
6027 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6029 23. Changed "disc" in the source to "disk" to conform to the documentation and
6030 the book and for uniformity.
6032 24. Ignore Sendmail's -Ooption=value command line item.
6034 25. When execve() failed while trying to run a command in a pipe transport,
6035 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6036 could be confused with a return value of 69 from the command itself. This
6037 has been changed to 127, the value the shell returns if it is asked to run
6038 a non-existent command. The wording for the related log line suggests a
6039 non-existent command as the problem.
6041 26. If received_header_text expands to an empty string, do not add a Received:
6042 header line to the message. (Well, it adds a token one on the spool, but
6043 marks it "old" so that it doesn't get used or transmitted.)
6045 27. Installed eximstats 1.28 (addition of -nt option).
6047 28. There was no check for failure on the call to getsockname() in the daemon
6048 code. This can fail if there is a shortage of resources on the system, with
6049 ENOMEM, for example. A temporary error is now given on failure.
6051 29. Contrary to the C standard, it seems that in some environments, the
6052 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6053 program. Exim now does this explicitly; it affects the formatting of
6054 timestamps using strftime().
6056 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6057 complaints. I've now initialized all the variables, to avoid this.
6059 32. Header lines added by a system filter were not being "seen" during
6060 transport-time rewrites.
6062 33. The info_callback() function passed to OpenSSL is set up with type void
6063 (*)(SSL *, int, int), as described somewhere. However, when calling the
6064 function (actually a macro) that sets it up, the type void(*)() is
6065 expected. I've put in a cast to prevent warnings from picky compilers.
6067 34. If a DNS black list lookup found a CNAME record, but there were no A
6068 records associated with the domain it pointed at, Exim crashed.
6070 35. If a DNS black list lookup returned more than one A record, Exim ignored
6071 all but the first. It now scans all returned addresses if a particular IP
6072 value is being sought. In this situation, the contents of the
6073 $dnslist_value variable are a list of all the addresses, separated by a
6076 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6077 to accept a host name and all its aliases if the forward lookup for any of
6078 them yielded the IP address of the incoming connection. Now it accepts only
6079 those names whose forward lookup yields the correct IP address. Any other
6080 names are discarded. This closes a loophole whereby a rogue DNS
6081 administrator could create reverse DNS records to break through a
6082 wildcarded host restriction in an ACL.
6084 37. If a user filter or a system filter that ran in a subprocess used any of
6085 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6086 the wrong values were passed to the pipe command ($thisaddress had the
6087 value of $0, $0 had the value of $1, etc). This bug was introduced by
6088 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6090 38. Improved the line breaking for long SMTP error messages from ACLs.
6091 Previously, if there was no break point between 40 and 75 characters, Exim
6092 left the rest of the message alone. Two changes have been made: (a) I've
6093 reduced the minimum length to 35 characters; (b) if it can't find a break
6094 point between 35 and 75 characters, it looks ahead and uses the first one
6095 that it finds. This may give the occasional overlong line, but at least the
6096 remaining text gets split now.
6098 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6099 file descriptors might be low, and that setting 1000 would always raise it.
6100 It turns out that in some environments, the limit is already over 1000 and
6101 that lowering it causes trouble. So now Exim takes care not to decrease it.
6103 40. When delivering a message, the value of $return_path is set to $sender_
6104 address at the start of routing (routers may change the value). By an
6105 oversight, this default was not being set up when an address was tested by
6106 -bt or -bv, which affected the outcome if any router or filter referred to
6109 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6110 log only when "message" or "log_message" are set. However, if one of the
6111 conditions was an address verification, or a call to a nested ACL, the
6112 messages generated by the underlying test were being passed through. This
6113 no longer happens. The underlying message is available in $acl_verify_
6114 message for both "message" and "log_message" expansions, so it can be
6115 passed through if needed.
6117 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6118 new expansion $bh_ to give the encoded byte string without charset
6119 translation. Translation happens only if iconv() is available; HAVE_ICONV
6120 indicates this at build time. HEADERS_CHARSET gives the charset to
6121 translate to; headers_charset can change it in the configuration, and
6122 "headers charset" can change it in an individual filter file.
6124 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6125 that creates RFC 2047 encoded "words" labels them as that charset instead
6126 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6127 expansion operator; (ii) when Exim creates a From: line for a local
6128 message; (iii) when a header line is rewritten to include a "phrase" part.
6130 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6131 buggy, causing it to skip the first lines of messages whose message ID
6132 ended in 'D'. This would not have bitten before Exim release 4.14, because
6133 message IDs were unlikely to end in 'D' before then. The effect was to have
6134 incorrect size information for certain domains.
6136 45. #include "config.h" was missing at the start of the crypt16.c module. This
6137 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6140 46. If there was a timeout during a "random" callout check, Exim treated it as
6141 a failure of the random address, and carried on sending RSET and the real
6142 address. If the delay was just some slowness somewhere, the response to the
6143 original RCPT would be taken as a response to RSET and so on, causing
6144 mayhem of various kinds.
6146 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6147 when I implemented it. It didn't allow for the fact that some option values
6148 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6149 the right test for positive values.
6151 48. Domain names in DNS records are case-independent. Exim always looks them up
6152 in lower case. Some resolvers return domain names in exactly the case they
6153 appear in the zone file, that is, they may contain uppercase letters. Not
6154 all resolvers do this - some return always lower case. Exim was treating a
6155 change of case by a resolver as a change of domain, similar to a widening
6156 of a domain abbreviation. This triggered its re-routing code and so it was
6157 trying to route what was effectively the same domain again. This normally
6158 caused routing to fail (because the router wouldn't handle the domain
6159 twice). Now Exim checks for this case specially, and just changes the
6160 casing of the domain that it ultimately uses when it transmits the message
6163 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6166 50. If a filter generated a file delivery with a non-absolute name (possible if
6167 no home directory exists for the router), the forbid_file option was not
6170 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6171 the existing equality matching.
6173 52. Exim was using ints instead of ino_t variables in some places where it was
6174 dealing with inode numbers.
6176 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6177 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6178 called TMPDIR, and if it finds it is different, it changes its value.
6180 54. The smtp_printf() function is now made available to local_scan() so
6181 additional output lines can be written before returning. There is also an
6182 smtp_fflush() function to enable the detection of a dropped connection.
6183 The variables smtp_input and smtp_batched_input are exported to
6186 55. Changed the default runtime configuration: the message "Unknown user"
6187 has been removed from the ACL, and instead placed on the localuser router,
6188 using the cannot_route_message feature. This means that any verification
6189 failures that generate their own messages won't get overridden. Similarly,
6190 the "Unrouteable address" message that was in the ACL for unverifiable
6191 relay addresses has also been removed.
6193 56. Added hosts_avoid_esmtp to the smtp transport.
6195 57. The exicyclog script was not checking for the esoteric option
6196 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6197 will work only if exicyclog is run under the appropriate euid.
6199 58. Following a discussion on the list, the rules by which Exim recognises line
6200 endings on incoming messages have been changed. The -dropcr and drop_cr
6201 options are now no-ops, retained only for backwards compatibility. The
6202 following line terminators are recognized: LF CRLF CR. However, special
6203 processing applies to CR:
6205 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6206 nor a local message in the state where . is a terminator.
6208 (ii) If a bare CR is encountered in a header line, an extra space is added
6209 after the line terminator so as not to end the header. The reasoning
6210 behind this is that bare CRs in header lines are most likely either
6211 to be mistakes, or people trying to play silly games.
6213 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6214 was being incorrectly given as 18 bytes larger than it should have been.
6215 This is a VOB (very old bug).
6217 60. This may never have affected anything current, but just in case it has:
6218 When the local host is found other than at the start of a list of hosts,
6219 the local host, those with the same MX, and any that follow, are discarded.
6220 When the list in question was part of a longer list of hosts, the following
6221 hosts (not currently being processed) were also being discarded. This no
6222 longer happens. I'm not sure if this situation could ever has previously
6225 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6228 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6229 header. This is something that is recommended in a new Internet Draft, and
6230 is something that is documented as being done by Sendmail. There are two
6231 possible values. For messages generated by the autoreply transport, Exim
6234 Auto-Submitted: auto-replied
6236 whereas for all other generated messages (e.g. bounces) it adds
6238 Auto-Submitted: auto-generated
6240 63. The "personal" condition in filters now includes a test for the
6241 Auto-Submitted: header. If it contains the string "auto-" the message it
6242 not considered personal.
6244 64. Added rcpt_include_affixes as a generic transport option.
6246 65. Added queue_only_override (default true).
6248 66. Added the syslog_duplication option.
6250 67. If what should have been the first header line of a message consisted of
6251 a space followed by a colon, Exim was mis-interpreting it as a header line.
6252 It isn't of course - it is syntactically invalid and should therefore be
6253 treated as the start of the message body. The misbehaviour could have
6254 caused a number of strange effects, including loss of data in subsequent
6255 header lines, and spool format errors.
6257 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6258 client host had authenticated. This control can now be exercised by an ACL
6259 for more flexibility.
6261 69. By default, callouts do not happen when testing with -bh. There is now a
6262 variant, -bhc, which does actually run the callout code, including
6263 consulting and updating the callout cache.
6265 70. Added support for saslauthd authentication, courtesy of Alexander
6268 71. If statvfs() failed on the spool or log directories while checking their
6269 size for availability, Exim confusingly gave the error "space shortage".
6270 Furthermore, in debugging mode it crashed with a floating point exception.
6271 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6272 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6273 serious problem, Exim now writes to the main and panic logs when this
6274 happens, with details of the failure. It then refuses to accept the
6275 incoming message, giving the message "spool directory problem" or "log
6276 directory problem" with a 421 code for SMTP messages.
6278 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6279 0, 1, and 2 exist, because some OS complain for execs without them (see
6280 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6281 descriptors. However, the code omitted to check that the open succeeded,
6282 causing mysterious errors if for some reason the permissions on /dev/null
6283 got screwed. Now Exim writes a message to the main and panic logs, and
6284 bombs out if it can't open /dev/null.
6286 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6287 interact so that it is all more flexible. It is supposed to remain
6288 backwards compatible. Also added extra_local_interfaces.
6290 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6291 to bomb out with an assertion failure - to the client this appears as a
6292 connection drop. This problem occurs in the part of the code that was taken
6293 from the Samba project. Fortunately, the assertion is in a very simple
6294 function, so I have fixed this by reproducing the function inline in the
6295 one place where it is called, and arranging for authentication to fail
6296 instead of killing the process with assert().
6298 75. The SPA client code was not working when the server requested OEM rather
6299 than Unicode encoding.
6301 76. Added code to make require_files with a specific uid setting more usable in
6302 the case where statting the file as root fails - usually a non-root-mounted
6303 NFS file system. When this happens and the failure is EACCES, Exim now
6304 forks a subprocess and does the per-uid checking as the relevant uid.
6306 77. Added process_log_path.
6308 78. If log_file_path was not explicitly set, a setting of check_log_space or
6309 check_log_inodes was ignored.
6311 79. If a space check for the spool or log partitions fails, the incident is now
6312 logged. Of course, in the latter case the data may get lost...
6314 80. Added the %p formatting code to string_format() so that it can be used to
6315 print addresses in debug_print(). Adjusted all the address printing in the
6316 debugging in store.c to use %p rather than %d.
6318 81. There was a concern that a line of code in smtp_in.c could overflow a
6319 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6320 initially expressed, the concern was not well-founded, because trailing
6321 spaces are removed early. However, if the trailing spaces were followed by
6322 a NULL, they did not get removed, so the overflow was possible. Two fixes
6325 (a) I re-wrote the offending code in a cleaner fashion.
6326 (b) If an incoming SMTP command contains a NULL character, it is rejected
6329 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6330 runs initgroups(), so that if the Exim user is in any additional groups,
6331 they will be used during message reception.
6337 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob