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.
37 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
38 to DSN fail messages (bounces): remote IP, remote greeting, remote response
39 to HELO, local diagnostic string.
41 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
42 log line to debug. Even when configured with a tls authenticator many
43 client connections are expected to not authenticate in this way, so
44 an authenticate fail is not an error.
46 HS/02 Add the Exim version string to the process info. This way exiwhat
47 gives some more detail about the running daemon.
49 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
50 matter for fast-change records such as DNSBLs.
52 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
53 retry record, even if constant. There may be multiple transports with
54 different interface settings and the retry behaviour needs to be kept
57 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
59 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
61 JH/10 Bug 840: fix log_defer_output option of pipe transport
63 JH/11 Bug 830: use same host for all RCPTS of a message, even under
64 hosts_randomize. This matters a lot when combined with mua_wrapper.
66 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
67 ${quote_pgsql:<string>} operator.
69 JH/13 Bug 1708: avoid misaligned access in cached lookup.
71 JH/14 Change header file name for freeradius-client. Relevant if compiling
72 with Radius support; from the Gentoo tree and checked under Fedora.
74 JH/15 Bug 1712: Introduce $prdr_requested flag variable
76 JH/16 Bug 1714: Permit an empty string as expansion result for transport
77 option transport_filter, meaning no filtering.
79 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
81 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
82 defaults to "*" (all hosts). The variable is now available when not built
83 with TLS, default unset, mainly to enable keeping the testuite sane.
84 If a server certificate is not supplied (via tls_certificate) an error is
85 logged, and clients will find TLS connections fail on startup. Presumably
86 they will retry in-clear.
87 Packagers of Exim are strongly encouraged to create a server certificate
93 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
96 JH/02 The smtp transport option "multi_domain" is now expanded.
98 JH/03 The smtp transport now requests PRDR by default, if the server offers
101 JH/04 Certificate name checking on server certificates, when exim is a client,
102 is now done by default. The transport option tls_verify_cert_hostnames
103 can be used to disable this per-host. The build option
104 EXPERIMENTAL_CERTNAMES is withdrawn.
106 JH/05 The value of the tls_verify_certificates smtp transport and main options
107 default to the word "system" to access the system default CA bundle.
108 For GnuTLS, only version 3.0.20 or later.
110 JH/06 Verification of the server certificate for a TLS connection is now tried
111 (but not required) by default. The verification status is now logged by
112 default, for both outbound TLS and client-certificate supplying inbound
115 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
118 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
119 Status Notification (bounce) messages are now MIME format per RFC 3464.
120 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
121 under the control of the dsn_advertise_hosts option, and routers may
122 have a dsn_lasthop option.
124 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
125 default, modifiable by a malware= option. The list separator for
126 the options can now be changed in the usual way. Bug 68.
128 JH/10 The smtp_receive_timeout main option is now expanded before use.
130 JH/11 The incoming_interface log option now also enables logging of the
131 local interface on delivery outgoing connections.
133 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
134 if the interface and destination host and port all match.
136 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
139 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
140 Patch from Andrew Lewis.
142 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
143 now supports optional time-restrictions, weighting, and priority
144 modifiers per server. Patch originally by <rommer@active.by>.
146 JH/16 The spamd_address main option now supports a mixed list of local
147 and remote servers. Remote servers can be IPv6 addresses, and
148 specify a port-range.
150 JH/17 Bug 68: The spamd_address main option now supports an optional
151 timeout value per server.
153 JH/18 Bug 1581: Router and transport options headers_add/remove can
154 now have the list separator specified.
156 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
159 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
162 JH/21 Support for the A6 type of dns record is withdrawn.
164 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
165 rather than the verbs used.
167 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
168 from 255 to 1024 chars.
170 JH/24 Verification callouts now attempt to use TLS by default.
172 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
173 are generic router options now. The defaults didn't change.
175 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
176 Original patch from Alexander Shikoff, worked over by JH.
178 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
179 files if ZCAT_COMMAND is not executable.
181 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
183 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
185 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
186 Normally benign, it bites when the pair was led to by a CNAME;
187 modern usage is to not canoicalize the domain to a CNAME target
188 (and we were inconsistent anyway for A-only vs AAAA+A).
190 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
192 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
193 when evaluating $sender_host_dnssec.
195 JH/31 Check the HELO verification lookup for DNSSEC, adding new
196 $sender_helo_dnssec variable.
198 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
200 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
202 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
204 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
205 documented as working, but never had. Support all but $spam_report.
207 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
208 added for tls authenticator.
213 TL/01 When running the test suite, the README says that variables such as
214 no_msglog_check are global and can be placed anywhere in a specific
215 test's script, however it was observed that placement needed to be near
216 the beginning for it to behave that way. Changed the runtest perl
217 script to read through the entire script once to detect and set these
218 variables, reset to the beginning of the script, and then run through
219 the script parsing/test process like normal.
221 TL/02 The BSD's have an arc4random API. One of the functions to induce
222 adding randomness was arc4random_stir(), but it has been removed in
223 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
224 function when detected.
226 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
227 cause callback expansion.
229 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
230 syntax errors in an expansion can be treated as a string instead of
231 logging or causing an error, due to the internal use of bool_lax
232 instead of bool when processing it.
234 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
235 server certificates when making smtp deliveries.
237 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
239 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
241 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
243 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
244 Merged patch from Sebastian Wiedenroth.
246 JH/05 Fix results-pipe from transport process. Several recipients, combined
247 with certificate use, exposed issues where response data items split
248 over buffer boundaries were not parsed properly. This eventually
249 resulted in duplicates being sent. This issue only became common enough
250 to notice due to the introduction of conection certificate information,
251 the item size being so much larger. Found and fixed by Wolfgang Breyha.
253 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
254 size buffer was used, resulting in syntax errors when an expansion
257 JH/07 Add support for directories of certificates when compiled with a GnuTLS
258 version 3.3.6 or later.
260 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
261 is EXPERIMENTAL_EVENT, the main-configuration and transport options
262 both become "event_action", the variables become $event_name, $event_data
263 and $event_defer_errno. There is a new variable $verify_mode, usable in
264 routers, transports and related events. The tls:cert event is now also
265 raised for inbound connections, if the main configuration event_action
268 TL/06 In test suite, disable OCSP for old versions of openssl which contained
269 early OCSP support, but no stapling (appears to be less than 1.0.0).
271 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
272 server certificate names available under the smtp transport option
273 "tls_verify_cert_hostname" now do not permit multi-component wildcard
276 JH/10 Time-related extraction expansions from certificates now use the main
277 option "timezone" setting for output formatting, and are consistent
278 between OpenSSL and GnuTLS compilations. Bug 1541.
280 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
281 encoded parameter in the incoming message. Bug 1558.
283 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
284 include certificate info, eximon was claiming there were spoolfile
287 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
289 JH/14 Log delivery-related information more consistently, using the sequence
290 "H=<name> [<ip>]" wherever possible.
292 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
293 are problematic for Debian distribution, omit them from the release
296 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
298 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
301 JH/17 Fix a null-indirection in certextract expansions when a nondefault
302 output list separator was used.
307 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
308 checkers that were complaining about end of non-void function with no
311 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
312 This was a regression intruduced in 4.83 by another bugfix.
314 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
316 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
317 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
323 TF/01 Correctly close the server side of TLS when forking for delivery.
325 When a message was received over SMTP with TLS, Exim failed to clear up
326 the incoming connection properly after forking off the child process to
327 deliver the message. In some situations the subsequent outgoing
328 delivery connection happened to have the same fd number as the incoming
329 connection previously had. Exim would try to use TLS and fail, logging
330 a "Bad file descriptor" error.
332 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
333 utilities have not been installed.
335 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
336 temporary space as the ACL may create new global variables.
338 TL/01 LDAP support uses per connection or global context settings, depending
339 upon the detected version of the libraries at build time.
341 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
342 to extract and use the src ip:port in logging and expansions as if it
343 were a direct connection from the outside internet. PPv2 support was
344 updated based on HAProxy spec change in May 2014.
346 JH/02 Add ${listextract {number}{list}{success}{fail}}.
348 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
349 Properly escape header and check for NULL return.
351 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
354 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
356 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
357 characters in header names, implemented as a verify condition.
358 Contributed by Michael Fischer v. Mollard.
360 TL/05 Rename SPF condition results err_perm and err_temp to standardized
361 results permerror and temperror. Previous values are deprecated but
362 still accepted. In a future release, err_perm and err_temp will be
363 completely removed, which will be a backward incompatibility if the
364 ACL tests for either of these two old results. Patch contributed by
365 user bes-internal on the mailing list.
367 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
369 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
370 selectors, in both main and reject logs.
372 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
375 JH/07 Add malware type "sock" for talking to simple daemon.
377 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
379 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
380 routers/transports under cutthrough routing.
382 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
383 numbers. Touch up "bool" conditional to keep the same definition.
385 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
387 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
389 JH/12 Expand items in router/transport headers_add or headers_remove lists
390 individually rather than the list as a whole. Bug 1452.
392 Required for reasonable handling of multiple headers_ options when
393 they may be empty; requires that headers_remove items with embedded
394 colons must have them doubled (or the list-separator changed).
396 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
397 view the policy declared in the DMARC record. Currently, $dmarc_status
398 is a combined value of both the record presence and the result of the
401 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
403 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
404 dnslookup router and the smtp transport (applying to the forward
407 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
408 of ldap servers used for a specific lookup. Patch provided by Heiko
411 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
412 New variable $lookup_dnssec_authenticated for observability.
414 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
415 Patch submitted by Lars Timman.
417 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
419 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
420 Requires trusted mode and valid format message id, aborts otherwise.
421 Patch contributed by Heiko Schlichting.
423 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
424 certextract with support for various fields. Bug 1358.
426 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
427 is requested by default, modifiable by smtp transport option
430 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
431 operate on certificate variables to give certificate fingerprints
432 Also new ${sha256:cert_variable}.
434 JH/23 The PRDR feature is moved from being Experimental into the mainline.
436 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
437 Christian Aistleitner.
439 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
441 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
442 file. Patch from Wolfgang Breyha.
444 JH/25 Expand the coverage of the delivery $host and $host_address to
445 client authenticators run in verify callout. Bug 1476.
447 JH/26 Port service names are now accepted for tls_on_connect_ports, to
448 align with daemon_smtp_ports. Bug 72.
450 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
451 support and error reporting did not work properly.
453 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
454 and is readable. Patch from Andrew Colin Kissa.
456 TL/14 Enhance documentation of ${run expansion and how it parses the
457 commandline after expansion, particularly in the case when an
458 unquoted variable expansion results in an empty value.
460 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
462 PP/02 Fix internal collision of T_APL on systems which support RFC3123
463 by renaming away from it. Addresses GH issue 15, reported by
466 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
468 TL/15 SECURITY: prevent double expansion in math comparison functions
469 (can expand unsanitized data). Not remotely exploitable.
476 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
478 PP/02 Make -n do something, by making it not do something.
479 When combined with -bP, the name of an option is not output.
481 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
484 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
485 $sender_host_name and config options to manage this, and basic check
488 PP/05 DSCP support for outbound connections and control modifier for inbound.
490 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
491 (Only plugin which currently uses this is kerberos4, which nobody should
492 be using, but we should make it available and other future plugins might
493 conceivably use it, even though it would break NAT; stuff *should* be
494 using channel bindings instead).
496 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
497 name; added for Sendmail compatibility; requires admin caller.
498 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
499 just ignore it); requires trusted caller.
500 Also parse but ignore: -Ac -Am -X<logfile>
503 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
505 TL/02 Add +smtp_confirmation as a default logging option.
507 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
508 Patch by Magnus Holmgren from 2007-02-20.
510 TL/04 Bugzilla 1281 - Spec typo.
511 Bugzilla 1283 - Spec typo.
512 Bugzilla 1290 - Spec grammar fixes.
514 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
516 TL/06 Add Experimental DMARC support using libopendmarc libraries.
518 TL/07 Fix an out of order global option causing a segfault. Reported to dev
519 mailing list by by Dmitry Isaikin.
521 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
523 JH/02 Support "G" suffix to numbers in ${if comparisons.
525 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
527 NM/01 Bugzilla 1197 - Spec typo
528 Bugzilla 1196 - Spec examples corrections
530 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
532 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
533 gnutls_enable_pkcs11, but renamed to more accurately indicate its
536 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
537 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
539 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
540 "acl {{name}{arg}...}", and optional args on acl condition
543 JH/05 Permit multiple router/transport headers_add/remove lines.
545 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
547 JH/07 Avoid using a waiting database for a single-message-only transport.
548 Performance patch from Paul Fisher. Bugzilla 1262.
550 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
553 JH/09 Add $headers_added variable, with content from use of ACL modifier
554 add_header (but not yet added to the message). Bugzilla 199.
556 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
557 Pulled from Bugzilla 817 by Wolfgang Breyha.
559 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
561 (nb: this is the same fix as in Exim 4.80.1)
563 JH/11 Add A= logging on delivery lines, and a client_set_id option on
566 JH/12 Add optional authenticated_sender logging to A= and a log_selector
569 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
571 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
572 advertise SMTP AUTH mechanism to us, instead of a generic
573 protocol violation error. Also, make Exim more robust to bad
574 data from the Dovecot auth socket.
576 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
578 When a queue runner is handling a message, Exim first routes the
579 recipient addresses, during which it prunes them based on the retry
580 hints database. After that it attempts to deliver the message to
581 any remaining recipients. It then updates the hints database using
584 So if a recipient address works intermittently, it can get repeatedly
585 deferred at routing time. The retry hints record remains fresh so the
586 address never reaches the final cutoff time.
588 This is a fairly common occurrence when a user is bumping up against
589 their storage quota. Exim had some logic in its local delivery code
590 to deal with this. However it did not apply to per-recipient defers
591 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
593 This change adds a proper retry rule check during routing so that the
594 final cutoff time is checked against the message's age. We only do
595 this check if there is an address retry record and there is not a
596 domain retry record; this implies that previous attempts to handle
597 the address had the retry_use_local_parts option turned on. We use
598 this as an approximation for the destination being like a local
599 delivery, as in LMTP.
601 I suspect this new check makes the old local delivery cutoff check
602 redundant, but I have not verified this so I left the code in place.
604 TF/02 Correct gecos expansion when From: is a prefix of the username.
606 Test 0254 submits a message to Exim with the header
610 When I ran the test suite under the user fanf2, Exim expanded
611 the header to contain my full name, whereas it should have added
612 a Resent-Sender: header. It erroneously treats any prefix of the
613 username as equal to the username.
615 This change corrects that bug.
617 GF/01 DCC debug and logging tidyup
618 Error conditions log to paniclog rather than rejectlog.
619 Debug lines prefixed by "DCC: " to remove any ambiguity.
621 TF/03 Avoid unnecessary rebuilds of lookup-related code.
623 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
624 Bug spotted by Jeremy Harris; was flawed since initial commit.
625 Would have resulted in OCSP responses post-SNI triggering an Exim
626 NULL dereference and crash.
628 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
630 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
631 Bug detection, analysis and fix by Samuel Thibault.
632 Bugzilla 1331, Debian bug #698092.
634 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
636 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
637 Server implementation by Todd Lyons, client by JH.
638 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
639 config variable "prdr_enable" controls whether the server
640 advertises the facility. If the client requests PRDR a new
641 acl_data_smtp_prdr ACL is called once for each recipient, after
642 the body content is received and before the acl_smtp_data ACL.
643 The client is controlled by bolth of: a hosts_try_prdr option
644 on the smtp transport, and the server advertisement.
645 Default client logging of deliveries and rejections involving
646 PRDR are flagged with the string "PRDR".
648 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
649 fclose(). Diagnosis by Todd Lyons.
651 PP/17 Update configure.default to handle IPv6 localhost better.
652 Patch by Alain Williams (plus minor tweaks).
655 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
656 This is now consistent with GnuTLS, and is now documented: the
657 previous undocumented portable approach to treating the option as
658 unset was to force an expansion failure. That still works, and
659 an empty string is now equivalent.
661 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
662 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
663 not performing validation itself.
665 PP/20 Added force_command boolean option to pipe transport.
666 Patch from Nick Koston, of cPanel Inc.
668 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
671 TF/04 Added udpsend ACL modifer and hexquote expansion operator
673 PP/21 Fix eximon continuous updating with timestamped log-files.
674 Broken in a format-string cleanup in 4.80, missed when I repaired the
675 other false fix of the same issue.
676 Report and fix from Heiko Schlichting.
679 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
680 Report from Prashanth Katuri.
682 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
683 It's SecureTransport, so affects any MacOS clients which use the
684 system-integrated TLS libraries, including email clients.
686 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
687 using a MIME ACL for non-SMTP local injection.
688 Report and assistance in diagnosis by Warren Baker.
690 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
692 JH/16 Fix comparisons for 64b. Bugzilla 1385.
694 TL/09 Add expansion variable $authenticated_fail_id to keep track of
695 last id that failed so it may be referenced in subsequent ACL's.
697 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
700 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
701 ldap library initialization, allowing self-signed CA's to be
702 used. Also properly sets require_cert option later in code by
703 using NULL (global ldap config) instead of ldap handle (per
704 session). Bug diagnosis and testing by alxgomz.
706 TL/12 Enhanced documentation in the ratelimit.pl script provided in
707 the src/util/ subdirectory.
709 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
710 renamed to Transport Post Delivery Action by Jeremy Harris, as
713 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
714 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
715 redis_servers = needs to be configured which will be used by the redis
716 lookup. Patch from Warren Baker, of The Packet Hub.
718 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
720 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
721 hostname or reverse DNS when processing a host list. Used suggestions
722 from multiple comments on this bug.
724 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
726 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
727 Missed a few lines, added it to make the runtest require no keyboard
730 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
731 contains upper case chars. Make router use caseful_local_part.
733 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
734 support when GnuTLS has been built with p11-kit.
740 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
742 This, or similar/improved, will also be change PP/11 of 4.82.
748 PP/01 Handle short writes when writing local log-files.
749 In practice, only affects FreeBSD (8 onwards).
750 Bugzilla 1053, with thanks to Dmitry Isaikin.
752 NM/01 Bugzilla 949 - Documentation tweak
754 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
757 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
759 PP/02 Implemented gsasl authenticator.
761 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
763 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
764 `pkg-config foo` for cflags/libs.
766 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
767 with rest of GSASL and with heimdal_gssapi.
769 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
770 `pkg-config foo` for cflags/libs for the TLS implementation.
772 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
773 properties get this fed in as external SSF. A number of robustness
774 and debugging improvements to the cyrus_sasl authenticator.
776 PP/08 cyrus_sasl server now expands the server_realm option.
778 PP/09 Bugzilla 1214 - Log authentication information in reject log.
779 Patch by Jeremy Harris.
781 PP/10 Added dbmjz lookup type.
783 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
785 PP/12 MAIL args handles TAB as well as SP, for better interop with
786 non-compliant senders.
787 Analysis and variant patch by Todd Lyons.
789 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
790 Bug report from Lars Müller <lars@samba.org> (via SUSE),
791 Patch from Dirk Mueller <dmueller@suse.com>
793 PP/13 tls_peerdn now print-escaped for spool files.
794 Observed some $tls_peerdn in wild which contained \n, which resulted
795 in spool file corruption.
797 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
798 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
799 or write after TLS renegotiation, which otherwise led to messages
802 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
803 as a tracking header (ie: a signed header comes before the signature).
804 Patch from Wolfgang Breyha.
806 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
807 comma-sep list; embedded commas doubled.
809 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
811 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
813 Report and patch from Dmitry Banschikov.
815 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
816 Removed SSL_clear() after SSL_new() which led to protocol negotiation
817 failures. We appear to now support TLS1.1+ with Exim.
819 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
820 lets Exim select keys and certificates based upon TLS SNI from client.
821 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
822 before an outbound SMTP session. New log_selector, +tls_sni.
824 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
825 NULL dereference. Report and patch from Alun Jones.
827 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
828 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
829 Not seeing resolver debug output on NetBSD, but suspect this is a
830 resolver implementation change.
832 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
833 Left warnings. Added "eximon gdb" invocation mode.
835 PP/21 Defaulting "accept_8bitmime" to true, not false.
837 PP/22 Added -bw for inetd wait mode support.
839 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
840 locate the relevant includes and libraries. Made this the default.
842 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
843 Bugzilla 1246, report and most of solution from Tomasz Kusy.
845 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
846 This may cause build issues on older platforms.
848 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
849 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
850 gnutls_require_mac & gnutls_require_protocols (no longer supported).
851 Added SNI support via GnuTLS too.
852 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
854 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
856 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
857 Applied second patch from Janne, implementing suggestion to default
858 multiple-strings-in-record handling to match SPF spec.
860 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
862 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
863 read-only, out of scope).
864 Patch from Wolfgang Breyha, report from Stuart Northfield.
866 PP/29 Fix three issues highlighted by clang analyser static analysis.
867 Only crash-plausible issue would require the Cambridge-specific
868 iplookup router and a misconfiguration.
869 Report from Marcin Mirosław.
871 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
873 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
874 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
875 As part of this, removing so much warning spew let me fix some minor
876 real issues in debug logging.
878 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
879 assignment on my part. Fixed.
881 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
882 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
883 Janne Snabb (who went above and beyond: thank you).
885 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
886 string otherwise requires a connection and a bunch more work and it's
887 relatively easy to get wrong. Should also expose TLS library linkage
890 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
891 64-bit ${eval} (JH/03).
893 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
894 GNU libc to support some of the 64-bit stuff, should not lead to
895 conflicts. Defined before os.h is pulled in, so if a given platform
896 needs to override this, it can.
898 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
899 protection layer was required, which is not implemented.
900 Bugzilla 1254, patch from Wolfgang Breyha.
902 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
903 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
904 tls_dhparam take prime identifiers. Also unbreak combination of
905 OpenSSL+DH_params+TLSSNI.
907 PP/39 Disable SSLv2 by default in OpenSSL support.
913 PP/01 Solaris build fix for Oracle's LDAP libraries.
914 Bugzilla 1109, patch from Stephen Usher.
916 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
918 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
921 TF/02 Fix a couple more cases where we did not log the error message
922 when unlink() failed. See also change 4.74-TF/03.
924 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
925 lock up or crash if it happened to be inside a call to libc when it
926 got a SIGUSR1 from exiwhat.
928 The SIGUSR1 handler appends the current process status to the process
929 log which is later printed by exiwhat. It used to use the general
930 purpose logging code to do this, but several functions it calls are
931 not safe for signals.
933 The new output code in the SIGUSR1 handler is specific to the process
934 log, and simple enough that it's easy to inspect for signal safety.
935 Removing some special cases also simplifies the general logging code.
936 Removing the spurious timestamps from the process log simplifies
939 TF/04 Improved ratelimit ACL condition.
941 The /noupdate option has been deprecated in favour of /readonly which
942 has clearer semantics. The /leaky, /strict, and /readonly update modes
943 are mutually exclusive. The update mode is no longer included in the
944 database key; it just determines when the database is updated. (This
945 means that when you upgrde Exim will forget old rate measurements.)
947 Exim now checks that the per_* options are used with an update mode that
948 makes sense for the current ACL. For example, when Exim is processing a
949 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
950 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
951 must specify per_mail/readonly. If you omit the update mode it defaults to
952 /leaky where that makes sense (as before) or /readonly where required.
954 The /noupdate option is now undocumented but still supported for
955 backwards compatibility. It is equivalent to /readonly except that in
956 ACLs where /readonly is required you may specify /leaky/noupdate or
957 /strict/noupdate which are treated the same as /readonly.
959 A useful new feature is the /count= option. This is a generalization
960 of the per_byte option, so that you can measure the throughput of other
961 aggregate values. For example, the per_byte option is now equivalent
962 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
964 The per_rcpt option has been generalized using the /count= mechanism
965 (though it's more complicated than the per_byte equivalence). When it is
966 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
967 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
968 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
969 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
970 Note that using per_rcpt with a non-readonly update mode in more than
971 one ACL will cause the recipients to be double-counted. (The per_mail
972 and per_byte options don't have this problem.)
974 The handling of very low rates has changed slightly. If the computed rate
975 is less than the event's count (usually one) then this event is the first
976 after a long gap. In this case the rate is set to the same as this event's
977 count, so that the first message of a spam run is counted properly.
979 The major new feature is a mechanism for counting the rate of unique
980 events. The new per_addr option counts the number of different
981 recipients that someone has sent messages to in the last time period. It
982 behaves like per_rcpt if all the recipient addresses are different, but
983 duplicate recipient addresses do not increase the measured rate. Like
984 the /count= option this is a general mechanism, so the per_addr option
985 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
986 example, measure the rate that a client uses different sender addresses
987 with the options per_mail/unique=$sender_address. There are further
988 details in the main documentation.
990 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
992 TF/06 Removed a few PCRE remnants.
994 TF/07 Automatically extract Exim's version number from tags in the git
995 repository when doing development or release builds.
997 PP/02 Raise smtp_cmd_buffer_size to 16kB.
998 Bugzilla 879. Patch from Paul Fisher.
1000 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1001 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1004 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1006 PP/05 Variable $av_failed, true if the AV scanner deferred.
1007 Bugzilla 1078. Patch from John Horne.
1009 PP/06 Stop make process more reliably on build failure.
1010 Bugzilla 1087. Patch from Heiko Schlittermann.
1012 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1013 Bugzilla 1089. Patch from Heiko Schlittermann.
1015 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1016 Bugzilla 1131. Patch from Holger Weiß.
1018 PP/09 Handle IPv6 addresses with SPF.
1019 Bugzilla 860. Patch from Wolfgang Breyha.
1021 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1023 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1026 PP/11 match_* no longer expand right-hand-side by default.
1027 New compile-time build option, EXPAND_LISTMATCH_RHS.
1028 New expansion conditions, "inlist", "inlisti".
1030 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1032 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1034 PP/14 fix log_write() format string regression from TF/03.
1035 Bugzilla 1152. Patch from Dmitry Isaikin.
1041 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1043 PP/02 Harmonised TLS library version reporting; only show if debugging.
1044 Layout now matches that introduced for other libraries in 4.74 PP/03.
1046 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1048 PP/04 New "dns_use_edns0" global option.
1050 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1053 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1054 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1056 TK/01 Updated PolarSSL code to 0.14.2.
1057 Bugzilla 1097. Patch from Andreas Metzler.
1059 PP/07 Catch divide-by-zero in ${eval:...}.
1060 Fixes bugzilla 1102.
1062 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1065 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1066 format-string attack -- SECURITY: remote arbitrary code execution.
1068 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1069 time unintentionally subject to list matching rules, letting the header
1070 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1071 arbitrary string expansion). This allowed for information disclosure.
1073 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1074 INT_MIN/-1 -- value coerced to INT_MAX.
1080 NM/01 Workround for PCRE version dependency in version reporting
1083 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1084 This fixes portability to compilers other than gcc, notably
1085 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1087 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1088 makefiles for portability to HP-UX and POSIX correctness.
1090 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1091 Also via indented variable definition in the Makefile.
1092 (Debugging by Oliver Heesakkers).
1094 PP/02 Restore caching of spamd results with expanded spamd_address.
1095 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1097 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1098 Improves build reliability. Fix from: Frank Elsner
1100 NM/02 Fix wide character breakage in the rfc2047 coding
1101 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1103 NM/03 Allow underscore in dnslist lookups
1104 Fixes bug 1026. Patch from Graeme Fowler
1106 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1107 Code patches from Adam Ciarcinski of NetBSD.
1109 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1112 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1113 is logged, to avoid truncation. Patch from John Horne.
1115 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1116 Patch from Jakob Hirsch.
1118 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1119 SQL string expansion failure details.
1120 Patch from Andrey Oktyabrski.
1122 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1123 Patch from Simon Arlott.
1125 PP/09 New lookups functionality failed to compile on old gcc which rejects
1126 extern declarations in function scope.
1127 Patch from Oliver Fleischmann
1129 PP/10 Use sig_atomic_t for flags set from signal handlers.
1130 Check getgroups() return and improve debugging.
1131 Fixed developed for diagnosis in bug 927 (which turned out to be
1134 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1135 Patch from Mark Zealey.
1137 PP/12 Bugzilla 1056: Improved spamd server selection.
1138 Patch from Mark Zealey.
1140 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1141 Based on patch from Heiko Schlittermann.
1143 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1144 Patch from Uwe Doering, sign-off by Michael Haardt.
1146 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1147 variable declaration deep within a block. Bug and patch from
1150 PP/15 lookups-Makefile IRIX compatibilty coercion.
1152 PP/16 Make DISABLE_DKIM build knob functional.
1154 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1155 Patch by Simon Arlott
1157 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1158 variable argument macros. Our copy now differs from upstream.
1164 TF/01 Failure to get a lock on a hints database can have serious
1165 consequences so log it to the panic log.
1167 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1168 controlled using the smtp_confirmation log selector.
1170 TF/03 Include the error message when we fail to unlink a spool file.
1172 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1173 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1174 for maintaining out-of-tree patches for some time.
1176 PP/01 Bugzilla 139: Documentation and portability issues.
1177 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1178 Handle per-OS dynamic-module compilation flags.
1180 PP/02 Let /dev/null have normal permissions.
1181 The 4.73 fixes were a little too stringent and complained about the
1182 permissions on /dev/null. Exempt it from some checks.
1183 Reported by Andreas M. Kirchwitz.
1185 PP/03 Report version information for many libraries, including
1186 Exim version information for dynamically loaded libraries. Created
1187 version.h, now support a version extension string for distributors
1188 who patch heavily. Dynamic module ABI change.
1190 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1191 privilege escalation vulnerability whereby the Exim run-time user
1192 can cause root to append content of the attacker's choosing to
1195 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1198 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1199 If dropping privileges for untrusted macros, we disabled normal logging
1200 on the basis that it would fail; for the Exim run-time user, this is not
1201 the case, and it resulted in successful deliveries going unlogged.
1202 Fixed. Reported by Andreas Metzler.
1208 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1209 only prepend for the Resent-* case. Fixes regression introduced in
1210 Exim 4.70 by NM/22 for Bugzilla 607.
1212 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1213 increasing numbers of administrators be bitten by this.
1215 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1217 PP/03 Bugzilla 994: added openssl_options main configuration option.
1219 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1221 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1223 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1225 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1226 without a peer certificate, leading to a segfault because of an
1227 assumption that peers always have certificates. Be a little more
1228 paranoid. Problem reported by Martin Tscholak.
1230 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1231 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1232 NB: ClamAV planning to remove STREAM in "middle of 2010".
1233 CL also introduces -bmalware, various -d+acl logging additions and
1234 more caution in buffer sizes.
1236 PP/09 Implemented reverse_ip expansion operator.
1238 PP/10 Bugzilla 937: provide a "debug" ACL control.
1240 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1242 PP/12 Bugzilla 973: Implement --version.
1244 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1246 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1248 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1250 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1251 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1252 ignore trailing whitespace.
1254 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1256 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1259 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1260 Notification from Dr Andrew Aitchison.
1262 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1263 ExtendedDetectionInfo response format.
1264 Notification from John Horne.
1266 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1269 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1270 XSL and documented dependency on system catalogs, with examples of how
1273 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1276 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1277 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1278 configuration file which is writeable by the Exim user or group.
1280 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1281 of configuration files to cover files specified with the -C option if
1282 they are going to be used with root privileges, not just the default
1285 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1286 option (effectively making it always true).
1288 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1289 files to be used while preserving root privileges.
1291 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1292 that rogue child processes cannot use them.
1294 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1295 run-time user, instead of root.
1297 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1298 Exim run-time user without dropping privileges.
1300 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1301 result string, instead of calling string_vformat() twice with the same
1304 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1305 for other users. Others should always drop root privileges if they use
1306 -C on the command line, even for a whitelisted configure file.
1308 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1310 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1316 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1317 $data_path, and $header_path variables; fixed documentation bugs and
1320 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1321 exipick to access non-standard spools, including the "frozen" queue
1324 NM/01 Bugzilla 965: Support mysql stored procedures.
1325 Patch from Alain Williams
1327 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1329 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1330 Patch from Andreas Metzler
1332 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1333 Patch from Kirill Miazine
1335 NM/05 Bugzilla 671: Added umask to procmail example.
1337 JJ/03 installed exipick 20100323.0, fixing doc bug
1339 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1340 directory. Notification and patch from Dan Rosenberg.
1342 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1344 TK/02 Improve log output when DKIM signing operation fails.
1346 MH/01 Treat the transport option dkim_domain as a colon separated
1347 list, not as a single string, and sign the message with each element,
1348 omitting multiple occurences of the same signer.
1350 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1351 Bugzilla 985, 986. Patch by Simon Arlott
1353 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1354 Patch by Simon Arlott
1356 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1357 MBX locking. Notification from Dan Rosenberg.
1363 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1365 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1367 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1369 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1371 NM/04 Bugzilla 915: Flag broken perl installation during build.
1377 TK/01 Added patch by Johannes Berg that expands the main option
1378 "spamd_address" if it starts with a dollar sign.
1380 TK/02 Write list of recipients to X-Envelope-Sender header when building
1381 the mbox-format spool file for content scanning (suggested by Jakob
1384 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1385 (http://www.dcc-servers.net/) support via dccifd. Activated by
1386 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1388 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1389 by Mark Daniel Reidel <mr@df.eu>.
1391 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1392 When building exim an external PCRE library is now needed -
1393 PCRE is a system library on the majority of modern systems.
1394 See entry on PCRE_LIBS in EDITME file.
1396 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1397 conversation. Added nologin parameter to request.
1398 Patch contributed by Kirill Miazine.
1400 TF/01 Do not log submission mode rewrites if they do not change the address.
1402 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1404 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1405 log files in place. Contributed by Roberto Lima.
1407 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1409 TF/03 Bugzilla 615: When checking the local_parts router precondition
1410 after a local_part_suffix or local_part_prefix option, Exim now
1411 does not use the address's named list lookup cache, since this
1412 contains cached lookups for the whole local part.
1414 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1415 Robert Millan. Documentation is in experimental-spec.txt.
1417 TF/04 Bugzilla 668: Fix parallel build (make -j).
1419 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1421 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1422 Patch provided by Jan Srzednicki.
1424 TF/05 Leading white space used to be stripped from $spam_report which
1425 wrecked the formatting. Now it is preserved.
1427 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1428 that they are available at delivery time.
1430 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1432 TF/08 TLS error reporting now respects the incoming_interface and
1433 incoming_port log selectors.
1435 TF/09 Produce a more useful error message if an SMTP transport's hosts
1436 setting expands to an empty string.
1438 NM/06 Bugzilla 744: EXPN did not work under TLS.
1439 Patch provided by Phil Pennock.
1441 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1442 Patch provided by Richard Godbee.
1444 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1445 acl_smtp_notquit, added index entry.
1447 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1448 Patch provided by Eugene Bujak.
1450 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1451 accept(). Patch provided by Maxim Dounin.
1453 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1454 Patch provided by Phil Pennock.
1456 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1458 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1459 Patch provided by Brad "anomie" Jorsch.
1461 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1462 Patch provided by Dean Brooks.
1464 TK/05 Add native DKIM support (does not depend on external libraries).
1466 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1467 Patch provided by Graeme Fowler.
1469 NM/16 Bugzilla 851: Documentation example syntax fix.
1471 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1473 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1476 NM/19 Bugzilla 745: TLS version reporting.
1477 Patch provided by Phil Pennock.
1479 NM/20 Bugzilla 167: bool: condition support.
1480 Patch provided by Phil Pennock.
1482 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1483 clients. Patch provided by Phil Pennock.
1485 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1486 Patch provided by Brad "anomie" Jorsch.
1488 NM/23 Bugzilla 687: Fix misparses in eximstats.
1489 Patch provided by Heiko Schlittermann.
1491 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1492 Patch provided by Heiko Schlittermann.
1494 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1495 plus update to original patch.
1497 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1499 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1500 Patch provided by David Brownlee.
1502 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1504 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1506 NM/30 Bugzilla 888: TLS documentation bugfixes.
1508 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1510 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1511 Unlike the original bugzilla I have changed all shell scripts in src tree.
1513 NM/33 Bugzilla 898: Transport filter timeout fix.
1514 Patch by Todd Rinaldo.
1516 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1517 Patch by Serge Demonchaux.
1519 NM/35 Bugzilla 39: Base64 decode bug fixes.
1520 Patch by Jakob Hirsch.
1522 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1524 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1526 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1528 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1534 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1535 ALT-N's libdkim that I have put here:
1536 http://duncanthrax.net/exim-experimental/
1538 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1539 were called 'true' and it seems that C99 defines that as a reserved
1540 keyword to be used with 'bool' variable types. That means you could
1541 not include C99-style headers which use bools without triggering
1542 build errors in sieve.c.
1544 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1545 as mailq or other aliases. Changed the --help handling significantly
1546 to do whats expected. exim_usage() emits usage/help information.
1548 SC/01 Added the -bylocaldomain option to eximstats.
1550 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1552 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1554 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1560 PH/01 Another patch from the Sieve maintainer.
1562 PH/02 When an IPv6 address is converted to a string for single-key lookup
1563 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1564 dots are used instead of colons so that keys in lsearch files need not
1565 contain colons. This was done some time before quoting was made available
1566 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1567 (notated using the quote facility) so as to distinguish them from IPv4
1568 keys. This meant that lookups for IP addresses in host lists did not work
1569 for iplsearch lookups.
1571 This has been fixed by arranging for IPv6 addresses to be expressed with
1572 colons if the lookup type is iplsearch. This is not incompatible, because
1573 previously such lookups could never work.
1575 The situation is now rather anomolous, since one *can* have colons in
1576 ordinary lsearch keys. However, making the change in all cases is
1577 incompatible and would probably break a number of configurations.
1579 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1582 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1583 conversion specification without a maximum field width, thereby enabling
1584 a rogue spamd server to cause a buffer overflow. While nobody in their
1585 right mind would setup Exim to query an untrusted spamd server, an
1586 attacker that gains access to a server running spamd could potentially
1587 exploit this vulnerability to run arbitrary code as the Exim user.
1589 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1590 $primary_hostname instead of what libspf2 thinks the hosts name is.
1592 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1593 a directory entry by the name of the lookup key. Previously, if a
1594 symlink pointed to a non-existing file or a file in a directory that
1595 Exim lacked permissions to read, a lookup for a key matching that
1596 symlink would fail. Now it is enough that a matching directory entry
1597 exists, symlink or not. (Bugzilla 503.)
1599 PH/03 The body_linecount and body_zerocount variables are now exported in the
1602 PH/04 Added the $dnslist_matched variable.
1604 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1605 This means they are set thereafter only if the connection becomes
1608 PH/06 Added the client_condition to authenticators so that some can be skipped
1609 by clients under certain conditions.
1611 PH/07 The error message for a badly-placed control=no_multiline_responses left
1612 "_responses" off the end of the name.
1614 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1616 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1617 (without spaces) instead of just copying the configuration text.
1619 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1621 PH/11 Added $max_received_linelength.
1623 PH/12 Added +ignore_defer and +include_defer to host lists.
1625 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1626 way in which PCRE > 7.0 is built.
1628 PH/14 Implemented queue_only_load_latch.
1630 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1631 MAIL command. The effect was to mangle the value on 64-bit systems.
1633 PH/16 Another patch from the Sieve maintainer.
1635 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1637 PH/18 If a system quota error occurred while trying to create the file for
1638 a maildir delivery, the message "Mailbox is full" was not appended to the
1639 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1640 only to a quota excession during the actual writing of the file.
1642 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1643 characters?) which causes problems in log lines. The DN values are now
1644 passed through string_printing() before being added to log lines.
1646 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1647 and InterBase are left for another time.)
1649 PH/21 Added message_body_newlines option.
1651 PH/22 Guard against possible overflow in moan_check_errorcopy().
1653 PH/23 POSIX allows open() to be a macro; guard against that.
1655 PH/24 If the recipient of an error message contained an @ in the local part
1656 (suitably quoted, of course), incorrect values were put in $domain and
1657 $local_part during the evaluation of errors_copy.
1663 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1664 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1667 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1668 issue a MAIL command.
1670 PH/02 In an ACL statement such as
1672 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1674 if a client was not listed at all, or was listed with a value other than
1675 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1676 the condition was not true (as it should be), so access was not denied.
1677 The bug was that the ! inversion was incorrectly passed on to the second
1678 item. This has been fixed.
1680 PH/03 Added additional dnslists conditions == and =& which are different from
1681 = and & when the dns lookup returns more than one IP address.
1683 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1684 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1686 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1687 FSYNC, which compiles an option called disable_fsync that allows for
1688 bypassing fsync(). The documentation is heavily laced with warnings.
1690 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1692 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1693 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1694 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1695 including adding "make clean"; (3) Added -fPIC when compiling the test
1696 dynamically loaded module, to get rid of a warning.
1698 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1699 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1700 = 0s. The bug is otherwise harmless.
1702 PH/07 There was a bug in the dovecot authenticator such that the value of
1703 $auth1 could be overwritten, and so not correctly preserved, after a
1704 successful authentication. This usually meant that the value preserved by
1705 the server_setid option was incorrect.
1707 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1709 PH/09 Installed PCRE release 7.0.
1711 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1712 run for batched SMTP input. It is now run at the start of every message
1713 in the batch. While fixing this I discovered that the process information
1714 (output by running exiwhat) was not always getting set for -bs and -bS
1715 input. This is fixed, and it now also says "batched" for BSMTP.
1717 PH/11 Added control=no_pipelining.
1719 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1720 patch, slightly modified), and move the expansion of helo_data till after
1721 the connection is made in the smtp transport (so it can use these
1724 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1726 PH/14 Added log_selector = +pid.
1728 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1730 PH/16 Add ${if forany and ${if forall.
1732 PH/17 Added dsn_from option to vary the From: line in DSNs.
1734 PH/18 Flush SMTP output before performing a callout, unless control =
1735 no_callout_flush is set.
1737 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1738 was true (the default) a successful delivery failed to delete the retry
1739 item, thus causing premature timeout of the address. The bug is now
1742 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1744 PH/21 Long custom messages for fakedefer and fakereject are now split up
1745 into multiline reponses in the same way that messages for "deny" and
1746 other ACL rejections are.
1748 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1749 with slight modification.
1751 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1752 draft, changing the syntax and factoring some duplicate code".
1754 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1755 for deliveries of the second and subsequent messages over the same SMTP
1758 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1759 ${reduce, with only minor "tidies".
1761 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1763 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1764 expansion side effects.
1766 PH/27 When a message times out after an over-quota error from an Exim-imposed
1767 quota, the bounce message says "mailbox is full". This message was not
1768 being given when it was a system quota that was exceeded. It now should
1771 MH/03 Made $recipients available in local_scan(). local_scan() already has
1772 better access to the recipient list through recipients_list[], but
1773 $recipients can be useful in postmaster-provided expansion strings.
1775 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1776 in the case of a MAIL command with additional options following the
1777 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1778 were accidentally chopped off.
1780 PH/29 SMTP synchronization checks are implemented when a command is read -
1781 there is a check that no more input is waiting when there shouldn't be
1782 any. However, for some commands, a delay in an ACL can mean that it is
1783 some time before the response is written. In this time, more input might
1784 arrive, invalidly. So now there are extra checks after an ACL has run for
1785 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1786 pipelining has not been advertised.
1788 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1790 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1791 returned in the SMTP response when smtp_return_error_details was false.
1792 This has been fixed.
1794 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1795 instead of the C I/O that was originally supplied, because problems were
1796 reported on Solaris.
1798 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1799 Exim which did not show up earlier: it was assuming that a call to
1800 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1801 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1802 was a macro that became an assignment, so it seemed to work. This has
1803 changed to a proper function call with a void return, hence the compile
1804 error. Exim's code has been fixed.
1806 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1809 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1811 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1813 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1814 because it turns out to be too big or there is a timeout) there is no
1815 "Completed" line in the log. When some messages of this type were
1816 selected by exigrep, they were listed as "not completed". Others were
1817 picked up by some special patterns. I have improved the selection
1818 criteria to be more general.
1820 PH/38 The host_find_failed option in the manualroute router can now be set
1821 to "ignore", to completely ignore a host whose IP address cannot be
1822 found. If all hosts are ignored, the behaviour is controlled by the new
1823 host_all_ignored option.
1825 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1826 homing or because of multiple MX records with /mx) generated more than
1827 one IP address, and the following item turned out to be the local host,
1828 all the secondary addresses of the first item were incorrectly removed
1829 from the list, along with the local host and any following hosts (which
1830 is what is supposed to happen).
1832 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1833 whoever called Exim into the -H file. In the case of the daemon it was
1834 behaving confusingly. When first started, it used values for whoever
1835 started the daemon, but after a SIGHUP it used the Exim user (because it
1836 calls itself on a restart). I have changed the code so that it now always
1839 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1840 message are rejected with the same error (e.g. no authentication or bad
1841 sender address), and a DATA command is nevertheless sent (as can happen
1842 with PIPELINING or a stupid MUA), the error message that was given to the
1843 RCPT commands is included in the rejection of the DATA command. This is
1844 intended to be helpful for MUAs that show only the final error to their
1847 PH/42 Another patch from the Sieve maintainer.
1849 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1850 Eximstats - Fixed some broken HTML links and added missing column headers
1852 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1853 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1855 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1857 PH/43 Yet another patch from the Sieve maintainer.
1859 PH/44 I found a way to check for a TCP/IP connection going away before sending
1860 the response to the final '.' that terminates a message, but only in the
1861 case where the client has not sent further data following the '.'
1862 (unfortunately, this is allowed). However, in many cases there won't be
1863 any further data because there won't be any more messages to send. A call
1864 to select() can be used: if it shows that the input is "ready", there is
1865 either input waiting, or the socket has been closed. An attempt to read
1866 the next input character can distinguish the two cases. Previously, Exim
1867 would have sent an OK response which the client would never have see.
1868 This could lead to message repetition. This fix should cure that, at
1869 least in a lot of common cases.
1871 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1872 advertised in response to EHLO.
1878 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1879 fixed by 4.65/MH/01 (is this a record?) are fixed:
1881 (i) An empty string was always treated as zero by the numeric comparison
1882 operators. This behaviour has been restored.
1884 (ii) It is documented that the numeric comparison operators always treat
1885 their arguments as decimal numbers. This was broken in that numbers
1886 starting with 0 were being interpreted as octal.
1888 While fixing these problems I realized that there was another issue that
1889 hadn't been noticed. Values of message_size_limit (both the global option
1890 and the transport option) were treated as octal if they started with 0.
1891 The documentation was vague. These values are now always treated as
1892 decimal, and I will make that clear in the documentation.
1898 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1899 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1902 MH/01 Don't check that the operands of numeric comparison operators are
1903 integers when their expansion is in "skipping" mode (fixes bug
1904 introduced by 4.64-PH/07).
1906 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1907 child addresses, Exim now panics and dies. Previously, because the count
1908 is held in a short int, deliveries were likely to be lost. As such a
1909 large number of recipients for a single message is ridiculous
1910 (performance will be very, very poor), I have chosen to impose a limit
1911 rather than extend the field.
1917 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1918 leftover -K file (the existence of which was triggered by #402).
1919 While we were at it, introduced process PID as part of the -K
1920 filename. This should rule out race conditions when creating
1923 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1924 processing considerably. Previous code took too long for large mails,
1925 triggering a timeout which in turn triggers #401.
1927 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1928 in the DK code in transports.c. sendfile() is not really portable,
1929 hence the _LINUX specificness.
1931 TF/01 In the add_headers option to the mail command in an Exim filter,
1932 there was a bug that Exim would claim a syntax error in any
1933 header after the first one which had an odd number of characters
1936 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1937 callout verification, Exim cached a "reject" for the entire domain. This
1938 is correct for most verifications, but it is not correct for a recipient
1939 verification with use_sender or use_postmaster set, because in that case
1940 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1941 case of MAIL FROM:<> rejection from other early rejections (e.g.
1942 rejection of HELO). When verifying a recipient using a non-null MAIL
1943 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1944 Whatever the result of the callout, the value of the domain cache is
1945 left unchanged (for any other kind of callout, getting as far as trying
1946 RCPT means that the domain itself is ok).
1948 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1951 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1952 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1953 session, the daemon ignores ECONNECT errors and logs others; it now
1954 ignores EPIPE as well.
1956 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1957 (quoted-printable decoding).
1959 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1960 later the small subsequent patch to fix an introduced bug.
1962 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1964 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1966 PH/08 An error is now given if message_size_limit is specified negative.
1968 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1969 to be given (somewhat) arbitrary names.
1971 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1974 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1975 miscellaneous code fixes
1977 PH/10 Added the log_reject_target ACL modifier to specify where to log
1980 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1981 hostname. This is wrong, because it relates to the incoming message (and
1982 probably the interface on which it is arriving) and not to the outgoing
1983 callout (which could be using a different interface). This has been
1984 changed to use the value of the helo_data option from the smtp transport
1985 instead - this is what is used when a message is actually being sent. If
1986 there is no remote transport (possible with a router that sets up host
1987 addresses), $smtp_active_hostname is used.
1989 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1990 tweaks were necessary in order to get it to work (see also 21 below):
1991 (a) The code assumed that strncpy() returns a negative number on buffer
1992 overflow, which isn't the case. Replaced with Exim's string_format()
1994 (b) There were several signed/unsigned issues. I just did the minimum
1995 hacking in of casts. There is scope for a larger refactoring.
1996 (c) The code used strcasecmp() which is not a standard C function.
1997 Replaced with Exim's strcmpic() function.
1998 (d) The code set only $1; it now sets $auth1 as well.
1999 (e) A simple test gave the error "authentication client didn't specify
2000 service in request". It would seem that Dovecot has changed its
2001 interface. Fortunately there's a specification; I followed it and
2002 changed what the client sends and it appears to be working now.
2004 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2007 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2008 address A is aliased to B and C, where B exists and C does not. Without
2009 -v the output is "A verified" because verification stops after a
2010 successful redirection if more than one address is generated. However,
2011 with -v the child addresses are also verified. Exim was outputting "A
2012 failed to verify" and then showing the successful verification for C,
2013 with its parentage. It now outputs "B failed to verify", showing B's
2014 parentage before showing the successful verification of C.
2016 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2017 look up a TXT record in a specific list after matching in a combined
2020 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2021 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2022 they consult the DNS. I had assumed they would set it the way they
2023 wanted; and indeed my experiments on Linux seem to show that in some
2024 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2025 To be on the safe side, however, I have now made the interface to
2026 host_find_byname() similar to host_find_bydns(), with an argument
2027 containing the DNS resolver options. The host_find_byname() function now
2028 sets these options at its start, just as host_find_bydns() does. The smtp
2029 transport options dns_qualify_single and dns_search_parents are passed to
2030 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2031 of host_find_byname() use the default settings of RES_DEFNAMES
2032 (qualify_single) but not RES_DNSRCH (search_parents).
2034 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2035 spool_read_header() do less string testing, by means of a preliminary
2036 switch on the second character of optional "-foo" lines. (This is
2037 overdue, caused by the large number of possibilities that now exist.
2038 Originally there were few.) While I was there, I also converted the
2039 str(n)cmp tests so they don't re-test the leading "-" and the first
2040 character, in the hope this might squeeze out yet more improvement.
2042 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2043 flag allowing group syntax was set by the header_syntax check but not
2044 turned off, possible causing trouble later; (2) The flag was not being
2045 set at all for the header_verify test, causing "group"-style headers to
2046 be rejected. I have now set it in this case, and also caused header_
2047 verify to ignore an empty address taken from a group. While doing this, I
2048 came across some other cases where the code for allowing group syntax
2049 while scanning a header line wasn't quite right (mostly, not resetting
2050 the flag correctly in the right place). These bugs could have caused
2051 trouble for malformed header lines. I hope it is now all correct.
2053 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2054 with the "reply" argument non-NULL. The code, however (which originally
2055 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2056 but it didn't always do it. This confused somebody who was copying the
2057 code for some other use. I have removed all the tests.
2059 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2060 feature that was used to support insecure browsers during the U.S. crypto
2061 embargo. It requires special client support, and Exim is probably the
2062 only MTA that supported it -- and would never use it because real RSA is
2063 always available. This code has been removed, because it had the bad
2064 effect of slowing Exim down by computing (never used) parameters for the
2065 RSA_EXPORT functionality.
2067 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2068 authenticator to fail if there's a tab character in the incoming data
2069 (there should never be unless someone is messing about, as it's supposed
2070 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2071 if the connection is using TLS or if the remote IP is the same as the
2072 local IP, and the "valid-client-cert option" if a client certificate has
2075 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2076 authenticators. This can be used for authorization after authentication
2077 succeeds. (In the case of plaintext, it servers for both authentication
2080 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2081 if any retry times were supplied.
2083 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2084 connection, where there is no client IP address to check. In this
2085 situation, the verify now always succeeds.
2087 PH/25 Applied John Jetmore's -Mset patch.
2089 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2091 PH/27 In a string expansion for a processed (not raw) header when multiple
2092 headers of the same name were present, leading whitespace was being
2093 removed from all of them, but trailing whitespace was being removed only
2094 from the last one. Now trailing whitespace is removed from each header
2095 before concatenation. Completely empty headers in a concatenation (as
2096 before) are ignored.
2098 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2099 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2101 PH/29 [Removed. This was a change that I later backed out, and forgot to
2102 correct the ChangeLog entry (that I had efficiently created) before
2103 committing the later change.]
2105 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2106 address errors (4xx response to RCPT) over the same connection as other
2107 messages routed to the same hosts. Such deliveries are always "forced",
2108 so retry times are not inspected. This resulted in far too many retries
2109 for the affected addresses. The effect occurred only when there were more
2110 hosts than the hosts_max_try setting in the smtp transport when it had
2111 the 4xx errors. Those hosts that it had tried were not added to the list
2112 of hosts for which the message was waiting, so if all were tried, there
2113 was no problem. Two fixes have been applied:
2115 (i) If there are any address or message errors in an SMTP delivery, none
2116 of the hosts (tried or untried) are now added to the list of hosts
2117 for which the message is waiting, so the message should not be a
2118 candidate for sending over the same connection that was used for a
2119 successful delivery of some other message. This seems entirely
2120 reasonable: after all the message is NOT "waiting for some host".
2121 This is so "obvious" that I'm not sure why it wasn't done
2122 previously. Hope I haven't missed anything, but it can't do any
2123 harm, as the worst effect is to miss an optimization.
2125 (ii) If, despite (i), such a delivery is accidentally attempted, the
2126 routing retry time is respected, so at least it doesn't keep
2127 hammering the server.
2129 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2130 in ${readsocket because some servers need this prod.
2132 PH/32 Added some extra debug output when updating a wait-xxx database.
2134 PH/33 The hint "could be header name not terminated by colon", which has been
2135 given for certain expansion errors for a long time, was not being given
2136 for the ${if def:h_colon_omitted{... case.
2138 PH/34 The spec says: "With one important exception, whenever a domain list is
2139 being scanned, $domain contains the subject domain." There was at least
2140 one case where this was not true.
2142 PH/35 The error "getsockname() failed: connection reset by peer" was being
2143 written to the panic log as well as the main log, but it isn't really
2144 panic-worthy as it just means the connection died rather early on. I have
2145 removed the panic log writing for the ECONNRESET error when getsockname()
2148 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2149 runs only) independently of the message's sender address. This meant
2150 that, if the 4xx error was in fact related to the sender, a different
2151 message to the same recipient with a different sender could confuse
2152 things. In particualar, this can happen when sending to a greylisting
2153 server, but other circumstances could also provoke similar problems.
2154 I have changed the default so that the retry time for these errors is now
2155 based a combination of the sender and recipient addresses. This change
2156 can be overridden by setting address_retry_include_sender=false in the
2159 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2160 remote server are returned as part of bounce messages. This was not
2161 happening for LMTP over a pipe (the lmtp transport), but now it is the
2162 same for both kinds of LMTP.
2164 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2165 in header lines that were in fact CNAMEs. This is no longer the case.
2167 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2168 and queue runs started by the daemon processed all messages. This has
2169 been fixed so that -R and -S can now usefully be given with -q<time>.
2171 PH/40 Import PCRE release 6.7 (fixes some bugs).
2173 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2175 PH/42 Give an error if -q is specified more than once.
2177 PH/43 Renamed the variables $interface_address and $interface_port as
2178 $received_ip_address and $received_port, to make it clear that these
2179 values apply to message reception, and not to the outgoing interface when
2180 a message is delivered. (The old names remain recognized, of course.)
2182 PH/44 There was no timeout on the connect() call when using a Unix domain
2183 socket in the ${readsocket expansion. There now is.
2185 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2186 be meaningful with "accept".
2188 SC/01 Eximstats V1.43
2189 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2191 SC/02 Eximstats V1.44
2192 Use a glob alias rather than an array ref in the generated
2193 parser. This improves both readability and performance.
2195 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2196 Collect SpamAssassin and rejection statistics.
2197 Don't display local sender or destination tables unless
2198 there is data to show.
2199 Added average volumes into the top table text output.
2201 SC/04 Eximstats V1.46
2202 Collect data on the number of addresses (recipients)
2203 as well as the number of messages.
2205 SC/05 Eximstats V1.47
2206 Added 'Message too big' to the list of mail rejection
2207 reasons (thanks to Marco Gaiarin).
2209 SC/06 Eximstats V1.48
2210 Mainlog lines which have GMT offsets and are too short to
2211 have a flag are now skipped.
2213 SC/07 Eximstats V1.49 (Alain Williams)
2214 Added the -emptyok flag.
2216 SC/08 Eximstats V1.50
2217 Fixes for obtaining the IP address from reject messages.
2219 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2220 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2221 whitesspace changes from 4.64-PH/27
2223 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2226 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2227 are found, allow negative numbers in numeric criteria)
2229 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2231 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2232 to match changes made in 4.64-PH/43
2234 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2236 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2237 "domain" is actually an IP address, and force a failure. This locks out
2238 those revolvers/nameservers that support "A-for-A" lookups, in
2239 contravention of the specifications.
2241 PH/48 When a host name was looked up from an IP address, and the subsequent
2242 forward lookup of the name timed out, the host name was left in
2243 $sender_host_name, contrary to the specification.
2245 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2246 restricted to single-key lookups, Exim was not diagnosing an error if
2247 * or *@ was used with a query-style lookup.
2249 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2251 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2252 long ago, but noone interested enough thought of it. Let's just say that
2253 the "1.1" means that there are some new functions that weren't there at
2254 some point in the past.
2256 PH/51 Error processing for expansion failure of helo_data from an smtp
2257 transport during callout processing was broken.
2259 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2260 tested/used via the -bh/-bhc/-bs options.
2262 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2263 bug, fixed in subsequent PCRE releases).
2265 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2266 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2268 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2274 SC/01 Use a glob alias rather than an array ref in eximstats generated
2275 parser. This improves both readability and performance.
2277 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2278 Don't display local sender or destination tables in eximstats unless
2279 there is data to show.
2280 Added average volumes into the eximstats top table text output.
2282 SC/03 Collect data on the number of addresses (recipients) as well
2283 as the number of messages in eximstats.
2285 TF/01 Correct an error in the documentation for the redirect router. Exim
2286 does (usually) call initgroups() when daemonizing.
2288 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2289 with consistent privilege compared to when running as a daemon.
2291 TF/03 Note in the spec that $authenticated_id is not set for local
2292 submissions from trusted users.
2294 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2295 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2297 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2298 by adding some example configuration directives to the default
2299 configuration file. A little bit of work is required to uncomment the
2300 directives and define how usernames and passwords are checked, but
2301 there is now a framework to start from.
2303 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2304 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2305 without this. I don't know how relevant this is to other LDAP libraries.
2307 PH/02 Add the verb name to the "unknown ACL verb" error.
2309 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2311 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2313 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2314 directory not expanded when it should be if an expanded home directory
2315 was set for the address (which is overridden by the transport).
2317 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2320 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2321 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2322 because it is too late at that time, and has no effect.
2324 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2325 security issue with \' (bugzilla #107). I could not use the
2326 PQescapeStringConn() function, because it needs a PGconn value as one of
2329 PH/08 When testing addresses using -bt, indicate those final addresses that
2330 are duplicates that would not cause an additional delivery. At least one
2331 person was confused, thinking that -bt output corresponded to deliveries.
2332 (Suppressing duplicates isn't a good idea as you lose the information
2333 about possibly different redirections that led to the duplicates.)
2335 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2336 systems where poll() doesn't work, in particular OS X.
2338 PH/10 Added more information to debugging output for retry time not reached.
2340 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2341 operations in malware.c.
2343 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2346 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2347 syslog_duplication set false, log lines that would normally be written
2348 both the the main log and to the reject log were not written to syslog at
2351 PH/14 In the default configuration, change the use of "message" in ACL warn
2352 statements to "add_header".
2354 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2355 not followed by a command (e.g. "seen endif").
2357 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2358 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2361 PH/17 Added extra conditions to the default value of delay_warning_condition
2365 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2366 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2367 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2370 The Auto-Submitted: and various List- headers are standardised, whereas I
2371 don't think Precedence: ever was.
2373 PH/18 Refactored debugging code in route_finduser() to show more information,
2374 in particular, the error code if getpwnam() issues one.
2376 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2377 This is apparently needed in addition to the PH/07 change above to avoid
2378 any possible encoding problems.
2380 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2381 but not after initializing Perl.
2383 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2384 output them only if debugging. By default they are written stderr,
2385 apparently, which is not desirable.
2387 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2390 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2393 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2395 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2396 authenticated or an ident call has been made. Suppress the default
2397 values for $authenticated_id and $authenticated_sender (but permit -oMai
2398 and -oMas) when testing with -bh.
2400 PH/24 Re-jigged the order of the tests in the default configuration so that the
2401 tests for valid domains and recipients precede the DNS black list and CSA
2402 tests, on the grounds that those ones are more expensive.
2404 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2405 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2406 This bug exists in every version of Exim that I still have, right back to
2409 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2410 However, an attempt to turn on case-sensitivity in a regex key by
2411 including (?-i) didn't work because the subject string was already
2412 lowercased, and the effects were non-intuitive. It turns out that a
2413 one-line patch can be used to allow (?-i) to work as expected.
2419 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2420 other effects) broke the use of negated acl sub-conditions.
2422 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2425 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2426 "Deny" causes Exim to reject the incoming connection with a 554 error.
2427 Unfortunately, if there is a major crisis, such as a disk failure,
2428 tcp-wrappers gives "deny", whereas what one would like would be some
2429 kind of temporary error. A kludge has been added to help with this.
2430 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2431 554 error is used if errno is still zero or contains ENOENT (which occurs
2432 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2435 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2437 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2438 errors. Otherwise a message that provokes a temporary error (when other
2439 messages do not) can cause a whole host to time out.
2441 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2442 addresses were routed directly to files or pipes from a redirect router.
2443 File deliveries just didn't batch; pipe deliveries might have suffered
2446 PH/06 A failure to get a lock for a hints database would erroneously always say
2447 "Failed to get write lock", even when it was really a read lock.
2449 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2450 of 0600. This has been changed to use the value of the lockfile_mode
2451 option (which defaults to 0600).
2453 PH/08 Applied small patch from the Sieve maintainer.
2455 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2456 folder from quota calculations, a direct delivery into this folder messed
2457 up the contents of the maildirsize file. This was because the regex was
2458 used only to exclude .Trash (or whatever) when the size of the mailbox
2459 was calculated. There was no check that a delivery was happening into an
2460 excluded directory. This bug has been fixed by ignoring all quota
2461 processing for deliveries into excluded directories.
2463 PH/10 Added the maildirfolder_create_regex option to appendfile.
2469 PH/01 The code for finding all the local interface addresses on a FreeBSD
2470 system running IPv6 was broken. This may well have applied to all BSD
2471 systems, as well as to others that have similar system calls. The broken
2472 code found IPv4 interfaces correctly, but gave incorrect values for the
2473 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2474 that it would not match correctly against @[] and not recognize the IPv6
2477 PH/02 The ipliteral router was not recognizing addresses of the form user@
2478 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2480 PH/03 Added disable_ipv6.
2482 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2483 decoded form, because it is most often used to construct To: headers
2484 lines in autoreplies, and the decoded form may well be syntactically
2485 invalid. However, $reply_address has leading white space removed, and all
2486 newlines turned into spaces so that the autoreply transport does not
2489 PH/05 If group was specified without a user on a router, and no group or user
2490 was specified on a transport, the group from the router was ignored.
2492 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2493 for visible compile-time settings that can be used to change these
2494 numbers, for those that want even more. Backwards compatibility with old
2495 spool files has been maintained. However, going back to a previous Exim
2496 release will lost any variables that are in spool files.
2498 PH/07 Two small changes when running in the test harness: increase delay when
2499 passing a TCP/IP connection to a new process, in case the original
2500 process has to generate a bounce, and remove special handling of
2501 127.0.0.2 (sic), which is no longer necessary.
2503 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2504 be the same on different OS.
2506 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2509 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2510 whether --show-vars was specified or not
2512 JJ/02 exipick: Added support for new ACL variable spool format introduced
2515 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2516 syntactically invalid From: or Reply-to: line, and a filter used this to
2517 generate an autoreply, and therefore failed to obtain an address for the
2518 autoreply, Exim could try to deliver to a non-existent relative file
2519 name, causing unrelated and misleading errors. What now happens is that
2520 it logs this as a hard delivery error, but does not attempt to create a
2523 PH/11 The exinext utility has a -C option for testing purposes, but although
2524 the given file was scanned by exinext itself; it wasn't being passed on
2525 when Exim was called.
2527 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2528 an end-of-file indication when reading a command response.
2530 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2531 compiled. In many other places in Exim, IPv6 addresses are always
2532 recognized, so I have changed this. It also means that IPv4 domain
2533 literals of the form [IPV4:n.n.n.n] are now always recognized.
2535 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2536 used if the router is not running as root, for example, when verifying at
2537 ACL time, or when using -bh. The debugging output from this situation was
2538 non-existent - all you got was a failure to exec. I have made two
2541 (a) Failures to set uid/gid, the current directory, or a process leader
2542 in a subprocess such as that created by queryprogram now generate
2543 suitable debugging ouput when -d is set.
2545 (b) The queryprogram router detects when it is not running as root,
2546 outputs suitable debugging information if -d is set, and then runs
2547 the subprocess without attempting to change uid/gid.
2549 PH/15 Minor change to Makefile for building test_host (undocumented testing
2552 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2553 additional section of a DNS packet that returns MX or SRV records.
2554 Instead, it always explicitly searches for A/AAAA records. This avoids
2555 major problems that occur when a DNS server includes only records of one
2556 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2557 fixed another bug: if SRV records were looked up and the corresponding
2558 address records were *not* found in the additional section, the port
2559 values from the SRV records were lost.
2561 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2562 using the correct key (the original address) when searching the retry
2563 rules in order to find which one to use for generating the retry hint.
2565 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2566 adding the default one. Similarly, if it contains a Reply-To: header, the
2567 errors_reply_to option, if set, is not used.
2569 PH/19 When calculating a retry time, Exim used to measure the "time since
2570 failure" by looking at the "first failed" field in the retry record. Now
2571 it does not use this if it is later than than the arrival time of the
2572 message. Instead it uses the arrival time. This makes for better
2573 behaviour in cases where some deliveries succeed, thus re-setting the
2574 "first failed" field. An example is a quota failure for a huge message
2575 when small messages continue to be delivered. Without this change, the
2576 "time since failure" will always be short, possible causing more frequent
2577 delivery attempts for the huge message than are intended.
2578 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2580 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2581 $1, $2, $3) because the numerical variables can be reset during some
2582 expansion items (e.g. "match"), thereby losing the authentication data.
2584 PH/21 Make -bV show the size of off_t variables so that the test suite can
2585 decide whether to run tests for quotas > 2G.
2587 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2588 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2589 filecount value is greater than 2G or if a quota value is greater than 2G
2590 on a system where the size of off_t is not greater than 4, a panic error
2593 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2594 never match. The debug and -bh output now contains an explicit error
2595 message indicating a malformed IPv4 address or mask.
2597 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2598 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2599 PH/23 above applies.
2601 PH/25 Do not write to syslog when running in the test harness. The only
2602 occasion when this arises is a failure to open the main or panic logs
2603 (for which there is an explicit test).
2605 PH/26 Added the /no_tell option to "control=freeze".
2607 PH/27 If a host name lookup failed very early in a connection, for example, if
2608 the IP address matched host_lookup and the reverse lookup yielded a name
2609 that did not have a forward lookup, an error message of the form "no IP
2610 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2611 could be logged. Now it outputs the IP address instead of "NULL".
2613 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2614 allows the sender and the authenticated sender to be set when
2615 submitting a message from within Exim. Since child_open_exim() is
2616 documented for local_scan(), the new function should be too.
2618 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2619 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2620 results in an empty string is now treated as unset.
2622 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2624 PH/31 Added sender_verify_fail logging option.
2626 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2627 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2628 tidied the source and removed it altogether.
2630 PH/33 When a queue run was abandoned because the load average was too high, a
2631 log line was always written; now it is written only if the queue_run log
2632 selector is set. In addition, the log line for abandonment now contains
2633 information about the queue run such as the pid. This is always present
2634 in "start" and "stop" lines but was omitted from the "abandon" line.
2636 PH/34 Omit spaces between a header name and the colon in the error message that
2637 is given when verify = headers_syntax fails (if there are lots of them,
2638 the message gets confusing).
2640 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2641 names, as there are now some PTR records that contain slashes. This check
2642 is only to protect against broken name servers that fall over on strange
2643 characters, so the fact that it applies to all lookups doesn't matter.
2645 PH/36 Now that the new test suite is complete, we can remove some of the
2646 special code in Exim that was needed for the old test suite. For example,
2647 sorting DNS records because real resolvers return them in an arbitrary
2648 order. The new test suite's fake resolver always returns records in the
2651 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2652 bounces) except when queue_only is set, to avoid logging races between
2653 the different processes.
2655 PH/38 Panic-die if .include specifies a non-absolute path.
2657 PH/39 A tweak to the "H" retry rule from its user.
2659 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2660 a label. They prevented compilation on older perls.
2662 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2663 a warning to be raised on newish perls.
2665 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2666 on queue. Changes to match documented behaviour of showing count of
2667 messages matching specified criteria.
2669 PH/40 Changed the default ident timeout from 30s to 5s.
2671 PH/41 Added support for the use of login_cap features, on those BSD systems
2672 that have them, for controlling the resources used by pipe deliveries.
2674 PH/42 The content-scanning code uses fopen() to create files in which to put
2675 message data. Previously it was not paying any attention to the mode of
2676 the files. Exim runs with umask(0) because the rest of the code creates
2677 files with open(), and sets the required mode explicitly. Thus, these
2678 files were ending up world-writeable. This was not a big issue, because,
2679 being within the spool directory, they were not world-accessible. I have
2680 created a function called modefopen, which takes an additional mode
2681 argument. It sets umask(777), creates the file, chmods it to the required
2682 mode, then resets the umask. All the relevant calls to fopen() in the
2683 content scanning code have been changed to use this function.
2685 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2686 to 24 hours. This avoids potential overflow problems when processing G
2687 and H retry rules. I suspect nobody ever tinkers with this value.
2689 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2691 PH/45 When the plaintext authenticator is running as a client, the server's
2692 challenges are checked to ensure they are valid base64 strings. By
2693 default, the authentication attempt is cancelled if an invalid string is
2694 received. Setting client_ignore_invalid_base64 true ignores these errors.
2695 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2696 they are received. Thus, the responses can be made to depend on the
2697 challenges. If an invalid string is ignored, an empty string is placed in
2700 PH/46 Messages that are created by the autoreply transport now contains a
2701 References: header, in accordance with RFCs 2822 and 3834.
2703 PH/47 Added authenticated_sender_force to the smtp transport.
2705 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2707 PH/49 Installed latest patch from the Sieve maintainer.
2709 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2710 was also set, the appendfile transport was unnecessarily scanning a
2711 directory of message files (e.g. for maildir delivery) to find the count
2712 of files (along with the size), even though it did not need this
2713 information. It now does the scan only if it needs to find either the
2714 size of the count of files.
2716 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2718 PH/52 Two bugs concerned with error handling when the smtp transport is
2721 (i) Exim was not creating retry information for temporary errors given
2722 for individual recipients after the DATA command when the smtp transport
2723 was used in LMTP mode. This meant that they could be retried too
2724 frequently, and not timed out correctly.
2726 (ii) Exim was setting the flag that allows error details to be returned
2727 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2728 recipients that were returned after the DATA command.
2730 PH/53 This is related to PH/52, but is more general: for any failing address,
2731 when detailed error information was permitted to be returned to the
2732 sender, but the error was temporary, then after the final timeout, only
2733 "retry timeout exceeded" was returned. Now it returns the full error as
2734 well as "retry timeout exceeded".
2736 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2737 do this, and (what is worse) MTAs that accept it.
2739 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2740 will now be deprecated.
2742 PH/56 New os.c-cygwin from the Cygwin maintainer.
2744 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2745 formats (previously only available in exim formats via -bpr, -bpru,
2746 and -bpra. Now also available in native and exiqgrep formats)
2748 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2749 with very large, slow to parse queues
2751 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2753 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2755 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2756 responses to authentication challenges, though it was showing the
2757 challenges; (ii) I've removed the CR characters from the debug output for
2760 PH/58 Allow for the insertion of a newline as well as a space when a string
2761 is turned into more than one encoded-word during RFC 2047 encoding. The
2762 Sieve code now uses this.
2764 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2765 data_4xx, lost_connection, tls_required.
2767 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2768 message was being sent as an SMTP response.
2770 PH/61 Add -l and -k options to exicyclog.
2772 PH/62 When verifying, if an address was redirected to one new address, so that
2773 verification continued, and the new address failed or deferred after
2774 having set something in $address_data, the value of $address_data was not
2775 passed back to the ACL. This was different to the case when no
2776 redirection occurred. The value is now passed back in both cases.
2778 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2779 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2780 use login_cap.h, so on its own it isn't the distinguishing feature. The
2781 new name refers directly to the setclassresources() function.
2783 PH/65 Added configuration files for NetBSD3.
2785 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2787 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2788 is preferred over IPv4.
2790 PH/68 The bounce_return_message and bounce_return_body options were not being
2791 honoured for bounces generated during the reception of non-SMTP messages.
2792 In particular, this applied to messages rejected by the ACL. This bug has
2793 been fixed. However, if bounce_return_message is true and bounce_return_
2794 body is false, the headers that are returned for a non-SMTP message
2795 include only those that have been read before the error was detected.
2796 (In the case of an ACL rejection, they have all been read.)
2798 PH/69 The HTML version of the specification is now built in a directory called
2799 spec_html instead of spec.html, because the latter looks like a path with
2800 a MIME-type, and this confuses some software.
2802 PH/70 Catch two compiler warnings in sieve.c.
2804 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2805 function verify_get_ident() calls ip_connect() to connect a socket, but
2806 if the "connect()" function timed out, ip_connect() used to close the
2807 socket. However, verify_get_ident() also closes the socket later, and in
2808 between Exim writes to the log, which may get opened at this point. When
2809 the socket was closed in ip_connect(), the log could get the same file
2810 descriptor number as the socket. This naturally causes chaos. The fix is
2811 not to close the socket in ip_connect(); the socket should be closed by
2812 the function that creates it. There was only one place in the code where
2813 this was missing, in the iplookup router, which I don't think anybody now
2814 uses, but I've fixed it anyway.
2816 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2817 well as to direct DNS lookups. Otherwise the handling of names in host
2818 lists is inconsistent and therefore confusing.
2824 PH/01 Two changes to the default runtime configuration:
2826 (1) Move the checks for relay_from_hosts and authenticated clients from
2827 after to before the (commented out) DNS black list checks.
2829 (2) Add control=submission to the relay_from_hosts and authenticated
2830 clients checks, on the grounds that messages accepted by these
2831 statements are most likely to be submissions.
2833 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2835 (1) Generate an error if the third argument for the ${prvs expansion is
2838 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2841 (3) Reset the variables that are obtained from the first argument of
2842 ${prvscheck and used in the second argument before leaving the code,
2843 because their memory is reclaimed, so using them afterwards may do
2846 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2847 one (it's much easier than Tom thought :-).
2849 (5) Because of (4), we can now allow for the use of $prvscheck_result
2850 inside the third argument.
2852 PH/03 For some reason, the default setting of PATH when running a command from
2853 a pipe transport was just "/usr/bin". I have changed it to
2856 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2857 anything to be listed in the output from -bV.
2859 PH/05 When a filter generated an autoreply, the entire To: header line was
2860 quoted in the delivery log line, like this:
2862 => >A.N.Other <ano@some.domain> <original@ddress> ...
2864 This has been changed so that it extracts the operative address. There
2865 may be more than one such address. If so, they are comma-separated, like
2868 => >ano@some.domain,ona@other.domain <original@ddress> ...
2870 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2871 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2872 not being looked up in the rDNS to get a host name, Exim was showing the
2873 IP address twice in Received: lines, even though the IP addresses were
2874 identical. For example:
2876 Received: from [1.2.3.4] (helo=[1.2.3.4])
2878 However, if the real host name was known, it was omitting the HELO data
2879 if it matched the actual IP address. This has been tidied up so that it
2880 doesn't show the same IP address twice.
2882 PH/07 When both +timestamp and +memory debugging was on, the value given by
2883 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2884 called by the debug printing, thereby overwriting the timestamp buffer.
2885 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2887 PH/08 When the original message was included in an autoreply transport, it
2888 always said "this is a copy of the message, including all the headers",
2889 even if body_only or headers_only was set. It now gives an appropriate
2892 PH/09 Applied a patch from the Sieve maintainer which:
2894 o fixes some comments
2895 o adds the (disabled) notify extension core
2896 o adds some debug output for the result of if/elsif tests
2897 o points to the current vacation draft in the documentation
2898 and documents the missing references header update
2902 o fixes a bug in processing the envelope test (when testing
2903 multiple envelope elements, the last element determinted the
2906 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2907 Electronic Mail") by including:
2909 Auto-submitted: auto-generated
2911 in the messages that it generates (bounce messages and others, such as
2912 warnings). In the case of bounce messages for non-SMTP mesages, there was
2913 also a typo: it was using "Auto_submitted" (underscore instead of
2914 hyphen). Since every message generated by Exim is necessarily in response
2915 to another message, thes have all been changed to:
2917 Auto-Submitted: auto-replied
2919 in accordance with these statements in the RFC:
2921 The auto-replied keyword:
2923 - SHOULD be used on messages sent in direct response to another
2924 message by an automatic process,
2926 - MUST NOT be used on manually-generated messages,
2928 - MAY be used on Delivery Status Notifications (DSNs) and Message
2929 Disposition Notifications (MDNs),
2931 - MUST NOT be used on messages generated by automatic or periodic
2932 processes, except for messages which are automatic responses to
2935 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2936 to the default Received: header definition.
2938 PH/12 Added log selector acl_warn_skipped (default on).
2940 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2941 variables because (a) they are in the wrong storage pool and (b) even if
2942 they were copied, it wouldn't work properly because of the caching.
2944 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2945 checking when decoding. Apparently there are clients that generate
2946 overlong encoded strings. Why am I not surprised?
2948 PH/15 If the first argument of "${if match_address" was not empty, but did not
2949 contain an "@" character, Exim crashed. Now it writes a panic log message
2950 and treats the condition as false.
2952 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2954 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2955 extension "envelope-auth". The code is finished and in agreement with
2956 other implementations, but there is no documentation so far and in fact,
2957 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2958 not changing the active code.
2960 Print executed "if" and "elsif" statements when debugging is used. This
2961 helps a great deal to understand what a filter does.
2963 Document more things not specified clearly in RFC3028. I had all this
2964 sorted out, when out of a sudden new issues came to my mind. Oops."
2966 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2969 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2970 canonical form (as documented). However, after a host name lookup from
2971 the IP address, check_host() was doing a simple string comparison with
2972 addresses acquired from the DNS when checking that the found name did
2973 have the original IP as one of its addresses. Since any found IPv6
2974 addresses are likely to be in abbreviated form, the comparison could
2975 fail. Luckily, there already exists a function for doing the comparison
2976 by converting both addresses to binary, so now that is used instead of
2977 the text comparison.
2979 PH/20 There was another similar case to PH/19, when a complete host name was
2980 given in a host list; looking up its IP address could give an abbreviated
2981 form, whereas the current host's name might or might not be abbreviated.
2982 The same fix has been applied.
2988 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2989 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2992 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2993 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2995 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2997 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2998 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2999 Also, for "Auto-Submitted", treat anything other than "no" as
3000 non-personal, in accordance with RFC 3834. (Previously it treated
3001 anything starting "auto-" as non-personal.)
3003 TF/01 The control=submission/name=... option had a problem with syntax
3004 errors if the name included a slash character. The /name= option
3005 now slurps the rest of the string, so it can include any characters
3006 but it must come last in the list of options (after /sender_retain
3009 PH/05 Some modifications to the interface to the fake nameserver for the new
3017 TK/01 Added the "success_on_redirect" address verification option. See
3018 NewStuff for rationale and an example.
3020 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3022 PH/02 Patch to exigrep to allow it to work on syslog lines.
3024 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3025 fread() to skip over the body file's header line, because in Cygwin the
3026 header line is locked and is inaccessible.
3028 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3029 co-exist for some time) to make it clear that it is the Exim ID that is
3030 referenced, not the Message-ID: header line.
3032 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3033 string_format() function, because snprintf() does not exist on all
3036 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3037 ${stat: expansion item.
3039 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3040 protocol synchronization error", to keep the pedants happy.
3042 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3043 well as for IRIX systems, when gcc is being used. See the host.c source
3046 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3048 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3051 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3052 transport and to the smtp transport in LMTP mode.
3054 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3056 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3057 run a filter in a subprocess. This could lead to confusion in subsequent
3058 lookups in the parent process. There should also be a search_tidyup() at
3059 the end of the subprocess.
3061 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3062 only if the host matched helo_try_verify_hosts, which caused the
3063 verification to occur when the EHLO/HELO command was issued. The ACL just
3064 tested the remembered result. Now, if a previous verification attempt has
3065 not happened, "verify = helo" does it there and then.
3067 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3069 TK/03 Fix log output including CR from clamd.
3071 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3072 exist provoked a memory error which could cause a segfault.
3074 PH/15 Installed PCRE 6.2
3076 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3078 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3079 of the problem. Specifically, suggested +O2 rather than +O1 for the
3082 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3084 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3085 record, the port was not correctly logged when the outgoing_port log
3086 selector was set (it logged the transort's default port).
3088 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3089 fallback_hosts, and "hosts" in the smtp transport.
3091 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3092 host errors such as "Connection refused".
3094 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3095 authentication with radiusclient 0.4.9:
3097 - Error returned from rc_read_config was caught wrongly
3098 - Username/password not passed on to radius server due to wrong length.
3100 The presumption is that some radiusclient API changes for 4.51/PH/17
3101 were not taken care of correctly. The code is still untested by me (my
3102 Linux distribution still has 0.3.2 of radiusclient), but it was
3103 contributed by a Radius user.
3105 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3106 expanding the "port" option of the smtp transport.
3108 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3109 while reading a MIME header. Thanks to Tom Hughes for a patch.
3111 PH/24 Include config.h inside local_scan.h so that configuration settings are
3114 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3115 that in an ACL for RCPT (for example), you can examine exactly what was
3118 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3119 commands, but it was not correctly comparing the address with the actual
3120 client host address. Thus, it would show the EHLO address in Received:
3121 header lines when this was not necessary.
3123 PH/27 Added the % operator to ${eval:}.
3125 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3126 it should be ignoring failures (because with -C, for example, it has lost
3127 privilege). It wasn't ignoring creation failures other than "already
3130 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3133 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3134 because an input line was too long, either on its own, or by virtue of
3135 too many continuations, the temporary file was not being removed, and the
3136 return code was incorrect.
3138 PH/31 Missing "BOOL" in function definition in filtertest.c.
3140 PH/32 Applied Sieve patches from the maintainer.
3142 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3144 PH/33 Added "verify = not_blind".
3146 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3147 Local/Makefile (with some defaults set). These are used in built scripts
3148 such as exicyclog, but they have never been used in the exim_install
3149 script (though there are many overriding facilities there). I have
3150 arranged that the exim_install script now takes note of these two
3153 PH/35 Installed configuration files for Dragonfly.
3155 PH/36 When a locally submitted message by a trusted user did not contain a
3156 From: header, and the sender address was obtained from -f or from an SMTP
3157 MAIL command, and the trusted user did not use -F to supply a sender
3158 name, $originator_name was incorrectly used when constructing a From:
3159 header. Furthermore, $originator_name was used for submission mode
3160 messages from external hosts without From: headers in a similar way,
3161 which is clearly wrong.
3163 PH/37 Added control=suppress_local_fixups.
3165 PH/38 When log_selector = +received_sender was set, and the addition of the
3166 sender made the log line's construction buffer exactly full, or one byte
3167 less than full, an overflow happened when the terminating "\n" was
3170 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3171 when the result of a list match is failure because a DNS lookup failed.
3173 PH/40 RM_COMMAND is now used in the building process.
3175 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3176 the "build-* directories that it finds.
3178 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3179 address in a domain literal was a prefix of an interface address.
3181 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3182 when verifying a sender address, unless rewrite_headers is false.
3184 PH/44 Wrote a long comment about why errors_to addresses are verified as
3185 recipients, not senders.
3187 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3188 the ratelimit ACL was added.
3190 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3192 PH/46 Added extra information about PostgreSQL errors to the error string.
3194 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3195 avoiding the need to install special zones in a real server. This is
3196 backwards compatible; if it can't find the fake resolver, it drops back.
3197 Thus, both old and new test suites can be run.
3199 TF/02 Added util/ratelimit.pl
3201 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3202 clock is set back in time.
3204 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3205 Candler <B.Candler@pobox.com>.
3207 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3208 OK for addresses that are the result of redirections.
3210 PH/48 A number of further additions for the benefit of the new test suite,
3211 including a fake gethostbyname() that interfaces to the fake DNS resolver
3214 TF/06 The fix for widen_domains has also been applied to qualify_single and
3215 search_parents which are the other dnslookup options that can cause
3218 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3221 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3223 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3224 service" if it encountered an MX record with an empty target hostname.
3225 The message is now "an MX or SRV record indicated no SMTP service".
3227 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3228 if the DNS of the sending site is misconfigured. This is quite a
3229 common situation. This change restores the behaviour of treating a
3230 helo verification defer as a failure.
3232 PH/51 If self=fail was set on a router, the bounce message did not include the
3233 actual error message.
3239 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3241 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3242 command itself ended in error, the underlying message about the transport
3243 filter timeout was being overwritten with the pipe command error. Now the
3244 underlying error message should be appended to the second error message.
3246 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3248 PH/02 Reduce the amount of output that "make" produces by default. Full output
3249 can still be requested.
3251 PH/03 The warning log line about a condition test deferring for a "warn" verb
3252 was being output only once per connection, rather than after each
3253 occurrence (because it was using the same function as for successful
3254 "warn" verbs). This seems wrong, so I have changed it.
3256 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3257 it should not have, which might have caused a crash in the right
3258 circumstances, but probably never did.
3260 PH/04 Installed a modified version of Tony Finch's patch to make submission
3261 mode fix the return path as well as the Sender: header line, and to
3262 add a /name= option so that you can make the user's friendly name appear
3265 TF/03 Added the control = fakedefer ACL modifier.
3267 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3268 Mark Lowes for thorough testing.
3270 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3272 TK/03 Merged latest SRS patch from Miles Wilton.
3274 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3275 with the definition in sysexits.h (which is #included earlier).
3276 Fortunately, Exim does not actually use EX_OK. The code used to try to
3277 preserve the sysexits.h value, by assumimg that macro definitions were
3278 scanned for macro replacements. I have been disabused of this notion,
3279 so now the code just undefines EX_OK before #including unistd.h.
3281 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3282 in the smtp transport. When a block could not be written in a single
3283 write() function, the timeout was being re-applied to each part-write.
3284 This seems wrong - if the receiver was accepting one byte at a time it
3285 would take for ever. The timeout is now adjusted when this happens. It
3286 doesn't have to be particularly precise.
3288 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3289 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3291 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3292 without a domain if the check to <postmaster@domain> fails.
3294 SC/01 Eximstats: added -xls and the ability to specify output files
3295 (patch written by Frank Heydlauf).
3297 SC/02 Eximstats: use FileHandles for outputing results.
3299 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3301 SC/04 Eximstats: fixed display of large numbers with -nvr option
3303 SC/05 Eximstats: fixed merging of reports with empty tables.
3305 SC/06 Eximstats: added the -include_original_destination flag
3307 SC/07 Eximstats: removed tabs and trailing whitespace.
3309 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3311 TK/06 MBOX spool code: Add real "From " MBOX separator line
3312 so the .eml file is really in mbox format (even though
3313 most programs do not really care). Patch from Alex Miller.
3315 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3316 The latter is generated from $received_to and is only set if the
3317 message has one envelope recipient. SA can use these headers,
3318 obviously out-of-the-box. Patch from Alex Miller.
3320 PH/08 The ${def test on a variable was returning false if the variable's
3321 value was "0", contrary to what the specification has always said!
3322 The result should be true unless the variable is empty.
3324 PH/09 The syntax error of a character other than { following "${if
3325 def:variable_name" (after optional whitespace) was not being diagnosed.
3326 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3327 accidental colon was present, for example, could give incorrect results.
3329 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3330 result is used (not including appendfile, where other changes are about
3333 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3334 This involved changing a lot of size variables from int to off_t. It
3335 should work with maildirs and everything.
3337 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3338 spamd dying while we are connected to it.
3340 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3343 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3344 function for BDB 4.3.
3346 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3348 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3349 the sources for PCRE have been re-arranged and more files are now
3352 PH/15 The code I had for printing potentially long long variables in PH/11
3353 above was not the best (it lost precision). The length of off_t variables
3354 is now inspected at build time, and an appropriate printing format (%ld
3355 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3356 to be "long long int" or "long int". This is needed for the internal
3357 formatting function string_vformat().
3359 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3360 the configuration file to be ":syslog", then the script "guesses" where
3361 the logs files are, rather than using the compiled in default. In our
3362 case the guess is not the same as the compiled default, so the script
3363 suddenly stopped working when I started to use syslog. The patch checks
3364 to see if log_file_path is "". If so, it attempts to read it from exim
3365 with no configuration file to get the compiled in version, before it
3366 falls back to the previous guessing code."
3368 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3369 implementing BATV in an Exim configuration. See NewStuff for the gory
3372 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3373 Makefile that are specific to HP-UX.
3375 PH/18 If the "use_postmaster" option was set for a recipient callout together
3376 with the "random" option, the postmaster address was used as the MAIL
3377 FROM address for the random test, but not for the subsequent recipient
3378 test. It is now used for both.
3380 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3381 patch removes a few documentation additions to RFC 3028, because the
3382 latest draft now contains them. It adds the new en;ascii-case comparator
3383 and a new error check for 8bit text in MIME parts. Comparator and
3384 require names are now matched exactly. I enabled the subaddress
3385 extension, but it is not well tested yet (read: it works for me)."
3387 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3388 rework some of the code of TK/09 above to avoid the hardwired use of
3389 "%lld" and "long long". Replaced the call to snprintf() with a call to
3392 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3393 records point to non-existent hosts", "retry timeout exceeded", and
3394 "retry time not reached for any host after a long failure period".
3396 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3397 experimental DomainKeys support:
3399 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3400 (2) On an error such as an illegally used "control", the wrong name for
3401 the control was given.
3403 These problems did NOT occur unless DomainKeys support was compiled.
3405 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3407 PH/24 Added ${if match_ip condition.
3409 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3410 databases so that it will be absolutely obvious if a crash occurs in the
3411 DB library. This is a regular occurrence (often caused by mis-matched
3414 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3415 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3416 picked up on a user's system that detects such things. There doesn't seem
3417 to be a gcc warning option for this - only an attribute that has to be
3418 put on the function's prototype. It seems that in Fedora Core 4 they have
3419 set this on a number of new functions. No doubt there will be more in due
3422 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3423 specify a transport. However, if an address that was verified by such a
3424 router was the subject of a callout, Exim crashed because it tried to
3425 read the rcpt_include_affixes from the non-existent transport. Now it
3426 just assumes that the setting of that option is false. This bug was
3427 introduced by 4.51/PH/31.
3429 PH/28 Changed -d+all to exclude +memory, because that information is very
3430 rarely of interest, but it makes the output a lot bigger. People tend to
3431 do -d+all out of habit.
3433 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3434 code in os-type was giving problems when libc.so lives in lib64, like on
3437 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3438 aren't the modern standard, and it seems that some systems' include files
3439 don't always have them. Exim was already checking for some of the newer
3440 ones like T_AAAA, and defining it itself. I've added checks for all the
3441 record types that Exim uses.
3443 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3444 not automatically generating a new one, as it is supposed to. This
3445 prevented TLS from working. If the file did exist, but contained invalid
3446 data, a new version was generated, as expected. It was only the case of a
3447 non-existent file that was broken.
3449 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3450 with a change in libdomainkeys > 0.64.
3452 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3453 from DNS. If the selector record carries the flag, it now has
3454 precedence over the domain-wide flag.
3456 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3458 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3459 the use of a port name that isn't defined in /etc/services) occurred, the
3460 message was deferred as in a normal delivery, and thus remained on the
3461 spool, instead of being failed because of the mua_wrapper setting. This
3462 is now fixed, and I tidied up some of the mua_wrapper messages at the
3465 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3466 the messages in a hash of arrays rather than using individual hashes.
3467 This is a bit cleaner and results in dramatic memory savings, albeit
3468 at a slight CPU cost.
3470 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3471 as requested by Marc Sherman.
3473 SC/10 Eximstats: added histograms for user specified patterns as requested
3476 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3478 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3479 fopen() in the content-scanning modules that did not already have it.
3485 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3486 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3488 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3490 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3492 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3493 to test Sieve filters that use "vacation".
3495 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3496 that changes the way the GnuTLS parameters are stored in the cache file.
3497 The new format can be generated externally. For backward compatibility,
3498 if the data in the cache doesn't make sense, Exim assumes it has read an
3499 old-format file, and it generates new data and writes a new file. This
3500 means that you can't go back to an older release without removing the
3503 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3504 work if there are any delivery delays because "one_time" forces the
3505 parent to be marked "delivered", so its unseen clone is never tried
3506 again. For this reason, Exim now forbids the simultaneous setting of
3509 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3510 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3511 entry if you want to know the details. The fix, however, neglected to
3512 consider the case when local delivery batching is involved. The test for
3513 "previously delivered" was not happening when checking to see if an
3514 address could be batched with a previous (undelivered) one; under
3515 certain circumstances this could lead to multiple deliveries to the same
3518 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3519 in its include files, and this causes problems building Exim.
3521 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3522 header_syntax) but Exim was just ignoring anything given after a slash.
3523 In particular, this caused confusion with an attempt to use "verify =
3524 reverse_host_lookup/defer_ok". An error is now given when options are
3525 supplied for verify items that do not have them. (Maybe reverse_host_
3526 lookup should have a defer_ok option, but that's a different point.)
3528 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3529 defined by RFC 821) to 2048, because there were problems with some AUTH
3530 commands, and RFC 1869 says the size should be increased for extended
3531 SMTP commands that take arguments.
3533 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3536 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3537 "unknown" error; now it says that the functionality isn't in the binary.
3539 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3540 an address' error message when a string expansion fails (syntax or
3541 whatever). Otherwise the password may appear in the log. Following change
3542 PH/42 below, there is no longer a chance of it appearing in a bounce
3545 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3547 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3548 of its addresses was ever tried. (Bugzilla bug #2.)
3550 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3551 the result incorrectly in the debug output. (It correctly added a newline
3552 to what was transported.)
3554 TF/01 Added $received_time.
3556 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3557 commented out examples of how to interface to a virus scanner and to
3558 SpamAssassin. Also added commented examples of av_scanner and
3559 spamd_address settings.
3561 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3562 and controls are allowed in which ACLs. There were a couple of minor
3563 errors. Some of the entries in the conditions table (which is a table of
3564 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3565 negation of where the condition IS allowed.
3567 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3569 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3570 header file does not have a version number, so I've had to invent a new
3571 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3572 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3573 radiusclient), but it was contributed by a Radius user.
3575 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3576 files or directories, for OpenSSL.
3578 PH/19 When an Exim process that is running as root has to create an Exim log
3579 file, it does so in a subprocess that runs as exim:exim so as to get the
3580 ownership right at creation (otherwise, other Exim processes might see
3581 the file with the wrong ownership). There was no test for failure of this
3582 fork() call, which would lead to the process getting stuck as it waited
3583 for a non-existent subprocess. Forks do occasionally fail when resources
3584 run out. I reviewed all the other calls to fork(); they all seem to check
3587 PH/20 When checking for unexpected SMTP input at connect time (before writing
3588 the banner), Exim was not dealing correctly with a non-positive return
3589 from the read() function. If the client had disconnected by this time,
3590 the result was a log entry for a synchronization error with an empty
3591 string after "input=" when read() returned zero. If read() returned -1
3592 (an event I could not check), uninitialized data bytes were printed.
3593 There were reports of junk text (parts of files, etc) appearing after
3596 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3598 PH/22 Added support for macro redefinition, and (re)definition in between
3599 driver and ACL definitions.
3601 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3602 forgetting to use the resulting value; it was using the unexpanded value.
3604 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3605 hadn't been configured. The fix is from Juergen Kreileder, who
3606 understands it better than I do:
3608 "Here's what I see happening with three configured cyrus_sasl
3609 authenticators configured (plain, login, cram-md5):
3611 On startup auth_cyrus_sasl_init() gets called for each of these.
3612 This means three calls to sasl_listmech() without a specified mech_list.
3613 => SASL tests which mechs of all available mechs actually work
3614 => three warnings about OTP not working
3615 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3617 With the patch, sasl_listmech() also gets called three times. But now
3618 SASL's mech_list option is set to the server_mech specified in the the
3619 authenticator. Or in other words, the answer from sasl_listmech()
3620 gets limited to just the mech you're testing for (which is different
3622 => the return list contains just 'plain' or 'login', 'cram-md5' or
3623 nothing depending on the value of ob->server_mech.
3625 I've just tested the patch: Authentication still works fine,
3626 unavailable mechs specified in the exim configuration are still
3627 caught, and the auth.log warnings about OTP are gone."
3629 PH/25 When debugging is enabled, the contents of the command line are added
3630 to the debugging output, even when log_selector=+arguments is not
3633 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3634 answer is "GNU", and only if the return is "GNU/something" is the answer
3637 PH/27 $acl_verify_message is now set immediately after the failure of a
3638 verification in an ACL, and so is available in subsequent modifiers. In
3639 particular, the message can be preserved by coding like this:
3641 warn !verify = sender
3642 set acl_m0 = $acl_verify_message
3644 Previously, $acl_verify_message was set only while expanding "message"
3645 and "log_message" when a very denied access.
3647 PH/28 Modified OS/os.c-Linux with
3649 -#ifndef OS_LOAD_AVERAGE
3650 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3652 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3653 nomenclature these days.)
3655 PH/29 Installed patch from the Sieve maintainer that adds the options
3656 sieve_useraddress and sieve_subaddress to the redirect router.
3658 PH/30 In these circumstances:
3659 . Two addresses routed to the same list of hosts;
3660 . First host does not offer TLS;
3661 . First host accepts first address;
3662 . First host gives temporary error to second address;
3663 . Second host offers TLS and a TLS session is established;
3664 . Second host accepts second address.
3665 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3666 and peerdn, if requested) that were in fact used only for the second
3669 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3670 attention to any local part prefix or suffix that was matched by the
3671 router that accepted the address. It now behaves in the same way as it
3672 does for delivery: the affixes are removed from the local part unless
3673 rcpt_include_affixes is set on the transport.
3675 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3676 timeout during the DATA phase of an incoming message.
3678 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3679 applied a patch sanctioned by the Sieve maintainer.
3681 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3682 the uid or gid is negative. A case of a negative gid caused this to be
3683 noticed. The fix allows for either to be negative.
3685 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3686 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3689 PH/36 The change PH/12 above was broken. Fixed it.
3691 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3692 the grounds that routing the same address twice would always produce the
3693 same answer. This might have been true once, but it is certainly no
3694 longer true now. Routing a child address may depend on the previous
3695 routing that produced that child. Some complicated redirection strategies
3696 went wrong when messages had multiple recipients, and made Exim's
3697 behaviour dependent on the order in which the addresses were given.
3699 I have moved the duplicate checking until after the routing is complete.
3700 Exim scans the addresses that are assigned to local and remote
3701 transports, and removes any duplicates. This means that more work will be
3702 done, as duplicates will always all be routed, but duplicates are
3703 presumably rare, so I don't expect this is of any significance.
3705 For deliveries to pipes, files, and autoreplies, the duplicate checking
3706 still happens during the routing process, since they are not going to be
3709 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3710 It corrects a timeout issue with spamd. This is Ian's comment: "The
3711 background is that sometimes spamd either never reads data from a
3712 connection it has accepted, or it never writes response data. The exiscan
3713 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3714 blindly assumes that writes won't block so it may never time out."
3716 PH/39 Allow G after quota size as well as K and M.
3718 PH/40 The value set for $authenticated_id in an authenticator may not contain
3719 binary zeroes or newlines because the value is written to log lines and
3720 to spool files. There was no check on this. Now the value is run through
3721 the string_printing() function so that such characters are converted to
3722 printable escape sequences.
3724 PH/41 $message_linecount is a new variable that contains the total number of
3725 lines in the message. Compare $body_linecount, which is the count for the
3728 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3729 bounce and delay warning messages, except in certain special cases, which
3732 (a) An SMTP error message from a remote host;
3733 (b) A message specified in a :fail: redirection;
3734 (c) A message specified in a "fail" command in a system filter;
3735 (d) A message specified in a FAIL return from the queryprogram router;
3736 (e) A message specified by the cannot_route_message router option.
3738 In these cases only, Exim does include the error details in bounce and
3739 warning messages. There are also a few cases where bland messages such
3740 as "unrouteable address" or "local delivery error" are given.
3742 PH/43 $value is now also set for the "else" part of a ${run expansion.
3744 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3745 being worked on, but at least Exim now implements the latest version to
3748 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3749 process to complete was treated as a delivery failure, a timeout while
3750 writing the message to the pipe was logged, but erroneously treated as a
3751 successful delivery. Such timeouts include transport filter timeouts. For
3752 consistency with the overall process timeout, these timeouts are now
3753 treated as errors, giving rise to delivery failures by default. However,
3754 there is now a new Boolean option for the pipe transport called
3755 timeout_defer, which, if set TRUE, converts the failures into defers for
3756 both kinds of timeout. A transport filter timeout is now identified in
3759 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3760 systems where "make" and "gmake" are different, calling "gmake" at top
3761 level broke things. I've arranged for the value of $(MAKE) to be passed
3762 from the Makefile to this script so that it can call the same version of
3766 A note about Exim versions 4.44 and 4.50
3767 ----------------------------------------
3769 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3770 changes of various kinds. As a consequence, a big documentation update was
3771 needed. This delayed the release for rather longer than seemed good, especially
3772 in the light of a couple of (minor) security issues. Therefore, the changes
3773 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3774 release. So 4.44 and 4.50 are in effect two different branches that both start
3777 I have left the 4.50 change log unchanged; it contains all the changes since
3778 4.43. The change log for 4.44 is below; many of its items are identical to
3779 those for 4.50. This seems to be the most sensible way to preserve the
3780 historical information.
3786 1. Minor wording change to the doc/README.SIEVE file.
3788 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3789 computation of the current number of files was incorrect.
3791 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3792 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3795 4. Give more explanation in the error message when the command for a transport
3796 filter fails to execute.
3798 5. There are several places where Exim runs a non-Exim command in a
3799 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3800 was being done only for the command run by the queryprogram router. It is
3801 now done for all such subprocesses. The other cases are: ${run, transport
3802 filters, and the commands run by the lmtp and pipe transports.
3804 6. Added CONFIGURE_GROUP build-time option.
3806 7. Some older OS have a limit of 256 on the maximum number of file
3807 descriptors. Exim was using setrlimit() to set 1000 as a large value
3808 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3809 systems. I've change it so that if it can't get 1000, it tries for 256.
3811 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3812 was an oversight, and furthermore, ever since the addition of extra
3813 controls (e.g. 4.43/32), the checks on when to allow different forms of
3814 "control" were broken. There should now be diagnostics for all cases when a
3815 control that does not make sense is encountered.
3817 9. Added the /retain_sender option to "control=submission".
3819 10. $recipients is now available in the predata ACL (oversight).
3821 11. Tidy the search cache before the fork to do a delivery from a message
3822 received from the command line. Otherwise the child will trigger a lookup
3823 failure and thereby defer the delivery if it tries to use (for example) a
3824 cached ldap connection that the parent has called unbind on.
3826 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3827 of $address_data from the recipient verification was clobbered by the
3828 sender verification.
3830 13. The value of address_data from a sender verification is now available in
3831 $sender_address_data in subsequent conditions in the ACL statement.
3833 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3835 15. Added a new option "connect=<time>" to callout options, to set a different
3838 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3839 was its contents. (It was OK if the option was not defined at all.)
3841 17. A "Completed" log line is now written for messages that are removed from
3842 the spool by the -Mrm option.
3844 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3845 information about exactly what failed.
3847 19. Added -dd to debug only the daemon process.
3849 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3850 handles timeouts, both on the server side and network timeouts. Renamed the
3851 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3853 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3854 It is now set to "smtps".
3856 22. $host_address is now set to the target address during the checking of
3857 ignore_target_hosts.
3859 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3860 being passed; this would have caused $sender_host_name to have been used if
3861 matching the list had actually called for a host name (not very likely,
3862 since this list is usually IP addresses). A host name is now passed as
3865 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3866 code that specifies a non-restarting handler (typically sigaction() in
3867 modern systems) in an attempt to fix a rare and obscure crash bug.
3869 25. Narrowed the window for a race in the daemon that could cause it to ignore
3870 SIGCHLD signals. This is not a major problem, because they are used only to
3871 wake it up if nothing else does.
3873 26. A malformed maildirsize file could cause Exim to calculate negative values
3874 for the mailbox size or file count. Odd effects could occur as a result.
3875 The maildirsize information is now recalculated if the size or filecount
3878 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3879 support for a long time. Removed HAVE_SYS_VFS_H.
3881 28. Installed the latest version of exipick from John Jetmore.
3883 29. In an address list, if the pattern was not a regular expression, an empty
3884 subject address (from a bounce message) matched only if the pattern was an
3885 empty string. Non-empty patterns were not even tested. This was the wrong
3886 because it is perfectly reasonable to use an empty address as part of a
3887 database query. An empty address is now tested by patterns that are
3888 lookups. However, all the other forms of pattern expect the subject to
3889 contain a local part and a domain, and therefore, for them, an empty
3890 address still always fails if the pattern is not itself empty.
3892 30. Exim went into a mad DNS loop when attempting to do a callout where the
3893 host was specified on an smtp transport, and looking it up yielded more
3894 than one IP address.
3896 31. Re-factored the code for checking spool and log partition space into a
3897 function that finds that data and another that does the check. The former
3898 is then used to implement four new variables: $spool_space, $log_space,
3899 $spool_inodes, and $log_inodes.
3901 32. The RFC2047 encoding function was originally intended for short strings
3902 such as real names; it was not keeping to the 75-character limit for
3903 encoded words that the RFC imposes. It now respects the limit, and
3904 generates multiple encoded words if necessary. To be on the safe side, I
3905 have increased the buffer size for the ${rfc2047: expansion operator from
3908 33. It is now permitted to omit both strings after an "if" condition; if the
3909 condition is true, the result is "true". As before, when the second string
3910 is omitted, a false condition yields an empty string. This makes it less
3911 cumbersome to write custom ACL and router conditions.
3913 34. Failure to deliver a bounce message always caused it to be frozen, even if
3914 there was an errors_to setting on the router. The errors_to setting is now
3917 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3918 canonical form (fully expanded) before being placed in
3919 $sender_host_address.
3921 36. The table in the code that translates DNS record types into text (T_A to
3922 "A" for instance) was missing entries for NS and CNAME. It is just possible
3923 that this could have caused confusion if both these types were looked up
3924 for the same domain, because the text type is used as part of Exim's
3925 per-process caching. But the chance of anyone hitting this buglet seems
3928 37. The dnsdb lookup has been extended in a number of ways.
3930 (1) There is a new type, "zns", which walks up the domain tree until it
3931 finds some nameserver records. It should be used with care.
3933 (2) There is a new type, "mxh", which is like "mx" except that it returns
3934 just the host names, not the priorities.
3936 (3) It is now possible to give a list of domains (or IP addresses) to be
3937 looked up. The behaviour when one of the lookups defers can be
3938 controlled by a keyword.
3940 (4) It is now possible to specify the separator character for use when
3941 multiple records are returned.
3943 38. The dnslists ACL condition has been extended: it is now possible to supply
3944 a list of IP addresses and/or domains to be looked up in a particular DNS
3947 39. Added log_selector=+queue_time_overall.
3949 40. When running the queue in the test harness, wait just a tad after forking a
3950 delivery process, to get repeatability of debugging output.
3952 41. Include certificate and key file names in error message when GnuTLS fails
3953 to set them up, because the GnuTLS error message doesn't include the name
3954 of the failing file when there is a problem reading it.
3956 42. Allow both -bf and -bF in the same test run.
3958 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3960 44. The "Exiscan patch" is now merged into the mainline Exim source.
3962 45. Sometimes the final signoff response after QUIT could fail to get
3963 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3964 < 0 before doing a fflush(). This bug looks as though it goes back to the
3965 introduction of TLS in release 3.20, but "sometimes" must have been rare
3966 because the tests only now provoked it.
3968 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3969 (this can affect the format of dates).
3971 47. exim_tidydb, when checking for the continued existence of a message for
3972 which it has found a message-specific retry record, was not finding
3973 messages that were in split spool directories. Consequently, it was
3974 deleting retry records that should have stayed in existence.
3976 48. Steve fixed some bugs in eximstats.
3978 49. The SPA authentication driver was not abandoning authentication and moving
3979 on to the next authenticator when an expansion was forced to fail,
3980 contradicting the general specification for all authenticators. Instead it
3981 was generating a temporary error. It now behaves as specified.
3983 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3984 (the order specifies the preference for clients). The order is now AES256,
3985 AES128, 3DES, ARCFOUR128.
3987 51. Small patch to Sieve code - explicitly set From: when generating an
3990 52. Exim crashed if a remote delivery caused a very long error message to be
3991 recorded - for instance if somebody sent an entire SpamAssassin report back
3992 as a large number of 550 error lines. This bug was coincidentally fixed by
3993 increasing the size of one of Exim's internal buffers (big_buffer) that
3994 happened as part of the Exiscan merge. However, to be on the safe side, I
3995 have made the code more robust (and fixed the comments that describe what
3998 53. Now that there can be additional text after "Completed" in log lines (if
3999 the queue_time_overall log selector is set), a one-byte patch to exigrep
4000 was needed to allow it to recognize "Completed" as not the last thing in
4003 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4004 patch that reportedly fixes this has been added. I am not expert enough to
4005 create a test for it. This is what the patch creator wrote:
4007 "I found a little strange behaviour of ldap code when working with
4008 Windows 2003 AD Domain, where users was placed in more than one
4009 Organization Units. When I tried to give exim partial DN, the exit code
4010 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4011 But simultaneously result of request was absolutely normal ldap result,
4012 so I produce this patch..."
4014 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4015 REFERENCE, so I have modified the code to exclude the patch when that macro
4018 55. Some experimental protocols are using DNS PTR records for new purposes. The
4019 keys for these records are domain names, not reversed IP addresses. The
4020 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4021 leaves it alone. Component reversal etc. now happens only for IP addresses.
4024 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4026 57. Double the size of the debug message buffer (to 2048) so that more of very
4027 long debug lines gets shown.
4029 58. The exicyclog utility now does better if the number of log files to keep
4030 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4032 59. Two changes related to the smtp_active_hostname option:
4034 (1) $smtp_active_hostname is now available as a variable.
4035 (2) The default for smtp_banner uses $smtp_active_hostname instead
4036 of $primary_hostname.
4038 60. The host_aton() function is supposed to be passed a string that is known
4039 to be a valid IP address. However, in the case of IPv6 addresses, it was
4040 not checking this. This is a hostage to fortune. Exim now panics and dies
4041 if the condition is not met. A case was found where this could be provoked
4042 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4043 components; fortuitously, this particular loophole had already been fixed
4044 by change 4.50/55 above.
4046 If there are any other similar loopholes, the new check in host_aton()
4047 itself should stop them being exploited. The report I received stated that
4048 data on the command line could provoke the exploit when Exim was running as
4049 exim, but did not say which command line option was involved. All I could
4050 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4051 running as the user.
4054 61. There was a buffer overflow vulnerability in the SPA authentication code
4055 (which came originally from the Samba project). I have added a test to the
4056 spa_base64_to_bits() function which I hope fixes it.
4059 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4060 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4062 63. The daemon start-up calls getloadavg() while still root for those OS that
4063 need the first call to be done as root, but it missed one case: when
4064 deliver_queue_load_max is set with deliver_drop_privilege. This is
4065 necessary for the benefit of the queue runner, because there is no re-exec
4066 when deliver_drop_privilege is set.
4068 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4069 This has been fixed.
4071 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4072 was in use, was not putting the data itself into the right store pool;
4073 consequently, it could be overwritten for a subsequent message in the same
4074 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4077 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4079 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4080 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4081 the function were treating the return as a boolean value, which happened to
4082 work because 0=false and not-0=true, but is not correct code.
4084 68. The host_aton() function was not handling scoped IPv6 addresses (those
4085 with, for example, "%eth0" on the end) correctly.
4087 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4088 negated items (that is, ~something) in unsigned ints. Some compilers
4089 apparently mutter when there is no cast.
4091 70. If an address verification called from an ACL failed, and did not produce a
4092 user-specific message (i.e. there was only a "system" message), nothing was
4093 put in $acl_verify_message. In this situation, it now puts the system
4096 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4097 session; change 4.31/43 added the unwanted input to the log line - except
4098 that it did not do this in the start of session case. It now does.
4100 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4101 This is wrong and can cause the other end to generate a synchronization
4102 error if it is another Exim or anything else that does the synchronization
4103 check. A QUIT command is no longer sent after a timeout.
4105 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4106 during host lookups.
4108 74. The defer_ok option of callout verification was not working if it was used
4109 when verifying addresses in header lines, that is, for this case:
4111 verify = header_sender/callout=defer_ok
4113 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4114 those file descriptors could be used for SMTP connections. If anything
4115 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4116 could be sent to the SMTP client, causing chaos. The daemon now opens
4117 stdin, stdout, and stderr to /dev/null when it puts itself into the
4120 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4121 log by default. The user can override this with suitable Perl magic.
4123 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4124 the log message when discard triggers, was not working for the DATA ACL or
4125 for the non-SMTP ACL.
4127 78. Error message wording change in sieve.c.
4129 79. If smtp_accept_max_per_host was set, the number of connections could be
4130 restricted to fewer than expected, because the daemon was trying to set up
4131 a new connection before checking whether the processes handling previous
4132 connections had finished. The check for completed processes is now done
4133 earlier. On busy systems, this bug wouldn't be noticed because something
4134 else would have woken the daemon, and it would have reaped the completed
4137 80. If a message was submitted locally by a user whose login name contained one
4138 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4139 It caused a spool format error. I have fixed the spool reading code. A
4140 related problem was that the "from" clause in the Received: line became
4141 illegal because of the space(s). It is now covered by ${quote_local_part.
4143 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4146 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4148 83. Patch from Sieve maintainer to fix unterminated string problem in
4149 "vacation" handling.
4151 84. Some minor changes to the Linux configuration files to help with other
4152 OS variants using glibc.
4154 85. One more patch for Sieve to update vacation handling to latest spec.
4157 ----------------------------------------------------
4158 See the note above about the 4.44 and 4.50 releases.
4159 ----------------------------------------------------
4165 1. Change 4.43/35 introduced a bug that caused file counts to be
4166 incorrectly computed when quota_filecount was set in an appendfile
4169 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4170 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4173 3. Give more explanation in the error message when the command for a transport
4174 filter fails to execute.
4176 4. There are several places where Exim runs a non-Exim command in a
4177 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4178 was being done only for the command run by the queryprogram router. It is
4179 now done for all such subprocesses. The other cases are: ${run, transport
4180 filters, and the commands run by the lmtp and pipe transports.
4182 5. Some older OS have a limit of 256 on the maximum number of file
4183 descriptors. Exim was using setrlimit() to set 1000 as a large value
4184 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4185 systems. I've change it so that if it can't get 1000, it tries for 256.
4187 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4188 was an oversight, and furthermore, ever since the addition of extra
4189 controls (e.g. 4.43/32), the checks on when to allow different forms of
4190 "control" were broken. There should now be diagnostics for all cases when a
4191 control that does not make sense is encountered.
4193 7. $recipients is now available in the predata ACL (oversight).
4195 8. Tidy the search cache before the fork to do a delivery from a message
4196 received from the command line. Otherwise the child will trigger a lookup
4197 failure and thereby defer the delivery if it tries to use (for example) a
4198 cached ldap connection that the parent has called unbind on.
4200 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4201 of $address_data from the recipient verification was clobbered by the
4202 sender verification.
4204 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4205 was its contents. (It was OK if the option was not defined at all.)
4207 11. A "Completed" log line is now written for messages that are removed from
4208 the spool by the -Mrm option.
4210 12. $host_address is now set to the target address during the checking of
4211 ignore_target_hosts.
4213 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4214 being passed; this would have caused $sender_host_name to have been used if
4215 matching the list had actually called for a host name (not very likely,
4216 since this list is usually IP addresses). A host name is now passed as
4219 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4220 code that specifies a non-restarting handler (typically sigaction() in
4221 modern systems) in an attempt to fix a rare and obscure crash bug.
4223 15. Narrowed the window for a race in the daemon that could cause it to ignore
4224 SIGCHLD signals. This is not a major problem, because they are used only to
4225 wake it up if nothing else does.
4227 16. A malformed maildirsize file could cause Exim to calculate negative values
4228 for the mailbox size or file count. Odd effects could occur as a result.
4229 The maildirsize information is now recalculated if the size or filecount
4232 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4233 support for a long time. Removed HAVE_SYS_VFS_H.
4235 18. Updated exipick to current release from John Jetmore.
4237 19. Allow an empty sender to be matched against a lookup in an address list.
4238 Previously the only cases considered were a regular expression, or an
4241 20. Exim went into a mad DNS lookup loop when doing a callout where the
4242 host was specified on the transport, if the DNS lookup yielded more than
4245 21. The RFC2047 encoding function was originally intended for short strings
4246 such as real names; it was not keeping to the 75-character limit for
4247 encoded words that the RFC imposes. It now respects the limit, and
4248 generates multiple encoded words if necessary. To be on the safe side, I
4249 have increased the buffer size for the ${rfc2047: expansion operator from
4252 22. Failure to deliver a bounce message always caused it to be frozen, even if
4253 there was an errors_to setting on the router. The errors_to setting is now
4256 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4257 canonical form (fully expanded) before being placed in
4258 $sender_host_address.
4260 24. Updated eximstats to version 1.33
4262 25. Include certificate and key file names in error message when GnuTLS fails
4263 to set them up, because the GnuTLS error message doesn't include the name
4264 of the failing file when there is a problem reading it.
4266 26. Expand error message when OpenSSL has problems setting up cert/key files.
4269 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4270 (this can affect the format of dates).
4272 28. exim_tidydb, when checking for the continued existence of a message for
4273 which it has found a message-specific retry record, was not finding
4274 messages that were in split spool directories. Consequently, it was
4275 deleting retry records that should have stayed in existence.
4277 29. eximstats updated to version 1.35
4278 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4279 1.35 - bugfix such that pie charts by volume are generated correctly
4281 30. The SPA authentication driver was not abandoning authentication and moving
4282 on to the next authenticator when an expansion was forced to fail,
4283 contradicting the general specification for all authenticators. Instead it
4284 was generating a temporary error. It now behaves as specified.
4286 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4287 (the order specifies the preference for clients). The order is now AES256,
4288 AES128, 3DES, ARCFOUR128.
4290 31. Small patch to Sieve code - explicitly set From: when generating an
4293 32. Exim crashed if a remote delivery caused a very long error message to be
4294 recorded - for instance if somebody sent an entire SpamAssassin report back
4295 as a large number of 550 error lines. This bug was coincidentally fixed by
4296 increasing the size of one of Exim's internal buffers (big_buffer) that
4297 happened as part of the Exiscan merge. However, to be on the safe side, I
4298 have made the code more robust (and fixed the comments that describe what
4301 33. Some experimental protocols are using DNS PTR records for new purposes. The
4302 keys for these records are domain names, not reversed IP addresses. The
4303 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4304 leaves it alone. Component reversal etc. now happens only for IP addresses.
4307 34. The host_aton() function is supposed to be passed a string that is known
4308 to be a valid IP address. However, in the case of IPv6 addresses, it was
4309 not checking this. This is a hostage to fortune. Exim now panics and dies
4310 if the condition is not met. A case was found where this could be provoked
4311 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4312 components; fortuitously, this particular loophole had already been fixed
4313 by change 4.50/55 or 4.44/33 above.
4315 If there are any other similar loopholes, the new check in host_aton()
4316 itself should stop them being exploited. The report I received stated that
4317 data on the command line could provoke the exploit when Exim was running as
4318 exim, but did not say which command line option was involved. All I could
4319 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4320 running as the user.
4323 35. There was a buffer overflow vulnerability in the SPA authentication code
4324 (which came originally from the Samba project). I have added a test to the
4325 spa_base64_to_bits() function which I hope fixes it.
4328 36. The daemon start-up calls getloadavg() while still root for those OS that
4329 need the first call to be done as root, but it missed one case: when
4330 deliver_queue_load_max is set with deliver_drop_privilege. This is
4331 necessary for the benefit of the queue runner, because there is no re-exec
4332 when deliver_drop_privilege is set.
4334 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4335 was in use, was not putting the data itself into the right store pool;
4336 consequently, it could be overwritten for a subsequent message in the same
4337 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4340 38. Sometimes the final signoff response after QUIT could fail to get
4341 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4342 < 0 before doing a fflush(). This bug looks as though it goes back to the
4343 introduction of TLS in release 3.20, but "sometimes" must have been rare
4344 because the tests only now provoked it.
4350 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4351 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4352 Now it returns an integer. A number of calls were still expecting a T/F
4353 return. Fortuitously, in all cases, the tests worked in OK situations,
4354 which is the norm. However, things would have gone wrong on any write
4355 failures on the smtp file descriptor. This function is used when sending
4356 messages over SMTP and also when doing verify callouts.
4358 2. When Exim is called to do synchronous delivery of a locally submitted
4359 message (the -odf or -odi options), it no longer closes stderr before doing
4362 3. Implemented the mua_wrapper option.
4364 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4366 5. Implemented the functions header_remove(), header_testname(),
4367 header_add_at_position(), and receive_remove_recipient(), and exported them
4370 6. If an ACL "warn" statement specified the addition of headers, Exim already
4371 inserted X-ACL-Warn: at the start if there was no header name. However, it
4372 was not making this test for the second and subsequent header lines if
4373 there were newlines in the string. This meant that an invalid header could
4374 be inserted if Exim was badly configured.
4376 7. Allow an ACL "warn" statement to add header lines at the start or after all
4377 the Received: headers, as well as at the end.
4379 8. Added the rcpt_4xx retry error code.
4381 9. Added postmaster_mailfrom=xxx to callout verification option.
4383 10. Added mailfrom=xxxx to the callout verification option, for verify=
4386 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4387 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4389 12. Inserted some casts to stop certain compilers warning when using pointer
4390 differences as field lengths or precisions in printf-type calls (mostly
4391 affecting debugging statements).
4393 13. Added optional readline() support for -be (dynamically loaded).
4395 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4396 same clock tick as a message's arrival, so that its received time was the
4397 same as the "first fail" time on the retry record, and that message
4398 remained on the queue past the ultimate address timeout, every queue runner
4399 would try a delivery (because it was past the ultimate address timeout) but
4400 after another failure, the ultimate address timeout, which should have then
4401 bounced the address, did not kick in. This was a "< instead of <=" error;
4402 in most cases the first failure would have been in the next clock tick
4403 after the received time, and all would be well.
4405 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4406 being recognized when the domain list was tested by the match_domain
4407 condition in an expansion string.
4409 16. Added the ${str2b64: operator.
4411 17. Exim was always calling setrlimit() to set a large limit for the number of
4412 processes, without checking whether the existing limit was already
4413 adequate. (It did check for the limit on file descriptors.) Furthermore,
4414 errors from getrlimit() and setrlimit() were being ignored. Now they are
4415 logged to the main and panic logs, but Exim does carry on, to try to do its
4416 job under whatever limits there are.
4418 18. Imported PCRE 5.0.
4420 19. Trivial typo in log message " temporarily refused connection" (the leading
4423 20. If the log selector return_path_on_delivery was set and an address was
4424 redirected to /dev/null, the delivery process crashed because it assumed
4425 that a return path would always be set for a "successful" delivery. In this
4426 case, the whole delivery is bypassed as an optimization, and therefore no
4429 21. Internal re-arrangement: the function for sending a challenge and reading
4430 a response while authentication was assuming a zero-terminated challenge
4431 string. It's now changed to take a pointer and a length, to allow for
4432 binary data in such strings.
4434 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4436 23. Exim was not respecting finduser_retries when seeking the login of the
4437 uid under which it was called; it was always trying 10 times. (The default
4438 setting of finduser_retries is zero.) Also, it was sleeping after the final
4439 failure, which is pointless.
4441 24. Implemented tls_on_connect_ports.
4443 25. Implemented acl_smtp_predata.
4445 26. If the domain in control=submission is set empty, Exim assumes that the
4446 authenticated id is a complete email address when it generates From: or
4447 Sender: header lines.
4449 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4450 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4451 chown and chgrp in /bin and hostname in /usr/bin.
4453 28. Exim was keeping the "process log" file open after each use, just as it
4454 does for the main log. This opens the possibility of it remaining open for
4455 long periods when the USR1 signal hits a daemon. Occasional processlog
4456 errors were reported, that could have been caused by this. Anyway, it seems
4457 much more sensible not to leave this file open at all, so that is what now
4460 29. The long-running daemon process does not normally write to the log once it
4461 has entered its main loop, and it closes the log before doing so. This is
4462 so that log files can straightforwardly be renamed and moved. However,
4463 there are a couple of unusual error situations where the daemon does write
4464 log entries, and I had neglected to close the log afterwards.
4466 30. The text of an SMTP error response that was received during a remote
4467 delivery was being truncated at 512 bytes. This is too short for some of
4468 the long messages that one sometimes sees. I've increased the limit to
4471 31. It is now possible to make retry rules that apply only when a message has a
4472 specific sender, in particular, an empty sender.
4474 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4475 it possible to be selective about when SMTP synchronization is enforced.
4477 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4479 32. Implemented hosts_connection_nolog.
4481 33. Added an ACL for QUIT.
4483 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4486 35. Added mailbox_size and mailbox_filecount to appendfile.
4488 36. Added control = no_multiline_responses to ACLs.
4490 37. There was a bug in the logic of the code that waits for the clock to tick
4491 in the case where the clock went backwards by a substantial amount such
4492 that the microsecond fraction of "now" was more than the microsecond
4493 fraction of "then" (but the whole seconds number was less).
4495 38. Added support for the libradius Radius client library this is found on
4496 FreeBSD (previously only the radiusclient library was supported).
4502 1. When certain lookups returned multiple values in the form name=value, the
4503 quoting of the values was not always being done properly. Specifically:
4504 (a) If the value started with a double quote, but contained no whitespace,
4506 (b) If the value contained whitespace other than a space character (i.e.
4507 tabs or newlines or carriage returns) it was not quoted.
4508 This fix has been applied to the mysql and pgsql lookups by writing a
4509 separate quoting function and calling it from the lookup code. The fix
4510 should probably also be applied to nisplus, ibase and oracle lookups, but
4511 since I cannot test any of those, I have not disturbed their existing code.
4513 2. A hit in the callout cache for a specific address caused a log line with no
4514 reason for rejecting RCPT. Now it says "Previous (cached) callout
4515 verification failure".
4517 3. There was an off-by-one bug in the queryprogram router. An over-long
4518 return line was truncated at 256 instead of 255 characters, thereby
4519 overflowing its buffer with the terminating zero. As well as fixing this, I
4520 have increased the buffer size to 1024 (and made a note to document this).
4522 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4523 when Exim is waiting for an SMTP response from a remote server, Exim
4524 restarts its select() call on the socket, thereby resetting its timeout.
4525 This is not a problem when such interrupts are rare. Somebody set up a cron
4526 job to run exiwhat every 2 minutes, which is less than the normal select()
4527 timeout (5 or 10 minutes). This meant that the select() timeout never
4528 kicked in because it was always reset. I have fixed this by comparing the
4529 time when an interrupt arrives with the time at the start of the first call
4530 to select(). If more time than the timeout has elapsed, the interrupt is
4531 treated as a timeout.
4533 5. Some internal re-factoring in preparation for the addition of Sieve
4534 extensions (by MH). In particular, the "personal" test is moved to a
4535 separate function, and given an option for scanning Cc: and Bcc: (which is
4536 not set for Exim filters).
4538 6. When Exim created an email address using the login of the caller as the
4539 local part (e.g. when creating a From: or Sender: header line), it was not
4540 quoting the local part when it contained special characters such as @.
4542 7. Installed new OpenBSD configuration files.
4544 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4545 try to make them clearer.
4547 9. Callout options, other than the timeout value, were being ignored when
4548 verifying sender addresses in header lines. For example, when using
4550 verify = header_sender/callout=no_cache
4552 the cache was (incorrectly) being used.
4554 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4555 only the Cygwin environment.
4557 11. When return_path_on_delivery was set as a log selector, if different remote
4558 addresses in the same message used different return paths and parallel
4559 remote delivery occurred, the wrong values would sometimes be logged.
4560 (Whenever a remote delivery process finished, the return path value from
4561 the most recently started remote delivery process was logged.)
4563 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4564 lines when AUTH and/or TLS are in use. This is the "received protocol"
4565 field. Exim used to use "asmtp" for authenticated SMTP, without any
4566 indication (in the protocol name) for TLS use. Now it follows the RFC and
4567 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4568 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4569 names appear in log lines as well as in Received: header lines.
4571 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4572 extensions, and comparison tests, and to fix some bugs.
4574 14. Changes to the "personal" filter test:
4576 (1) The test was buggy in that it was just doing the equivalent of
4577 "contains" tests on header lines. For example, if a user's address was
4578 anne@some.where, the "personal" test would incorrectly be true for
4580 To: susanne@some.where
4582 This test is now done by extracting each address from the header in turn,
4583 and checking the entire address. Other tests that are part of "personal"
4584 are now done using regular expressions (for example, to check local parts
4585 of addresses in From: header lines).
4587 (2) The list of non-personal local parts in From: addresses has been
4588 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4589 taken from the Sieve specification recommendations.
4591 (3) If the message contains any header line starting with "List-" it is
4592 treated as non-personal.
4594 (4) The test for "circular" in the Subject: header line has been removed
4595 because it now seems ill-conceived.
4597 15. Minor typos in src/EDITME comments corrected.
4599 16. Installed latest exipick from John Jetmore.
4601 17. If headers_add on a router specified a text string that was too long for
4602 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4603 of string_sprintf() is now avoided.
4605 18. $message_body_size was not set (it was always zero) when running the DATA
4606 ACL and the local_scan() function.
4608 19. For the "mail" command in an Exim filter, no default was being set for
4609 the once_repeat time, causing a random time value to be used if "once" was
4610 specified. (If the value happened to be <= 0, no repeat happened.) The
4611 default is now 0s, meaning "never repeat". The "vacation" command was OK
4612 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4613 (I found it when inspecting the code).
4615 20. There is now an overall timeout for performing a callout verification. It
4616 defaults to 4 times the callout timeout, which applies to individual SMTP
4617 commands during the callout. The overall timeout applies when there is more
4618 than one host that can be tried. The timeout is checked before trying the
4619 next host. This prevents very long delays if there are a large number of
4620 hosts and all are timing out (e.g. when the network connections are timing
4621 out). The value of the overall timeout can be changed by specifying an
4622 additional sub-option for "callout", called "maxwait". For example:
4624 verify = sender/callout=5s,maxwait=20s
4626 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4627 to the end before writing, but this should make it even safer).
4629 22. Exim was forgetting that it had advertised PIPELINING for the second and
4630 subsequent messages on an SMTP connection. It was also not resetting its
4631 memory on STARTTLS and an internal HELO.
4633 23. When Exim logs an SMTP synchronization error within a session, it now
4634 records whether PIPELINING has been advertised or not.
4636 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4637 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4638 rather than long int.
4640 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4642 26. Added the never_mail option to autoreply.
4648 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4649 crash if the getsockname() call failed; this can happen if a connection is
4650 closed very soon after it is established. The problem was simply in the
4651 order in which certain operations were done, causing Exim to try to write
4652 to the SMTP stream before it had set up the file descriptor. The bug has
4653 been fixed by making things happen in the correct order.
4659 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4660 before the connection was closed, thus losing the rejection response.
4662 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4663 some early Solaris releases, but causes trouble in current releases where
4664 socklen_t is defined.
4666 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4669 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4672 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4673 that releasing the top of it at the end releases what was used for sub-
4674 expansions (unless the block got too big). However, discard this block if
4675 the first thing is a variable or header, so that we can use its block when
4676 it is dynamic (useful for very large $message_headers, for example).
4678 6. Lookups now cache *every* query, not just the most recent. A new, separate
4679 store pool is used for this. It can be recovered when all lookup caches are
4680 flushed. Lookups now release memory at the end of their result strings.
4681 This has involved some general refactoring of the lookup sources.
4683 7. Some code has been added to the store_xxx() functions to reduce the amount
4684 of flapping under certain conditions.
4686 8. log_incoming_interface used to affect only the <= reception log lines. Now
4687 it causes the local interface and port to be added to several more SMTP log
4688 lines, for example "SMTP connection from", and rejection lines.
4690 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4692 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4694 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4695 could be overwritten at the end of the current message (or the start of a
4696 new message if it was set in a HELO ACL). The value is now preserved for
4697 the duration of the SMTP connection.
4699 12. If a transport had a headers_rewrite setting, and a matching header line
4700 contained an unqualified address, that address was qualified, even if it
4701 did not match any rewriting rules. The underlying bug was that the values
4702 of the flags that permit the existence of unqualified sender and recipient
4703 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4704 non-local messages, and by -bnq for local messages) were not being
4705 preserved with the message after it was received.
4707 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4708 "next input=" as part of the text comprising the host identity instead of
4709 the correct text. The code was using the same buffer for two different
4710 strings. However, depending on which order the printing function evaluated
4711 its arguments, the bug did not always show up. Under Linux, for example, my
4712 test suite worked just fine.
4714 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4715 doesn't work with some older versions of Perl. It has been changed to "my",
4716 which in any case is probably the better facility to use.
4718 15. A really picky compiler found some instances of statements for creating
4719 error messages that either had too many or two few arguments for the format
4722 16. The size of the buffer for calls to the DNS resolver has been increased
4723 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4724 for addresses that have a lot of PTR records. This alleviates a problem; it
4725 does not fully solve it.
4727 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4728 buffer now truncates the list and logs the incident, which is the same
4729 action as happens when Exim is looking up a host name and its aliases.
4730 Previously in this situation something unpredictable would happen;
4731 sometimes it was "internal error: store_reset failed".
4733 18. If a server dropped the connection unexpectedly when an Exim client was
4734 using GnuTLS and trying to read a response, the client delivery process
4735 crashed while trying to generate an error log message.
4737 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4738 string, for example:
4740 warn message = H1: something\nH2: something
4742 the text was added as a single header line from Exim's point of view
4743 though it ended up OK in the delivered message. However, searching for the
4744 second and subsequent header lines using $h_h2: did not work. This has been
4745 fixed. Similarly, if a system filter added multiple headers in this way,
4746 the routers could not see them.
4748 20. Expanded the error message when iplsearch is called with an invalid key to
4749 suggest using net-iplsearch in a host list.
4751 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4752 ACLs are no longer actually imposed (and a message to that effect is
4755 22. If a "gecos" field in a passwd entry contained escaped characters, in
4756 particular, if it contained a \" sequence, Exim got it wrong when building
4757 a From: or a Sender: header from that name. A second bug also caused
4758 incorrect handling when an unquoted " was present following a character
4759 that needed quoting.
4761 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4762 was not being matched caselessly.
4764 24. Arranged for all hyphens in the exim.8 source to be escaped with
4767 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4768 properly. Recipient callout cache records were still being keyed to include
4769 the sender, even when use_sender was set false. This led to far more
4770 callouts that were necessary. The sender is no longer included in the key
4771 when use_sender is false.
4773 26. Added "control = submission" modifier to ACLs.
4775 27. Added the ${base62d: operator to decode base 62 numbers.
4777 28. dnsdb lookups can now access SRV records.
4779 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4780 the configuration file.
4782 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4783 (-v) mode. This makes the output for a verbose queue run more intelligible.
4785 31. Added a use_postmaster feature to recipient callouts.
4787 32. Added the $body_zerocount variable, containing the number of binary zero
4788 bytes in the message body.
4790 33. The time of last modification of the "new" subdirectory is now used as the
4791 "mailbox time last read" when there is a quota error for a maildir
4794 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4796 35. Added +ignore_unknown as a special item in host lists.
4798 36. Code for decoding IPv6 addresses in host lists is now included, even if
4799 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4800 address was recognized as an IP address, but was then not correctly decoded
4801 into binary, causing unexpected and incorrect effects when compared with
4808 1. Very minor rewording of debugging text in manualroute to say "list of
4809 hosts" instead of "hostlist".
4811 2. If verify=header_syntax was set, and a header line with an unqualified
4812 address (no domain) and a large number of spaces between the end of the
4813 name and the colon was received, the reception process suffered a buffer
4814 overflow, and (when I tested it) crashed. This was caused by some obsolete
4815 code that should have been removed. The fix is to remove it!
4817 3. When running in the test harness, delay a bit after writing a bounce
4818 message to get a bit more predictability in the log output.
4820 4. Added a call to search_tidyup() just before forking a reception process. In
4821 theory, someone could use a lookup in the expansion of smtp_accept_max_
4822 per_host which, without the tidyup, could leave open a database connection.
4824 5. Added the variables $recipient_data and $sender_data which get set from a
4825 lookup success in an ACL "recipients" or "senders" condition, or a router
4826 "senders" option, similar to $domain_data and $local_part_data.
4828 6. Moved the writing of debug_print from before to after the "senders" test
4831 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4832 problems for message scanning, either using a data ACL, or using
4833 local_scan() because the Received: header was not generated till after they
4834 were called (in order to set the time as the time of reception completion).
4835 I have revised the way this works. The header is now generated after the
4836 body is received, but before the ACL or local_scan() are called. After they
4837 are run, the timestamp in the header is updated.
4843 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4844 before starting a queue runner without re-exec. This happened only when
4845 deliver_drop_privilege was set or when the Exim user was set to root. The
4846 effect of the bug was that timeouts during subsequent deliveries caused
4847 crashes instead of being properly handled. The handler is now left at its
4848 default (and expected) setting.
4850 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4851 message, again when deliver_drop_privilege is set or Exim is run as root.
4852 The bug described in (1) was not present in this case, but the tidying up
4853 of the other signals was missing. I have made the two cases consistent.
4855 3. The ignore_target_hosts setting on a manualroute router was being ignored
4856 for hosts that were looked up using the /MX notation.
4858 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4861 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4862 operated on the sender address. After changing the $sender_address to <>
4863 for the sender address verify, Exim was re-instated it as the original
4864 (before rewriting) address, but remembering that it had rewritten it, so it
4865 wasn't rewriting it again. This bug also had the effect of breaking the
4866 sender address verification caching when the sender address was rewritten.
4868 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4869 This has been changed so that if the ip literal address matches
4870 ignore_target_hosts, the router declines.
4872 7. Added expansion conditions match_domain, match_address, and match_local_
4873 part (NOT match_host).
4875 8. The placeholder for the Received: header didn't have a length field set.
4877 9. Added code to Exim itself and to exim_lock to test for a specific race
4878 condition that could lead to file corruption when using MBX delivery. The
4879 issue is with the lockfile that is created in /tmp. If this file is removed
4880 after a process has opened it but before that process has acquired a lock,
4881 there is the potential for a second process to recreate the file and also
4882 acquire a lock. This could lead to two Exim processes writing to the file
4883 at the same time. The added code performs the same test as UW imapd; it
4884 checks after acquiring the lock that its file descriptor still refers to
4885 the same named file.
4887 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4888 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4891 11. Added the smtp_active_hostname option. If used, this will typically be made
4892 to depend on the incoming interface address. Because $interface_address is
4893 not set up until the daemon has forked a reception process, error responses
4894 that can happen earlier (such as "too many connections") no longer contain
4897 12. If an expansion in a condition on a "warn" statement fails because a lookup
4898 defers, the "warn" statement is abandoned, and the next ACL statement is
4899 processed. Previously this caused the whole ACL to be aborted.
4901 13. Added the iplsearch lookup type.
4903 14. Added ident_timeout as a log selector.
4905 15. Added tls_certificate_verified as a log selector.
4907 16. Added a global option tls_require_ciphers (compare the smtp transport
4908 option of the same name). This controls incoming TLS connections.
4910 17. I finally figured out how to make tls_require_ciphers do a similar thing
4911 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4912 before starting the TLS session.
4914 18. Tabs are now shown as \t in -bP output.
4916 19. If the log selector return_path_on_delivery was set, Exim crashed when
4917 bouncing a message because it had too many Received: header lines.
4919 20. If two routers both had headers_remove settings, and the first one included
4920 a superfluous trailing colon, the final name in the first list and the
4921 first name in the second list were incorrectly joined into one item (with a
4922 colon in the middle).
4928 1. Added -C and -D options to the exinext utility, mainly to make it easier
4929 to include in the automated testing, but these could be helpful when
4930 multiple configurations are in use.
4932 2. The exinext utility was not formatting the output nicely when there was
4933 an alternate port involved in the retry record key, nor when there was a
4934 message id as well (for retries that were specific to a specific message
4935 and a specific host). It was also confused by IPv6 addresses, because of
4936 the additional colons they contain. I have fixed the IPv4 problem, and
4937 patched it up to do a reasonable job for IPv6.
4939 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4940 delivery, the log line now contains "pipelined" if PIPELINING was used.
4942 4. An SMTP transport process used to panic and die if the bind() call to set
4943 an explicit outgoing interface failed. This has been changed; it is now
4944 treated in the same way as a connect() failure.
4946 5. A reference to $sender_host_name in the part of a conditional expansion
4947 that was being skipped was still causing a DNS lookup. This no longer
4950 6. The def: expansion condition was not recognizing references to header lines
4951 that used bh_ and bheader_.
4953 7. Added the _cache feature to named lists.
4955 8. The code for checking quota_filecount in the appendfile transport was
4956 allowing one more file than it should have been.
4958 9. For compatibility with Sendmail, the command line option
4966 and sets the incoming protocol and host name (for trusted callers). The
4967 host name and its colon can be omitted when only the protocol is to be set.
4968 Note the Exim already has two private options, -pd and -ps, that refer to
4969 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4970 "s", but I don't think that's a major issue.
4972 10. A number of refactoring changes to the code, none of which should affect
4975 (a) The number of logging options was getting close to filling up the
4976 32-bit word that was used as a bit map. I have split them into two classes:
4977 those that are passed in the argument to log_write(), and those that are
4978 only ever tested independently outside of that function. These are now in
4979 separate 32-bit words, so there is plenty of room for expansion again.
4980 There is no change in the user interface or the logging behaviour.
4982 (b) When building, for example, log lines, the code previously used a
4983 macro that called string_cat() twice, in order to add two strings. This is
4984 not really sufficiently general. Furthermore, there was one instance where
4985 it was actually wrong because one of the argument was used twice, and in
4986 one call a function was used. (As it happened, calling the function twice
4987 did not affect the overall behaviour.) The macro has been replaced by a
4988 function that can join an arbitrary number of extra strings onto a growing
4991 (c) The code for expansion conditions now uses a table and a binary chop
4992 instead of a serial search (which was left over from when there were very
4993 few conditions). Also, it now recognizes conditions like "pam" even when
4994 the relevant support is not compiled in: a suitably worded error message is
4995 given if an attempt is made to use such a condition.
4997 11. Added ${time_interval:xxxxx}.
4999 12. A bug was causing one of the ddress fields not to be passed back correctly
5000 from remote delivery subprocesses. The field in question was not being
5001 subsequently used, so this caused to problems in practice.
5003 13. Added new log selectors queue_time and deliver_time.
5005 14. Might have fixed a bug in maildirsizefile handling that threw up
5006 "unexpected character" debug warnings, and recalculated the data
5007 unnecessarily. In any case, I expanded the warning message to give more
5010 15. Added the message "Restricted characters in address" to the statements in
5011 the default ACL that block characters like @ and % in local parts.
5013 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5014 Three changes have been made:
5016 (a) There was a serious bug; a negative response to MAIL caused the whole
5017 recipient domain to be cached as invalid, thereby blocking all messages
5018 to all local parts at the same domain, from all senders. This bug has
5019 been fixed. The domain is no longer cached after a negative response to
5020 MAIL if the sender used is not empty.
5022 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5025 (c) A new callout option, "use_sender" has been added for people who want
5026 the modified behaviour.
5032 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5035 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5036 indeed breaks things for older releases.
5038 3. Added additional logging to the case where there is a problem reading data
5039 from a filter that is running in a subprocess using a pipe, in order to
5040 try to track down a specific problem.
5042 4. Testing facility fudge: when running in the test harness and attempting
5043 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5044 getting "No route to host". Convert this to a timeout.
5046 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5049 6. Some OS don't have socklen_t but use size_t instead. This affects the
5050 fifth argument of getsockopt() amongst other things. This is now
5051 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5052 can be set for individual OS. I have set it for SunOS5, OSF1, and
5053 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5054 some earlier ones do not.
5056 7. Change 4.30/15 was not doing the test caselessly.
5058 8. The standard form for an IPv6 address literal was being rejected by address
5059 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5060 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5061 this, as well as the form without the "IPv6" on the front (but only when
5062 address literals are enabled, of course).
5064 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5066 10. Exim crashed if a message with an empty sender address specified by -f
5067 encountered a router with an errors_to setting. This could be provoked only
5068 by a command such as
5072 where an empty string was supplied; "<>" did not hit this bug.
5074 11. Installed PCRE release 4.5.
5076 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5077 remained set. It is now erased.
5079 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5080 times from message ids (which are base 36 rather than the normal 62).
5082 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5083 were being counted as actual protocol errors, and logged if the log
5084 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5085 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5086 and DATA following a set of rejected RCPTs do not count as protocol errors.
5087 In other words, Exim assumes they were pipelined, though this may not
5088 actually be the case. Of course, in all cases the client gets an
5089 appropriate error code.
5091 15. If a lookup fails in an ACL condition, a message about the failure may
5092 be available; it is used if testing the ACL cannot continue, because most
5093 such messages specify what the cause of the deferral is. However, some
5094 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5095 that caused an old message to be retained and used if a later statement
5096 caused a defer, replacing the real cause of the deferral.
5098 16. If an IP address had so many PTR records that the DNS lookup buffer
5099 was not large enough to hold them, Exim could crash while trying to process
5100 the truncated data. It now detects and logs this case.
5102 17. Further to 4.21/58, another change has been made: if (and only if) the
5103 first line of a message (the first header line) ends with CRLF, a bare LF
5104 in a subsequent header line has a space inserted after it, so as not to
5105 terminate the header.
5107 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5108 unnecessarily, used atoi() instead of strtol(), and didn't check the
5109 termination when getting file sizes from file names by regex.
5111 19. Completely re-implemented the support for maildirsize files, in the light
5112 of a number of problems with the previous contributed implementation
5113 (4.30/29). In particular:
5115 . If the quota is zero, the maildirsize file is maintained, but no quota is
5118 . If the maildir directory does not exist, it is created before any attempt
5119 to write a maildirsize file.
5121 . The quota value in the file is just a cache; if the quota is changed in
5122 the transport, the new value overrides.
5124 . A regular expression is available for excluding directories from the
5127 20. The autoreply transport checks the characters in options that define the
5128 message's headers; it allows continued headers, but it was checking with
5129 isspace() after an embedded newline instead of explicitly looking for a
5132 21. If all the "regular" hosts to which an address was routed had passed their
5133 expiry times, and had not reached their retry times, the address was
5134 bounced, even if fallback hosts were defined. Now Exim should go on to try
5137 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5138 equivalent code in the SMTP transport. Some hosts send humungous responses
5139 to HELO/EHLO, more than 1024 it seems.
5141 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5142 gives compiler warnings in some environments. I've now done it "properly",
5145 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5146 (because of problems with the built-in one) was declared to return uschar *
5147 instead of char *, causing compiler failure.
5149 25. Fixed a file descriptor leak when processing alias/forward files.
5151 26. Fixed a minor format string issue in dbfn.c.
5153 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5155 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5156 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5159 29. When a sender address is verified, it is cached, to save repeating the test
5160 when there is more than one recipient in a message. However, when the
5161 verification involves a callout, it is possible for different callout
5162 options to be set for different recipients. It is too complicated to keep
5163 track of this in the cache, so now Exim always runs a verification when a
5164 callout is required, relying on the callout cache for the optimization.
5165 The overhead is duplication of the address routing, but this should not be
5168 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5169 to be verified with callout=postmaster, and the main callout worked but the
5170 postmaster check failed, the verification correctly failed. However, if a
5171 subsequent RCPT command asked for sender verification *without* the
5172 postmaster check, incorrect caching caused this verification also to fail,
5175 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5176 it was not caching the DNS options (qualify_single, search_parents) that
5177 were used when the lookup failed. A subsequent lookup with different
5178 options therefore always gave the same answer, though there were cases
5179 where it should not have. (Example: a "domains = !$mx_any" option on a
5180 dnslookup router: the "domains" option is always processed without any
5181 widening, but the router might have qualify_single set.) Now Exim uses the
5182 cached value only when the same options are set.
5184 32. Added John Jetmore's "exipick" utility to the distribution.
5186 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5187 than a timeout (e.g. a certificate is required, and is not provided), an
5188 Exim server now closes the connection immediately. Previously it waited for
5189 the client to close - but if the client is SSL, it seems that they each
5190 wait for each other, leading to a delay before one of them times out.
5192 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5193 maintained 0.8.x compatibility because I don't think many are using it, and
5194 it is clearly obsolete.
5196 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5199 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5200 client certificate was expired. A simple patch fixes this, though I don't
5201 understand the full logic of why the verify callback is called multiple
5204 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5205 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5206 which causes problems with some clients (such as the Certicom SSL Plus
5207 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5208 disables the coutermeasure allowing Eudora to connect."
5210 38. Exim was not checking that a write() to a log file succeeded. This could
5211 lead to Bad Things if a log got too big, in particular if it hit a file
5212 size limit. Exim now panics and dies if it cannot write to a log file, just
5213 as it does if it cannot open a log file.
5215 39. Modified OS/Makefile-Linux so that it now contains
5217 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5219 The two -D definitions ensure that Exim is compiled with large file
5220 support, which makes it possible to handle log files that are bigger than
5223 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5224 instance) a domain was checked against a named list that involved a lookup,
5225 causing $domain_data to be set, then another domain was checked against the
5226 same list, then the first domain was re-checked, the value of $domain_data
5227 after the final check could be wrong. In particular, if the second check
5228 failed, it could be set empty. This bug probably also applied to
5231 41. The strip_trailing_dot option was not being applied to the address given
5232 with the -f command-line option.
5234 42. The code for reading a message's header from the spool was incrementing
5235 $received_count, but never initializing it. This meant that the value was
5236 incorrect (doubled) while delivering a message in the same process in which
5237 it was received. In the most common configuration of Exim, this never
5238 happens - a fresh exec is done - but it can happen when
5239 deliver_drop_privilege is set.
5241 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5242 it now includes up to 150 characters of the unexpected data in the log
5245 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5246 and building data strings. The size of both of these buffers was 10 000
5247 bytes - far larger than anybody would *ever* want, thought I. Needless to
5248 say, somebody hit the limit. I have increased the maximum line length to
5249 20 000 and the maximum data length of concatenated lines to 100 000. I have
5250 also fixed two bugs, because there was no checking on these buffers. Tsk,
5251 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5252 buffer is too small.
5254 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5255 lsearch lookups. Now it does.
5257 46. When parsing a route_list item in a manualroute router, a fixed-length
5258 buffer was used for the list of hosts. I made this 1024 bytes long,
5259 thinking that nobody would ever have a list of hosts that long. Wrong.
5260 Somebody had a whole pile of complicated expansion conditions, and the
5261 string was silently truncated, leading to an expansion error. It turns out
5262 that it is easier to change to an unlimited length (owing to other changes
5263 that have happened since this code was originally written) than to build
5264 structure for giving a limitation error. The length of the item that
5265 expands into the list of hosts is now unlimited.
5267 47. The lsearch lookup could not handle data where the length of text line was
5268 more than 4095 characters. Such lines were truncated, leading to shortened
5269 data being returned. It should now handle lines of any length.
5271 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5272 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5275 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5276 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5277 no longer happens in comment lines. A list of replacements is now placed
5278 at the head of all of the source files, except those whose only change is
5279 to replace PERL_COMMAND in the very first #! line.
5281 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5282 messages on the queue, with a bottom-up merge sort, using code contributed
5283 by Michael Haardt. This should make operations like -bp somewhat faster on
5284 large queues. It won't affect queue runners, except when queue_run_in_order
5287 51. Installed eximstats 1.31 in the distribution.
5289 52. Added support for SRV lookups to the dnslookup router.
5291 53. If an ACL referred to $message_body or $message_body_end, the value was not
5292 reset for any messages that followed in the same SMTP session.
5294 54. The store-handling optimization for building very long strings was not
5295 differentiating between the different store pools. I don't think this
5296 actually made any difference in practice, but I've tidied it.
5298 55. While running the routers to verify a sender address, $sender_address
5299 was still set to the sender address. This is wrong, because when routing to
5300 send a bounce to the sender, it would be empty. Therefore, I have changed
5301 it so that, while verifying a sender address, $sender_address is set to <>.
5302 (There is no change to what happens when verifying a recipient address.)
5304 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5305 target A or AAAA records (if not already returned) without resetting the
5306 qualify_single or search_parents options of the DNS resolver. These are
5307 inappropriate in this case because the targets of MX and SRV records must
5308 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5309 target that, when qualified, matched something in the local domain. These
5310 two options are now turned off when doing these lookups.
5312 57. It seems that at least some releases of Reiserfs (which does not have the
5313 concept of a fixed number of inodes) returns zero and not -1 for the
5314 number of available inodes. This interacted badly with check_spool_inodes,
5315 which assumed that -1 was the "no such thing" setting. What I have done is
5316 to check that the total number of inodes is greater than zero before doing
5317 the test of how many are available.
5319 58. When a "warn" ACL statement has a log_message modifier, the message is
5320 remembered, and not repeated. This is to avoid a lot of repetition when a
5321 message has many recipients that cause the same warning to be written.
5322 Howewer, Exim was preserving the list of already written lines for an
5323 entire SMTP session, which doesn't seem right. The memory is now reset if a
5324 new message is started.
5326 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5327 debugging output unless log_rewrite was also set.
5329 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5330 of (int)(handle) when we know that handle contains (void *)(-1).
5332 61. The Exim daemon panic-logs an error return when it closes the incoming
5333 connection. However "connection reset by peer" seems to be common, and
5334 isn't really an error worthy of noting specially, so that particular error
5337 62. When Exim is trying to find all the local interfaces, it used to panic and
5338 die if the ioctl to get the interface flags failed. However, it seems that
5339 on at least one OS (Solaris 9) it is possible to have an interface that is
5340 included in the list of interfaces, but for which you get a failure error
5341 for this call. This happens when the interface is not "plumbed" into a
5342 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5343 failure of the "get flags" call assumes that the interface is down.
5345 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5346 makes life easier for people who are doing arithmetic on fields extracted
5347 from dates, where you often get leading zeros that should not be
5348 interpreted as octal.
5350 64. Added qualify_domain to the redirect router, to override the global
5353 65. If a pathologically long header line contained very many addresses (the
5354 report of this problem mentioned 10 000) and each of them was rewritten,
5355 Exim could use up a very large amount of memory. (It kept on making new
5356 copies of the header line as it rewrote, and never released the old ones.)
5357 At the expense of a bit more processing, the header rewriting function has
5358 been changed so that it no longer eats memory in this way.
5360 66. The generation of the Received: header has been moved from the time that a
5361 message starts to be received, to the time that it finishes. The timestamp
5362 in the Received: header should now be very close to that of the <= log
5363 line. There are two side-effects of this change:
5365 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5366 logged header lines no longer include the local Received: line, because
5367 it has not yet been created. The same applies to a copy of the message
5368 that is returned to a non-SMTP sender when a message is rejected.
5370 (b) When a filter file is tested using -bf, no additional Received: header
5371 is added to the test message. After some thought, I decided that this
5374 This change does not affect the value of $received_for. It is still set
5375 after address rewriting, but before local_scan() is called.
5377 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5379 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5380 gave an unhelpful panic error message, and a defer error. I have managed to
5381 change this behaviour so that it now rejects any supplied certificate,
5382 which seems right, as the list of acceptable certificates is empty.
5384 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5385 gave an unhelpful defer error. I have not managed to make this reject any
5386 supplied certificates, but the error message it gives is "no certificate
5387 supplied", which is not helpful.
5389 70. exigrep's output now also includes lines that are not associated with any
5390 message, but which match the given pattern. Implemented by a patch from
5391 Martin Sluka, which also tidied up the Perl a bit.
5393 71. Recipient callout verification, like sender verification, was using <> in
5394 the MAIL FROM command. This isn't really the right thing, since the actual
5395 sender may affect whether the remote host accepts the recipient or not. I
5396 have changed it to use the actual sender in the callout; this means that
5397 the cache record is now keyed on a recipient/sender pair, not just the
5398 recipient address. There doesn't seem to be a real danger of callout loops,
5399 since a callout by the remote host to check the sender would use <>.
5400 [SEE ABOVE: changed after hitting problems.]
5402 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5403 temporary errors. However, in the case of such a code being given after
5404 the end of a data transmission (i.e. after ".") Exim was failing to write
5405 a retry record for the message. (Yes, there was some broken host that was
5406 actually sending 8xx at this point.)
5408 73. An unknown lookup type in a host list could cause Exim to panic-die when
5409 the list was checked. (An example that provoked this was putting <; in the
5410 middle of a list instead of at the start.) If this happened during a DATA
5411 ACL check, a -D file could be left lying around. This kind of configuration
5412 error no longer causes Exim to die; instead it causes a defer errror. The
5413 incident is still logged to the main and panic logs.
5415 74. Buglet left over from Exim 3 conversion. The message "too many messages
5416 in one connection" was written to the rejectlog but not the mainlog, except
5417 when address rewriting (yes!) was being logged.
5419 75. Added write_rejectlog option.
5421 76. When a system filter was run not as root (that is, when system_filter_user
5422 was set), the values of the $n variables were not being returned to the
5423 main process; thus, they were not subsequently available in the $sn
5426 77. Added +return_path_on_delivery log selector.
5428 78. A connection timeout was being treated differently from recipients deferred
5429 when testing hosts_max_try with a message that was older than the host's
5430 retry timeout. (The host should not be counted, thus allowing all hosts to
5431 be tried at least once before bouncing.) This may have been the cause of an
5432 occasionally reported bug whereby a message would remain on the queue
5433 longer than the retry timeout, but would be bounced if a delivery was
5434 forced. I say "may" because I never totally pinned down the problem;
5435 setting up timeout/retry tests is difficult. See also the next item.
5437 79. The ultimate address timeout was not being applied to errors that involved
5438 a combination of host plus message (for example, a timeout on a MAIL
5439 command). When an address resolved to a number of possible hosts, and they
5440 were not all tried for each delivery (e.g. because of hosts_max_try), a
5441 message could remain on the queue longer than the retry timeout.
5443 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5446 81. Fixed an obscure SMTP outgoing bug which required at least the following
5447 conditions: (a) there was another message waiting for the same server;
5448 (b) the server returned 5xx to all RCPT commands in the first message so
5449 that the message was not completed; (c) the server dropped the connection
5450 or gave a negative response to the RSET that Exim sends to abort the
5451 transaction. The observed case was a dropped connection after DATA that had
5452 been sent in pipelining mode. That is, the server had advertised PIPELINING
5453 but was not implementing it correctly. The effect of the bug was incorrect
5454 behaviour, such as trying another host, and this could lead to a crash.
5460 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5461 and daemon.c were passed as pointers to ints; they should have been
5462 pointers to socklen_t variables (which are typically unsigned ints).
5464 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5467 3. Fixed a really odd bug that affected only the testing scheme; patching a
5468 certain fixed string in the binary changed the value of another string that
5469 happened to be identical to the end of the original first string.
5471 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5472 name", it returns that address as the IP address. On some operating
5473 systems (e.g. Solaris), it also passes back the IP address string as the
5474 "host name". However, on others (e.g. Linux), it passes back an empty
5475 string. Exim wasn't checking for this, and was changing the host name to an
5476 empty string, assuming it had been canonicized.
5478 5. Although rare, it is permitted to have more than one PTR record for a given
5479 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5480 all the names associated with an address, because they do in Solaris.
5481 However, it seems that they do not in Linux for data that comes from the
5482 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5483 I found this out when I moved to a new Linux workstation and tried to run
5484 the Exim test suite.
5486 To get round this problem I have changed the code so that it now does its
5487 own call to the DNS to look up PTR records when searching for a host name.
5488 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5489 addresses that are only in /etc/hosts are still found.
5491 This behaviour is, however, controlled by an option called host_lookup_
5492 order, which defaults to "bydns:byaddr". If people want to use the other
5493 order, or indeed, just use one or the other means of lookup, they can
5494 specify it in this variable.
5496 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5497 some operating systems, this comes back from gethostbyaddr() as an empty
5498 string, and this is what Exim used to test for. However, it seems that in
5499 other systems, "." is yielded. Exim now tests for this case too.
5501 7. The values of check_spool_space and check_log_space are now held internally
5502 as a number of kilobytes instead of an absolute number of bytes. If a
5503 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5504 kilobyte. This means that much larger values can be stored.
5506 8. Exim monitor: an attempt to get the action menu when not actually pointing
5507 at a message produces an empty menu entitled "No message selected". This
5508 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5509 no entries in it ("Shell widget menu has zero width and/or height"). So I
5510 have added a single, blank menu entry in this case.
5512 9. Added ${quote_local_part.
5514 10. MIME decoding is now applied to the contents of Subject: header lines when
5517 11. Now that a reference to $sender_host_address automatically causes a reverse
5518 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5519 host lookup before query-style lookups in lists that might use this
5520 variable. This has therefore been abolished, and the "net-" prefix is no
5521 longer necessary for query-style lookups.
5523 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5524 have been a typo for LFLAGS, so it has been changed.
5526 13. The install script calls Exim with "-C /dev/null" in order to find the
5527 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5528 to be output. Howeve, since Exim outputs its version number before the
5529 error, it didn't break the script. It just looked ugly. I fixed this by
5530 always allowing "-C /dev/null" if the caller is root.
5532 14. Ignore overlarge ACL variable number when reading spool file - insurance
5533 against a later release with more variables having written the file.
5535 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5536 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5537 this, as well as the form without the "IPv6" on the front.
5539 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5540 OS/Makefile-Darwin file.
5542 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5543 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5545 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5546 variable. However, some OS use ints for inodes. I've added cast to long int
5547 to get rid of the compiler warning.
5549 19. I had forgotten to lock out "/../" in configuration file names when
5550 ALT_CONFIG_PREFIX was set.
5552 20. Routers used for verification do not need to specify transports. However,
5553 if such a router generated a host list, and callout was configured, Exim
5554 crashed, because it could not find a port number from the (non-existent)
5555 transport. It now assumes port 25 in this circumstance.
5557 21. Added the -t option to exigrep.
5559 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5560 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5561 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5562 them is set, LOOKUP_LSEARCH is forced.
5564 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5566 24. Added sender and host information to the "rejected by local_scan()" log
5567 line; previously there was no indication of these.
5569 25. Added .include_if_exists.
5571 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5572 Linux. It turns out that not all file systems support this. Apparently some
5573 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5574 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5575 not supported on the file descriptor, is now ignored when Exim is trying to
5576 sync a directory. This applies only to Linux.
5578 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5580 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5581 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5584 29. Added support for maildirsize files from supplied patch (modified a bit).
5586 30. The use of :fail: followed by an empty string could lead Exim to respond to
5587 sender verification failures with (e.g.):
5589 550 Verification failed for <xxx>
5590 550 Sender verify failed
5592 where the first response line was missing the '-' that indicates it is not
5593 the final line of the response.
5595 31. The loop for finding the name of the user that called Exim had a hardwired
5596 limit of 10; it now uses the value of finduser_retries, which is used for
5597 all other user lookups.
5599 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5602 33. Exim was neglecting to zero errno before one call of strtol() when
5603 expanding a string and expecting an integer value. On some systems this
5604 resulted in spurious "integer overflow" errors. Also, it was casting the
5605 result into an int without checking.
5607 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5608 did not work. The code looks as if it has *never* worked, though it appears
5609 to have been documented since at least releast 1.62. I have made it work.
5611 35. The "timeout_DNS" error in retry rules, also documented since at least
5612 1.62, also never worked. As it isn't clear exactly what this means, and
5613 clearly it isn't a major issue, I have abolished the feature by treating it
5614 as "timeout", and writing a warning to the main and panic logs.
5616 36. The display of retry rules for -brt wasn't always showing the error code
5619 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5620 timeout_connect_A, timeout_connect_MX.
5622 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5623 to the empty sender.
5625 39. The daemon was not analysing the content of -oX till after it had closed
5626 stderr and disconnected from the controlling terminal. This meant that any
5627 syntax errors were only noted on the panic log, and the return code from
5628 the command was 0. By re-arranging the code a little, I've made the
5629 decoding happen first, so such errors now appear on stderr, and the return
5630 code is 1. However, the actual setting up of the sockets still happens in
5631 the disconnected process, so errors there are still only recorded on the
5634 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5635 connections (as happens on some IP stacks) was logged at start up time as
5636 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5637 it from "IPv6 and IPv4", which means that two separate sockets are being
5640 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5641 says whether AF_INET or AF_INET6 was passed as an argument.
5643 42. Exiwhat output was messed up when time zones were included in log
5646 43. Exiwhat now gives more information about the daemon's listening ports,
5647 and whether -tls-on-connect was used.
5649 44. The "port" option of the smtp transport is now expanded.
5651 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5652 silently ignored. Now an error message is written to the main and panic
5655 46. There's a new ACL modifier called "logwrite" which writes to a log file
5656 as soon as it is encountered.
5658 47. Added $local_user_uid and $local_user_gid at routing time.
5660 48. Exim crashed when trying to verify a sender address that was being
5663 49. Exim was recognizing only a space character after ".include". It now also
5664 recognizes a tab character.
5666 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5667 extracting the relevant information from the specification. The man page no
5668 longer contains scrambled data for the -d option, and I've added a section
5669 at the front about calling Exim under different names.
5671 51. Added "extra_headers" argument to the "mail" command in filter files.
5673 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5676 53. Installed eximstats 1.29.
5678 54. Added transport_filter_timeout as a generic transport option.
5680 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5681 Cc: header lines. This was required by RFC 822, but it not required by RFC
5684 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5685 incoming messages that did not have them. Now it does so only if the
5686 message originates locally, that is, if there is no associated remote host
5687 address. When Resent- header lines are present, this applies to the Resent-
5688 lines rather than the non-Resent- lines.
5690 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5691 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5693 58. Messages for configuration errors now include the name of the main
5694 configuration file - useful now that there may be more than one file in a
5695 list (.included file names were always shown).
5697 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5698 for those rare installations that do not start the daemon as root or run it
5699 setuid root. I've cut out the call to initgroups() if the daemon is not
5702 60. The Exim user and group can now be bound into the binary as text strings
5703 that are looked up at the start of Exim's processing.
5705 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5707 62. Added $mailstore_basename variable.
5709 63. Installed patch to sieve.c from Michael Haardt.
5711 64. When Exim failed to open the panic log after failing to open the main log,
5712 the original message it was trying to log was written to stderr and debug
5713 output, but if they were not available (the usual case in production), it
5714 was lost. Now it is written to syslog before the two lines that record the
5715 failures to open the logs.
5717 65. Users' Exim filters run in subprocesses under the user's uid. It is
5718 possible for a "deliver" command or an alias in a "personal" command to
5719 provoke an address rewrite. If logging of address rewriting is configured,
5720 this fails because the process is not running as root or exim. There may be
5721 a better way of dealing with this, but for the moment (because 4.30 needs
5722 to be released), I have disabled address rewrite logging when running a
5723 filter in a non-root, non-exim process.
5729 1. The buildconfig auxiliary program wasn't quoting the value set for
5730 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5731 not defined. This bug was masked in 4.22 by the effect that was fixed in
5734 2. Some messages that were rejected after a message id was allocated were
5735 shown as "incomplete" by exigrep. It no longer does this for messages that
5736 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5738 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5739 have allow_domain_literals set, the ID did not get logged in the <= line.
5740 Domain literals are now always recognized in Message-ID: header lines.
5742 4. The first argument for a ${extract expansion item is the key name or field
5743 number. Leading and trailing spaces in this item were not being ignored,
5744 causing some misleading effects.
5746 5. When deliver_drop_privilege was set, single queue runner processes started
5747 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5748 same command in the process it spins off) were not dropping privilege.
5750 6. When the daemon running as "exim" started a queue runner, it always
5751 re-executed Exim in the spun-off process. This is a waste of effort when
5752 deliver_drop_privilege is set. The new process now just calls the
5753 queue-runner function directly.
5759 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5762 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5763 ignored. Though the use of -oP was forcing the writing of a pid file, it
5764 was always written to the default place.
5766 3. If the message "no IP address found for host xxxx" is generated during
5767 incoming verification, it is now followed by identification of the incoming
5768 connection (so you can more easily find what provoked it).
5770 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5772 5. Added some features to "harden" Exim a bit more against certain attacks:
5774 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5775 be put in Local/Makefile. This is like the never_users runtime option,
5776 but it cannot be overridden. The default setting is "root".
5778 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5779 prefix string with which any file named in a -C command line option
5782 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5783 is retained for -C and -D only if the caller of Exim is root. Without
5784 it, the exim user may also use -C and -D and retain privilege.
5786 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5787 command line option is disabled.
5789 6. Macro names set by the -D option must start with an upper case letter, just
5790 like macro names defined in the configuration file.
5792 7. Added "dereference=" facility to LDAP.
5794 8. Two instances of the typo "uknown" in the source files are fixed.
5796 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5797 the Configure-Makefile script screwed up while processing it.
5799 10. Incorporated PCRE 4.4.
5801 11. The SMTP synchronization check was not operating right at the start of an
5802 SMTP session. For example, it could not catch a HELO sent before the client
5803 waited for the greeting. There is now a check for outstanding input at the
5804 point when the greeting is written. Because of the duplex, asynchronous
5805 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5806 way, but not yet received, when the check is performed.
5808 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5809 on TCP/IP sockets, because this apparently causes some broken clients to
5812 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5813 unchanged) from the Cygwin maintainer.
5815 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5816 was supported instead of testing for mbx. Effectively a typo.
5818 15. The spa authenticator server code was not checking that the input it
5819 received was valid base64.
5821 16. The debug output line for the "set" modifier in ACLs was not showing the
5822 name of the variable that was being set.
5824 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5826 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5827 lookup on its own. Something else was needed to trigger the lookup. For
5828 example, a match in host_lookup or the need for a host name in a host list.
5829 Now, if $sender_host_name is referenced and the host name has not yet been
5830 looked up, a lookup is performed. If the lookup fails, the variable remains
5831 empty, and $host_lookup_failed is set to "1".
5833 19. Added "eqi" as a case-independent comparison operator.
5835 20. The saslauthd authentication condition could segfault if neither service
5836 nor realm was specified.
5838 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5839 error message that was logged was misleading, and incoming SMTP
5840 connections were dropped. The message is now more accurate, and temporary
5841 errors are given to SMTP connections.
5843 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5844 (except RSET) with a 421 error, until QUIT is received. However, it was
5845 failing to send a response to QUIT.
5847 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5848 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5849 ACL is now run after the helo_try_verify_hosts code.
5851 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5852 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5853 case-independent, so other case variants are also recognized). Apparently
5854 some systems use these upper case variants.
5856 25. If more than two messages were waiting for the same host, and a transport
5857 filter was specified for the transport, Exim sent two messages over the
5858 same TCP/IP connection, and then failed with "socket operation on non-
5859 socket" when it tried to send the third.
5861 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5863 27. The extern definition of crypt16() in expand.c was not being excluded when
5864 the OS had its own crypt16() function.
5866 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5867 as a preferred synonym for return_size_limit, both as an option and as an
5870 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5872 30. Changed the default configuration ACL to relax the local part checking rule
5873 for addresses that are not in any local domains. For these addresses,
5874 slashes and pipe symbols are allowed within local parts, but the sequence
5875 /../ is explicitly forbidden.
5877 31. SPA server authentication was not clearing the challenge buffer before
5880 32. log_message in a "warn" ACL statement was writing to the reject log as
5881 well as to the main log, which contradicts the documentation and doesn't
5882 seem right (because no rejection is happening). So I have stopped it.
5884 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5885 However, I am unable to do any testing of this.
5887 34. Fixed an infelicity in the appendfile transport. When checking directories
5888 for a mailbox, to see if any needed to be created, it was accidentally
5889 using path names with one or more superfluous leading slashes; tracing
5890 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5892 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5893 contents are added to the log line that is written for every discarded
5894 recipient. (Previously a log_message setting was ignored.)
5896 36. The ${quote: operator now quotes the string if it is empty.
5898 37. The install script runs exim in order to find its version number. If for
5899 some reason other than non-existence or emptiness, which it checks, it
5900 could not run './exim', it was installing it with an empty version number,
5901 i.e. as "exim-". This error state is now caught, and the installation is
5904 38. An argument was missing from the function that creates an error message
5905 when Exim fails to connect to the socket for saslauthd authentication.
5906 This could cause Exim to crash, or give a corrupted message.
5908 39. Added isip, isip4, and isip6 to ${if conditions.
5910 40. The ACL variables $acl_xx are now saved with the message, and can be
5911 accessed later in routers, transports, and filters.
5913 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5914 strings in the file are not string-expanded.
5916 42. If a MAIL command specified a SIZE value that was too large to fit into an
5917 int variable, the check against message_size_limit failed. Such values are
5918 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5919 day this will have to be increased, but I don't think I want to be around
5920 when emails are that large.
5927 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5928 iconv() is not standard in FreeBSD.
5930 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5931 IPv6 enabled. The observed symptom was a segmentation fault on return from
5932 the function os_common_find_running_interfaces() in src/os.c.
5934 3. In the check_special_case() function in daemon.c I had used "errno" as an
5935 argument name, which causes warnings on some systems. This was basically a
5936 typo, since it was named "eno" in the comments!
5938 4. The code that waits for the clock to tick (at a resolution of some fraction
5939 of a second) so as to ensure message-id uniqueness was always waiting for
5940 at least one whole tick, when it could have waited for less. [This is
5941 almost certainly not relevant at current processor speeds, where it is
5942 unlikely to ever wait at all. But we try to future-proof.]
5944 5. The function that sleeps for a time interval that includes fractions of a
5945 second contained a race. It did not block SIGALRM between setting the
5946 timer, and suspending (a couple of lines later). If the interval was short
5947 and the sigsuspend() was delayed until after it had expired, the suspension
5948 never ended. On busy systems this could lead to processes getting stuck for
5951 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5952 process, before it forks any delivery processes. The open lookup caching
5953 mechanism meant that the open file or database connection was passed into
5954 the delivery process. The problem was that delivery processes always tidy
5955 up cached lookup data. This could cause a problem for the next delivery
5956 process started by the queue runner, because the external queue runner
5957 process does not know about the closure. So the next delivery process
5958 still has data in the lookup cache. In the case of a file lookup, there was
5959 no problem because closing a file descriptor in a subprocess doesn't affect
5960 the parent. However, if the lookup was caching a connection to a database,
5961 the connection was closed, and the second delivery process was likely to
5962 see errors such as "PGSQL: query failed: server closed the connection
5963 unexpectedly". The problem has been fixed by closing all cached lookups
5964 in a queue runner before running a delivery process.
5966 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5967 seem to have the "const" qualifier which it has on other OS. I've
5970 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5971 *of the same type* (client or server) with the same public name that an
5972 error should be diagnosed.
5974 9. When Exim looked up a host name for an IP address, but failed to find the
5975 original IP address when looking up the host name (a safety check), it
5976 output the message "<ip address> does not match any IP for NULL", which was
5977 confusing, to say the least. The bug was that the host name should have
5978 appeared instead of "NULL".
5980 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5981 user that is built into the binary, and the -C or -D options is used, root
5982 privilege is dropped before the configuration file is read. In addition,
5983 logging is switched to stderr instead of the normal log files. If the
5984 configuration then re-defines the Exim user, the unprivileged environment
5985 is probably not what is expected, so Exim logs a panic warning message (but
5988 However, if deliver_drop_privilege is set, the unprivileged state may well
5989 be exactly what is intended, so the warning has been cut out in that case,
5990 and Exim is allowed to try to write to its normal log files.
5996 1. smtp_return_error_details was not giving details for temporary sender
5997 or receiver verification errors.
5999 2. Diagnose a configuration error if two authenticators have the same public
6002 3. Exim used not to create the message log file for a message until the first
6003 delivery attempt. This could be confusing when incoming messages were held
6004 for policy or load reasons. The message log file is now created at the time
6005 the message is received, and an initial "Received" line is written to it.
6007 4. The automatically generated man page for command line options had a minor
6008 bug that caused no ill effects; however, a more serious problem was that
6009 the procedure for building the man page automatically didn't always
6010 operate. Consequently, release 4.20 contains an out-of-date version. This
6011 shouldn't happen again.
6013 5. When building Exim with embedded Perl support, the script that builds the
6014 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6015 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6017 6. The freeze_tell option was not being used for messages that were frozen on
6018 arrival, either by an ACL or by local_scan().
6020 7. Added the smtp_incomplete_transaction log selector.
6022 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6023 was accepting AUTH without a new EHLO.
6025 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6026 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6029 10. Logging of TCP/IP connections (when configured) now happens in the main
6030 daemon process instead of the child process, so that the TCP/IP connection
6031 count is more accurate (but it can never be perfect).
6033 11. The use of "drop" in a nested ACL was not being handled correctly in the
6034 outer ACL. Now, if condition failure induced by the nested "drop" causes
6035 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6036 or "require"), the connection is dropped.
6038 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6039 that yield "discard" can now be used with an "accept" or a "discard" verb,
6040 but an error is generated for any others (because I can't see a useful way
6041 to define what should happen).
6043 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6044 are now processed in the same way as lines in the Exim configuration file.
6045 In particular, continuation lines are supported.
6047 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6049 15. Added -ti meaning -t -i.
6051 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6052 domains, and warn by logging if others are found.
6054 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6055 returned from GIFCONF when Exim is trying to find the list of interfaces on
6056 a host. The code in os.c has been modified to copy each ifreq to an aligned
6057 structure in all cases.
6059 Also, in some cases, the returned ifreq's were being copied to a 'struct
6060 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6061 means the last couple of bytes of an IPv6 address could be chopped if the
6062 ifreq contained only a normal sockaddr (14 bytes storage).
6064 18. Named domain lists were not supported in the hosts_treat_as_local option.
6065 An entry such as +xxxx was not recognized, and was treated as a literal
6068 19. Ensure that header lines added by a DATA ACL are included in the reject log
6069 if the ACL subsequently rejects the message.
6071 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6072 MD5 (which is deprecated).
6074 21. When testing a filter file using -bf, Exim was writing a message when it
6075 took the sender from a "From " line in the message, but it was not doing so
6076 when it took $return_path from a Return-Path: header line. It now does.
6078 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6079 with a valid header line field name (a series of printing characters
6080 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6082 23. Changed "disc" in the source to "disk" to conform to the documentation and
6083 the book and for uniformity.
6085 24. Ignore Sendmail's -Ooption=value command line item.
6087 25. When execve() failed while trying to run a command in a pipe transport,
6088 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6089 could be confused with a return value of 69 from the command itself. This
6090 has been changed to 127, the value the shell returns if it is asked to run
6091 a non-existent command. The wording for the related log line suggests a
6092 non-existent command as the problem.
6094 26. If received_header_text expands to an empty string, do not add a Received:
6095 header line to the message. (Well, it adds a token one on the spool, but
6096 marks it "old" so that it doesn't get used or transmitted.)
6098 27. Installed eximstats 1.28 (addition of -nt option).
6100 28. There was no check for failure on the call to getsockname() in the daemon
6101 code. This can fail if there is a shortage of resources on the system, with
6102 ENOMEM, for example. A temporary error is now given on failure.
6104 29. Contrary to the C standard, it seems that in some environments, the
6105 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6106 program. Exim now does this explicitly; it affects the formatting of
6107 timestamps using strftime().
6109 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6110 complaints. I've now initialized all the variables, to avoid this.
6112 32. Header lines added by a system filter were not being "seen" during
6113 transport-time rewrites.
6115 33. The info_callback() function passed to OpenSSL is set up with type void
6116 (*)(SSL *, int, int), as described somewhere. However, when calling the
6117 function (actually a macro) that sets it up, the type void(*)() is
6118 expected. I've put in a cast to prevent warnings from picky compilers.
6120 34. If a DNS black list lookup found a CNAME record, but there were no A
6121 records associated with the domain it pointed at, Exim crashed.
6123 35. If a DNS black list lookup returned more than one A record, Exim ignored
6124 all but the first. It now scans all returned addresses if a particular IP
6125 value is being sought. In this situation, the contents of the
6126 $dnslist_value variable are a list of all the addresses, separated by a
6129 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6130 to accept a host name and all its aliases if the forward lookup for any of
6131 them yielded the IP address of the incoming connection. Now it accepts only
6132 those names whose forward lookup yields the correct IP address. Any other
6133 names are discarded. This closes a loophole whereby a rogue DNS
6134 administrator could create reverse DNS records to break through a
6135 wildcarded host restriction in an ACL.
6137 37. If a user filter or a system filter that ran in a subprocess used any of
6138 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6139 the wrong values were passed to the pipe command ($thisaddress had the
6140 value of $0, $0 had the value of $1, etc). This bug was introduced by
6141 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6143 38. Improved the line breaking for long SMTP error messages from ACLs.
6144 Previously, if there was no break point between 40 and 75 characters, Exim
6145 left the rest of the message alone. Two changes have been made: (a) I've
6146 reduced the minimum length to 35 characters; (b) if it can't find a break
6147 point between 35 and 75 characters, it looks ahead and uses the first one
6148 that it finds. This may give the occasional overlong line, but at least the
6149 remaining text gets split now.
6151 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6152 file descriptors might be low, and that setting 1000 would always raise it.
6153 It turns out that in some environments, the limit is already over 1000 and
6154 that lowering it causes trouble. So now Exim takes care not to decrease it.
6156 40. When delivering a message, the value of $return_path is set to $sender_
6157 address at the start of routing (routers may change the value). By an
6158 oversight, this default was not being set up when an address was tested by
6159 -bt or -bv, which affected the outcome if any router or filter referred to
6162 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6163 log only when "message" or "log_message" are set. However, if one of the
6164 conditions was an address verification, or a call to a nested ACL, the
6165 messages generated by the underlying test were being passed through. This
6166 no longer happens. The underlying message is available in $acl_verify_
6167 message for both "message" and "log_message" expansions, so it can be
6168 passed through if needed.
6170 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6171 new expansion $bh_ to give the encoded byte string without charset
6172 translation. Translation happens only if iconv() is available; HAVE_ICONV
6173 indicates this at build time. HEADERS_CHARSET gives the charset to
6174 translate to; headers_charset can change it in the configuration, and
6175 "headers charset" can change it in an individual filter file.
6177 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6178 that creates RFC 2047 encoded "words" labels them as that charset instead
6179 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6180 expansion operator; (ii) when Exim creates a From: line for a local
6181 message; (iii) when a header line is rewritten to include a "phrase" part.
6183 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6184 buggy, causing it to skip the first lines of messages whose message ID
6185 ended in 'D'. This would not have bitten before Exim release 4.14, because
6186 message IDs were unlikely to end in 'D' before then. The effect was to have
6187 incorrect size information for certain domains.
6189 45. #include "config.h" was missing at the start of the crypt16.c module. This
6190 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6193 46. If there was a timeout during a "random" callout check, Exim treated it as
6194 a failure of the random address, and carried on sending RSET and the real
6195 address. If the delay was just some slowness somewhere, the response to the
6196 original RCPT would be taken as a response to RSET and so on, causing
6197 mayhem of various kinds.
6199 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6200 when I implemented it. It didn't allow for the fact that some option values
6201 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6202 the right test for positive values.
6204 48. Domain names in DNS records are case-independent. Exim always looks them up
6205 in lower case. Some resolvers return domain names in exactly the case they
6206 appear in the zone file, that is, they may contain uppercase letters. Not
6207 all resolvers do this - some return always lower case. Exim was treating a
6208 change of case by a resolver as a change of domain, similar to a widening
6209 of a domain abbreviation. This triggered its re-routing code and so it was
6210 trying to route what was effectively the same domain again. This normally
6211 caused routing to fail (because the router wouldn't handle the domain
6212 twice). Now Exim checks for this case specially, and just changes the
6213 casing of the domain that it ultimately uses when it transmits the message
6216 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6219 50. If a filter generated a file delivery with a non-absolute name (possible if
6220 no home directory exists for the router), the forbid_file option was not
6223 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6224 the existing equality matching.
6226 52. Exim was using ints instead of ino_t variables in some places where it was
6227 dealing with inode numbers.
6229 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6230 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6231 called TMPDIR, and if it finds it is different, it changes its value.
6233 54. The smtp_printf() function is now made available to local_scan() so
6234 additional output lines can be written before returning. There is also an
6235 smtp_fflush() function to enable the detection of a dropped connection.
6236 The variables smtp_input and smtp_batched_input are exported to
6239 55. Changed the default runtime configuration: the message "Unknown user"
6240 has been removed from the ACL, and instead placed on the localuser router,
6241 using the cannot_route_message feature. This means that any verification
6242 failures that generate their own messages won't get overridden. Similarly,
6243 the "Unrouteable address" message that was in the ACL for unverifiable
6244 relay addresses has also been removed.
6246 56. Added hosts_avoid_esmtp to the smtp transport.
6248 57. The exicyclog script was not checking for the esoteric option
6249 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6250 will work only if exicyclog is run under the appropriate euid.
6252 58. Following a discussion on the list, the rules by which Exim recognises line
6253 endings on incoming messages have been changed. The -dropcr and drop_cr
6254 options are now no-ops, retained only for backwards compatibility. The
6255 following line terminators are recognized: LF CRLF CR. However, special
6256 processing applies to CR:
6258 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6259 nor a local message in the state where . is a terminator.
6261 (ii) If a bare CR is encountered in a header line, an extra space is added
6262 after the line terminator so as not to end the header. The reasoning
6263 behind this is that bare CRs in header lines are most likely either
6264 to be mistakes, or people trying to play silly games.
6266 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6267 was being incorrectly given as 18 bytes larger than it should have been.
6268 This is a VOB (very old bug).
6270 60. This may never have affected anything current, but just in case it has:
6271 When the local host is found other than at the start of a list of hosts,
6272 the local host, those with the same MX, and any that follow, are discarded.
6273 When the list in question was part of a longer list of hosts, the following
6274 hosts (not currently being processed) were also being discarded. This no
6275 longer happens. I'm not sure if this situation could ever has previously
6278 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6281 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6282 header. This is something that is recommended in a new Internet Draft, and
6283 is something that is documented as being done by Sendmail. There are two
6284 possible values. For messages generated by the autoreply transport, Exim
6287 Auto-Submitted: auto-replied
6289 whereas for all other generated messages (e.g. bounces) it adds
6291 Auto-Submitted: auto-generated
6293 63. The "personal" condition in filters now includes a test for the
6294 Auto-Submitted: header. If it contains the string "auto-" the message it
6295 not considered personal.
6297 64. Added rcpt_include_affixes as a generic transport option.
6299 65. Added queue_only_override (default true).
6301 66. Added the syslog_duplication option.
6303 67. If what should have been the first header line of a message consisted of
6304 a space followed by a colon, Exim was mis-interpreting it as a header line.
6305 It isn't of course - it is syntactically invalid and should therefore be
6306 treated as the start of the message body. The misbehaviour could have
6307 caused a number of strange effects, including loss of data in subsequent
6308 header lines, and spool format errors.
6310 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6311 client host had authenticated. This control can now be exercised by an ACL
6312 for more flexibility.
6314 69. By default, callouts do not happen when testing with -bh. There is now a
6315 variant, -bhc, which does actually run the callout code, including
6316 consulting and updating the callout cache.
6318 70. Added support for saslauthd authentication, courtesy of Alexander
6321 71. If statvfs() failed on the spool or log directories while checking their
6322 size for availability, Exim confusingly gave the error "space shortage".
6323 Furthermore, in debugging mode it crashed with a floating point exception.
6324 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6325 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6326 serious problem, Exim now writes to the main and panic logs when this
6327 happens, with details of the failure. It then refuses to accept the
6328 incoming message, giving the message "spool directory problem" or "log
6329 directory problem" with a 421 code for SMTP messages.
6331 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6332 0, 1, and 2 exist, because some OS complain for execs without them (see
6333 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6334 descriptors. However, the code omitted to check that the open succeeded,
6335 causing mysterious errors if for some reason the permissions on /dev/null
6336 got screwed. Now Exim writes a message to the main and panic logs, and
6337 bombs out if it can't open /dev/null.
6339 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6340 interact so that it is all more flexible. It is supposed to remain
6341 backwards compatible. Also added extra_local_interfaces.
6343 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6344 to bomb out with an assertion failure - to the client this appears as a
6345 connection drop. This problem occurs in the part of the code that was taken
6346 from the Samba project. Fortunately, the assertion is in a very simple
6347 function, so I have fixed this by reproducing the function inline in the
6348 one place where it is called, and arranging for authentication to fail
6349 instead of killing the process with assert().
6351 75. The SPA client code was not working when the server requested OEM rather
6352 than Unicode encoding.
6354 76. Added code to make require_files with a specific uid setting more usable in
6355 the case where statting the file as root fails - usually a non-root-mounted
6356 NFS file system. When this happens and the failure is EACCES, Exim now
6357 forks a subprocess and does the per-uid checking as the relevant uid.
6359 77. Added process_log_path.
6361 78. If log_file_path was not explicitly set, a setting of check_log_space or
6362 check_log_inodes was ignored.
6364 79. If a space check for the spool or log partitions fails, the incident is now
6365 logged. Of course, in the latter case the data may get lost...
6367 80. Added the %p formatting code to string_format() so that it can be used to
6368 print addresses in debug_print(). Adjusted all the address printing in the
6369 debugging in store.c to use %p rather than %d.
6371 81. There was a concern that a line of code in smtp_in.c could overflow a
6372 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6373 initially expressed, the concern was not well-founded, because trailing
6374 spaces are removed early. However, if the trailing spaces were followed by
6375 a NULL, they did not get removed, so the overflow was possible. Two fixes
6378 (a) I re-wrote the offending code in a cleaner fashion.
6379 (b) If an incoming SMTP command contains a NULL character, it is rejected
6382 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6383 runs initgroups(), so that if the Exim user is in any additional groups,
6384 they will be used during message reception.
6390 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob