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
90 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
91 with the $config_file variable.
93 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
94 in transport context, after the attempt, and per-recipient. The latter type
95 is per host attempted. The event data is the error message, and the errno
96 information encodes the lookup type (A vs. MX) used for the (first) host,
97 and the trailing two digits of the smtp 4xx reponse.
99 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
100 to write to mainlog (or rejectlog, paniclog) in the window between file
101 creation and permissions/ownership being changed. Particularly affects
102 installations where exicyclog is run as root, rather than exim user;
103 result is that the running daemon panics and dies.
105 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
107 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
108 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
109 "pri" and "weight". Note that the previous implicit priority given by the
110 list order is no longer honoured.
112 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalisation
115 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
116 by defining SUPPORT_SOCKS.
118 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
119 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
120 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
121 variables are renamed to proxy_{local,external}_{address,port}.
123 JH/27 Move Internationalisation support from Experimental to mainline, enabled
124 for a build by defining SUPPORT_I18N
126 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
127 of the query string, and make ${quote_redis:} do that quoting.
129 JH/29 Move Events support from Experimental to mainline, enabled by default
130 and removable for a build by defining DISABLE_EVENT.
132 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
134 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
135 cached by the daemon.
137 JH/32 Move Redis support from Experimental to mainline, enabled for a build
138 by defining LOOKUP_REDIS. The libhiredis library is required.
140 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
141 keys are given for lookup.
143 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
144 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
145 only supported when built with TLS support.
147 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
148 openssl_options), for security. OpenSSL forces this from version 1.1.0
149 server-side so match that on older versions.
151 JH/36 Fix a longstanding bug in memory use by the ${run } expansion: A fresh
152 allocation for $value could be released as the expansion processing
153 concluded, but leaving the global pointer active for it. Possibly
154 involved in Bug 1778.
159 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
162 JH/02 The smtp transport option "multi_domain" is now expanded.
164 JH/03 The smtp transport now requests PRDR by default, if the server offers
167 JH/04 Certificate name checking on server certificates, when exim is a client,
168 is now done by default. The transport option tls_verify_cert_hostnames
169 can be used to disable this per-host. The build option
170 EXPERIMENTAL_CERTNAMES is withdrawn.
172 JH/05 The value of the tls_verify_certificates smtp transport and main options
173 default to the word "system" to access the system default CA bundle.
174 For GnuTLS, only version 3.0.20 or later.
176 JH/06 Verification of the server certificate for a TLS connection is now tried
177 (but not required) by default. The verification status is now logged by
178 default, for both outbound TLS and client-certificate supplying inbound
181 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
184 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
185 Status Notification (bounce) messages are now MIME format per RFC 3464.
186 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
187 under the control of the dsn_advertise_hosts option, and routers may
188 have a dsn_lasthop option.
190 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
191 default, modifiable by a malware= option. The list separator for
192 the options can now be changed in the usual way. Bug 68.
194 JH/10 The smtp_receive_timeout main option is now expanded before use.
196 JH/11 The incoming_interface log option now also enables logging of the
197 local interface on delivery outgoing connections.
199 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
200 if the interface and destination host and port all match.
202 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
205 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
206 Patch from Andrew Lewis.
208 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
209 now supports optional time-restrictions, weighting, and priority
210 modifiers per server. Patch originally by <rommer@active.by>.
212 JH/16 The spamd_address main option now supports a mixed list of local
213 and remote servers. Remote servers can be IPv6 addresses, and
214 specify a port-range.
216 JH/17 Bug 68: The spamd_address main option now supports an optional
217 timeout value per server.
219 JH/18 Bug 1581: Router and transport options headers_add/remove can
220 now have the list separator specified.
222 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
225 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
228 JH/21 Support for the A6 type of dns record is withdrawn.
230 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
231 rather than the verbs used.
233 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
234 from 255 to 1024 chars.
236 JH/24 Verification callouts now attempt to use TLS by default.
238 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
239 are generic router options now. The defaults didn't change.
241 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
242 Original patch from Alexander Shikoff, worked over by JH.
244 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
245 files if ZCAT_COMMAND is not executable.
247 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
249 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
251 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
252 Normally benign, it bites when the pair was led to by a CNAME;
253 modern usage is to not canoicalize the domain to a CNAME target
254 (and we were inconsistent anyway for A-only vs AAAA+A).
256 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
258 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
259 when evaluating $sender_host_dnssec.
261 JH/31 Check the HELO verification lookup for DNSSEC, adding new
262 $sender_helo_dnssec variable.
264 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
266 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
268 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
270 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
271 documented as working, but never had. Support all but $spam_report.
273 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
274 added for tls authenticator.
279 TL/01 When running the test suite, the README says that variables such as
280 no_msglog_check are global and can be placed anywhere in a specific
281 test's script, however it was observed that placement needed to be near
282 the beginning for it to behave that way. Changed the runtest perl
283 script to read through the entire script once to detect and set these
284 variables, reset to the beginning of the script, and then run through
285 the script parsing/test process like normal.
287 TL/02 The BSD's have an arc4random API. One of the functions to induce
288 adding randomness was arc4random_stir(), but it has been removed in
289 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
290 function when detected.
292 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
293 cause callback expansion.
295 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
296 syntax errors in an expansion can be treated as a string instead of
297 logging or causing an error, due to the internal use of bool_lax
298 instead of bool when processing it.
300 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
301 server certificates when making smtp deliveries.
303 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
305 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
307 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
309 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
310 Merged patch from Sebastian Wiedenroth.
312 JH/05 Fix results-pipe from transport process. Several recipients, combined
313 with certificate use, exposed issues where response data items split
314 over buffer boundaries were not parsed properly. This eventually
315 resulted in duplicates being sent. This issue only became common enough
316 to notice due to the introduction of conection certificate information,
317 the item size being so much larger. Found and fixed by Wolfgang Breyha.
319 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
320 size buffer was used, resulting in syntax errors when an expansion
323 JH/07 Add support for directories of certificates when compiled with a GnuTLS
324 version 3.3.6 or later.
326 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
327 is EXPERIMENTAL_EVENT, the main-configuration and transport options
328 both become "event_action", the variables become $event_name, $event_data
329 and $event_defer_errno. There is a new variable $verify_mode, usable in
330 routers, transports and related events. The tls:cert event is now also
331 raised for inbound connections, if the main configuration event_action
334 TL/06 In test suite, disable OCSP for old versions of openssl which contained
335 early OCSP support, but no stapling (appears to be less than 1.0.0).
337 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
338 server certificate names available under the smtp transport option
339 "tls_verify_cert_hostname" now do not permit multi-component wildcard
342 JH/10 Time-related extraction expansions from certificates now use the main
343 option "timezone" setting for output formatting, and are consistent
344 between OpenSSL and GnuTLS compilations. Bug 1541.
346 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
347 encoded parameter in the incoming message. Bug 1558.
349 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
350 include certificate info, eximon was claiming there were spoolfile
353 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
355 JH/14 Log delivery-related information more consistently, using the sequence
356 "H=<name> [<ip>]" wherever possible.
358 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
359 are problematic for Debian distribution, omit them from the release
362 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
364 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
367 JH/17 Fix a null-indirection in certextract expansions when a nondefault
368 output list separator was used.
373 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
374 checkers that were complaining about end of non-void function with no
377 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
378 This was a regression intruduced in 4.83 by another bugfix.
380 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
382 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
383 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
389 TF/01 Correctly close the server side of TLS when forking for delivery.
391 When a message was received over SMTP with TLS, Exim failed to clear up
392 the incoming connection properly after forking off the child process to
393 deliver the message. In some situations the subsequent outgoing
394 delivery connection happened to have the same fd number as the incoming
395 connection previously had. Exim would try to use TLS and fail, logging
396 a "Bad file descriptor" error.
398 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
399 utilities have not been installed.
401 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
402 temporary space as the ACL may create new global variables.
404 TL/01 LDAP support uses per connection or global context settings, depending
405 upon the detected version of the libraries at build time.
407 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
408 to extract and use the src ip:port in logging and expansions as if it
409 were a direct connection from the outside internet. PPv2 support was
410 updated based on HAProxy spec change in May 2014.
412 JH/02 Add ${listextract {number}{list}{success}{fail}}.
414 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
415 Properly escape header and check for NULL return.
417 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
420 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
422 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
423 characters in header names, implemented as a verify condition.
424 Contributed by Michael Fischer v. Mollard.
426 TL/05 Rename SPF condition results err_perm and err_temp to standardized
427 results permerror and temperror. Previous values are deprecated but
428 still accepted. In a future release, err_perm and err_temp will be
429 completely removed, which will be a backward incompatibility if the
430 ACL tests for either of these two old results. Patch contributed by
431 user bes-internal on the mailing list.
433 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
435 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
436 selectors, in both main and reject logs.
438 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
441 JH/07 Add malware type "sock" for talking to simple daemon.
443 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
445 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
446 routers/transports under cutthrough routing.
448 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
449 numbers. Touch up "bool" conditional to keep the same definition.
451 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
453 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
455 JH/12 Expand items in router/transport headers_add or headers_remove lists
456 individually rather than the list as a whole. Bug 1452.
458 Required for reasonable handling of multiple headers_ options when
459 they may be empty; requires that headers_remove items with embedded
460 colons must have them doubled (or the list-separator changed).
462 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
463 view the policy declared in the DMARC record. Currently, $dmarc_status
464 is a combined value of both the record presence and the result of the
467 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
469 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
470 dnslookup router and the smtp transport (applying to the forward
473 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
474 of ldap servers used for a specific lookup. Patch provided by Heiko
477 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
478 New variable $lookup_dnssec_authenticated for observability.
480 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
481 Patch submitted by Lars Timman.
483 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
485 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
486 Requires trusted mode and valid format message id, aborts otherwise.
487 Patch contributed by Heiko Schlichting.
489 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
490 certextract with support for various fields. Bug 1358.
492 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
493 is requested by default, modifiable by smtp transport option
496 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
497 operate on certificate variables to give certificate fingerprints
498 Also new ${sha256:cert_variable}.
500 JH/23 The PRDR feature is moved from being Experimental into the mainline.
502 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
503 Christian Aistleitner.
505 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
507 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
508 file. Patch from Wolfgang Breyha.
510 JH/25 Expand the coverage of the delivery $host and $host_address to
511 client authenticators run in verify callout. Bug 1476.
513 JH/26 Port service names are now accepted for tls_on_connect_ports, to
514 align with daemon_smtp_ports. Bug 72.
516 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
517 support and error reporting did not work properly.
519 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
520 and is readable. Patch from Andrew Colin Kissa.
522 TL/14 Enhance documentation of ${run expansion and how it parses the
523 commandline after expansion, particularly in the case when an
524 unquoted variable expansion results in an empty value.
526 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
528 PP/02 Fix internal collision of T_APL on systems which support RFC3123
529 by renaming away from it. Addresses GH issue 15, reported by
532 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
534 TL/15 SECURITY: prevent double expansion in math comparison functions
535 (can expand unsanitized data). Not remotely exploitable.
542 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
544 PP/02 Make -n do something, by making it not do something.
545 When combined with -bP, the name of an option is not output.
547 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
550 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
551 $sender_host_name and config options to manage this, and basic check
554 PP/05 DSCP support for outbound connections and control modifier for inbound.
556 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
557 (Only plugin which currently uses this is kerberos4, which nobody should
558 be using, but we should make it available and other future plugins might
559 conceivably use it, even though it would break NAT; stuff *should* be
560 using channel bindings instead).
562 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
563 name; added for Sendmail compatibility; requires admin caller.
564 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
565 just ignore it); requires trusted caller.
566 Also parse but ignore: -Ac -Am -X<logfile>
569 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
571 TL/02 Add +smtp_confirmation as a default logging option.
573 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
574 Patch by Magnus Holmgren from 2007-02-20.
576 TL/04 Bugzilla 1281 - Spec typo.
577 Bugzilla 1283 - Spec typo.
578 Bugzilla 1290 - Spec grammar fixes.
580 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
582 TL/06 Add Experimental DMARC support using libopendmarc libraries.
584 TL/07 Fix an out of order global option causing a segfault. Reported to dev
585 mailing list by by Dmitry Isaikin.
587 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
589 JH/02 Support "G" suffix to numbers in ${if comparisons.
591 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
593 NM/01 Bugzilla 1197 - Spec typo
594 Bugzilla 1196 - Spec examples corrections
596 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
598 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
599 gnutls_enable_pkcs11, but renamed to more accurately indicate its
602 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
603 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
605 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
606 "acl {{name}{arg}...}", and optional args on acl condition
609 JH/05 Permit multiple router/transport headers_add/remove lines.
611 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
613 JH/07 Avoid using a waiting database for a single-message-only transport.
614 Performance patch from Paul Fisher. Bugzilla 1262.
616 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
619 JH/09 Add $headers_added variable, with content from use of ACL modifier
620 add_header (but not yet added to the message). Bugzilla 199.
622 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
623 Pulled from Bugzilla 817 by Wolfgang Breyha.
625 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
627 (nb: this is the same fix as in Exim 4.80.1)
629 JH/11 Add A= logging on delivery lines, and a client_set_id option on
632 JH/12 Add optional authenticated_sender logging to A= and a log_selector
635 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
637 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
638 advertise SMTP AUTH mechanism to us, instead of a generic
639 protocol violation error. Also, make Exim more robust to bad
640 data from the Dovecot auth socket.
642 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
644 When a queue runner is handling a message, Exim first routes the
645 recipient addresses, during which it prunes them based on the retry
646 hints database. After that it attempts to deliver the message to
647 any remaining recipients. It then updates the hints database using
650 So if a recipient address works intermittently, it can get repeatedly
651 deferred at routing time. The retry hints record remains fresh so the
652 address never reaches the final cutoff time.
654 This is a fairly common occurrence when a user is bumping up against
655 their storage quota. Exim had some logic in its local delivery code
656 to deal with this. However it did not apply to per-recipient defers
657 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
659 This change adds a proper retry rule check during routing so that the
660 final cutoff time is checked against the message's age. We only do
661 this check if there is an address retry record and there is not a
662 domain retry record; this implies that previous attempts to handle
663 the address had the retry_use_local_parts option turned on. We use
664 this as an approximation for the destination being like a local
665 delivery, as in LMTP.
667 I suspect this new check makes the old local delivery cutoff check
668 redundant, but I have not verified this so I left the code in place.
670 TF/02 Correct gecos expansion when From: is a prefix of the username.
672 Test 0254 submits a message to Exim with the header
676 When I ran the test suite under the user fanf2, Exim expanded
677 the header to contain my full name, whereas it should have added
678 a Resent-Sender: header. It erroneously treats any prefix of the
679 username as equal to the username.
681 This change corrects that bug.
683 GF/01 DCC debug and logging tidyup
684 Error conditions log to paniclog rather than rejectlog.
685 Debug lines prefixed by "DCC: " to remove any ambiguity.
687 TF/03 Avoid unnecessary rebuilds of lookup-related code.
689 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
690 Bug spotted by Jeremy Harris; was flawed since initial commit.
691 Would have resulted in OCSP responses post-SNI triggering an Exim
692 NULL dereference and crash.
694 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
696 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
697 Bug detection, analysis and fix by Samuel Thibault.
698 Bugzilla 1331, Debian bug #698092.
700 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
702 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
703 Server implementation by Todd Lyons, client by JH.
704 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
705 config variable "prdr_enable" controls whether the server
706 advertises the facility. If the client requests PRDR a new
707 acl_data_smtp_prdr ACL is called once for each recipient, after
708 the body content is received and before the acl_smtp_data ACL.
709 The client is controlled by bolth of: a hosts_try_prdr option
710 on the smtp transport, and the server advertisement.
711 Default client logging of deliveries and rejections involving
712 PRDR are flagged with the string "PRDR".
714 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
715 fclose(). Diagnosis by Todd Lyons.
717 PP/17 Update configure.default to handle IPv6 localhost better.
718 Patch by Alain Williams (plus minor tweaks).
721 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
722 This is now consistent with GnuTLS, and is now documented: the
723 previous undocumented portable approach to treating the option as
724 unset was to force an expansion failure. That still works, and
725 an empty string is now equivalent.
727 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
728 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
729 not performing validation itself.
731 PP/20 Added force_command boolean option to pipe transport.
732 Patch from Nick Koston, of cPanel Inc.
734 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
737 TF/04 Added udpsend ACL modifer and hexquote expansion operator
739 PP/21 Fix eximon continuous updating with timestamped log-files.
740 Broken in a format-string cleanup in 4.80, missed when I repaired the
741 other false fix of the same issue.
742 Report and fix from Heiko Schlichting.
745 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
746 Report from Prashanth Katuri.
748 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
749 It's SecureTransport, so affects any MacOS clients which use the
750 system-integrated TLS libraries, including email clients.
752 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
753 using a MIME ACL for non-SMTP local injection.
754 Report and assistance in diagnosis by Warren Baker.
756 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
758 JH/16 Fix comparisons for 64b. Bugzilla 1385.
760 TL/09 Add expansion variable $authenticated_fail_id to keep track of
761 last id that failed so it may be referenced in subsequent ACL's.
763 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
766 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
767 ldap library initialization, allowing self-signed CA's to be
768 used. Also properly sets require_cert option later in code by
769 using NULL (global ldap config) instead of ldap handle (per
770 session). Bug diagnosis and testing by alxgomz.
772 TL/12 Enhanced documentation in the ratelimit.pl script provided in
773 the src/util/ subdirectory.
775 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
776 renamed to Transport Post Delivery Action by Jeremy Harris, as
779 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
780 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
781 redis_servers = needs to be configured which will be used by the redis
782 lookup. Patch from Warren Baker, of The Packet Hub.
784 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
786 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
787 hostname or reverse DNS when processing a host list. Used suggestions
788 from multiple comments on this bug.
790 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
792 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
793 Missed a few lines, added it to make the runtest require no keyboard
796 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
797 contains upper case chars. Make router use caseful_local_part.
799 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
800 support when GnuTLS has been built with p11-kit.
806 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
808 This, or similar/improved, will also be change PP/11 of 4.82.
814 PP/01 Handle short writes when writing local log-files.
815 In practice, only affects FreeBSD (8 onwards).
816 Bugzilla 1053, with thanks to Dmitry Isaikin.
818 NM/01 Bugzilla 949 - Documentation tweak
820 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
823 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
825 PP/02 Implemented gsasl authenticator.
827 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
829 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
830 `pkg-config foo` for cflags/libs.
832 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
833 with rest of GSASL and with heimdal_gssapi.
835 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
836 `pkg-config foo` for cflags/libs for the TLS implementation.
838 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
839 properties get this fed in as external SSF. A number of robustness
840 and debugging improvements to the cyrus_sasl authenticator.
842 PP/08 cyrus_sasl server now expands the server_realm option.
844 PP/09 Bugzilla 1214 - Log authentication information in reject log.
845 Patch by Jeremy Harris.
847 PP/10 Added dbmjz lookup type.
849 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
851 PP/12 MAIL args handles TAB as well as SP, for better interop with
852 non-compliant senders.
853 Analysis and variant patch by Todd Lyons.
855 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
856 Bug report from Lars Müller <lars@samba.org> (via SUSE),
857 Patch from Dirk Mueller <dmueller@suse.com>
859 PP/13 tls_peerdn now print-escaped for spool files.
860 Observed some $tls_peerdn in wild which contained \n, which resulted
861 in spool file corruption.
863 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
864 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
865 or write after TLS renegotiation, which otherwise led to messages
868 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
869 as a tracking header (ie: a signed header comes before the signature).
870 Patch from Wolfgang Breyha.
872 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
873 comma-sep list; embedded commas doubled.
875 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
877 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
879 Report and patch from Dmitry Banschikov.
881 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
882 Removed SSL_clear() after SSL_new() which led to protocol negotiation
883 failures. We appear to now support TLS1.1+ with Exim.
885 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
886 lets Exim select keys and certificates based upon TLS SNI from client.
887 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
888 before an outbound SMTP session. New log_selector, +tls_sni.
890 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
891 NULL dereference. Report and patch from Alun Jones.
893 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
894 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
895 Not seeing resolver debug output on NetBSD, but suspect this is a
896 resolver implementation change.
898 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
899 Left warnings. Added "eximon gdb" invocation mode.
901 PP/21 Defaulting "accept_8bitmime" to true, not false.
903 PP/22 Added -bw for inetd wait mode support.
905 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
906 locate the relevant includes and libraries. Made this the default.
908 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
909 Bugzilla 1246, report and most of solution from Tomasz Kusy.
911 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
912 This may cause build issues on older platforms.
914 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
915 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
916 gnutls_require_mac & gnutls_require_protocols (no longer supported).
917 Added SNI support via GnuTLS too.
918 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
920 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
922 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
923 Applied second patch from Janne, implementing suggestion to default
924 multiple-strings-in-record handling to match SPF spec.
926 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
928 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
929 read-only, out of scope).
930 Patch from Wolfgang Breyha, report from Stuart Northfield.
932 PP/29 Fix three issues highlighted by clang analyser static analysis.
933 Only crash-plausible issue would require the Cambridge-specific
934 iplookup router and a misconfiguration.
935 Report from Marcin Mirosław.
937 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
939 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
940 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
941 As part of this, removing so much warning spew let me fix some minor
942 real issues in debug logging.
944 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
945 assignment on my part. Fixed.
947 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
948 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
949 Janne Snabb (who went above and beyond: thank you).
951 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
952 string otherwise requires a connection and a bunch more work and it's
953 relatively easy to get wrong. Should also expose TLS library linkage
956 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
957 64-bit ${eval} (JH/03).
959 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
960 GNU libc to support some of the 64-bit stuff, should not lead to
961 conflicts. Defined before os.h is pulled in, so if a given platform
962 needs to override this, it can.
964 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
965 protection layer was required, which is not implemented.
966 Bugzilla 1254, patch from Wolfgang Breyha.
968 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
969 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
970 tls_dhparam take prime identifiers. Also unbreak combination of
971 OpenSSL+DH_params+TLSSNI.
973 PP/39 Disable SSLv2 by default in OpenSSL support.
979 PP/01 Solaris build fix for Oracle's LDAP libraries.
980 Bugzilla 1109, patch from Stephen Usher.
982 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
984 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
987 TF/02 Fix a couple more cases where we did not log the error message
988 when unlink() failed. See also change 4.74-TF/03.
990 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
991 lock up or crash if it happened to be inside a call to libc when it
992 got a SIGUSR1 from exiwhat.
994 The SIGUSR1 handler appends the current process status to the process
995 log which is later printed by exiwhat. It used to use the general
996 purpose logging code to do this, but several functions it calls are
997 not safe for signals.
999 The new output code in the SIGUSR1 handler is specific to the process
1000 log, and simple enough that it's easy to inspect for signal safety.
1001 Removing some special cases also simplifies the general logging code.
1002 Removing the spurious timestamps from the process log simplifies
1005 TF/04 Improved ratelimit ACL condition.
1007 The /noupdate option has been deprecated in favour of /readonly which
1008 has clearer semantics. The /leaky, /strict, and /readonly update modes
1009 are mutually exclusive. The update mode is no longer included in the
1010 database key; it just determines when the database is updated. (This
1011 means that when you upgrde Exim will forget old rate measurements.)
1013 Exim now checks that the per_* options are used with an update mode that
1014 makes sense for the current ACL. For example, when Exim is processing a
1015 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1016 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1017 must specify per_mail/readonly. If you omit the update mode it defaults to
1018 /leaky where that makes sense (as before) or /readonly where required.
1020 The /noupdate option is now undocumented but still supported for
1021 backwards compatibility. It is equivalent to /readonly except that in
1022 ACLs where /readonly is required you may specify /leaky/noupdate or
1023 /strict/noupdate which are treated the same as /readonly.
1025 A useful new feature is the /count= option. This is a generalization
1026 of the per_byte option, so that you can measure the throughput of other
1027 aggregate values. For example, the per_byte option is now equivalent
1028 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1030 The per_rcpt option has been generalized using the /count= mechanism
1031 (though it's more complicated than the per_byte equivalence). When it is
1032 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1033 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1034 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1035 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1036 Note that using per_rcpt with a non-readonly update mode in more than
1037 one ACL will cause the recipients to be double-counted. (The per_mail
1038 and per_byte options don't have this problem.)
1040 The handling of very low rates has changed slightly. If the computed rate
1041 is less than the event's count (usually one) then this event is the first
1042 after a long gap. In this case the rate is set to the same as this event's
1043 count, so that the first message of a spam run is counted properly.
1045 The major new feature is a mechanism for counting the rate of unique
1046 events. The new per_addr option counts the number of different
1047 recipients that someone has sent messages to in the last time period. It
1048 behaves like per_rcpt if all the recipient addresses are different, but
1049 duplicate recipient addresses do not increase the measured rate. Like
1050 the /count= option this is a general mechanism, so the per_addr option
1051 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1052 example, measure the rate that a client uses different sender addresses
1053 with the options per_mail/unique=$sender_address. There are further
1054 details in the main documentation.
1056 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1058 TF/06 Removed a few PCRE remnants.
1060 TF/07 Automatically extract Exim's version number from tags in the git
1061 repository when doing development or release builds.
1063 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1064 Bugzilla 879. Patch from Paul Fisher.
1066 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1067 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1070 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1072 PP/05 Variable $av_failed, true if the AV scanner deferred.
1073 Bugzilla 1078. Patch from John Horne.
1075 PP/06 Stop make process more reliably on build failure.
1076 Bugzilla 1087. Patch from Heiko Schlittermann.
1078 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1079 Bugzilla 1089. Patch from Heiko Schlittermann.
1081 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1082 Bugzilla 1131. Patch from Holger Weiß.
1084 PP/09 Handle IPv6 addresses with SPF.
1085 Bugzilla 860. Patch from Wolfgang Breyha.
1087 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1089 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1092 PP/11 match_* no longer expand right-hand-side by default.
1093 New compile-time build option, EXPAND_LISTMATCH_RHS.
1094 New expansion conditions, "inlist", "inlisti".
1096 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1098 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1100 PP/14 fix log_write() format string regression from TF/03.
1101 Bugzilla 1152. Patch from Dmitry Isaikin.
1107 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1109 PP/02 Harmonised TLS library version reporting; only show if debugging.
1110 Layout now matches that introduced for other libraries in 4.74 PP/03.
1112 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1114 PP/04 New "dns_use_edns0" global option.
1116 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1119 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1120 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1122 TK/01 Updated PolarSSL code to 0.14.2.
1123 Bugzilla 1097. Patch from Andreas Metzler.
1125 PP/07 Catch divide-by-zero in ${eval:...}.
1126 Fixes bugzilla 1102.
1128 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1131 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1132 format-string attack -- SECURITY: remote arbitrary code execution.
1134 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1135 time unintentionally subject to list matching rules, letting the header
1136 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1137 arbitrary string expansion). This allowed for information disclosure.
1139 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1140 INT_MIN/-1 -- value coerced to INT_MAX.
1146 NM/01 Workround for PCRE version dependency in version reporting
1149 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1150 This fixes portability to compilers other than gcc, notably
1151 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1153 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1154 makefiles for portability to HP-UX and POSIX correctness.
1156 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1157 Also via indented variable definition in the Makefile.
1158 (Debugging by Oliver Heesakkers).
1160 PP/02 Restore caching of spamd results with expanded spamd_address.
1161 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1163 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1164 Improves build reliability. Fix from: Frank Elsner
1166 NM/02 Fix wide character breakage in the rfc2047 coding
1167 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1169 NM/03 Allow underscore in dnslist lookups
1170 Fixes bug 1026. Patch from Graeme Fowler
1172 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1173 Code patches from Adam Ciarcinski of NetBSD.
1175 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1178 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1179 is logged, to avoid truncation. Patch from John Horne.
1181 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1182 Patch from Jakob Hirsch.
1184 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1185 SQL string expansion failure details.
1186 Patch from Andrey Oktyabrski.
1188 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1189 Patch from Simon Arlott.
1191 PP/09 New lookups functionality failed to compile on old gcc which rejects
1192 extern declarations in function scope.
1193 Patch from Oliver Fleischmann
1195 PP/10 Use sig_atomic_t for flags set from signal handlers.
1196 Check getgroups() return and improve debugging.
1197 Fixed developed for diagnosis in bug 927 (which turned out to be
1200 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1201 Patch from Mark Zealey.
1203 PP/12 Bugzilla 1056: Improved spamd server selection.
1204 Patch from Mark Zealey.
1206 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1207 Based on patch from Heiko Schlittermann.
1209 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1210 Patch from Uwe Doering, sign-off by Michael Haardt.
1212 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1213 variable declaration deep within a block. Bug and patch from
1216 PP/15 lookups-Makefile IRIX compatibilty coercion.
1218 PP/16 Make DISABLE_DKIM build knob functional.
1220 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1221 Patch by Simon Arlott
1223 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1224 variable argument macros. Our copy now differs from upstream.
1230 TF/01 Failure to get a lock on a hints database can have serious
1231 consequences so log it to the panic log.
1233 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1234 controlled using the smtp_confirmation log selector.
1236 TF/03 Include the error message when we fail to unlink a spool file.
1238 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1239 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1240 for maintaining out-of-tree patches for some time.
1242 PP/01 Bugzilla 139: Documentation and portability issues.
1243 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1244 Handle per-OS dynamic-module compilation flags.
1246 PP/02 Let /dev/null have normal permissions.
1247 The 4.73 fixes were a little too stringent and complained about the
1248 permissions on /dev/null. Exempt it from some checks.
1249 Reported by Andreas M. Kirchwitz.
1251 PP/03 Report version information for many libraries, including
1252 Exim version information for dynamically loaded libraries. Created
1253 version.h, now support a version extension string for distributors
1254 who patch heavily. Dynamic module ABI change.
1256 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1257 privilege escalation vulnerability whereby the Exim run-time user
1258 can cause root to append content of the attacker's choosing to
1261 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1264 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1265 If dropping privileges for untrusted macros, we disabled normal logging
1266 on the basis that it would fail; for the Exim run-time user, this is not
1267 the case, and it resulted in successful deliveries going unlogged.
1268 Fixed. Reported by Andreas Metzler.
1274 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1275 only prepend for the Resent-* case. Fixes regression introduced in
1276 Exim 4.70 by NM/22 for Bugzilla 607.
1278 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1279 increasing numbers of administrators be bitten by this.
1281 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1283 PP/03 Bugzilla 994: added openssl_options main configuration option.
1285 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1287 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1289 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1291 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1292 without a peer certificate, leading to a segfault because of an
1293 assumption that peers always have certificates. Be a little more
1294 paranoid. Problem reported by Martin Tscholak.
1296 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1297 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1298 NB: ClamAV planning to remove STREAM in "middle of 2010".
1299 CL also introduces -bmalware, various -d+acl logging additions and
1300 more caution in buffer sizes.
1302 PP/09 Implemented reverse_ip expansion operator.
1304 PP/10 Bugzilla 937: provide a "debug" ACL control.
1306 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1308 PP/12 Bugzilla 973: Implement --version.
1310 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1312 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1314 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1316 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1317 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1318 ignore trailing whitespace.
1320 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1322 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1325 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1326 Notification from Dr Andrew Aitchison.
1328 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1329 ExtendedDetectionInfo response format.
1330 Notification from John Horne.
1332 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1335 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1336 XSL and documented dependency on system catalogs, with examples of how
1339 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1342 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1343 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1344 configuration file which is writeable by the Exim user or group.
1346 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1347 of configuration files to cover files specified with the -C option if
1348 they are going to be used with root privileges, not just the default
1351 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1352 option (effectively making it always true).
1354 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1355 files to be used while preserving root privileges.
1357 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1358 that rogue child processes cannot use them.
1360 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1361 run-time user, instead of root.
1363 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1364 Exim run-time user without dropping privileges.
1366 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1367 result string, instead of calling string_vformat() twice with the same
1370 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1371 for other users. Others should always drop root privileges if they use
1372 -C on the command line, even for a whitelisted configure file.
1374 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1376 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1382 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1383 $data_path, and $header_path variables; fixed documentation bugs and
1386 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1387 exipick to access non-standard spools, including the "frozen" queue
1390 NM/01 Bugzilla 965: Support mysql stored procedures.
1391 Patch from Alain Williams
1393 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1395 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1396 Patch from Andreas Metzler
1398 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1399 Patch from Kirill Miazine
1401 NM/05 Bugzilla 671: Added umask to procmail example.
1403 JJ/03 installed exipick 20100323.0, fixing doc bug
1405 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1406 directory. Notification and patch from Dan Rosenberg.
1408 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1410 TK/02 Improve log output when DKIM signing operation fails.
1412 MH/01 Treat the transport option dkim_domain as a colon separated
1413 list, not as a single string, and sign the message with each element,
1414 omitting multiple occurences of the same signer.
1416 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1417 Bugzilla 985, 986. Patch by Simon Arlott
1419 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1420 Patch by Simon Arlott
1422 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1423 MBX locking. Notification from Dan Rosenberg.
1429 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1431 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1433 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1435 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1437 NM/04 Bugzilla 915: Flag broken perl installation during build.
1443 TK/01 Added patch by Johannes Berg that expands the main option
1444 "spamd_address" if it starts with a dollar sign.
1446 TK/02 Write list of recipients to X-Envelope-Sender header when building
1447 the mbox-format spool file for content scanning (suggested by Jakob
1450 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1451 (http://www.dcc-servers.net/) support via dccifd. Activated by
1452 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1454 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1455 by Mark Daniel Reidel <mr@df.eu>.
1457 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1458 When building exim an external PCRE library is now needed -
1459 PCRE is a system library on the majority of modern systems.
1460 See entry on PCRE_LIBS in EDITME file.
1462 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1463 conversation. Added nologin parameter to request.
1464 Patch contributed by Kirill Miazine.
1466 TF/01 Do not log submission mode rewrites if they do not change the address.
1468 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1470 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1471 log files in place. Contributed by Roberto Lima.
1473 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1475 TF/03 Bugzilla 615: When checking the local_parts router precondition
1476 after a local_part_suffix or local_part_prefix option, Exim now
1477 does not use the address's named list lookup cache, since this
1478 contains cached lookups for the whole local part.
1480 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1481 Robert Millan. Documentation is in experimental-spec.txt.
1483 TF/04 Bugzilla 668: Fix parallel build (make -j).
1485 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1487 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1488 Patch provided by Jan Srzednicki.
1490 TF/05 Leading white space used to be stripped from $spam_report which
1491 wrecked the formatting. Now it is preserved.
1493 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1494 that they are available at delivery time.
1496 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1498 TF/08 TLS error reporting now respects the incoming_interface and
1499 incoming_port log selectors.
1501 TF/09 Produce a more useful error message if an SMTP transport's hosts
1502 setting expands to an empty string.
1504 NM/06 Bugzilla 744: EXPN did not work under TLS.
1505 Patch provided by Phil Pennock.
1507 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1508 Patch provided by Richard Godbee.
1510 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1511 acl_smtp_notquit, added index entry.
1513 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1514 Patch provided by Eugene Bujak.
1516 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1517 accept(). Patch provided by Maxim Dounin.
1519 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1520 Patch provided by Phil Pennock.
1522 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1524 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1525 Patch provided by Brad "anomie" Jorsch.
1527 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1528 Patch provided by Dean Brooks.
1530 TK/05 Add native DKIM support (does not depend on external libraries).
1532 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1533 Patch provided by Graeme Fowler.
1535 NM/16 Bugzilla 851: Documentation example syntax fix.
1537 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1539 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1542 NM/19 Bugzilla 745: TLS version reporting.
1543 Patch provided by Phil Pennock.
1545 NM/20 Bugzilla 167: bool: condition support.
1546 Patch provided by Phil Pennock.
1548 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1549 clients. Patch provided by Phil Pennock.
1551 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1552 Patch provided by Brad "anomie" Jorsch.
1554 NM/23 Bugzilla 687: Fix misparses in eximstats.
1555 Patch provided by Heiko Schlittermann.
1557 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1558 Patch provided by Heiko Schlittermann.
1560 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1561 plus update to original patch.
1563 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1565 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1566 Patch provided by David Brownlee.
1568 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1570 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1572 NM/30 Bugzilla 888: TLS documentation bugfixes.
1574 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1576 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1577 Unlike the original bugzilla I have changed all shell scripts in src tree.
1579 NM/33 Bugzilla 898: Transport filter timeout fix.
1580 Patch by Todd Rinaldo.
1582 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1583 Patch by Serge Demonchaux.
1585 NM/35 Bugzilla 39: Base64 decode bug fixes.
1586 Patch by Jakob Hirsch.
1588 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1590 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1592 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1594 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1600 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1601 ALT-N's libdkim that I have put here:
1602 http://duncanthrax.net/exim-experimental/
1604 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1605 were called 'true' and it seems that C99 defines that as a reserved
1606 keyword to be used with 'bool' variable types. That means you could
1607 not include C99-style headers which use bools without triggering
1608 build errors in sieve.c.
1610 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1611 as mailq or other aliases. Changed the --help handling significantly
1612 to do whats expected. exim_usage() emits usage/help information.
1614 SC/01 Added the -bylocaldomain option to eximstats.
1616 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1618 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1620 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1626 PH/01 Another patch from the Sieve maintainer.
1628 PH/02 When an IPv6 address is converted to a string for single-key lookup
1629 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1630 dots are used instead of colons so that keys in lsearch files need not
1631 contain colons. This was done some time before quoting was made available
1632 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1633 (notated using the quote facility) so as to distinguish them from IPv4
1634 keys. This meant that lookups for IP addresses in host lists did not work
1635 for iplsearch lookups.
1637 This has been fixed by arranging for IPv6 addresses to be expressed with
1638 colons if the lookup type is iplsearch. This is not incompatible, because
1639 previously such lookups could never work.
1641 The situation is now rather anomolous, since one *can* have colons in
1642 ordinary lsearch keys. However, making the change in all cases is
1643 incompatible and would probably break a number of configurations.
1645 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1648 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1649 conversion specification without a maximum field width, thereby enabling
1650 a rogue spamd server to cause a buffer overflow. While nobody in their
1651 right mind would setup Exim to query an untrusted spamd server, an
1652 attacker that gains access to a server running spamd could potentially
1653 exploit this vulnerability to run arbitrary code as the Exim user.
1655 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1656 $primary_hostname instead of what libspf2 thinks the hosts name is.
1658 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1659 a directory entry by the name of the lookup key. Previously, if a
1660 symlink pointed to a non-existing file or a file in a directory that
1661 Exim lacked permissions to read, a lookup for a key matching that
1662 symlink would fail. Now it is enough that a matching directory entry
1663 exists, symlink or not. (Bugzilla 503.)
1665 PH/03 The body_linecount and body_zerocount variables are now exported in the
1668 PH/04 Added the $dnslist_matched variable.
1670 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1671 This means they are set thereafter only if the connection becomes
1674 PH/06 Added the client_condition to authenticators so that some can be skipped
1675 by clients under certain conditions.
1677 PH/07 The error message for a badly-placed control=no_multiline_responses left
1678 "_responses" off the end of the name.
1680 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1682 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1683 (without spaces) instead of just copying the configuration text.
1685 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1687 PH/11 Added $max_received_linelength.
1689 PH/12 Added +ignore_defer and +include_defer to host lists.
1691 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1692 way in which PCRE > 7.0 is built.
1694 PH/14 Implemented queue_only_load_latch.
1696 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1697 MAIL command. The effect was to mangle the value on 64-bit systems.
1699 PH/16 Another patch from the Sieve maintainer.
1701 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1703 PH/18 If a system quota error occurred while trying to create the file for
1704 a maildir delivery, the message "Mailbox is full" was not appended to the
1705 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1706 only to a quota excession during the actual writing of the file.
1708 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1709 characters?) which causes problems in log lines. The DN values are now
1710 passed through string_printing() before being added to log lines.
1712 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1713 and InterBase are left for another time.)
1715 PH/21 Added message_body_newlines option.
1717 PH/22 Guard against possible overflow in moan_check_errorcopy().
1719 PH/23 POSIX allows open() to be a macro; guard against that.
1721 PH/24 If the recipient of an error message contained an @ in the local part
1722 (suitably quoted, of course), incorrect values were put in $domain and
1723 $local_part during the evaluation of errors_copy.
1729 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1730 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1733 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1734 issue a MAIL command.
1736 PH/02 In an ACL statement such as
1738 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1740 if a client was not listed at all, or was listed with a value other than
1741 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1742 the condition was not true (as it should be), so access was not denied.
1743 The bug was that the ! inversion was incorrectly passed on to the second
1744 item. This has been fixed.
1746 PH/03 Added additional dnslists conditions == and =& which are different from
1747 = and & when the dns lookup returns more than one IP address.
1749 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1750 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1752 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1753 FSYNC, which compiles an option called disable_fsync that allows for
1754 bypassing fsync(). The documentation is heavily laced with warnings.
1756 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1758 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1759 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1760 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1761 including adding "make clean"; (3) Added -fPIC when compiling the test
1762 dynamically loaded module, to get rid of a warning.
1764 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1765 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1766 = 0s. The bug is otherwise harmless.
1768 PH/07 There was a bug in the dovecot authenticator such that the value of
1769 $auth1 could be overwritten, and so not correctly preserved, after a
1770 successful authentication. This usually meant that the value preserved by
1771 the server_setid option was incorrect.
1773 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1775 PH/09 Installed PCRE release 7.0.
1777 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1778 run for batched SMTP input. It is now run at the start of every message
1779 in the batch. While fixing this I discovered that the process information
1780 (output by running exiwhat) was not always getting set for -bs and -bS
1781 input. This is fixed, and it now also says "batched" for BSMTP.
1783 PH/11 Added control=no_pipelining.
1785 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1786 patch, slightly modified), and move the expansion of helo_data till after
1787 the connection is made in the smtp transport (so it can use these
1790 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1792 PH/14 Added log_selector = +pid.
1794 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1796 PH/16 Add ${if forany and ${if forall.
1798 PH/17 Added dsn_from option to vary the From: line in DSNs.
1800 PH/18 Flush SMTP output before performing a callout, unless control =
1801 no_callout_flush is set.
1803 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1804 was true (the default) a successful delivery failed to delete the retry
1805 item, thus causing premature timeout of the address. The bug is now
1808 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1810 PH/21 Long custom messages for fakedefer and fakereject are now split up
1811 into multiline reponses in the same way that messages for "deny" and
1812 other ACL rejections are.
1814 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1815 with slight modification.
1817 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1818 draft, changing the syntax and factoring some duplicate code".
1820 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1821 for deliveries of the second and subsequent messages over the same SMTP
1824 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1825 ${reduce, with only minor "tidies".
1827 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1829 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1830 expansion side effects.
1832 PH/27 When a message times out after an over-quota error from an Exim-imposed
1833 quota, the bounce message says "mailbox is full". This message was not
1834 being given when it was a system quota that was exceeded. It now should
1837 MH/03 Made $recipients available in local_scan(). local_scan() already has
1838 better access to the recipient list through recipients_list[], but
1839 $recipients can be useful in postmaster-provided expansion strings.
1841 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1842 in the case of a MAIL command with additional options following the
1843 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1844 were accidentally chopped off.
1846 PH/29 SMTP synchronization checks are implemented when a command is read -
1847 there is a check that no more input is waiting when there shouldn't be
1848 any. However, for some commands, a delay in an ACL can mean that it is
1849 some time before the response is written. In this time, more input might
1850 arrive, invalidly. So now there are extra checks after an ACL has run for
1851 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1852 pipelining has not been advertised.
1854 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1856 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1857 returned in the SMTP response when smtp_return_error_details was false.
1858 This has been fixed.
1860 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1861 instead of the C I/O that was originally supplied, because problems were
1862 reported on Solaris.
1864 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1865 Exim which did not show up earlier: it was assuming that a call to
1866 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1867 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1868 was a macro that became an assignment, so it seemed to work. This has
1869 changed to a proper function call with a void return, hence the compile
1870 error. Exim's code has been fixed.
1872 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1875 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1877 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1879 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1880 because it turns out to be too big or there is a timeout) there is no
1881 "Completed" line in the log. When some messages of this type were
1882 selected by exigrep, they were listed as "not completed". Others were
1883 picked up by some special patterns. I have improved the selection
1884 criteria to be more general.
1886 PH/38 The host_find_failed option in the manualroute router can now be set
1887 to "ignore", to completely ignore a host whose IP address cannot be
1888 found. If all hosts are ignored, the behaviour is controlled by the new
1889 host_all_ignored option.
1891 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1892 homing or because of multiple MX records with /mx) generated more than
1893 one IP address, and the following item turned out to be the local host,
1894 all the secondary addresses of the first item were incorrectly removed
1895 from the list, along with the local host and any following hosts (which
1896 is what is supposed to happen).
1898 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1899 whoever called Exim into the -H file. In the case of the daemon it was
1900 behaving confusingly. When first started, it used values for whoever
1901 started the daemon, but after a SIGHUP it used the Exim user (because it
1902 calls itself on a restart). I have changed the code so that it now always
1905 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1906 message are rejected with the same error (e.g. no authentication or bad
1907 sender address), and a DATA command is nevertheless sent (as can happen
1908 with PIPELINING or a stupid MUA), the error message that was given to the
1909 RCPT commands is included in the rejection of the DATA command. This is
1910 intended to be helpful for MUAs that show only the final error to their
1913 PH/42 Another patch from the Sieve maintainer.
1915 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1916 Eximstats - Fixed some broken HTML links and added missing column headers
1918 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1919 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1921 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1923 PH/43 Yet another patch from the Sieve maintainer.
1925 PH/44 I found a way to check for a TCP/IP connection going away before sending
1926 the response to the final '.' that terminates a message, but only in the
1927 case where the client has not sent further data following the '.'
1928 (unfortunately, this is allowed). However, in many cases there won't be
1929 any further data because there won't be any more messages to send. A call
1930 to select() can be used: if it shows that the input is "ready", there is
1931 either input waiting, or the socket has been closed. An attempt to read
1932 the next input character can distinguish the two cases. Previously, Exim
1933 would have sent an OK response which the client would never have see.
1934 This could lead to message repetition. This fix should cure that, at
1935 least in a lot of common cases.
1937 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1938 advertised in response to EHLO.
1944 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1945 fixed by 4.65/MH/01 (is this a record?) are fixed:
1947 (i) An empty string was always treated as zero by the numeric comparison
1948 operators. This behaviour has been restored.
1950 (ii) It is documented that the numeric comparison operators always treat
1951 their arguments as decimal numbers. This was broken in that numbers
1952 starting with 0 were being interpreted as octal.
1954 While fixing these problems I realized that there was another issue that
1955 hadn't been noticed. Values of message_size_limit (both the global option
1956 and the transport option) were treated as octal if they started with 0.
1957 The documentation was vague. These values are now always treated as
1958 decimal, and I will make that clear in the documentation.
1964 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1965 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1968 MH/01 Don't check that the operands of numeric comparison operators are
1969 integers when their expansion is in "skipping" mode (fixes bug
1970 introduced by 4.64-PH/07).
1972 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1973 child addresses, Exim now panics and dies. Previously, because the count
1974 is held in a short int, deliveries were likely to be lost. As such a
1975 large number of recipients for a single message is ridiculous
1976 (performance will be very, very poor), I have chosen to impose a limit
1977 rather than extend the field.
1983 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1984 leftover -K file (the existence of which was triggered by #402).
1985 While we were at it, introduced process PID as part of the -K
1986 filename. This should rule out race conditions when creating
1989 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1990 processing considerably. Previous code took too long for large mails,
1991 triggering a timeout which in turn triggers #401.
1993 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1994 in the DK code in transports.c. sendfile() is not really portable,
1995 hence the _LINUX specificness.
1997 TF/01 In the add_headers option to the mail command in an Exim filter,
1998 there was a bug that Exim would claim a syntax error in any
1999 header after the first one which had an odd number of characters
2002 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2003 callout verification, Exim cached a "reject" for the entire domain. This
2004 is correct for most verifications, but it is not correct for a recipient
2005 verification with use_sender or use_postmaster set, because in that case
2006 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2007 case of MAIL FROM:<> rejection from other early rejections (e.g.
2008 rejection of HELO). When verifying a recipient using a non-null MAIL
2009 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2010 Whatever the result of the callout, the value of the domain cache is
2011 left unchanged (for any other kind of callout, getting as far as trying
2012 RCPT means that the domain itself is ok).
2014 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2017 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2018 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2019 session, the daemon ignores ECONNECT errors and logs others; it now
2020 ignores EPIPE as well.
2022 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2023 (quoted-printable decoding).
2025 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2026 later the small subsequent patch to fix an introduced bug.
2028 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2030 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2032 PH/08 An error is now given if message_size_limit is specified negative.
2034 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2035 to be given (somewhat) arbitrary names.
2037 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2040 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2041 miscellaneous code fixes
2043 PH/10 Added the log_reject_target ACL modifier to specify where to log
2046 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2047 hostname. This is wrong, because it relates to the incoming message (and
2048 probably the interface on which it is arriving) and not to the outgoing
2049 callout (which could be using a different interface). This has been
2050 changed to use the value of the helo_data option from the smtp transport
2051 instead - this is what is used when a message is actually being sent. If
2052 there is no remote transport (possible with a router that sets up host
2053 addresses), $smtp_active_hostname is used.
2055 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2056 tweaks were necessary in order to get it to work (see also 21 below):
2057 (a) The code assumed that strncpy() returns a negative number on buffer
2058 overflow, which isn't the case. Replaced with Exim's string_format()
2060 (b) There were several signed/unsigned issues. I just did the minimum
2061 hacking in of casts. There is scope for a larger refactoring.
2062 (c) The code used strcasecmp() which is not a standard C function.
2063 Replaced with Exim's strcmpic() function.
2064 (d) The code set only $1; it now sets $auth1 as well.
2065 (e) A simple test gave the error "authentication client didn't specify
2066 service in request". It would seem that Dovecot has changed its
2067 interface. Fortunately there's a specification; I followed it and
2068 changed what the client sends and it appears to be working now.
2070 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2073 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2074 address A is aliased to B and C, where B exists and C does not. Without
2075 -v the output is "A verified" because verification stops after a
2076 successful redirection if more than one address is generated. However,
2077 with -v the child addresses are also verified. Exim was outputting "A
2078 failed to verify" and then showing the successful verification for C,
2079 with its parentage. It now outputs "B failed to verify", showing B's
2080 parentage before showing the successful verification of C.
2082 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2083 look up a TXT record in a specific list after matching in a combined
2086 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2087 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2088 they consult the DNS. I had assumed they would set it the way they
2089 wanted; and indeed my experiments on Linux seem to show that in some
2090 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2091 To be on the safe side, however, I have now made the interface to
2092 host_find_byname() similar to host_find_bydns(), with an argument
2093 containing the DNS resolver options. The host_find_byname() function now
2094 sets these options at its start, just as host_find_bydns() does. The smtp
2095 transport options dns_qualify_single and dns_search_parents are passed to
2096 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2097 of host_find_byname() use the default settings of RES_DEFNAMES
2098 (qualify_single) but not RES_DNSRCH (search_parents).
2100 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2101 spool_read_header() do less string testing, by means of a preliminary
2102 switch on the second character of optional "-foo" lines. (This is
2103 overdue, caused by the large number of possibilities that now exist.
2104 Originally there were few.) While I was there, I also converted the
2105 str(n)cmp tests so they don't re-test the leading "-" and the first
2106 character, in the hope this might squeeze out yet more improvement.
2108 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2109 flag allowing group syntax was set by the header_syntax check but not
2110 turned off, possible causing trouble later; (2) The flag was not being
2111 set at all for the header_verify test, causing "group"-style headers to
2112 be rejected. I have now set it in this case, and also caused header_
2113 verify to ignore an empty address taken from a group. While doing this, I
2114 came across some other cases where the code for allowing group syntax
2115 while scanning a header line wasn't quite right (mostly, not resetting
2116 the flag correctly in the right place). These bugs could have caused
2117 trouble for malformed header lines. I hope it is now all correct.
2119 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2120 with the "reply" argument non-NULL. The code, however (which originally
2121 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2122 but it didn't always do it. This confused somebody who was copying the
2123 code for some other use. I have removed all the tests.
2125 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2126 feature that was used to support insecure browsers during the U.S. crypto
2127 embargo. It requires special client support, and Exim is probably the
2128 only MTA that supported it -- and would never use it because real RSA is
2129 always available. This code has been removed, because it had the bad
2130 effect of slowing Exim down by computing (never used) parameters for the
2131 RSA_EXPORT functionality.
2133 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2134 authenticator to fail if there's a tab character in the incoming data
2135 (there should never be unless someone is messing about, as it's supposed
2136 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2137 if the connection is using TLS or if the remote IP is the same as the
2138 local IP, and the "valid-client-cert option" if a client certificate has
2141 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2142 authenticators. This can be used for authorization after authentication
2143 succeeds. (In the case of plaintext, it servers for both authentication
2146 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2147 if any retry times were supplied.
2149 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2150 connection, where there is no client IP address to check. In this
2151 situation, the verify now always succeeds.
2153 PH/25 Applied John Jetmore's -Mset patch.
2155 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2157 PH/27 In a string expansion for a processed (not raw) header when multiple
2158 headers of the same name were present, leading whitespace was being
2159 removed from all of them, but trailing whitespace was being removed only
2160 from the last one. Now trailing whitespace is removed from each header
2161 before concatenation. Completely empty headers in a concatenation (as
2162 before) are ignored.
2164 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2165 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2167 PH/29 [Removed. This was a change that I later backed out, and forgot to
2168 correct the ChangeLog entry (that I had efficiently created) before
2169 committing the later change.]
2171 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2172 address errors (4xx response to RCPT) over the same connection as other
2173 messages routed to the same hosts. Such deliveries are always "forced",
2174 so retry times are not inspected. This resulted in far too many retries
2175 for the affected addresses. The effect occurred only when there were more
2176 hosts than the hosts_max_try setting in the smtp transport when it had
2177 the 4xx errors. Those hosts that it had tried were not added to the list
2178 of hosts for which the message was waiting, so if all were tried, there
2179 was no problem. Two fixes have been applied:
2181 (i) If there are any address or message errors in an SMTP delivery, none
2182 of the hosts (tried or untried) are now added to the list of hosts
2183 for which the message is waiting, so the message should not be a
2184 candidate for sending over the same connection that was used for a
2185 successful delivery of some other message. This seems entirely
2186 reasonable: after all the message is NOT "waiting for some host".
2187 This is so "obvious" that I'm not sure why it wasn't done
2188 previously. Hope I haven't missed anything, but it can't do any
2189 harm, as the worst effect is to miss an optimization.
2191 (ii) If, despite (i), such a delivery is accidentally attempted, the
2192 routing retry time is respected, so at least it doesn't keep
2193 hammering the server.
2195 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2196 in ${readsocket because some servers need this prod.
2198 PH/32 Added some extra debug output when updating a wait-xxx database.
2200 PH/33 The hint "could be header name not terminated by colon", which has been
2201 given for certain expansion errors for a long time, was not being given
2202 for the ${if def:h_colon_omitted{... case.
2204 PH/34 The spec says: "With one important exception, whenever a domain list is
2205 being scanned, $domain contains the subject domain." There was at least
2206 one case where this was not true.
2208 PH/35 The error "getsockname() failed: connection reset by peer" was being
2209 written to the panic log as well as the main log, but it isn't really
2210 panic-worthy as it just means the connection died rather early on. I have
2211 removed the panic log writing for the ECONNRESET error when getsockname()
2214 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2215 runs only) independently of the message's sender address. This meant
2216 that, if the 4xx error was in fact related to the sender, a different
2217 message to the same recipient with a different sender could confuse
2218 things. In particualar, this can happen when sending to a greylisting
2219 server, but other circumstances could also provoke similar problems.
2220 I have changed the default so that the retry time for these errors is now
2221 based a combination of the sender and recipient addresses. This change
2222 can be overridden by setting address_retry_include_sender=false in the
2225 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2226 remote server are returned as part of bounce messages. This was not
2227 happening for LMTP over a pipe (the lmtp transport), but now it is the
2228 same for both kinds of LMTP.
2230 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2231 in header lines that were in fact CNAMEs. This is no longer the case.
2233 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2234 and queue runs started by the daemon processed all messages. This has
2235 been fixed so that -R and -S can now usefully be given with -q<time>.
2237 PH/40 Import PCRE release 6.7 (fixes some bugs).
2239 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2241 PH/42 Give an error if -q is specified more than once.
2243 PH/43 Renamed the variables $interface_address and $interface_port as
2244 $received_ip_address and $received_port, to make it clear that these
2245 values apply to message reception, and not to the outgoing interface when
2246 a message is delivered. (The old names remain recognized, of course.)
2248 PH/44 There was no timeout on the connect() call when using a Unix domain
2249 socket in the ${readsocket expansion. There now is.
2251 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2252 be meaningful with "accept".
2254 SC/01 Eximstats V1.43
2255 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2257 SC/02 Eximstats V1.44
2258 Use a glob alias rather than an array ref in the generated
2259 parser. This improves both readability and performance.
2261 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2262 Collect SpamAssassin and rejection statistics.
2263 Don't display local sender or destination tables unless
2264 there is data to show.
2265 Added average volumes into the top table text output.
2267 SC/04 Eximstats V1.46
2268 Collect data on the number of addresses (recipients)
2269 as well as the number of messages.
2271 SC/05 Eximstats V1.47
2272 Added 'Message too big' to the list of mail rejection
2273 reasons (thanks to Marco Gaiarin).
2275 SC/06 Eximstats V1.48
2276 Mainlog lines which have GMT offsets and are too short to
2277 have a flag are now skipped.
2279 SC/07 Eximstats V1.49 (Alain Williams)
2280 Added the -emptyok flag.
2282 SC/08 Eximstats V1.50
2283 Fixes for obtaining the IP address from reject messages.
2285 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2286 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2287 whitesspace changes from 4.64-PH/27
2289 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2292 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2293 are found, allow negative numbers in numeric criteria)
2295 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2297 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2298 to match changes made in 4.64-PH/43
2300 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2302 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2303 "domain" is actually an IP address, and force a failure. This locks out
2304 those revolvers/nameservers that support "A-for-A" lookups, in
2305 contravention of the specifications.
2307 PH/48 When a host name was looked up from an IP address, and the subsequent
2308 forward lookup of the name timed out, the host name was left in
2309 $sender_host_name, contrary to the specification.
2311 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2312 restricted to single-key lookups, Exim was not diagnosing an error if
2313 * or *@ was used with a query-style lookup.
2315 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2317 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2318 long ago, but noone interested enough thought of it. Let's just say that
2319 the "1.1" means that there are some new functions that weren't there at
2320 some point in the past.
2322 PH/51 Error processing for expansion failure of helo_data from an smtp
2323 transport during callout processing was broken.
2325 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2326 tested/used via the -bh/-bhc/-bs options.
2328 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2329 bug, fixed in subsequent PCRE releases).
2331 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2332 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2334 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2340 SC/01 Use a glob alias rather than an array ref in eximstats generated
2341 parser. This improves both readability and performance.
2343 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2344 Don't display local sender or destination tables in eximstats unless
2345 there is data to show.
2346 Added average volumes into the eximstats top table text output.
2348 SC/03 Collect data on the number of addresses (recipients) as well
2349 as the number of messages in eximstats.
2351 TF/01 Correct an error in the documentation for the redirect router. Exim
2352 does (usually) call initgroups() when daemonizing.
2354 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2355 with consistent privilege compared to when running as a daemon.
2357 TF/03 Note in the spec that $authenticated_id is not set for local
2358 submissions from trusted users.
2360 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2361 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2363 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2364 by adding some example configuration directives to the default
2365 configuration file. A little bit of work is required to uncomment the
2366 directives and define how usernames and passwords are checked, but
2367 there is now a framework to start from.
2369 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2370 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2371 without this. I don't know how relevant this is to other LDAP libraries.
2373 PH/02 Add the verb name to the "unknown ACL verb" error.
2375 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2377 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2379 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2380 directory not expanded when it should be if an expanded home directory
2381 was set for the address (which is overridden by the transport).
2383 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2386 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2387 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2388 because it is too late at that time, and has no effect.
2390 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2391 security issue with \' (bugzilla #107). I could not use the
2392 PQescapeStringConn() function, because it needs a PGconn value as one of
2395 PH/08 When testing addresses using -bt, indicate those final addresses that
2396 are duplicates that would not cause an additional delivery. At least one
2397 person was confused, thinking that -bt output corresponded to deliveries.
2398 (Suppressing duplicates isn't a good idea as you lose the information
2399 about possibly different redirections that led to the duplicates.)
2401 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2402 systems where poll() doesn't work, in particular OS X.
2404 PH/10 Added more information to debugging output for retry time not reached.
2406 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2407 operations in malware.c.
2409 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2412 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2413 syslog_duplication set false, log lines that would normally be written
2414 both the the main log and to the reject log were not written to syslog at
2417 PH/14 In the default configuration, change the use of "message" in ACL warn
2418 statements to "add_header".
2420 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2421 not followed by a command (e.g. "seen endif").
2423 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2424 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2427 PH/17 Added extra conditions to the default value of delay_warning_condition
2431 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2432 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2433 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2436 The Auto-Submitted: and various List- headers are standardised, whereas I
2437 don't think Precedence: ever was.
2439 PH/18 Refactored debugging code in route_finduser() to show more information,
2440 in particular, the error code if getpwnam() issues one.
2442 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2443 This is apparently needed in addition to the PH/07 change above to avoid
2444 any possible encoding problems.
2446 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2447 but not after initializing Perl.
2449 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2450 output them only if debugging. By default they are written stderr,
2451 apparently, which is not desirable.
2453 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2456 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2459 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2461 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2462 authenticated or an ident call has been made. Suppress the default
2463 values for $authenticated_id and $authenticated_sender (but permit -oMai
2464 and -oMas) when testing with -bh.
2466 PH/24 Re-jigged the order of the tests in the default configuration so that the
2467 tests for valid domains and recipients precede the DNS black list and CSA
2468 tests, on the grounds that those ones are more expensive.
2470 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2471 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2472 This bug exists in every version of Exim that I still have, right back to
2475 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2476 However, an attempt to turn on case-sensitivity in a regex key by
2477 including (?-i) didn't work because the subject string was already
2478 lowercased, and the effects were non-intuitive. It turns out that a
2479 one-line patch can be used to allow (?-i) to work as expected.
2485 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2486 other effects) broke the use of negated acl sub-conditions.
2488 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2491 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2492 "Deny" causes Exim to reject the incoming connection with a 554 error.
2493 Unfortunately, if there is a major crisis, such as a disk failure,
2494 tcp-wrappers gives "deny", whereas what one would like would be some
2495 kind of temporary error. A kludge has been added to help with this.
2496 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2497 554 error is used if errno is still zero or contains ENOENT (which occurs
2498 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2501 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2503 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2504 errors. Otherwise a message that provokes a temporary error (when other
2505 messages do not) can cause a whole host to time out.
2507 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2508 addresses were routed directly to files or pipes from a redirect router.
2509 File deliveries just didn't batch; pipe deliveries might have suffered
2512 PH/06 A failure to get a lock for a hints database would erroneously always say
2513 "Failed to get write lock", even when it was really a read lock.
2515 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2516 of 0600. This has been changed to use the value of the lockfile_mode
2517 option (which defaults to 0600).
2519 PH/08 Applied small patch from the Sieve maintainer.
2521 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2522 folder from quota calculations, a direct delivery into this folder messed
2523 up the contents of the maildirsize file. This was because the regex was
2524 used only to exclude .Trash (or whatever) when the size of the mailbox
2525 was calculated. There was no check that a delivery was happening into an
2526 excluded directory. This bug has been fixed by ignoring all quota
2527 processing for deliveries into excluded directories.
2529 PH/10 Added the maildirfolder_create_regex option to appendfile.
2535 PH/01 The code for finding all the local interface addresses on a FreeBSD
2536 system running IPv6 was broken. This may well have applied to all BSD
2537 systems, as well as to others that have similar system calls. The broken
2538 code found IPv4 interfaces correctly, but gave incorrect values for the
2539 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2540 that it would not match correctly against @[] and not recognize the IPv6
2543 PH/02 The ipliteral router was not recognizing addresses of the form user@
2544 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2546 PH/03 Added disable_ipv6.
2548 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2549 decoded form, because it is most often used to construct To: headers
2550 lines in autoreplies, and the decoded form may well be syntactically
2551 invalid. However, $reply_address has leading white space removed, and all
2552 newlines turned into spaces so that the autoreply transport does not
2555 PH/05 If group was specified without a user on a router, and no group or user
2556 was specified on a transport, the group from the router was ignored.
2558 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2559 for visible compile-time settings that can be used to change these
2560 numbers, for those that want even more. Backwards compatibility with old
2561 spool files has been maintained. However, going back to a previous Exim
2562 release will lost any variables that are in spool files.
2564 PH/07 Two small changes when running in the test harness: increase delay when
2565 passing a TCP/IP connection to a new process, in case the original
2566 process has to generate a bounce, and remove special handling of
2567 127.0.0.2 (sic), which is no longer necessary.
2569 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2570 be the same on different OS.
2572 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2575 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2576 whether --show-vars was specified or not
2578 JJ/02 exipick: Added support for new ACL variable spool format introduced
2581 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2582 syntactically invalid From: or Reply-to: line, and a filter used this to
2583 generate an autoreply, and therefore failed to obtain an address for the
2584 autoreply, Exim could try to deliver to a non-existent relative file
2585 name, causing unrelated and misleading errors. What now happens is that
2586 it logs this as a hard delivery error, but does not attempt to create a
2589 PH/11 The exinext utility has a -C option for testing purposes, but although
2590 the given file was scanned by exinext itself; it wasn't being passed on
2591 when Exim was called.
2593 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2594 an end-of-file indication when reading a command response.
2596 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2597 compiled. In many other places in Exim, IPv6 addresses are always
2598 recognized, so I have changed this. It also means that IPv4 domain
2599 literals of the form [IPV4:n.n.n.n] are now always recognized.
2601 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2602 used if the router is not running as root, for example, when verifying at
2603 ACL time, or when using -bh. The debugging output from this situation was
2604 non-existent - all you got was a failure to exec. I have made two
2607 (a) Failures to set uid/gid, the current directory, or a process leader
2608 in a subprocess such as that created by queryprogram now generate
2609 suitable debugging ouput when -d is set.
2611 (b) The queryprogram router detects when it is not running as root,
2612 outputs suitable debugging information if -d is set, and then runs
2613 the subprocess without attempting to change uid/gid.
2615 PH/15 Minor change to Makefile for building test_host (undocumented testing
2618 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2619 additional section of a DNS packet that returns MX or SRV records.
2620 Instead, it always explicitly searches for A/AAAA records. This avoids
2621 major problems that occur when a DNS server includes only records of one
2622 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2623 fixed another bug: if SRV records were looked up and the corresponding
2624 address records were *not* found in the additional section, the port
2625 values from the SRV records were lost.
2627 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2628 using the correct key (the original address) when searching the retry
2629 rules in order to find which one to use for generating the retry hint.
2631 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2632 adding the default one. Similarly, if it contains a Reply-To: header, the
2633 errors_reply_to option, if set, is not used.
2635 PH/19 When calculating a retry time, Exim used to measure the "time since
2636 failure" by looking at the "first failed" field in the retry record. Now
2637 it does not use this if it is later than than the arrival time of the
2638 message. Instead it uses the arrival time. This makes for better
2639 behaviour in cases where some deliveries succeed, thus re-setting the
2640 "first failed" field. An example is a quota failure for a huge message
2641 when small messages continue to be delivered. Without this change, the
2642 "time since failure" will always be short, possible causing more frequent
2643 delivery attempts for the huge message than are intended.
2644 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2646 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2647 $1, $2, $3) because the numerical variables can be reset during some
2648 expansion items (e.g. "match"), thereby losing the authentication data.
2650 PH/21 Make -bV show the size of off_t variables so that the test suite can
2651 decide whether to run tests for quotas > 2G.
2653 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2654 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2655 filecount value is greater than 2G or if a quota value is greater than 2G
2656 on a system where the size of off_t is not greater than 4, a panic error
2659 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2660 never match. The debug and -bh output now contains an explicit error
2661 message indicating a malformed IPv4 address or mask.
2663 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2664 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2665 PH/23 above applies.
2667 PH/25 Do not write to syslog when running in the test harness. The only
2668 occasion when this arises is a failure to open the main or panic logs
2669 (for which there is an explicit test).
2671 PH/26 Added the /no_tell option to "control=freeze".
2673 PH/27 If a host name lookup failed very early in a connection, for example, if
2674 the IP address matched host_lookup and the reverse lookup yielded a name
2675 that did not have a forward lookup, an error message of the form "no IP
2676 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2677 could be logged. Now it outputs the IP address instead of "NULL".
2679 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2680 allows the sender and the authenticated sender to be set when
2681 submitting a message from within Exim. Since child_open_exim() is
2682 documented for local_scan(), the new function should be too.
2684 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2685 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2686 results in an empty string is now treated as unset.
2688 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2690 PH/31 Added sender_verify_fail logging option.
2692 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2693 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2694 tidied the source and removed it altogether.
2696 PH/33 When a queue run was abandoned because the load average was too high, a
2697 log line was always written; now it is written only if the queue_run log
2698 selector is set. In addition, the log line for abandonment now contains
2699 information about the queue run such as the pid. This is always present
2700 in "start" and "stop" lines but was omitted from the "abandon" line.
2702 PH/34 Omit spaces between a header name and the colon in the error message that
2703 is given when verify = headers_syntax fails (if there are lots of them,
2704 the message gets confusing).
2706 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2707 names, as there are now some PTR records that contain slashes. This check
2708 is only to protect against broken name servers that fall over on strange
2709 characters, so the fact that it applies to all lookups doesn't matter.
2711 PH/36 Now that the new test suite is complete, we can remove some of the
2712 special code in Exim that was needed for the old test suite. For example,
2713 sorting DNS records because real resolvers return them in an arbitrary
2714 order. The new test suite's fake resolver always returns records in the
2717 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2718 bounces) except when queue_only is set, to avoid logging races between
2719 the different processes.
2721 PH/38 Panic-die if .include specifies a non-absolute path.
2723 PH/39 A tweak to the "H" retry rule from its user.
2725 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2726 a label. They prevented compilation on older perls.
2728 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2729 a warning to be raised on newish perls.
2731 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2732 on queue. Changes to match documented behaviour of showing count of
2733 messages matching specified criteria.
2735 PH/40 Changed the default ident timeout from 30s to 5s.
2737 PH/41 Added support for the use of login_cap features, on those BSD systems
2738 that have them, for controlling the resources used by pipe deliveries.
2740 PH/42 The content-scanning code uses fopen() to create files in which to put
2741 message data. Previously it was not paying any attention to the mode of
2742 the files. Exim runs with umask(0) because the rest of the code creates
2743 files with open(), and sets the required mode explicitly. Thus, these
2744 files were ending up world-writeable. This was not a big issue, because,
2745 being within the spool directory, they were not world-accessible. I have
2746 created a function called modefopen, which takes an additional mode
2747 argument. It sets umask(777), creates the file, chmods it to the required
2748 mode, then resets the umask. All the relevant calls to fopen() in the
2749 content scanning code have been changed to use this function.
2751 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2752 to 24 hours. This avoids potential overflow problems when processing G
2753 and H retry rules. I suspect nobody ever tinkers with this value.
2755 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2757 PH/45 When the plaintext authenticator is running as a client, the server's
2758 challenges are checked to ensure they are valid base64 strings. By
2759 default, the authentication attempt is cancelled if an invalid string is
2760 received. Setting client_ignore_invalid_base64 true ignores these errors.
2761 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2762 they are received. Thus, the responses can be made to depend on the
2763 challenges. If an invalid string is ignored, an empty string is placed in
2766 PH/46 Messages that are created by the autoreply transport now contains a
2767 References: header, in accordance with RFCs 2822 and 3834.
2769 PH/47 Added authenticated_sender_force to the smtp transport.
2771 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2773 PH/49 Installed latest patch from the Sieve maintainer.
2775 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2776 was also set, the appendfile transport was unnecessarily scanning a
2777 directory of message files (e.g. for maildir delivery) to find the count
2778 of files (along with the size), even though it did not need this
2779 information. It now does the scan only if it needs to find either the
2780 size of the count of files.
2782 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2784 PH/52 Two bugs concerned with error handling when the smtp transport is
2787 (i) Exim was not creating retry information for temporary errors given
2788 for individual recipients after the DATA command when the smtp transport
2789 was used in LMTP mode. This meant that they could be retried too
2790 frequently, and not timed out correctly.
2792 (ii) Exim was setting the flag that allows error details to be returned
2793 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2794 recipients that were returned after the DATA command.
2796 PH/53 This is related to PH/52, but is more general: for any failing address,
2797 when detailed error information was permitted to be returned to the
2798 sender, but the error was temporary, then after the final timeout, only
2799 "retry timeout exceeded" was returned. Now it returns the full error as
2800 well as "retry timeout exceeded".
2802 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2803 do this, and (what is worse) MTAs that accept it.
2805 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2806 will now be deprecated.
2808 PH/56 New os.c-cygwin from the Cygwin maintainer.
2810 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2811 formats (previously only available in exim formats via -bpr, -bpru,
2812 and -bpra. Now also available in native and exiqgrep formats)
2814 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2815 with very large, slow to parse queues
2817 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2819 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2821 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2822 responses to authentication challenges, though it was showing the
2823 challenges; (ii) I've removed the CR characters from the debug output for
2826 PH/58 Allow for the insertion of a newline as well as a space when a string
2827 is turned into more than one encoded-word during RFC 2047 encoding. The
2828 Sieve code now uses this.
2830 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2831 data_4xx, lost_connection, tls_required.
2833 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2834 message was being sent as an SMTP response.
2836 PH/61 Add -l and -k options to exicyclog.
2838 PH/62 When verifying, if an address was redirected to one new address, so that
2839 verification continued, and the new address failed or deferred after
2840 having set something in $address_data, the value of $address_data was not
2841 passed back to the ACL. This was different to the case when no
2842 redirection occurred. The value is now passed back in both cases.
2844 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2845 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2846 use login_cap.h, so on its own it isn't the distinguishing feature. The
2847 new name refers directly to the setclassresources() function.
2849 PH/65 Added configuration files for NetBSD3.
2851 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2853 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2854 is preferred over IPv4.
2856 PH/68 The bounce_return_message and bounce_return_body options were not being
2857 honoured for bounces generated during the reception of non-SMTP messages.
2858 In particular, this applied to messages rejected by the ACL. This bug has
2859 been fixed. However, if bounce_return_message is true and bounce_return_
2860 body is false, the headers that are returned for a non-SMTP message
2861 include only those that have been read before the error was detected.
2862 (In the case of an ACL rejection, they have all been read.)
2864 PH/69 The HTML version of the specification is now built in a directory called
2865 spec_html instead of spec.html, because the latter looks like a path with
2866 a MIME-type, and this confuses some software.
2868 PH/70 Catch two compiler warnings in sieve.c.
2870 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2871 function verify_get_ident() calls ip_connect() to connect a socket, but
2872 if the "connect()" function timed out, ip_connect() used to close the
2873 socket. However, verify_get_ident() also closes the socket later, and in
2874 between Exim writes to the log, which may get opened at this point. When
2875 the socket was closed in ip_connect(), the log could get the same file
2876 descriptor number as the socket. This naturally causes chaos. The fix is
2877 not to close the socket in ip_connect(); the socket should be closed by
2878 the function that creates it. There was only one place in the code where
2879 this was missing, in the iplookup router, which I don't think anybody now
2880 uses, but I've fixed it anyway.
2882 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2883 well as to direct DNS lookups. Otherwise the handling of names in host
2884 lists is inconsistent and therefore confusing.
2890 PH/01 Two changes to the default runtime configuration:
2892 (1) Move the checks for relay_from_hosts and authenticated clients from
2893 after to before the (commented out) DNS black list checks.
2895 (2) Add control=submission to the relay_from_hosts and authenticated
2896 clients checks, on the grounds that messages accepted by these
2897 statements are most likely to be submissions.
2899 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2901 (1) Generate an error if the third argument for the ${prvs expansion is
2904 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2907 (3) Reset the variables that are obtained from the first argument of
2908 ${prvscheck and used in the second argument before leaving the code,
2909 because their memory is reclaimed, so using them afterwards may do
2912 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2913 one (it's much easier than Tom thought :-).
2915 (5) Because of (4), we can now allow for the use of $prvscheck_result
2916 inside the third argument.
2918 PH/03 For some reason, the default setting of PATH when running a command from
2919 a pipe transport was just "/usr/bin". I have changed it to
2922 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2923 anything to be listed in the output from -bV.
2925 PH/05 When a filter generated an autoreply, the entire To: header line was
2926 quoted in the delivery log line, like this:
2928 => >A.N.Other <ano@some.domain> <original@ddress> ...
2930 This has been changed so that it extracts the operative address. There
2931 may be more than one such address. If so, they are comma-separated, like
2934 => >ano@some.domain,ona@other.domain <original@ddress> ...
2936 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2937 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2938 not being looked up in the rDNS to get a host name, Exim was showing the
2939 IP address twice in Received: lines, even though the IP addresses were
2940 identical. For example:
2942 Received: from [1.2.3.4] (helo=[1.2.3.4])
2944 However, if the real host name was known, it was omitting the HELO data
2945 if it matched the actual IP address. This has been tidied up so that it
2946 doesn't show the same IP address twice.
2948 PH/07 When both +timestamp and +memory debugging was on, the value given by
2949 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2950 called by the debug printing, thereby overwriting the timestamp buffer.
2951 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2953 PH/08 When the original message was included in an autoreply transport, it
2954 always said "this is a copy of the message, including all the headers",
2955 even if body_only or headers_only was set. It now gives an appropriate
2958 PH/09 Applied a patch from the Sieve maintainer which:
2960 o fixes some comments
2961 o adds the (disabled) notify extension core
2962 o adds some debug output for the result of if/elsif tests
2963 o points to the current vacation draft in the documentation
2964 and documents the missing references header update
2968 o fixes a bug in processing the envelope test (when testing
2969 multiple envelope elements, the last element determinted the
2972 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2973 Electronic Mail") by including:
2975 Auto-submitted: auto-generated
2977 in the messages that it generates (bounce messages and others, such as
2978 warnings). In the case of bounce messages for non-SMTP mesages, there was
2979 also a typo: it was using "Auto_submitted" (underscore instead of
2980 hyphen). Since every message generated by Exim is necessarily in response
2981 to another message, thes have all been changed to:
2983 Auto-Submitted: auto-replied
2985 in accordance with these statements in the RFC:
2987 The auto-replied keyword:
2989 - SHOULD be used on messages sent in direct response to another
2990 message by an automatic process,
2992 - MUST NOT be used on manually-generated messages,
2994 - MAY be used on Delivery Status Notifications (DSNs) and Message
2995 Disposition Notifications (MDNs),
2997 - MUST NOT be used on messages generated by automatic or periodic
2998 processes, except for messages which are automatic responses to
3001 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3002 to the default Received: header definition.
3004 PH/12 Added log selector acl_warn_skipped (default on).
3006 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3007 variables because (a) they are in the wrong storage pool and (b) even if
3008 they were copied, it wouldn't work properly because of the caching.
3010 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3011 checking when decoding. Apparently there are clients that generate
3012 overlong encoded strings. Why am I not surprised?
3014 PH/15 If the first argument of "${if match_address" was not empty, but did not
3015 contain an "@" character, Exim crashed. Now it writes a panic log message
3016 and treats the condition as false.
3018 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3020 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3021 extension "envelope-auth". The code is finished and in agreement with
3022 other implementations, but there is no documentation so far and in fact,
3023 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3024 not changing the active code.
3026 Print executed "if" and "elsif" statements when debugging is used. This
3027 helps a great deal to understand what a filter does.
3029 Document more things not specified clearly in RFC3028. I had all this
3030 sorted out, when out of a sudden new issues came to my mind. Oops."
3032 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3035 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3036 canonical form (as documented). However, after a host name lookup from
3037 the IP address, check_host() was doing a simple string comparison with
3038 addresses acquired from the DNS when checking that the found name did
3039 have the original IP as one of its addresses. Since any found IPv6
3040 addresses are likely to be in abbreviated form, the comparison could
3041 fail. Luckily, there already exists a function for doing the comparison
3042 by converting both addresses to binary, so now that is used instead of
3043 the text comparison.
3045 PH/20 There was another similar case to PH/19, when a complete host name was
3046 given in a host list; looking up its IP address could give an abbreviated
3047 form, whereas the current host's name might or might not be abbreviated.
3048 The same fix has been applied.
3054 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3055 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3058 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3059 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3061 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3063 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3064 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3065 Also, for "Auto-Submitted", treat anything other than "no" as
3066 non-personal, in accordance with RFC 3834. (Previously it treated
3067 anything starting "auto-" as non-personal.)
3069 TF/01 The control=submission/name=... option had a problem with syntax
3070 errors if the name included a slash character. The /name= option
3071 now slurps the rest of the string, so it can include any characters
3072 but it must come last in the list of options (after /sender_retain
3075 PH/05 Some modifications to the interface to the fake nameserver for the new
3083 TK/01 Added the "success_on_redirect" address verification option. See
3084 NewStuff for rationale and an example.
3086 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3088 PH/02 Patch to exigrep to allow it to work on syslog lines.
3090 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3091 fread() to skip over the body file's header line, because in Cygwin the
3092 header line is locked and is inaccessible.
3094 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3095 co-exist for some time) to make it clear that it is the Exim ID that is
3096 referenced, not the Message-ID: header line.
3098 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3099 string_format() function, because snprintf() does not exist on all
3102 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3103 ${stat: expansion item.
3105 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3106 protocol synchronization error", to keep the pedants happy.
3108 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3109 well as for IRIX systems, when gcc is being used. See the host.c source
3112 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3114 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3117 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3118 transport and to the smtp transport in LMTP mode.
3120 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3122 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3123 run a filter in a subprocess. This could lead to confusion in subsequent
3124 lookups in the parent process. There should also be a search_tidyup() at
3125 the end of the subprocess.
3127 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3128 only if the host matched helo_try_verify_hosts, which caused the
3129 verification to occur when the EHLO/HELO command was issued. The ACL just
3130 tested the remembered result. Now, if a previous verification attempt has
3131 not happened, "verify = helo" does it there and then.
3133 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3135 TK/03 Fix log output including CR from clamd.
3137 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3138 exist provoked a memory error which could cause a segfault.
3140 PH/15 Installed PCRE 6.2
3142 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3144 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3145 of the problem. Specifically, suggested +O2 rather than +O1 for the
3148 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3150 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3151 record, the port was not correctly logged when the outgoing_port log
3152 selector was set (it logged the transort's default port).
3154 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3155 fallback_hosts, and "hosts" in the smtp transport.
3157 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3158 host errors such as "Connection refused".
3160 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3161 authentication with radiusclient 0.4.9:
3163 - Error returned from rc_read_config was caught wrongly
3164 - Username/password not passed on to radius server due to wrong length.
3166 The presumption is that some radiusclient API changes for 4.51/PH/17
3167 were not taken care of correctly. The code is still untested by me (my
3168 Linux distribution still has 0.3.2 of radiusclient), but it was
3169 contributed by a Radius user.
3171 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3172 expanding the "port" option of the smtp transport.
3174 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3175 while reading a MIME header. Thanks to Tom Hughes for a patch.
3177 PH/24 Include config.h inside local_scan.h so that configuration settings are
3180 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3181 that in an ACL for RCPT (for example), you can examine exactly what was
3184 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3185 commands, but it was not correctly comparing the address with the actual
3186 client host address. Thus, it would show the EHLO address in Received:
3187 header lines when this was not necessary.
3189 PH/27 Added the % operator to ${eval:}.
3191 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3192 it should be ignoring failures (because with -C, for example, it has lost
3193 privilege). It wasn't ignoring creation failures other than "already
3196 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3199 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3200 because an input line was too long, either on its own, or by virtue of
3201 too many continuations, the temporary file was not being removed, and the
3202 return code was incorrect.
3204 PH/31 Missing "BOOL" in function definition in filtertest.c.
3206 PH/32 Applied Sieve patches from the maintainer.
3208 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3210 PH/33 Added "verify = not_blind".
3212 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3213 Local/Makefile (with some defaults set). These are used in built scripts
3214 such as exicyclog, but they have never been used in the exim_install
3215 script (though there are many overriding facilities there). I have
3216 arranged that the exim_install script now takes note of these two
3219 PH/35 Installed configuration files for Dragonfly.
3221 PH/36 When a locally submitted message by a trusted user did not contain a
3222 From: header, and the sender address was obtained from -f or from an SMTP
3223 MAIL command, and the trusted user did not use -F to supply a sender
3224 name, $originator_name was incorrectly used when constructing a From:
3225 header. Furthermore, $originator_name was used for submission mode
3226 messages from external hosts without From: headers in a similar way,
3227 which is clearly wrong.
3229 PH/37 Added control=suppress_local_fixups.
3231 PH/38 When log_selector = +received_sender was set, and the addition of the
3232 sender made the log line's construction buffer exactly full, or one byte
3233 less than full, an overflow happened when the terminating "\n" was
3236 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3237 when the result of a list match is failure because a DNS lookup failed.
3239 PH/40 RM_COMMAND is now used in the building process.
3241 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3242 the "build-* directories that it finds.
3244 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3245 address in a domain literal was a prefix of an interface address.
3247 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3248 when verifying a sender address, unless rewrite_headers is false.
3250 PH/44 Wrote a long comment about why errors_to addresses are verified as
3251 recipients, not senders.
3253 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3254 the ratelimit ACL was added.
3256 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3258 PH/46 Added extra information about PostgreSQL errors to the error string.
3260 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3261 avoiding the need to install special zones in a real server. This is
3262 backwards compatible; if it can't find the fake resolver, it drops back.
3263 Thus, both old and new test suites can be run.
3265 TF/02 Added util/ratelimit.pl
3267 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3268 clock is set back in time.
3270 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3271 Candler <B.Candler@pobox.com>.
3273 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3274 OK for addresses that are the result of redirections.
3276 PH/48 A number of further additions for the benefit of the new test suite,
3277 including a fake gethostbyname() that interfaces to the fake DNS resolver
3280 TF/06 The fix for widen_domains has also been applied to qualify_single and
3281 search_parents which are the other dnslookup options that can cause
3284 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3287 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3289 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3290 service" if it encountered an MX record with an empty target hostname.
3291 The message is now "an MX or SRV record indicated no SMTP service".
3293 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3294 if the DNS of the sending site is misconfigured. This is quite a
3295 common situation. This change restores the behaviour of treating a
3296 helo verification defer as a failure.
3298 PH/51 If self=fail was set on a router, the bounce message did not include the
3299 actual error message.
3305 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3307 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3308 command itself ended in error, the underlying message about the transport
3309 filter timeout was being overwritten with the pipe command error. Now the
3310 underlying error message should be appended to the second error message.
3312 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3314 PH/02 Reduce the amount of output that "make" produces by default. Full output
3315 can still be requested.
3317 PH/03 The warning log line about a condition test deferring for a "warn" verb
3318 was being output only once per connection, rather than after each
3319 occurrence (because it was using the same function as for successful
3320 "warn" verbs). This seems wrong, so I have changed it.
3322 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3323 it should not have, which might have caused a crash in the right
3324 circumstances, but probably never did.
3326 PH/04 Installed a modified version of Tony Finch's patch to make submission
3327 mode fix the return path as well as the Sender: header line, and to
3328 add a /name= option so that you can make the user's friendly name appear
3331 TF/03 Added the control = fakedefer ACL modifier.
3333 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3334 Mark Lowes for thorough testing.
3336 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3338 TK/03 Merged latest SRS patch from Miles Wilton.
3340 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3341 with the definition in sysexits.h (which is #included earlier).
3342 Fortunately, Exim does not actually use EX_OK. The code used to try to
3343 preserve the sysexits.h value, by assumimg that macro definitions were
3344 scanned for macro replacements. I have been disabused of this notion,
3345 so now the code just undefines EX_OK before #including unistd.h.
3347 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3348 in the smtp transport. When a block could not be written in a single
3349 write() function, the timeout was being re-applied to each part-write.
3350 This seems wrong - if the receiver was accepting one byte at a time it
3351 would take for ever. The timeout is now adjusted when this happens. It
3352 doesn't have to be particularly precise.
3354 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3355 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3357 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3358 without a domain if the check to <postmaster@domain> fails.
3360 SC/01 Eximstats: added -xls and the ability to specify output files
3361 (patch written by Frank Heydlauf).
3363 SC/02 Eximstats: use FileHandles for outputing results.
3365 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3367 SC/04 Eximstats: fixed display of large numbers with -nvr option
3369 SC/05 Eximstats: fixed merging of reports with empty tables.
3371 SC/06 Eximstats: added the -include_original_destination flag
3373 SC/07 Eximstats: removed tabs and trailing whitespace.
3375 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3377 TK/06 MBOX spool code: Add real "From " MBOX separator line
3378 so the .eml file is really in mbox format (even though
3379 most programs do not really care). Patch from Alex Miller.
3381 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3382 The latter is generated from $received_to and is only set if the
3383 message has one envelope recipient. SA can use these headers,
3384 obviously out-of-the-box. Patch from Alex Miller.
3386 PH/08 The ${def test on a variable was returning false if the variable's
3387 value was "0", contrary to what the specification has always said!
3388 The result should be true unless the variable is empty.
3390 PH/09 The syntax error of a character other than { following "${if
3391 def:variable_name" (after optional whitespace) was not being diagnosed.
3392 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3393 accidental colon was present, for example, could give incorrect results.
3395 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3396 result is used (not including appendfile, where other changes are about
3399 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3400 This involved changing a lot of size variables from int to off_t. It
3401 should work with maildirs and everything.
3403 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3404 spamd dying while we are connected to it.
3406 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3409 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3410 function for BDB 4.3.
3412 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3414 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3415 the sources for PCRE have been re-arranged and more files are now
3418 PH/15 The code I had for printing potentially long long variables in PH/11
3419 above was not the best (it lost precision). The length of off_t variables
3420 is now inspected at build time, and an appropriate printing format (%ld
3421 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3422 to be "long long int" or "long int". This is needed for the internal
3423 formatting function string_vformat().
3425 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3426 the configuration file to be ":syslog", then the script "guesses" where
3427 the logs files are, rather than using the compiled in default. In our
3428 case the guess is not the same as the compiled default, so the script
3429 suddenly stopped working when I started to use syslog. The patch checks
3430 to see if log_file_path is "". If so, it attempts to read it from exim
3431 with no configuration file to get the compiled in version, before it
3432 falls back to the previous guessing code."
3434 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3435 implementing BATV in an Exim configuration. See NewStuff for the gory
3438 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3439 Makefile that are specific to HP-UX.
3441 PH/18 If the "use_postmaster" option was set for a recipient callout together
3442 with the "random" option, the postmaster address was used as the MAIL
3443 FROM address for the random test, but not for the subsequent recipient
3444 test. It is now used for both.
3446 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3447 patch removes a few documentation additions to RFC 3028, because the
3448 latest draft now contains them. It adds the new en;ascii-case comparator
3449 and a new error check for 8bit text in MIME parts. Comparator and
3450 require names are now matched exactly. I enabled the subaddress
3451 extension, but it is not well tested yet (read: it works for me)."
3453 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3454 rework some of the code of TK/09 above to avoid the hardwired use of
3455 "%lld" and "long long". Replaced the call to snprintf() with a call to
3458 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3459 records point to non-existent hosts", "retry timeout exceeded", and
3460 "retry time not reached for any host after a long failure period".
3462 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3463 experimental DomainKeys support:
3465 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3466 (2) On an error such as an illegally used "control", the wrong name for
3467 the control was given.
3469 These problems did NOT occur unless DomainKeys support was compiled.
3471 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3473 PH/24 Added ${if match_ip condition.
3475 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3476 databases so that it will be absolutely obvious if a crash occurs in the
3477 DB library. This is a regular occurrence (often caused by mis-matched
3480 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3481 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3482 picked up on a user's system that detects such things. There doesn't seem
3483 to be a gcc warning option for this - only an attribute that has to be
3484 put on the function's prototype. It seems that in Fedora Core 4 they have
3485 set this on a number of new functions. No doubt there will be more in due
3488 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3489 specify a transport. However, if an address that was verified by such a
3490 router was the subject of a callout, Exim crashed because it tried to
3491 read the rcpt_include_affixes from the non-existent transport. Now it
3492 just assumes that the setting of that option is false. This bug was
3493 introduced by 4.51/PH/31.
3495 PH/28 Changed -d+all to exclude +memory, because that information is very
3496 rarely of interest, but it makes the output a lot bigger. People tend to
3497 do -d+all out of habit.
3499 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3500 code in os-type was giving problems when libc.so lives in lib64, like on
3503 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3504 aren't the modern standard, and it seems that some systems' include files
3505 don't always have them. Exim was already checking for some of the newer
3506 ones like T_AAAA, and defining it itself. I've added checks for all the
3507 record types that Exim uses.
3509 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3510 not automatically generating a new one, as it is supposed to. This
3511 prevented TLS from working. If the file did exist, but contained invalid
3512 data, a new version was generated, as expected. It was only the case of a
3513 non-existent file that was broken.
3515 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3516 with a change in libdomainkeys > 0.64.
3518 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3519 from DNS. If the selector record carries the flag, it now has
3520 precedence over the domain-wide flag.
3522 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3524 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3525 the use of a port name that isn't defined in /etc/services) occurred, the
3526 message was deferred as in a normal delivery, and thus remained on the
3527 spool, instead of being failed because of the mua_wrapper setting. This
3528 is now fixed, and I tidied up some of the mua_wrapper messages at the
3531 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3532 the messages in a hash of arrays rather than using individual hashes.
3533 This is a bit cleaner and results in dramatic memory savings, albeit
3534 at a slight CPU cost.
3536 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3537 as requested by Marc Sherman.
3539 SC/10 Eximstats: added histograms for user specified patterns as requested
3542 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3544 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3545 fopen() in the content-scanning modules that did not already have it.
3551 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3552 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3554 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3556 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3558 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3559 to test Sieve filters that use "vacation".
3561 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3562 that changes the way the GnuTLS parameters are stored in the cache file.
3563 The new format can be generated externally. For backward compatibility,
3564 if the data in the cache doesn't make sense, Exim assumes it has read an
3565 old-format file, and it generates new data and writes a new file. This
3566 means that you can't go back to an older release without removing the
3569 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3570 work if there are any delivery delays because "one_time" forces the
3571 parent to be marked "delivered", so its unseen clone is never tried
3572 again. For this reason, Exim now forbids the simultaneous setting of
3575 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3576 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3577 entry if you want to know the details. The fix, however, neglected to
3578 consider the case when local delivery batching is involved. The test for
3579 "previously delivered" was not happening when checking to see if an
3580 address could be batched with a previous (undelivered) one; under
3581 certain circumstances this could lead to multiple deliveries to the same
3584 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3585 in its include files, and this causes problems building Exim.
3587 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3588 header_syntax) but Exim was just ignoring anything given after a slash.
3589 In particular, this caused confusion with an attempt to use "verify =
3590 reverse_host_lookup/defer_ok". An error is now given when options are
3591 supplied for verify items that do not have them. (Maybe reverse_host_
3592 lookup should have a defer_ok option, but that's a different point.)
3594 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3595 defined by RFC 821) to 2048, because there were problems with some AUTH
3596 commands, and RFC 1869 says the size should be increased for extended
3597 SMTP commands that take arguments.
3599 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3602 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3603 "unknown" error; now it says that the functionality isn't in the binary.
3605 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3606 an address' error message when a string expansion fails (syntax or
3607 whatever). Otherwise the password may appear in the log. Following change
3608 PH/42 below, there is no longer a chance of it appearing in a bounce
3611 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3613 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3614 of its addresses was ever tried. (Bugzilla bug #2.)
3616 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3617 the result incorrectly in the debug output. (It correctly added a newline
3618 to what was transported.)
3620 TF/01 Added $received_time.
3622 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3623 commented out examples of how to interface to a virus scanner and to
3624 SpamAssassin. Also added commented examples of av_scanner and
3625 spamd_address settings.
3627 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3628 and controls are allowed in which ACLs. There were a couple of minor
3629 errors. Some of the entries in the conditions table (which is a table of
3630 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3631 negation of where the condition IS allowed.
3633 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3635 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3636 header file does not have a version number, so I've had to invent a new
3637 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3638 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3639 radiusclient), but it was contributed by a Radius user.
3641 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3642 files or directories, for OpenSSL.
3644 PH/19 When an Exim process that is running as root has to create an Exim log
3645 file, it does so in a subprocess that runs as exim:exim so as to get the
3646 ownership right at creation (otherwise, other Exim processes might see
3647 the file with the wrong ownership). There was no test for failure of this
3648 fork() call, which would lead to the process getting stuck as it waited
3649 for a non-existent subprocess. Forks do occasionally fail when resources
3650 run out. I reviewed all the other calls to fork(); they all seem to check
3653 PH/20 When checking for unexpected SMTP input at connect time (before writing
3654 the banner), Exim was not dealing correctly with a non-positive return
3655 from the read() function. If the client had disconnected by this time,
3656 the result was a log entry for a synchronization error with an empty
3657 string after "input=" when read() returned zero. If read() returned -1
3658 (an event I could not check), uninitialized data bytes were printed.
3659 There were reports of junk text (parts of files, etc) appearing after
3662 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3664 PH/22 Added support for macro redefinition, and (re)definition in between
3665 driver and ACL definitions.
3667 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3668 forgetting to use the resulting value; it was using the unexpanded value.
3670 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3671 hadn't been configured. The fix is from Juergen Kreileder, who
3672 understands it better than I do:
3674 "Here's what I see happening with three configured cyrus_sasl
3675 authenticators configured (plain, login, cram-md5):
3677 On startup auth_cyrus_sasl_init() gets called for each of these.
3678 This means three calls to sasl_listmech() without a specified mech_list.
3679 => SASL tests which mechs of all available mechs actually work
3680 => three warnings about OTP not working
3681 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3683 With the patch, sasl_listmech() also gets called three times. But now
3684 SASL's mech_list option is set to the server_mech specified in the the
3685 authenticator. Or in other words, the answer from sasl_listmech()
3686 gets limited to just the mech you're testing for (which is different
3688 => the return list contains just 'plain' or 'login', 'cram-md5' or
3689 nothing depending on the value of ob->server_mech.
3691 I've just tested the patch: Authentication still works fine,
3692 unavailable mechs specified in the exim configuration are still
3693 caught, and the auth.log warnings about OTP are gone."
3695 PH/25 When debugging is enabled, the contents of the command line are added
3696 to the debugging output, even when log_selector=+arguments is not
3699 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3700 answer is "GNU", and only if the return is "GNU/something" is the answer
3703 PH/27 $acl_verify_message is now set immediately after the failure of a
3704 verification in an ACL, and so is available in subsequent modifiers. In
3705 particular, the message can be preserved by coding like this:
3707 warn !verify = sender
3708 set acl_m0 = $acl_verify_message
3710 Previously, $acl_verify_message was set only while expanding "message"
3711 and "log_message" when a very denied access.
3713 PH/28 Modified OS/os.c-Linux with
3715 -#ifndef OS_LOAD_AVERAGE
3716 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3718 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3719 nomenclature these days.)
3721 PH/29 Installed patch from the Sieve maintainer that adds the options
3722 sieve_useraddress and sieve_subaddress to the redirect router.
3724 PH/30 In these circumstances:
3725 . Two addresses routed to the same list of hosts;
3726 . First host does not offer TLS;
3727 . First host accepts first address;
3728 . First host gives temporary error to second address;
3729 . Second host offers TLS and a TLS session is established;
3730 . Second host accepts second address.
3731 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3732 and peerdn, if requested) that were in fact used only for the second
3735 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3736 attention to any local part prefix or suffix that was matched by the
3737 router that accepted the address. It now behaves in the same way as it
3738 does for delivery: the affixes are removed from the local part unless
3739 rcpt_include_affixes is set on the transport.
3741 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3742 timeout during the DATA phase of an incoming message.
3744 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3745 applied a patch sanctioned by the Sieve maintainer.
3747 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3748 the uid or gid is negative. A case of a negative gid caused this to be
3749 noticed. The fix allows for either to be negative.
3751 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3752 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3755 PH/36 The change PH/12 above was broken. Fixed it.
3757 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3758 the grounds that routing the same address twice would always produce the
3759 same answer. This might have been true once, but it is certainly no
3760 longer true now. Routing a child address may depend on the previous
3761 routing that produced that child. Some complicated redirection strategies
3762 went wrong when messages had multiple recipients, and made Exim's
3763 behaviour dependent on the order in which the addresses were given.
3765 I have moved the duplicate checking until after the routing is complete.
3766 Exim scans the addresses that are assigned to local and remote
3767 transports, and removes any duplicates. This means that more work will be
3768 done, as duplicates will always all be routed, but duplicates are
3769 presumably rare, so I don't expect this is of any significance.
3771 For deliveries to pipes, files, and autoreplies, the duplicate checking
3772 still happens during the routing process, since they are not going to be
3775 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3776 It corrects a timeout issue with spamd. This is Ian's comment: "The
3777 background is that sometimes spamd either never reads data from a
3778 connection it has accepted, or it never writes response data. The exiscan
3779 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3780 blindly assumes that writes won't block so it may never time out."
3782 PH/39 Allow G after quota size as well as K and M.
3784 PH/40 The value set for $authenticated_id in an authenticator may not contain
3785 binary zeroes or newlines because the value is written to log lines and
3786 to spool files. There was no check on this. Now the value is run through
3787 the string_printing() function so that such characters are converted to
3788 printable escape sequences.
3790 PH/41 $message_linecount is a new variable that contains the total number of
3791 lines in the message. Compare $body_linecount, which is the count for the
3794 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3795 bounce and delay warning messages, except in certain special cases, which
3798 (a) An SMTP error message from a remote host;
3799 (b) A message specified in a :fail: redirection;
3800 (c) A message specified in a "fail" command in a system filter;
3801 (d) A message specified in a FAIL return from the queryprogram router;
3802 (e) A message specified by the cannot_route_message router option.
3804 In these cases only, Exim does include the error details in bounce and
3805 warning messages. There are also a few cases where bland messages such
3806 as "unrouteable address" or "local delivery error" are given.
3808 PH/43 $value is now also set for the "else" part of a ${run expansion.
3810 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3811 being worked on, but at least Exim now implements the latest version to
3814 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3815 process to complete was treated as a delivery failure, a timeout while
3816 writing the message to the pipe was logged, but erroneously treated as a
3817 successful delivery. Such timeouts include transport filter timeouts. For
3818 consistency with the overall process timeout, these timeouts are now
3819 treated as errors, giving rise to delivery failures by default. However,
3820 there is now a new Boolean option for the pipe transport called
3821 timeout_defer, which, if set TRUE, converts the failures into defers for
3822 both kinds of timeout. A transport filter timeout is now identified in
3825 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3826 systems where "make" and "gmake" are different, calling "gmake" at top
3827 level broke things. I've arranged for the value of $(MAKE) to be passed
3828 from the Makefile to this script so that it can call the same version of
3832 A note about Exim versions 4.44 and 4.50
3833 ----------------------------------------
3835 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3836 changes of various kinds. As a consequence, a big documentation update was
3837 needed. This delayed the release for rather longer than seemed good, especially
3838 in the light of a couple of (minor) security issues. Therefore, the changes
3839 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3840 release. So 4.44 and 4.50 are in effect two different branches that both start
3843 I have left the 4.50 change log unchanged; it contains all the changes since
3844 4.43. The change log for 4.44 is below; many of its items are identical to
3845 those for 4.50. This seems to be the most sensible way to preserve the
3846 historical information.
3852 1. Minor wording change to the doc/README.SIEVE file.
3854 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3855 computation of the current number of files was incorrect.
3857 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3858 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3861 4. Give more explanation in the error message when the command for a transport
3862 filter fails to execute.
3864 5. There are several places where Exim runs a non-Exim command in a
3865 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3866 was being done only for the command run by the queryprogram router. It is
3867 now done for all such subprocesses. The other cases are: ${run, transport
3868 filters, and the commands run by the lmtp and pipe transports.
3870 6. Added CONFIGURE_GROUP build-time option.
3872 7. Some older OS have a limit of 256 on the maximum number of file
3873 descriptors. Exim was using setrlimit() to set 1000 as a large value
3874 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3875 systems. I've change it so that if it can't get 1000, it tries for 256.
3877 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3878 was an oversight, and furthermore, ever since the addition of extra
3879 controls (e.g. 4.43/32), the checks on when to allow different forms of
3880 "control" were broken. There should now be diagnostics for all cases when a
3881 control that does not make sense is encountered.
3883 9. Added the /retain_sender option to "control=submission".
3885 10. $recipients is now available in the predata ACL (oversight).
3887 11. Tidy the search cache before the fork to do a delivery from a message
3888 received from the command line. Otherwise the child will trigger a lookup
3889 failure and thereby defer the delivery if it tries to use (for example) a
3890 cached ldap connection that the parent has called unbind on.
3892 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3893 of $address_data from the recipient verification was clobbered by the
3894 sender verification.
3896 13. The value of address_data from a sender verification is now available in
3897 $sender_address_data in subsequent conditions in the ACL statement.
3899 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3901 15. Added a new option "connect=<time>" to callout options, to set a different
3904 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3905 was its contents. (It was OK if the option was not defined at all.)
3907 17. A "Completed" log line is now written for messages that are removed from
3908 the spool by the -Mrm option.
3910 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3911 information about exactly what failed.
3913 19. Added -dd to debug only the daemon process.
3915 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3916 handles timeouts, both on the server side and network timeouts. Renamed the
3917 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3919 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3920 It is now set to "smtps".
3922 22. $host_address is now set to the target address during the checking of
3923 ignore_target_hosts.
3925 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3926 being passed; this would have caused $sender_host_name to have been used if
3927 matching the list had actually called for a host name (not very likely,
3928 since this list is usually IP addresses). A host name is now passed as
3931 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3932 code that specifies a non-restarting handler (typically sigaction() in
3933 modern systems) in an attempt to fix a rare and obscure crash bug.
3935 25. Narrowed the window for a race in the daemon that could cause it to ignore
3936 SIGCHLD signals. This is not a major problem, because they are used only to
3937 wake it up if nothing else does.
3939 26. A malformed maildirsize file could cause Exim to calculate negative values
3940 for the mailbox size or file count. Odd effects could occur as a result.
3941 The maildirsize information is now recalculated if the size or filecount
3944 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3945 support for a long time. Removed HAVE_SYS_VFS_H.
3947 28. Installed the latest version of exipick from John Jetmore.
3949 29. In an address list, if the pattern was not a regular expression, an empty
3950 subject address (from a bounce message) matched only if the pattern was an
3951 empty string. Non-empty patterns were not even tested. This was the wrong
3952 because it is perfectly reasonable to use an empty address as part of a
3953 database query. An empty address is now tested by patterns that are
3954 lookups. However, all the other forms of pattern expect the subject to
3955 contain a local part and a domain, and therefore, for them, an empty
3956 address still always fails if the pattern is not itself empty.
3958 30. Exim went into a mad DNS loop when attempting to do a callout where the
3959 host was specified on an smtp transport, and looking it up yielded more
3960 than one IP address.
3962 31. Re-factored the code for checking spool and log partition space into a
3963 function that finds that data and another that does the check. The former
3964 is then used to implement four new variables: $spool_space, $log_space,
3965 $spool_inodes, and $log_inodes.
3967 32. The RFC2047 encoding function was originally intended for short strings
3968 such as real names; it was not keeping to the 75-character limit for
3969 encoded words that the RFC imposes. It now respects the limit, and
3970 generates multiple encoded words if necessary. To be on the safe side, I
3971 have increased the buffer size for the ${rfc2047: expansion operator from
3974 33. It is now permitted to omit both strings after an "if" condition; if the
3975 condition is true, the result is "true". As before, when the second string
3976 is omitted, a false condition yields an empty string. This makes it less
3977 cumbersome to write custom ACL and router conditions.
3979 34. Failure to deliver a bounce message always caused it to be frozen, even if
3980 there was an errors_to setting on the router. The errors_to setting is now
3983 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3984 canonical form (fully expanded) before being placed in
3985 $sender_host_address.
3987 36. The table in the code that translates DNS record types into text (T_A to
3988 "A" for instance) was missing entries for NS and CNAME. It is just possible
3989 that this could have caused confusion if both these types were looked up
3990 for the same domain, because the text type is used as part of Exim's
3991 per-process caching. But the chance of anyone hitting this buglet seems
3994 37. The dnsdb lookup has been extended in a number of ways.
3996 (1) There is a new type, "zns", which walks up the domain tree until it
3997 finds some nameserver records. It should be used with care.
3999 (2) There is a new type, "mxh", which is like "mx" except that it returns
4000 just the host names, not the priorities.
4002 (3) It is now possible to give a list of domains (or IP addresses) to be
4003 looked up. The behaviour when one of the lookups defers can be
4004 controlled by a keyword.
4006 (4) It is now possible to specify the separator character for use when
4007 multiple records are returned.
4009 38. The dnslists ACL condition has been extended: it is now possible to supply
4010 a list of IP addresses and/or domains to be looked up in a particular DNS
4013 39. Added log_selector=+queue_time_overall.
4015 40. When running the queue in the test harness, wait just a tad after forking a
4016 delivery process, to get repeatability of debugging output.
4018 41. Include certificate and key file names in error message when GnuTLS fails
4019 to set them up, because the GnuTLS error message doesn't include the name
4020 of the failing file when there is a problem reading it.
4022 42. Allow both -bf and -bF in the same test run.
4024 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4026 44. The "Exiscan patch" is now merged into the mainline Exim source.
4028 45. Sometimes the final signoff response after QUIT could fail to get
4029 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4030 < 0 before doing a fflush(). This bug looks as though it goes back to the
4031 introduction of TLS in release 3.20, but "sometimes" must have been rare
4032 because the tests only now provoked it.
4034 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4035 (this can affect the format of dates).
4037 47. exim_tidydb, when checking for the continued existence of a message for
4038 which it has found a message-specific retry record, was not finding
4039 messages that were in split spool directories. Consequently, it was
4040 deleting retry records that should have stayed in existence.
4042 48. Steve fixed some bugs in eximstats.
4044 49. The SPA authentication driver was not abandoning authentication and moving
4045 on to the next authenticator when an expansion was forced to fail,
4046 contradicting the general specification for all authenticators. Instead it
4047 was generating a temporary error. It now behaves as specified.
4049 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4050 (the order specifies the preference for clients). The order is now AES256,
4051 AES128, 3DES, ARCFOUR128.
4053 51. Small patch to Sieve code - explicitly set From: when generating an
4056 52. Exim crashed if a remote delivery caused a very long error message to be
4057 recorded - for instance if somebody sent an entire SpamAssassin report back
4058 as a large number of 550 error lines. This bug was coincidentally fixed by
4059 increasing the size of one of Exim's internal buffers (big_buffer) that
4060 happened as part of the Exiscan merge. However, to be on the safe side, I
4061 have made the code more robust (and fixed the comments that describe what
4064 53. Now that there can be additional text after "Completed" in log lines (if
4065 the queue_time_overall log selector is set), a one-byte patch to exigrep
4066 was needed to allow it to recognize "Completed" as not the last thing in
4069 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4070 patch that reportedly fixes this has been added. I am not expert enough to
4071 create a test for it. This is what the patch creator wrote:
4073 "I found a little strange behaviour of ldap code when working with
4074 Windows 2003 AD Domain, where users was placed in more than one
4075 Organization Units. When I tried to give exim partial DN, the exit code
4076 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4077 But simultaneously result of request was absolutely normal ldap result,
4078 so I produce this patch..."
4080 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4081 REFERENCE, so I have modified the code to exclude the patch when that macro
4084 55. Some experimental protocols are using DNS PTR records for new purposes. The
4085 keys for these records are domain names, not reversed IP addresses. The
4086 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4087 leaves it alone. Component reversal etc. now happens only for IP addresses.
4090 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4092 57. Double the size of the debug message buffer (to 2048) so that more of very
4093 long debug lines gets shown.
4095 58. The exicyclog utility now does better if the number of log files to keep
4096 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4098 59. Two changes related to the smtp_active_hostname option:
4100 (1) $smtp_active_hostname is now available as a variable.
4101 (2) The default for smtp_banner uses $smtp_active_hostname instead
4102 of $primary_hostname.
4104 60. The host_aton() function is supposed to be passed a string that is known
4105 to be a valid IP address. However, in the case of IPv6 addresses, it was
4106 not checking this. This is a hostage to fortune. Exim now panics and dies
4107 if the condition is not met. A case was found where this could be provoked
4108 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4109 components; fortuitously, this particular loophole had already been fixed
4110 by change 4.50/55 above.
4112 If there are any other similar loopholes, the new check in host_aton()
4113 itself should stop them being exploited. The report I received stated that
4114 data on the command line could provoke the exploit when Exim was running as
4115 exim, but did not say which command line option was involved. All I could
4116 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4117 running as the user.
4120 61. There was a buffer overflow vulnerability in the SPA authentication code
4121 (which came originally from the Samba project). I have added a test to the
4122 spa_base64_to_bits() function which I hope fixes it.
4125 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4126 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4128 63. The daemon start-up calls getloadavg() while still root for those OS that
4129 need the first call to be done as root, but it missed one case: when
4130 deliver_queue_load_max is set with deliver_drop_privilege. This is
4131 necessary for the benefit of the queue runner, because there is no re-exec
4132 when deliver_drop_privilege is set.
4134 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4135 This has been fixed.
4137 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4138 was in use, was not putting the data itself into the right store pool;
4139 consequently, it could be overwritten for a subsequent message in the same
4140 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4143 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4145 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4146 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4147 the function were treating the return as a boolean value, which happened to
4148 work because 0=false and not-0=true, but is not correct code.
4150 68. The host_aton() function was not handling scoped IPv6 addresses (those
4151 with, for example, "%eth0" on the end) correctly.
4153 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4154 negated items (that is, ~something) in unsigned ints. Some compilers
4155 apparently mutter when there is no cast.
4157 70. If an address verification called from an ACL failed, and did not produce a
4158 user-specific message (i.e. there was only a "system" message), nothing was
4159 put in $acl_verify_message. In this situation, it now puts the system
4162 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4163 session; change 4.31/43 added the unwanted input to the log line - except
4164 that it did not do this in the start of session case. It now does.
4166 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4167 This is wrong and can cause the other end to generate a synchronization
4168 error if it is another Exim or anything else that does the synchronization
4169 check. A QUIT command is no longer sent after a timeout.
4171 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4172 during host lookups.
4174 74. The defer_ok option of callout verification was not working if it was used
4175 when verifying addresses in header lines, that is, for this case:
4177 verify = header_sender/callout=defer_ok
4179 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4180 those file descriptors could be used for SMTP connections. If anything
4181 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4182 could be sent to the SMTP client, causing chaos. The daemon now opens
4183 stdin, stdout, and stderr to /dev/null when it puts itself into the
4186 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4187 log by default. The user can override this with suitable Perl magic.
4189 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4190 the log message when discard triggers, was not working for the DATA ACL or
4191 for the non-SMTP ACL.
4193 78. Error message wording change in sieve.c.
4195 79. If smtp_accept_max_per_host was set, the number of connections could be
4196 restricted to fewer than expected, because the daemon was trying to set up
4197 a new connection before checking whether the processes handling previous
4198 connections had finished. The check for completed processes is now done
4199 earlier. On busy systems, this bug wouldn't be noticed because something
4200 else would have woken the daemon, and it would have reaped the completed
4203 80. If a message was submitted locally by a user whose login name contained one
4204 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4205 It caused a spool format error. I have fixed the spool reading code. A
4206 related problem was that the "from" clause in the Received: line became
4207 illegal because of the space(s). It is now covered by ${quote_local_part.
4209 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4212 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4214 83. Patch from Sieve maintainer to fix unterminated string problem in
4215 "vacation" handling.
4217 84. Some minor changes to the Linux configuration files to help with other
4218 OS variants using glibc.
4220 85. One more patch for Sieve to update vacation handling to latest spec.
4223 ----------------------------------------------------
4224 See the note above about the 4.44 and 4.50 releases.
4225 ----------------------------------------------------
4231 1. Change 4.43/35 introduced a bug that caused file counts to be
4232 incorrectly computed when quota_filecount was set in an appendfile
4235 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4236 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4239 3. Give more explanation in the error message when the command for a transport
4240 filter fails to execute.
4242 4. There are several places where Exim runs a non-Exim command in a
4243 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4244 was being done only for the command run by the queryprogram router. It is
4245 now done for all such subprocesses. The other cases are: ${run, transport
4246 filters, and the commands run by the lmtp and pipe transports.
4248 5. Some older OS have a limit of 256 on the maximum number of file
4249 descriptors. Exim was using setrlimit() to set 1000 as a large value
4250 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4251 systems. I've change it so that if it can't get 1000, it tries for 256.
4253 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4254 was an oversight, and furthermore, ever since the addition of extra
4255 controls (e.g. 4.43/32), the checks on when to allow different forms of
4256 "control" were broken. There should now be diagnostics for all cases when a
4257 control that does not make sense is encountered.
4259 7. $recipients is now available in the predata ACL (oversight).
4261 8. Tidy the search cache before the fork to do a delivery from a message
4262 received from the command line. Otherwise the child will trigger a lookup
4263 failure and thereby defer the delivery if it tries to use (for example) a
4264 cached ldap connection that the parent has called unbind on.
4266 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4267 of $address_data from the recipient verification was clobbered by the
4268 sender verification.
4270 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4271 was its contents. (It was OK if the option was not defined at all.)
4273 11. A "Completed" log line is now written for messages that are removed from
4274 the spool by the -Mrm option.
4276 12. $host_address is now set to the target address during the checking of
4277 ignore_target_hosts.
4279 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4280 being passed; this would have caused $sender_host_name to have been used if
4281 matching the list had actually called for a host name (not very likely,
4282 since this list is usually IP addresses). A host name is now passed as
4285 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4286 code that specifies a non-restarting handler (typically sigaction() in
4287 modern systems) in an attempt to fix a rare and obscure crash bug.
4289 15. Narrowed the window for a race in the daemon that could cause it to ignore
4290 SIGCHLD signals. This is not a major problem, because they are used only to
4291 wake it up if nothing else does.
4293 16. A malformed maildirsize file could cause Exim to calculate negative values
4294 for the mailbox size or file count. Odd effects could occur as a result.
4295 The maildirsize information is now recalculated if the size or filecount
4298 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4299 support for a long time. Removed HAVE_SYS_VFS_H.
4301 18. Updated exipick to current release from John Jetmore.
4303 19. Allow an empty sender to be matched against a lookup in an address list.
4304 Previously the only cases considered were a regular expression, or an
4307 20. Exim went into a mad DNS lookup loop when doing a callout where the
4308 host was specified on the transport, if the DNS lookup yielded more than
4311 21. The RFC2047 encoding function was originally intended for short strings
4312 such as real names; it was not keeping to the 75-character limit for
4313 encoded words that the RFC imposes. It now respects the limit, and
4314 generates multiple encoded words if necessary. To be on the safe side, I
4315 have increased the buffer size for the ${rfc2047: expansion operator from
4318 22. Failure to deliver a bounce message always caused it to be frozen, even if
4319 there was an errors_to setting on the router. The errors_to setting is now
4322 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4323 canonical form (fully expanded) before being placed in
4324 $sender_host_address.
4326 24. Updated eximstats to version 1.33
4328 25. Include certificate and key file names in error message when GnuTLS fails
4329 to set them up, because the GnuTLS error message doesn't include the name
4330 of the failing file when there is a problem reading it.
4332 26. Expand error message when OpenSSL has problems setting up cert/key files.
4335 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4336 (this can affect the format of dates).
4338 28. exim_tidydb, when checking for the continued existence of a message for
4339 which it has found a message-specific retry record, was not finding
4340 messages that were in split spool directories. Consequently, it was
4341 deleting retry records that should have stayed in existence.
4343 29. eximstats updated to version 1.35
4344 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4345 1.35 - bugfix such that pie charts by volume are generated correctly
4347 30. The SPA authentication driver was not abandoning authentication and moving
4348 on to the next authenticator when an expansion was forced to fail,
4349 contradicting the general specification for all authenticators. Instead it
4350 was generating a temporary error. It now behaves as specified.
4352 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4353 (the order specifies the preference for clients). The order is now AES256,
4354 AES128, 3DES, ARCFOUR128.
4356 31. Small patch to Sieve code - explicitly set From: when generating an
4359 32. Exim crashed if a remote delivery caused a very long error message to be
4360 recorded - for instance if somebody sent an entire SpamAssassin report back
4361 as a large number of 550 error lines. This bug was coincidentally fixed by
4362 increasing the size of one of Exim's internal buffers (big_buffer) that
4363 happened as part of the Exiscan merge. However, to be on the safe side, I
4364 have made the code more robust (and fixed the comments that describe what
4367 33. Some experimental protocols are using DNS PTR records for new purposes. The
4368 keys for these records are domain names, not reversed IP addresses. The
4369 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4370 leaves it alone. Component reversal etc. now happens only for IP addresses.
4373 34. The host_aton() function is supposed to be passed a string that is known
4374 to be a valid IP address. However, in the case of IPv6 addresses, it was
4375 not checking this. This is a hostage to fortune. Exim now panics and dies
4376 if the condition is not met. A case was found where this could be provoked
4377 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4378 components; fortuitously, this particular loophole had already been fixed
4379 by change 4.50/55 or 4.44/33 above.
4381 If there are any other similar loopholes, the new check in host_aton()
4382 itself should stop them being exploited. The report I received stated that
4383 data on the command line could provoke the exploit when Exim was running as
4384 exim, but did not say which command line option was involved. All I could
4385 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4386 running as the user.
4389 35. There was a buffer overflow vulnerability in the SPA authentication code
4390 (which came originally from the Samba project). I have added a test to the
4391 spa_base64_to_bits() function which I hope fixes it.
4394 36. The daemon start-up calls getloadavg() while still root for those OS that
4395 need the first call to be done as root, but it missed one case: when
4396 deliver_queue_load_max is set with deliver_drop_privilege. This is
4397 necessary for the benefit of the queue runner, because there is no re-exec
4398 when deliver_drop_privilege is set.
4400 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4401 was in use, was not putting the data itself into the right store pool;
4402 consequently, it could be overwritten for a subsequent message in the same
4403 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4406 38. Sometimes the final signoff response after QUIT could fail to get
4407 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4408 < 0 before doing a fflush(). This bug looks as though it goes back to the
4409 introduction of TLS in release 3.20, but "sometimes" must have been rare
4410 because the tests only now provoked it.
4416 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4417 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4418 Now it returns an integer. A number of calls were still expecting a T/F
4419 return. Fortuitously, in all cases, the tests worked in OK situations,
4420 which is the norm. However, things would have gone wrong on any write
4421 failures on the smtp file descriptor. This function is used when sending
4422 messages over SMTP and also when doing verify callouts.
4424 2. When Exim is called to do synchronous delivery of a locally submitted
4425 message (the -odf or -odi options), it no longer closes stderr before doing
4428 3. Implemented the mua_wrapper option.
4430 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4432 5. Implemented the functions header_remove(), header_testname(),
4433 header_add_at_position(), and receive_remove_recipient(), and exported them
4436 6. If an ACL "warn" statement specified the addition of headers, Exim already
4437 inserted X-ACL-Warn: at the start if there was no header name. However, it
4438 was not making this test for the second and subsequent header lines if
4439 there were newlines in the string. This meant that an invalid header could
4440 be inserted if Exim was badly configured.
4442 7. Allow an ACL "warn" statement to add header lines at the start or after all
4443 the Received: headers, as well as at the end.
4445 8. Added the rcpt_4xx retry error code.
4447 9. Added postmaster_mailfrom=xxx to callout verification option.
4449 10. Added mailfrom=xxxx to the callout verification option, for verify=
4452 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4453 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4455 12. Inserted some casts to stop certain compilers warning when using pointer
4456 differences as field lengths or precisions in printf-type calls (mostly
4457 affecting debugging statements).
4459 13. Added optional readline() support for -be (dynamically loaded).
4461 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4462 same clock tick as a message's arrival, so that its received time was the
4463 same as the "first fail" time on the retry record, and that message
4464 remained on the queue past the ultimate address timeout, every queue runner
4465 would try a delivery (because it was past the ultimate address timeout) but
4466 after another failure, the ultimate address timeout, which should have then
4467 bounced the address, did not kick in. This was a "< instead of <=" error;
4468 in most cases the first failure would have been in the next clock tick
4469 after the received time, and all would be well.
4471 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4472 being recognized when the domain list was tested by the match_domain
4473 condition in an expansion string.
4475 16. Added the ${str2b64: operator.
4477 17. Exim was always calling setrlimit() to set a large limit for the number of
4478 processes, without checking whether the existing limit was already
4479 adequate. (It did check for the limit on file descriptors.) Furthermore,
4480 errors from getrlimit() and setrlimit() were being ignored. Now they are
4481 logged to the main and panic logs, but Exim does carry on, to try to do its
4482 job under whatever limits there are.
4484 18. Imported PCRE 5.0.
4486 19. Trivial typo in log message " temporarily refused connection" (the leading
4489 20. If the log selector return_path_on_delivery was set and an address was
4490 redirected to /dev/null, the delivery process crashed because it assumed
4491 that a return path would always be set for a "successful" delivery. In this
4492 case, the whole delivery is bypassed as an optimization, and therefore no
4495 21. Internal re-arrangement: the function for sending a challenge and reading
4496 a response while authentication was assuming a zero-terminated challenge
4497 string. It's now changed to take a pointer and a length, to allow for
4498 binary data in such strings.
4500 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4502 23. Exim was not respecting finduser_retries when seeking the login of the
4503 uid under which it was called; it was always trying 10 times. (The default
4504 setting of finduser_retries is zero.) Also, it was sleeping after the final
4505 failure, which is pointless.
4507 24. Implemented tls_on_connect_ports.
4509 25. Implemented acl_smtp_predata.
4511 26. If the domain in control=submission is set empty, Exim assumes that the
4512 authenticated id is a complete email address when it generates From: or
4513 Sender: header lines.
4515 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4516 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4517 chown and chgrp in /bin and hostname in /usr/bin.
4519 28. Exim was keeping the "process log" file open after each use, just as it
4520 does for the main log. This opens the possibility of it remaining open for
4521 long periods when the USR1 signal hits a daemon. Occasional processlog
4522 errors were reported, that could have been caused by this. Anyway, it seems
4523 much more sensible not to leave this file open at all, so that is what now
4526 29. The long-running daemon process does not normally write to the log once it
4527 has entered its main loop, and it closes the log before doing so. This is
4528 so that log files can straightforwardly be renamed and moved. However,
4529 there are a couple of unusual error situations where the daemon does write
4530 log entries, and I had neglected to close the log afterwards.
4532 30. The text of an SMTP error response that was received during a remote
4533 delivery was being truncated at 512 bytes. This is too short for some of
4534 the long messages that one sometimes sees. I've increased the limit to
4537 31. It is now possible to make retry rules that apply only when a message has a
4538 specific sender, in particular, an empty sender.
4540 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4541 it possible to be selective about when SMTP synchronization is enforced.
4543 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4545 32. Implemented hosts_connection_nolog.
4547 33. Added an ACL for QUIT.
4549 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4552 35. Added mailbox_size and mailbox_filecount to appendfile.
4554 36. Added control = no_multiline_responses to ACLs.
4556 37. There was a bug in the logic of the code that waits for the clock to tick
4557 in the case where the clock went backwards by a substantial amount such
4558 that the microsecond fraction of "now" was more than the microsecond
4559 fraction of "then" (but the whole seconds number was less).
4561 38. Added support for the libradius Radius client library this is found on
4562 FreeBSD (previously only the radiusclient library was supported).
4568 1. When certain lookups returned multiple values in the form name=value, the
4569 quoting of the values was not always being done properly. Specifically:
4570 (a) If the value started with a double quote, but contained no whitespace,
4572 (b) If the value contained whitespace other than a space character (i.e.
4573 tabs or newlines or carriage returns) it was not quoted.
4574 This fix has been applied to the mysql and pgsql lookups by writing a
4575 separate quoting function and calling it from the lookup code. The fix
4576 should probably also be applied to nisplus, ibase and oracle lookups, but
4577 since I cannot test any of those, I have not disturbed their existing code.
4579 2. A hit in the callout cache for a specific address caused a log line with no
4580 reason for rejecting RCPT. Now it says "Previous (cached) callout
4581 verification failure".
4583 3. There was an off-by-one bug in the queryprogram router. An over-long
4584 return line was truncated at 256 instead of 255 characters, thereby
4585 overflowing its buffer with the terminating zero. As well as fixing this, I
4586 have increased the buffer size to 1024 (and made a note to document this).
4588 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4589 when Exim is waiting for an SMTP response from a remote server, Exim
4590 restarts its select() call on the socket, thereby resetting its timeout.
4591 This is not a problem when such interrupts are rare. Somebody set up a cron
4592 job to run exiwhat every 2 minutes, which is less than the normal select()
4593 timeout (5 or 10 minutes). This meant that the select() timeout never
4594 kicked in because it was always reset. I have fixed this by comparing the
4595 time when an interrupt arrives with the time at the start of the first call
4596 to select(). If more time than the timeout has elapsed, the interrupt is
4597 treated as a timeout.
4599 5. Some internal re-factoring in preparation for the addition of Sieve
4600 extensions (by MH). In particular, the "personal" test is moved to a
4601 separate function, and given an option for scanning Cc: and Bcc: (which is
4602 not set for Exim filters).
4604 6. When Exim created an email address using the login of the caller as the
4605 local part (e.g. when creating a From: or Sender: header line), it was not
4606 quoting the local part when it contained special characters such as @.
4608 7. Installed new OpenBSD configuration files.
4610 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4611 try to make them clearer.
4613 9. Callout options, other than the timeout value, were being ignored when
4614 verifying sender addresses in header lines. For example, when using
4616 verify = header_sender/callout=no_cache
4618 the cache was (incorrectly) being used.
4620 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4621 only the Cygwin environment.
4623 11. When return_path_on_delivery was set as a log selector, if different remote
4624 addresses in the same message used different return paths and parallel
4625 remote delivery occurred, the wrong values would sometimes be logged.
4626 (Whenever a remote delivery process finished, the return path value from
4627 the most recently started remote delivery process was logged.)
4629 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4630 lines when AUTH and/or TLS are in use. This is the "received protocol"
4631 field. Exim used to use "asmtp" for authenticated SMTP, without any
4632 indication (in the protocol name) for TLS use. Now it follows the RFC and
4633 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4634 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4635 names appear in log lines as well as in Received: header lines.
4637 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4638 extensions, and comparison tests, and to fix some bugs.
4640 14. Changes to the "personal" filter test:
4642 (1) The test was buggy in that it was just doing the equivalent of
4643 "contains" tests on header lines. For example, if a user's address was
4644 anne@some.where, the "personal" test would incorrectly be true for
4646 To: susanne@some.where
4648 This test is now done by extracting each address from the header in turn,
4649 and checking the entire address. Other tests that are part of "personal"
4650 are now done using regular expressions (for example, to check local parts
4651 of addresses in From: header lines).
4653 (2) The list of non-personal local parts in From: addresses has been
4654 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4655 taken from the Sieve specification recommendations.
4657 (3) If the message contains any header line starting with "List-" it is
4658 treated as non-personal.
4660 (4) The test for "circular" in the Subject: header line has been removed
4661 because it now seems ill-conceived.
4663 15. Minor typos in src/EDITME comments corrected.
4665 16. Installed latest exipick from John Jetmore.
4667 17. If headers_add on a router specified a text string that was too long for
4668 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4669 of string_sprintf() is now avoided.
4671 18. $message_body_size was not set (it was always zero) when running the DATA
4672 ACL and the local_scan() function.
4674 19. For the "mail" command in an Exim filter, no default was being set for
4675 the once_repeat time, causing a random time value to be used if "once" was
4676 specified. (If the value happened to be <= 0, no repeat happened.) The
4677 default is now 0s, meaning "never repeat". The "vacation" command was OK
4678 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4679 (I found it when inspecting the code).
4681 20. There is now an overall timeout for performing a callout verification. It
4682 defaults to 4 times the callout timeout, which applies to individual SMTP
4683 commands during the callout. The overall timeout applies when there is more
4684 than one host that can be tried. The timeout is checked before trying the
4685 next host. This prevents very long delays if there are a large number of
4686 hosts and all are timing out (e.g. when the network connections are timing
4687 out). The value of the overall timeout can be changed by specifying an
4688 additional sub-option for "callout", called "maxwait". For example:
4690 verify = sender/callout=5s,maxwait=20s
4692 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4693 to the end before writing, but this should make it even safer).
4695 22. Exim was forgetting that it had advertised PIPELINING for the second and
4696 subsequent messages on an SMTP connection. It was also not resetting its
4697 memory on STARTTLS and an internal HELO.
4699 23. When Exim logs an SMTP synchronization error within a session, it now
4700 records whether PIPELINING has been advertised or not.
4702 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4703 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4704 rather than long int.
4706 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4708 26. Added the never_mail option to autoreply.
4714 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4715 crash if the getsockname() call failed; this can happen if a connection is
4716 closed very soon after it is established. The problem was simply in the
4717 order in which certain operations were done, causing Exim to try to write
4718 to the SMTP stream before it had set up the file descriptor. The bug has
4719 been fixed by making things happen in the correct order.
4725 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4726 before the connection was closed, thus losing the rejection response.
4728 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4729 some early Solaris releases, but causes trouble in current releases where
4730 socklen_t is defined.
4732 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4735 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4738 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4739 that releasing the top of it at the end releases what was used for sub-
4740 expansions (unless the block got too big). However, discard this block if
4741 the first thing is a variable or header, so that we can use its block when
4742 it is dynamic (useful for very large $message_headers, for example).
4744 6. Lookups now cache *every* query, not just the most recent. A new, separate
4745 store pool is used for this. It can be recovered when all lookup caches are
4746 flushed. Lookups now release memory at the end of their result strings.
4747 This has involved some general refactoring of the lookup sources.
4749 7. Some code has been added to the store_xxx() functions to reduce the amount
4750 of flapping under certain conditions.
4752 8. log_incoming_interface used to affect only the <= reception log lines. Now
4753 it causes the local interface and port to be added to several more SMTP log
4754 lines, for example "SMTP connection from", and rejection lines.
4756 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4758 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4760 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4761 could be overwritten at the end of the current message (or the start of a
4762 new message if it was set in a HELO ACL). The value is now preserved for
4763 the duration of the SMTP connection.
4765 12. If a transport had a headers_rewrite setting, and a matching header line
4766 contained an unqualified address, that address was qualified, even if it
4767 did not match any rewriting rules. The underlying bug was that the values
4768 of the flags that permit the existence of unqualified sender and recipient
4769 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4770 non-local messages, and by -bnq for local messages) were not being
4771 preserved with the message after it was received.
4773 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4774 "next input=" as part of the text comprising the host identity instead of
4775 the correct text. The code was using the same buffer for two different
4776 strings. However, depending on which order the printing function evaluated
4777 its arguments, the bug did not always show up. Under Linux, for example, my
4778 test suite worked just fine.
4780 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4781 doesn't work with some older versions of Perl. It has been changed to "my",
4782 which in any case is probably the better facility to use.
4784 15. A really picky compiler found some instances of statements for creating
4785 error messages that either had too many or two few arguments for the format
4788 16. The size of the buffer for calls to the DNS resolver has been increased
4789 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4790 for addresses that have a lot of PTR records. This alleviates a problem; it
4791 does not fully solve it.
4793 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4794 buffer now truncates the list and logs the incident, which is the same
4795 action as happens when Exim is looking up a host name and its aliases.
4796 Previously in this situation something unpredictable would happen;
4797 sometimes it was "internal error: store_reset failed".
4799 18. If a server dropped the connection unexpectedly when an Exim client was
4800 using GnuTLS and trying to read a response, the client delivery process
4801 crashed while trying to generate an error log message.
4803 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4804 string, for example:
4806 warn message = H1: something\nH2: something
4808 the text was added as a single header line from Exim's point of view
4809 though it ended up OK in the delivered message. However, searching for the
4810 second and subsequent header lines using $h_h2: did not work. This has been
4811 fixed. Similarly, if a system filter added multiple headers in this way,
4812 the routers could not see them.
4814 20. Expanded the error message when iplsearch is called with an invalid key to
4815 suggest using net-iplsearch in a host list.
4817 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4818 ACLs are no longer actually imposed (and a message to that effect is
4821 22. If a "gecos" field in a passwd entry contained escaped characters, in
4822 particular, if it contained a \" sequence, Exim got it wrong when building
4823 a From: or a Sender: header from that name. A second bug also caused
4824 incorrect handling when an unquoted " was present following a character
4825 that needed quoting.
4827 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4828 was not being matched caselessly.
4830 24. Arranged for all hyphens in the exim.8 source to be escaped with
4833 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4834 properly. Recipient callout cache records were still being keyed to include
4835 the sender, even when use_sender was set false. This led to far more
4836 callouts that were necessary. The sender is no longer included in the key
4837 when use_sender is false.
4839 26. Added "control = submission" modifier to ACLs.
4841 27. Added the ${base62d: operator to decode base 62 numbers.
4843 28. dnsdb lookups can now access SRV records.
4845 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4846 the configuration file.
4848 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4849 (-v) mode. This makes the output for a verbose queue run more intelligible.
4851 31. Added a use_postmaster feature to recipient callouts.
4853 32. Added the $body_zerocount variable, containing the number of binary zero
4854 bytes in the message body.
4856 33. The time of last modification of the "new" subdirectory is now used as the
4857 "mailbox time last read" when there is a quota error for a maildir
4860 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4862 35. Added +ignore_unknown as a special item in host lists.
4864 36. Code for decoding IPv6 addresses in host lists is now included, even if
4865 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4866 address was recognized as an IP address, but was then not correctly decoded
4867 into binary, causing unexpected and incorrect effects when compared with
4874 1. Very minor rewording of debugging text in manualroute to say "list of
4875 hosts" instead of "hostlist".
4877 2. If verify=header_syntax was set, and a header line with an unqualified
4878 address (no domain) and a large number of spaces between the end of the
4879 name and the colon was received, the reception process suffered a buffer
4880 overflow, and (when I tested it) crashed. This was caused by some obsolete
4881 code that should have been removed. The fix is to remove it!
4883 3. When running in the test harness, delay a bit after writing a bounce
4884 message to get a bit more predictability in the log output.
4886 4. Added a call to search_tidyup() just before forking a reception process. In
4887 theory, someone could use a lookup in the expansion of smtp_accept_max_
4888 per_host which, without the tidyup, could leave open a database connection.
4890 5. Added the variables $recipient_data and $sender_data which get set from a
4891 lookup success in an ACL "recipients" or "senders" condition, or a router
4892 "senders" option, similar to $domain_data and $local_part_data.
4894 6. Moved the writing of debug_print from before to after the "senders" test
4897 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4898 problems for message scanning, either using a data ACL, or using
4899 local_scan() because the Received: header was not generated till after they
4900 were called (in order to set the time as the time of reception completion).
4901 I have revised the way this works. The header is now generated after the
4902 body is received, but before the ACL or local_scan() are called. After they
4903 are run, the timestamp in the header is updated.
4909 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4910 before starting a queue runner without re-exec. This happened only when
4911 deliver_drop_privilege was set or when the Exim user was set to root. The
4912 effect of the bug was that timeouts during subsequent deliveries caused
4913 crashes instead of being properly handled. The handler is now left at its
4914 default (and expected) setting.
4916 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4917 message, again when deliver_drop_privilege is set or Exim is run as root.
4918 The bug described in (1) was not present in this case, but the tidying up
4919 of the other signals was missing. I have made the two cases consistent.
4921 3. The ignore_target_hosts setting on a manualroute router was being ignored
4922 for hosts that were looked up using the /MX notation.
4924 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4927 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4928 operated on the sender address. After changing the $sender_address to <>
4929 for the sender address verify, Exim was re-instated it as the original
4930 (before rewriting) address, but remembering that it had rewritten it, so it
4931 wasn't rewriting it again. This bug also had the effect of breaking the
4932 sender address verification caching when the sender address was rewritten.
4934 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4935 This has been changed so that if the ip literal address matches
4936 ignore_target_hosts, the router declines.
4938 7. Added expansion conditions match_domain, match_address, and match_local_
4939 part (NOT match_host).
4941 8. The placeholder for the Received: header didn't have a length field set.
4943 9. Added code to Exim itself and to exim_lock to test for a specific race
4944 condition that could lead to file corruption when using MBX delivery. The
4945 issue is with the lockfile that is created in /tmp. If this file is removed
4946 after a process has opened it but before that process has acquired a lock,
4947 there is the potential for a second process to recreate the file and also
4948 acquire a lock. This could lead to two Exim processes writing to the file
4949 at the same time. The added code performs the same test as UW imapd; it
4950 checks after acquiring the lock that its file descriptor still refers to
4951 the same named file.
4953 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4954 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4957 11. Added the smtp_active_hostname option. If used, this will typically be made
4958 to depend on the incoming interface address. Because $interface_address is
4959 not set up until the daemon has forked a reception process, error responses
4960 that can happen earlier (such as "too many connections") no longer contain
4963 12. If an expansion in a condition on a "warn" statement fails because a lookup
4964 defers, the "warn" statement is abandoned, and the next ACL statement is
4965 processed. Previously this caused the whole ACL to be aborted.
4967 13. Added the iplsearch lookup type.
4969 14. Added ident_timeout as a log selector.
4971 15. Added tls_certificate_verified as a log selector.
4973 16. Added a global option tls_require_ciphers (compare the smtp transport
4974 option of the same name). This controls incoming TLS connections.
4976 17. I finally figured out how to make tls_require_ciphers do a similar thing
4977 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4978 before starting the TLS session.
4980 18. Tabs are now shown as \t in -bP output.
4982 19. If the log selector return_path_on_delivery was set, Exim crashed when
4983 bouncing a message because it had too many Received: header lines.
4985 20. If two routers both had headers_remove settings, and the first one included
4986 a superfluous trailing colon, the final name in the first list and the
4987 first name in the second list were incorrectly joined into one item (with a
4988 colon in the middle).
4994 1. Added -C and -D options to the exinext utility, mainly to make it easier
4995 to include in the automated testing, but these could be helpful when
4996 multiple configurations are in use.
4998 2. The exinext utility was not formatting the output nicely when there was
4999 an alternate port involved in the retry record key, nor when there was a
5000 message id as well (for retries that were specific to a specific message
5001 and a specific host). It was also confused by IPv6 addresses, because of
5002 the additional colons they contain. I have fixed the IPv4 problem, and
5003 patched it up to do a reasonable job for IPv6.
5005 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5006 delivery, the log line now contains "pipelined" if PIPELINING was used.
5008 4. An SMTP transport process used to panic and die if the bind() call to set
5009 an explicit outgoing interface failed. This has been changed; it is now
5010 treated in the same way as a connect() failure.
5012 5. A reference to $sender_host_name in the part of a conditional expansion
5013 that was being skipped was still causing a DNS lookup. This no longer
5016 6. The def: expansion condition was not recognizing references to header lines
5017 that used bh_ and bheader_.
5019 7. Added the _cache feature to named lists.
5021 8. The code for checking quota_filecount in the appendfile transport was
5022 allowing one more file than it should have been.
5024 9. For compatibility with Sendmail, the command line option
5032 and sets the incoming protocol and host name (for trusted callers). The
5033 host name and its colon can be omitted when only the protocol is to be set.
5034 Note the Exim already has two private options, -pd and -ps, that refer to
5035 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5036 "s", but I don't think that's a major issue.
5038 10. A number of refactoring changes to the code, none of which should affect
5041 (a) The number of logging options was getting close to filling up the
5042 32-bit word that was used as a bit map. I have split them into two classes:
5043 those that are passed in the argument to log_write(), and those that are
5044 only ever tested independently outside of that function. These are now in
5045 separate 32-bit words, so there is plenty of room for expansion again.
5046 There is no change in the user interface or the logging behaviour.
5048 (b) When building, for example, log lines, the code previously used a
5049 macro that called string_cat() twice, in order to add two strings. This is
5050 not really sufficiently general. Furthermore, there was one instance where
5051 it was actually wrong because one of the argument was used twice, and in
5052 one call a function was used. (As it happened, calling the function twice
5053 did not affect the overall behaviour.) The macro has been replaced by a
5054 function that can join an arbitrary number of extra strings onto a growing
5057 (c) The code for expansion conditions now uses a table and a binary chop
5058 instead of a serial search (which was left over from when there were very
5059 few conditions). Also, it now recognizes conditions like "pam" even when
5060 the relevant support is not compiled in: a suitably worded error message is
5061 given if an attempt is made to use such a condition.
5063 11. Added ${time_interval:xxxxx}.
5065 12. A bug was causing one of the ddress fields not to be passed back correctly
5066 from remote delivery subprocesses. The field in question was not being
5067 subsequently used, so this caused to problems in practice.
5069 13. Added new log selectors queue_time and deliver_time.
5071 14. Might have fixed a bug in maildirsizefile handling that threw up
5072 "unexpected character" debug warnings, and recalculated the data
5073 unnecessarily. In any case, I expanded the warning message to give more
5076 15. Added the message "Restricted characters in address" to the statements in
5077 the default ACL that block characters like @ and % in local parts.
5079 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5080 Three changes have been made:
5082 (a) There was a serious bug; a negative response to MAIL caused the whole
5083 recipient domain to be cached as invalid, thereby blocking all messages
5084 to all local parts at the same domain, from all senders. This bug has
5085 been fixed. The domain is no longer cached after a negative response to
5086 MAIL if the sender used is not empty.
5088 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5091 (c) A new callout option, "use_sender" has been added for people who want
5092 the modified behaviour.
5098 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5101 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5102 indeed breaks things for older releases.
5104 3. Added additional logging to the case where there is a problem reading data
5105 from a filter that is running in a subprocess using a pipe, in order to
5106 try to track down a specific problem.
5108 4. Testing facility fudge: when running in the test harness and attempting
5109 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5110 getting "No route to host". Convert this to a timeout.
5112 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5115 6. Some OS don't have socklen_t but use size_t instead. This affects the
5116 fifth argument of getsockopt() amongst other things. This is now
5117 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5118 can be set for individual OS. I have set it for SunOS5, OSF1, and
5119 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5120 some earlier ones do not.
5122 7. Change 4.30/15 was not doing the test caselessly.
5124 8. The standard form for an IPv6 address literal was being rejected by address
5125 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5126 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5127 this, as well as the form without the "IPv6" on the front (but only when
5128 address literals are enabled, of course).
5130 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5132 10. Exim crashed if a message with an empty sender address specified by -f
5133 encountered a router with an errors_to setting. This could be provoked only
5134 by a command such as
5138 where an empty string was supplied; "<>" did not hit this bug.
5140 11. Installed PCRE release 4.5.
5142 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5143 remained set. It is now erased.
5145 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5146 times from message ids (which are base 36 rather than the normal 62).
5148 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5149 were being counted as actual protocol errors, and logged if the log
5150 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5151 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5152 and DATA following a set of rejected RCPTs do not count as protocol errors.
5153 In other words, Exim assumes they were pipelined, though this may not
5154 actually be the case. Of course, in all cases the client gets an
5155 appropriate error code.
5157 15. If a lookup fails in an ACL condition, a message about the failure may
5158 be available; it is used if testing the ACL cannot continue, because most
5159 such messages specify what the cause of the deferral is. However, some
5160 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5161 that caused an old message to be retained and used if a later statement
5162 caused a defer, replacing the real cause of the deferral.
5164 16. If an IP address had so many PTR records that the DNS lookup buffer
5165 was not large enough to hold them, Exim could crash while trying to process
5166 the truncated data. It now detects and logs this case.
5168 17. Further to 4.21/58, another change has been made: if (and only if) the
5169 first line of a message (the first header line) ends with CRLF, a bare LF
5170 in a subsequent header line has a space inserted after it, so as not to
5171 terminate the header.
5173 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5174 unnecessarily, used atoi() instead of strtol(), and didn't check the
5175 termination when getting file sizes from file names by regex.
5177 19. Completely re-implemented the support for maildirsize files, in the light
5178 of a number of problems with the previous contributed implementation
5179 (4.30/29). In particular:
5181 . If the quota is zero, the maildirsize file is maintained, but no quota is
5184 . If the maildir directory does not exist, it is created before any attempt
5185 to write a maildirsize file.
5187 . The quota value in the file is just a cache; if the quota is changed in
5188 the transport, the new value overrides.
5190 . A regular expression is available for excluding directories from the
5193 20. The autoreply transport checks the characters in options that define the
5194 message's headers; it allows continued headers, but it was checking with
5195 isspace() after an embedded newline instead of explicitly looking for a
5198 21. If all the "regular" hosts to which an address was routed had passed their
5199 expiry times, and had not reached their retry times, the address was
5200 bounced, even if fallback hosts were defined. Now Exim should go on to try
5203 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5204 equivalent code in the SMTP transport. Some hosts send humungous responses
5205 to HELO/EHLO, more than 1024 it seems.
5207 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5208 gives compiler warnings in some environments. I've now done it "properly",
5211 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5212 (because of problems with the built-in one) was declared to return uschar *
5213 instead of char *, causing compiler failure.
5215 25. Fixed a file descriptor leak when processing alias/forward files.
5217 26. Fixed a minor format string issue in dbfn.c.
5219 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5221 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5222 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5225 29. When a sender address is verified, it is cached, to save repeating the test
5226 when there is more than one recipient in a message. However, when the
5227 verification involves a callout, it is possible for different callout
5228 options to be set for different recipients. It is too complicated to keep
5229 track of this in the cache, so now Exim always runs a verification when a
5230 callout is required, relying on the callout cache for the optimization.
5231 The overhead is duplication of the address routing, but this should not be
5234 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5235 to be verified with callout=postmaster, and the main callout worked but the
5236 postmaster check failed, the verification correctly failed. However, if a
5237 subsequent RCPT command asked for sender verification *without* the
5238 postmaster check, incorrect caching caused this verification also to fail,
5241 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5242 it was not caching the DNS options (qualify_single, search_parents) that
5243 were used when the lookup failed. A subsequent lookup with different
5244 options therefore always gave the same answer, though there were cases
5245 where it should not have. (Example: a "domains = !$mx_any" option on a
5246 dnslookup router: the "domains" option is always processed without any
5247 widening, but the router might have qualify_single set.) Now Exim uses the
5248 cached value only when the same options are set.
5250 32. Added John Jetmore's "exipick" utility to the distribution.
5252 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5253 than a timeout (e.g. a certificate is required, and is not provided), an
5254 Exim server now closes the connection immediately. Previously it waited for
5255 the client to close - but if the client is SSL, it seems that they each
5256 wait for each other, leading to a delay before one of them times out.
5258 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5259 maintained 0.8.x compatibility because I don't think many are using it, and
5260 it is clearly obsolete.
5262 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5265 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5266 client certificate was expired. A simple patch fixes this, though I don't
5267 understand the full logic of why the verify callback is called multiple
5270 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5271 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5272 which causes problems with some clients (such as the Certicom SSL Plus
5273 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5274 disables the coutermeasure allowing Eudora to connect."
5276 38. Exim was not checking that a write() to a log file succeeded. This could
5277 lead to Bad Things if a log got too big, in particular if it hit a file
5278 size limit. Exim now panics and dies if it cannot write to a log file, just
5279 as it does if it cannot open a log file.
5281 39. Modified OS/Makefile-Linux so that it now contains
5283 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5285 The two -D definitions ensure that Exim is compiled with large file
5286 support, which makes it possible to handle log files that are bigger than
5289 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5290 instance) a domain was checked against a named list that involved a lookup,
5291 causing $domain_data to be set, then another domain was checked against the
5292 same list, then the first domain was re-checked, the value of $domain_data
5293 after the final check could be wrong. In particular, if the second check
5294 failed, it could be set empty. This bug probably also applied to
5297 41. The strip_trailing_dot option was not being applied to the address given
5298 with the -f command-line option.
5300 42. The code for reading a message's header from the spool was incrementing
5301 $received_count, but never initializing it. This meant that the value was
5302 incorrect (doubled) while delivering a message in the same process in which
5303 it was received. In the most common configuration of Exim, this never
5304 happens - a fresh exec is done - but it can happen when
5305 deliver_drop_privilege is set.
5307 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5308 it now includes up to 150 characters of the unexpected data in the log
5311 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5312 and building data strings. The size of both of these buffers was 10 000
5313 bytes - far larger than anybody would *ever* want, thought I. Needless to
5314 say, somebody hit the limit. I have increased the maximum line length to
5315 20 000 and the maximum data length of concatenated lines to 100 000. I have
5316 also fixed two bugs, because there was no checking on these buffers. Tsk,
5317 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5318 buffer is too small.
5320 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5321 lsearch lookups. Now it does.
5323 46. When parsing a route_list item in a manualroute router, a fixed-length
5324 buffer was used for the list of hosts. I made this 1024 bytes long,
5325 thinking that nobody would ever have a list of hosts that long. Wrong.
5326 Somebody had a whole pile of complicated expansion conditions, and the
5327 string was silently truncated, leading to an expansion error. It turns out
5328 that it is easier to change to an unlimited length (owing to other changes
5329 that have happened since this code was originally written) than to build
5330 structure for giving a limitation error. The length of the item that
5331 expands into the list of hosts is now unlimited.
5333 47. The lsearch lookup could not handle data where the length of text line was
5334 more than 4095 characters. Such lines were truncated, leading to shortened
5335 data being returned. It should now handle lines of any length.
5337 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5338 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5341 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5342 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5343 no longer happens in comment lines. A list of replacements is now placed
5344 at the head of all of the source files, except those whose only change is
5345 to replace PERL_COMMAND in the very first #! line.
5347 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5348 messages on the queue, with a bottom-up merge sort, using code contributed
5349 by Michael Haardt. This should make operations like -bp somewhat faster on
5350 large queues. It won't affect queue runners, except when queue_run_in_order
5353 51. Installed eximstats 1.31 in the distribution.
5355 52. Added support for SRV lookups to the dnslookup router.
5357 53. If an ACL referred to $message_body or $message_body_end, the value was not
5358 reset for any messages that followed in the same SMTP session.
5360 54. The store-handling optimization for building very long strings was not
5361 differentiating between the different store pools. I don't think this
5362 actually made any difference in practice, but I've tidied it.
5364 55. While running the routers to verify a sender address, $sender_address
5365 was still set to the sender address. This is wrong, because when routing to
5366 send a bounce to the sender, it would be empty. Therefore, I have changed
5367 it so that, while verifying a sender address, $sender_address is set to <>.
5368 (There is no change to what happens when verifying a recipient address.)
5370 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5371 target A or AAAA records (if not already returned) without resetting the
5372 qualify_single or search_parents options of the DNS resolver. These are
5373 inappropriate in this case because the targets of MX and SRV records must
5374 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5375 target that, when qualified, matched something in the local domain. These
5376 two options are now turned off when doing these lookups.
5378 57. It seems that at least some releases of Reiserfs (which does not have the
5379 concept of a fixed number of inodes) returns zero and not -1 for the
5380 number of available inodes. This interacted badly with check_spool_inodes,
5381 which assumed that -1 was the "no such thing" setting. What I have done is
5382 to check that the total number of inodes is greater than zero before doing
5383 the test of how many are available.
5385 58. When a "warn" ACL statement has a log_message modifier, the message is
5386 remembered, and not repeated. This is to avoid a lot of repetition when a
5387 message has many recipients that cause the same warning to be written.
5388 Howewer, Exim was preserving the list of already written lines for an
5389 entire SMTP session, which doesn't seem right. The memory is now reset if a
5390 new message is started.
5392 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5393 debugging output unless log_rewrite was also set.
5395 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5396 of (int)(handle) when we know that handle contains (void *)(-1).
5398 61. The Exim daemon panic-logs an error return when it closes the incoming
5399 connection. However "connection reset by peer" seems to be common, and
5400 isn't really an error worthy of noting specially, so that particular error
5403 62. When Exim is trying to find all the local interfaces, it used to panic and
5404 die if the ioctl to get the interface flags failed. However, it seems that
5405 on at least one OS (Solaris 9) it is possible to have an interface that is
5406 included in the list of interfaces, but for which you get a failure error
5407 for this call. This happens when the interface is not "plumbed" into a
5408 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5409 failure of the "get flags" call assumes that the interface is down.
5411 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5412 makes life easier for people who are doing arithmetic on fields extracted
5413 from dates, where you often get leading zeros that should not be
5414 interpreted as octal.
5416 64. Added qualify_domain to the redirect router, to override the global
5419 65. If a pathologically long header line contained very many addresses (the
5420 report of this problem mentioned 10 000) and each of them was rewritten,
5421 Exim could use up a very large amount of memory. (It kept on making new
5422 copies of the header line as it rewrote, and never released the old ones.)
5423 At the expense of a bit more processing, the header rewriting function has
5424 been changed so that it no longer eats memory in this way.
5426 66. The generation of the Received: header has been moved from the time that a
5427 message starts to be received, to the time that it finishes. The timestamp
5428 in the Received: header should now be very close to that of the <= log
5429 line. There are two side-effects of this change:
5431 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5432 logged header lines no longer include the local Received: line, because
5433 it has not yet been created. The same applies to a copy of the message
5434 that is returned to a non-SMTP sender when a message is rejected.
5436 (b) When a filter file is tested using -bf, no additional Received: header
5437 is added to the test message. After some thought, I decided that this
5440 This change does not affect the value of $received_for. It is still set
5441 after address rewriting, but before local_scan() is called.
5443 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5445 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5446 gave an unhelpful panic error message, and a defer error. I have managed to
5447 change this behaviour so that it now rejects any supplied certificate,
5448 which seems right, as the list of acceptable certificates is empty.
5450 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5451 gave an unhelpful defer error. I have not managed to make this reject any
5452 supplied certificates, but the error message it gives is "no certificate
5453 supplied", which is not helpful.
5455 70. exigrep's output now also includes lines that are not associated with any
5456 message, but which match the given pattern. Implemented by a patch from
5457 Martin Sluka, which also tidied up the Perl a bit.
5459 71. Recipient callout verification, like sender verification, was using <> in
5460 the MAIL FROM command. This isn't really the right thing, since the actual
5461 sender may affect whether the remote host accepts the recipient or not. I
5462 have changed it to use the actual sender in the callout; this means that
5463 the cache record is now keyed on a recipient/sender pair, not just the
5464 recipient address. There doesn't seem to be a real danger of callout loops,
5465 since a callout by the remote host to check the sender would use <>.
5466 [SEE ABOVE: changed after hitting problems.]
5468 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5469 temporary errors. However, in the case of such a code being given after
5470 the end of a data transmission (i.e. after ".") Exim was failing to write
5471 a retry record for the message. (Yes, there was some broken host that was
5472 actually sending 8xx at this point.)
5474 73. An unknown lookup type in a host list could cause Exim to panic-die when
5475 the list was checked. (An example that provoked this was putting <; in the
5476 middle of a list instead of at the start.) If this happened during a DATA
5477 ACL check, a -D file could be left lying around. This kind of configuration
5478 error no longer causes Exim to die; instead it causes a defer errror. The
5479 incident is still logged to the main and panic logs.
5481 74. Buglet left over from Exim 3 conversion. The message "too many messages
5482 in one connection" was written to the rejectlog but not the mainlog, except
5483 when address rewriting (yes!) was being logged.
5485 75. Added write_rejectlog option.
5487 76. When a system filter was run not as root (that is, when system_filter_user
5488 was set), the values of the $n variables were not being returned to the
5489 main process; thus, they were not subsequently available in the $sn
5492 77. Added +return_path_on_delivery log selector.
5494 78. A connection timeout was being treated differently from recipients deferred
5495 when testing hosts_max_try with a message that was older than the host's
5496 retry timeout. (The host should not be counted, thus allowing all hosts to
5497 be tried at least once before bouncing.) This may have been the cause of an
5498 occasionally reported bug whereby a message would remain on the queue
5499 longer than the retry timeout, but would be bounced if a delivery was
5500 forced. I say "may" because I never totally pinned down the problem;
5501 setting up timeout/retry tests is difficult. See also the next item.
5503 79. The ultimate address timeout was not being applied to errors that involved
5504 a combination of host plus message (for example, a timeout on a MAIL
5505 command). When an address resolved to a number of possible hosts, and they
5506 were not all tried for each delivery (e.g. because of hosts_max_try), a
5507 message could remain on the queue longer than the retry timeout.
5509 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5512 81. Fixed an obscure SMTP outgoing bug which required at least the following
5513 conditions: (a) there was another message waiting for the same server;
5514 (b) the server returned 5xx to all RCPT commands in the first message so
5515 that the message was not completed; (c) the server dropped the connection
5516 or gave a negative response to the RSET that Exim sends to abort the
5517 transaction. The observed case was a dropped connection after DATA that had
5518 been sent in pipelining mode. That is, the server had advertised PIPELINING
5519 but was not implementing it correctly. The effect of the bug was incorrect
5520 behaviour, such as trying another host, and this could lead to a crash.
5526 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5527 and daemon.c were passed as pointers to ints; they should have been
5528 pointers to socklen_t variables (which are typically unsigned ints).
5530 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5533 3. Fixed a really odd bug that affected only the testing scheme; patching a
5534 certain fixed string in the binary changed the value of another string that
5535 happened to be identical to the end of the original first string.
5537 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5538 name", it returns that address as the IP address. On some operating
5539 systems (e.g. Solaris), it also passes back the IP address string as the
5540 "host name". However, on others (e.g. Linux), it passes back an empty
5541 string. Exim wasn't checking for this, and was changing the host name to an
5542 empty string, assuming it had been canonicized.
5544 5. Although rare, it is permitted to have more than one PTR record for a given
5545 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5546 all the names associated with an address, because they do in Solaris.
5547 However, it seems that they do not in Linux for data that comes from the
5548 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5549 I found this out when I moved to a new Linux workstation and tried to run
5550 the Exim test suite.
5552 To get round this problem I have changed the code so that it now does its
5553 own call to the DNS to look up PTR records when searching for a host name.
5554 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5555 addresses that are only in /etc/hosts are still found.
5557 This behaviour is, however, controlled by an option called host_lookup_
5558 order, which defaults to "bydns:byaddr". If people want to use the other
5559 order, or indeed, just use one or the other means of lookup, they can
5560 specify it in this variable.
5562 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5563 some operating systems, this comes back from gethostbyaddr() as an empty
5564 string, and this is what Exim used to test for. However, it seems that in
5565 other systems, "." is yielded. Exim now tests for this case too.
5567 7. The values of check_spool_space and check_log_space are now held internally
5568 as a number of kilobytes instead of an absolute number of bytes. If a
5569 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5570 kilobyte. This means that much larger values can be stored.
5572 8. Exim monitor: an attempt to get the action menu when not actually pointing
5573 at a message produces an empty menu entitled "No message selected". This
5574 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5575 no entries in it ("Shell widget menu has zero width and/or height"). So I
5576 have added a single, blank menu entry in this case.
5578 9. Added ${quote_local_part.
5580 10. MIME decoding is now applied to the contents of Subject: header lines when
5583 11. Now that a reference to $sender_host_address automatically causes a reverse
5584 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5585 host lookup before query-style lookups in lists that might use this
5586 variable. This has therefore been abolished, and the "net-" prefix is no
5587 longer necessary for query-style lookups.
5589 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5590 have been a typo for LFLAGS, so it has been changed.
5592 13. The install script calls Exim with "-C /dev/null" in order to find the
5593 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5594 to be output. Howeve, since Exim outputs its version number before the
5595 error, it didn't break the script. It just looked ugly. I fixed this by
5596 always allowing "-C /dev/null" if the caller is root.
5598 14. Ignore overlarge ACL variable number when reading spool file - insurance
5599 against a later release with more variables having written the file.
5601 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5602 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5603 this, as well as the form without the "IPv6" on the front.
5605 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5606 OS/Makefile-Darwin file.
5608 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5609 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5611 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5612 variable. However, some OS use ints for inodes. I've added cast to long int
5613 to get rid of the compiler warning.
5615 19. I had forgotten to lock out "/../" in configuration file names when
5616 ALT_CONFIG_PREFIX was set.
5618 20. Routers used for verification do not need to specify transports. However,
5619 if such a router generated a host list, and callout was configured, Exim
5620 crashed, because it could not find a port number from the (non-existent)
5621 transport. It now assumes port 25 in this circumstance.
5623 21. Added the -t option to exigrep.
5625 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5626 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5627 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5628 them is set, LOOKUP_LSEARCH is forced.
5630 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5632 24. Added sender and host information to the "rejected by local_scan()" log
5633 line; previously there was no indication of these.
5635 25. Added .include_if_exists.
5637 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5638 Linux. It turns out that not all file systems support this. Apparently some
5639 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5640 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5641 not supported on the file descriptor, is now ignored when Exim is trying to
5642 sync a directory. This applies only to Linux.
5644 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5646 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5647 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5650 29. Added support for maildirsize files from supplied patch (modified a bit).
5652 30. The use of :fail: followed by an empty string could lead Exim to respond to
5653 sender verification failures with (e.g.):
5655 550 Verification failed for <xxx>
5656 550 Sender verify failed
5658 where the first response line was missing the '-' that indicates it is not
5659 the final line of the response.
5661 31. The loop for finding the name of the user that called Exim had a hardwired
5662 limit of 10; it now uses the value of finduser_retries, which is used for
5663 all other user lookups.
5665 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5668 33. Exim was neglecting to zero errno before one call of strtol() when
5669 expanding a string and expecting an integer value. On some systems this
5670 resulted in spurious "integer overflow" errors. Also, it was casting the
5671 result into an int without checking.
5673 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5674 did not work. The code looks as if it has *never* worked, though it appears
5675 to have been documented since at least releast 1.62. I have made it work.
5677 35. The "timeout_DNS" error in retry rules, also documented since at least
5678 1.62, also never worked. As it isn't clear exactly what this means, and
5679 clearly it isn't a major issue, I have abolished the feature by treating it
5680 as "timeout", and writing a warning to the main and panic logs.
5682 36. The display of retry rules for -brt wasn't always showing the error code
5685 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5686 timeout_connect_A, timeout_connect_MX.
5688 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5689 to the empty sender.
5691 39. The daemon was not analysing the content of -oX till after it had closed
5692 stderr and disconnected from the controlling terminal. This meant that any
5693 syntax errors were only noted on the panic log, and the return code from
5694 the command was 0. By re-arranging the code a little, I've made the
5695 decoding happen first, so such errors now appear on stderr, and the return
5696 code is 1. However, the actual setting up of the sockets still happens in
5697 the disconnected process, so errors there are still only recorded on the
5700 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5701 connections (as happens on some IP stacks) was logged at start up time as
5702 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5703 it from "IPv6 and IPv4", which means that two separate sockets are being
5706 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5707 says whether AF_INET or AF_INET6 was passed as an argument.
5709 42. Exiwhat output was messed up when time zones were included in log
5712 43. Exiwhat now gives more information about the daemon's listening ports,
5713 and whether -tls-on-connect was used.
5715 44. The "port" option of the smtp transport is now expanded.
5717 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5718 silently ignored. Now an error message is written to the main and panic
5721 46. There's a new ACL modifier called "logwrite" which writes to a log file
5722 as soon as it is encountered.
5724 47. Added $local_user_uid and $local_user_gid at routing time.
5726 48. Exim crashed when trying to verify a sender address that was being
5729 49. Exim was recognizing only a space character after ".include". It now also
5730 recognizes a tab character.
5732 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5733 extracting the relevant information from the specification. The man page no
5734 longer contains scrambled data for the -d option, and I've added a section
5735 at the front about calling Exim under different names.
5737 51. Added "extra_headers" argument to the "mail" command in filter files.
5739 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5742 53. Installed eximstats 1.29.
5744 54. Added transport_filter_timeout as a generic transport option.
5746 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5747 Cc: header lines. This was required by RFC 822, but it not required by RFC
5750 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5751 incoming messages that did not have them. Now it does so only if the
5752 message originates locally, that is, if there is no associated remote host
5753 address. When Resent- header lines are present, this applies to the Resent-
5754 lines rather than the non-Resent- lines.
5756 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5757 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5759 58. Messages for configuration errors now include the name of the main
5760 configuration file - useful now that there may be more than one file in a
5761 list (.included file names were always shown).
5763 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5764 for those rare installations that do not start the daemon as root or run it
5765 setuid root. I've cut out the call to initgroups() if the daemon is not
5768 60. The Exim user and group can now be bound into the binary as text strings
5769 that are looked up at the start of Exim's processing.
5771 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5773 62. Added $mailstore_basename variable.
5775 63. Installed patch to sieve.c from Michael Haardt.
5777 64. When Exim failed to open the panic log after failing to open the main log,
5778 the original message it was trying to log was written to stderr and debug
5779 output, but if they were not available (the usual case in production), it
5780 was lost. Now it is written to syslog before the two lines that record the
5781 failures to open the logs.
5783 65. Users' Exim filters run in subprocesses under the user's uid. It is
5784 possible for a "deliver" command or an alias in a "personal" command to
5785 provoke an address rewrite. If logging of address rewriting is configured,
5786 this fails because the process is not running as root or exim. There may be
5787 a better way of dealing with this, but for the moment (because 4.30 needs
5788 to be released), I have disabled address rewrite logging when running a
5789 filter in a non-root, non-exim process.
5795 1. The buildconfig auxiliary program wasn't quoting the value set for
5796 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5797 not defined. This bug was masked in 4.22 by the effect that was fixed in
5800 2. Some messages that were rejected after a message id was allocated were
5801 shown as "incomplete" by exigrep. It no longer does this for messages that
5802 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5804 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5805 have allow_domain_literals set, the ID did not get logged in the <= line.
5806 Domain literals are now always recognized in Message-ID: header lines.
5808 4. The first argument for a ${extract expansion item is the key name or field
5809 number. Leading and trailing spaces in this item were not being ignored,
5810 causing some misleading effects.
5812 5. When deliver_drop_privilege was set, single queue runner processes started
5813 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5814 same command in the process it spins off) were not dropping privilege.
5816 6. When the daemon running as "exim" started a queue runner, it always
5817 re-executed Exim in the spun-off process. This is a waste of effort when
5818 deliver_drop_privilege is set. The new process now just calls the
5819 queue-runner function directly.
5825 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5828 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5829 ignored. Though the use of -oP was forcing the writing of a pid file, it
5830 was always written to the default place.
5832 3. If the message "no IP address found for host xxxx" is generated during
5833 incoming verification, it is now followed by identification of the incoming
5834 connection (so you can more easily find what provoked it).
5836 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5838 5. Added some features to "harden" Exim a bit more against certain attacks:
5840 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5841 be put in Local/Makefile. This is like the never_users runtime option,
5842 but it cannot be overridden. The default setting is "root".
5844 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5845 prefix string with which any file named in a -C command line option
5848 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5849 is retained for -C and -D only if the caller of Exim is root. Without
5850 it, the exim user may also use -C and -D and retain privilege.
5852 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5853 command line option is disabled.
5855 6. Macro names set by the -D option must start with an upper case letter, just
5856 like macro names defined in the configuration file.
5858 7. Added "dereference=" facility to LDAP.
5860 8. Two instances of the typo "uknown" in the source files are fixed.
5862 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5863 the Configure-Makefile script screwed up while processing it.
5865 10. Incorporated PCRE 4.4.
5867 11. The SMTP synchronization check was not operating right at the start of an
5868 SMTP session. For example, it could not catch a HELO sent before the client
5869 waited for the greeting. There is now a check for outstanding input at the
5870 point when the greeting is written. Because of the duplex, asynchronous
5871 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5872 way, but not yet received, when the check is performed.
5874 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5875 on TCP/IP sockets, because this apparently causes some broken clients to
5878 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5879 unchanged) from the Cygwin maintainer.
5881 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5882 was supported instead of testing for mbx. Effectively a typo.
5884 15. The spa authenticator server code was not checking that the input it
5885 received was valid base64.
5887 16. The debug output line for the "set" modifier in ACLs was not showing the
5888 name of the variable that was being set.
5890 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5892 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5893 lookup on its own. Something else was needed to trigger the lookup. For
5894 example, a match in host_lookup or the need for a host name in a host list.
5895 Now, if $sender_host_name is referenced and the host name has not yet been
5896 looked up, a lookup is performed. If the lookup fails, the variable remains
5897 empty, and $host_lookup_failed is set to "1".
5899 19. Added "eqi" as a case-independent comparison operator.
5901 20. The saslauthd authentication condition could segfault if neither service
5902 nor realm was specified.
5904 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5905 error message that was logged was misleading, and incoming SMTP
5906 connections were dropped. The message is now more accurate, and temporary
5907 errors are given to SMTP connections.
5909 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5910 (except RSET) with a 421 error, until QUIT is received. However, it was
5911 failing to send a response to QUIT.
5913 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5914 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5915 ACL is now run after the helo_try_verify_hosts code.
5917 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5918 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5919 case-independent, so other case variants are also recognized). Apparently
5920 some systems use these upper case variants.
5922 25. If more than two messages were waiting for the same host, and a transport
5923 filter was specified for the transport, Exim sent two messages over the
5924 same TCP/IP connection, and then failed with "socket operation on non-
5925 socket" when it tried to send the third.
5927 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5929 27. The extern definition of crypt16() in expand.c was not being excluded when
5930 the OS had its own crypt16() function.
5932 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5933 as a preferred synonym for return_size_limit, both as an option and as an
5936 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5938 30. Changed the default configuration ACL to relax the local part checking rule
5939 for addresses that are not in any local domains. For these addresses,
5940 slashes and pipe symbols are allowed within local parts, but the sequence
5941 /../ is explicitly forbidden.
5943 31. SPA server authentication was not clearing the challenge buffer before
5946 32. log_message in a "warn" ACL statement was writing to the reject log as
5947 well as to the main log, which contradicts the documentation and doesn't
5948 seem right (because no rejection is happening). So I have stopped it.
5950 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5951 However, I am unable to do any testing of this.
5953 34. Fixed an infelicity in the appendfile transport. When checking directories
5954 for a mailbox, to see if any needed to be created, it was accidentally
5955 using path names with one or more superfluous leading slashes; tracing
5956 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5958 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5959 contents are added to the log line that is written for every discarded
5960 recipient. (Previously a log_message setting was ignored.)
5962 36. The ${quote: operator now quotes the string if it is empty.
5964 37. The install script runs exim in order to find its version number. If for
5965 some reason other than non-existence or emptiness, which it checks, it
5966 could not run './exim', it was installing it with an empty version number,
5967 i.e. as "exim-". This error state is now caught, and the installation is
5970 38. An argument was missing from the function that creates an error message
5971 when Exim fails to connect to the socket for saslauthd authentication.
5972 This could cause Exim to crash, or give a corrupted message.
5974 39. Added isip, isip4, and isip6 to ${if conditions.
5976 40. The ACL variables $acl_xx are now saved with the message, and can be
5977 accessed later in routers, transports, and filters.
5979 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5980 strings in the file are not string-expanded.
5982 42. If a MAIL command specified a SIZE value that was too large to fit into an
5983 int variable, the check against message_size_limit failed. Such values are
5984 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5985 day this will have to be increased, but I don't think I want to be around
5986 when emails are that large.
5993 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5994 iconv() is not standard in FreeBSD.
5996 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5997 IPv6 enabled. The observed symptom was a segmentation fault on return from
5998 the function os_common_find_running_interfaces() in src/os.c.
6000 3. In the check_special_case() function in daemon.c I had used "errno" as an
6001 argument name, which causes warnings on some systems. This was basically a
6002 typo, since it was named "eno" in the comments!
6004 4. The code that waits for the clock to tick (at a resolution of some fraction
6005 of a second) so as to ensure message-id uniqueness was always waiting for
6006 at least one whole tick, when it could have waited for less. [This is
6007 almost certainly not relevant at current processor speeds, where it is
6008 unlikely to ever wait at all. But we try to future-proof.]
6010 5. The function that sleeps for a time interval that includes fractions of a
6011 second contained a race. It did not block SIGALRM between setting the
6012 timer, and suspending (a couple of lines later). If the interval was short
6013 and the sigsuspend() was delayed until after it had expired, the suspension
6014 never ended. On busy systems this could lead to processes getting stuck for
6017 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6018 process, before it forks any delivery processes. The open lookup caching
6019 mechanism meant that the open file or database connection was passed into
6020 the delivery process. The problem was that delivery processes always tidy
6021 up cached lookup data. This could cause a problem for the next delivery
6022 process started by the queue runner, because the external queue runner
6023 process does not know about the closure. So the next delivery process
6024 still has data in the lookup cache. In the case of a file lookup, there was
6025 no problem because closing a file descriptor in a subprocess doesn't affect
6026 the parent. However, if the lookup was caching a connection to a database,
6027 the connection was closed, and the second delivery process was likely to
6028 see errors such as "PGSQL: query failed: server closed the connection
6029 unexpectedly". The problem has been fixed by closing all cached lookups
6030 in a queue runner before running a delivery process.
6032 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6033 seem to have the "const" qualifier which it has on other OS. I've
6036 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6037 *of the same type* (client or server) with the same public name that an
6038 error should be diagnosed.
6040 9. When Exim looked up a host name for an IP address, but failed to find the
6041 original IP address when looking up the host name (a safety check), it
6042 output the message "<ip address> does not match any IP for NULL", which was
6043 confusing, to say the least. The bug was that the host name should have
6044 appeared instead of "NULL".
6046 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6047 user that is built into the binary, and the -C or -D options is used, root
6048 privilege is dropped before the configuration file is read. In addition,
6049 logging is switched to stderr instead of the normal log files. If the
6050 configuration then re-defines the Exim user, the unprivileged environment
6051 is probably not what is expected, so Exim logs a panic warning message (but
6054 However, if deliver_drop_privilege is set, the unprivileged state may well
6055 be exactly what is intended, so the warning has been cut out in that case,
6056 and Exim is allowed to try to write to its normal log files.
6062 1. smtp_return_error_details was not giving details for temporary sender
6063 or receiver verification errors.
6065 2. Diagnose a configuration error if two authenticators have the same public
6068 3. Exim used not to create the message log file for a message until the first
6069 delivery attempt. This could be confusing when incoming messages were held
6070 for policy or load reasons. The message log file is now created at the time
6071 the message is received, and an initial "Received" line is written to it.
6073 4. The automatically generated man page for command line options had a minor
6074 bug that caused no ill effects; however, a more serious problem was that
6075 the procedure for building the man page automatically didn't always
6076 operate. Consequently, release 4.20 contains an out-of-date version. This
6077 shouldn't happen again.
6079 5. When building Exim with embedded Perl support, the script that builds the
6080 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6081 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6083 6. The freeze_tell option was not being used for messages that were frozen on
6084 arrival, either by an ACL or by local_scan().
6086 7. Added the smtp_incomplete_transaction log selector.
6088 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6089 was accepting AUTH without a new EHLO.
6091 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6092 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6095 10. Logging of TCP/IP connections (when configured) now happens in the main
6096 daemon process instead of the child process, so that the TCP/IP connection
6097 count is more accurate (but it can never be perfect).
6099 11. The use of "drop" in a nested ACL was not being handled correctly in the
6100 outer ACL. Now, if condition failure induced by the nested "drop" causes
6101 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6102 or "require"), the connection is dropped.
6104 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6105 that yield "discard" can now be used with an "accept" or a "discard" verb,
6106 but an error is generated for any others (because I can't see a useful way
6107 to define what should happen).
6109 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6110 are now processed in the same way as lines in the Exim configuration file.
6111 In particular, continuation lines are supported.
6113 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6115 15. Added -ti meaning -t -i.
6117 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6118 domains, and warn by logging if others are found.
6120 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6121 returned from GIFCONF when Exim is trying to find the list of interfaces on
6122 a host. The code in os.c has been modified to copy each ifreq to an aligned
6123 structure in all cases.
6125 Also, in some cases, the returned ifreq's were being copied to a 'struct
6126 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6127 means the last couple of bytes of an IPv6 address could be chopped if the
6128 ifreq contained only a normal sockaddr (14 bytes storage).
6130 18. Named domain lists were not supported in the hosts_treat_as_local option.
6131 An entry such as +xxxx was not recognized, and was treated as a literal
6134 19. Ensure that header lines added by a DATA ACL are included in the reject log
6135 if the ACL subsequently rejects the message.
6137 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6138 MD5 (which is deprecated).
6140 21. When testing a filter file using -bf, Exim was writing a message when it
6141 took the sender from a "From " line in the message, but it was not doing so
6142 when it took $return_path from a Return-Path: header line. It now does.
6144 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6145 with a valid header line field name (a series of printing characters
6146 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6148 23. Changed "disc" in the source to "disk" to conform to the documentation and
6149 the book and for uniformity.
6151 24. Ignore Sendmail's -Ooption=value command line item.
6153 25. When execve() failed while trying to run a command in a pipe transport,
6154 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6155 could be confused with a return value of 69 from the command itself. This
6156 has been changed to 127, the value the shell returns if it is asked to run
6157 a non-existent command. The wording for the related log line suggests a
6158 non-existent command as the problem.
6160 26. If received_header_text expands to an empty string, do not add a Received:
6161 header line to the message. (Well, it adds a token one on the spool, but
6162 marks it "old" so that it doesn't get used or transmitted.)
6164 27. Installed eximstats 1.28 (addition of -nt option).
6166 28. There was no check for failure on the call to getsockname() in the daemon
6167 code. This can fail if there is a shortage of resources on the system, with
6168 ENOMEM, for example. A temporary error is now given on failure.
6170 29. Contrary to the C standard, it seems that in some environments, the
6171 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6172 program. Exim now does this explicitly; it affects the formatting of
6173 timestamps using strftime().
6175 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6176 complaints. I've now initialized all the variables, to avoid this.
6178 32. Header lines added by a system filter were not being "seen" during
6179 transport-time rewrites.
6181 33. The info_callback() function passed to OpenSSL is set up with type void
6182 (*)(SSL *, int, int), as described somewhere. However, when calling the
6183 function (actually a macro) that sets it up, the type void(*)() is
6184 expected. I've put in a cast to prevent warnings from picky compilers.
6186 34. If a DNS black list lookup found a CNAME record, but there were no A
6187 records associated with the domain it pointed at, Exim crashed.
6189 35. If a DNS black list lookup returned more than one A record, Exim ignored
6190 all but the first. It now scans all returned addresses if a particular IP
6191 value is being sought. In this situation, the contents of the
6192 $dnslist_value variable are a list of all the addresses, separated by a
6195 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6196 to accept a host name and all its aliases if the forward lookup for any of
6197 them yielded the IP address of the incoming connection. Now it accepts only
6198 those names whose forward lookup yields the correct IP address. Any other
6199 names are discarded. This closes a loophole whereby a rogue DNS
6200 administrator could create reverse DNS records to break through a
6201 wildcarded host restriction in an ACL.
6203 37. If a user filter or a system filter that ran in a subprocess used any of
6204 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6205 the wrong values were passed to the pipe command ($thisaddress had the
6206 value of $0, $0 had the value of $1, etc). This bug was introduced by
6207 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6209 38. Improved the line breaking for long SMTP error messages from ACLs.
6210 Previously, if there was no break point between 40 and 75 characters, Exim
6211 left the rest of the message alone. Two changes have been made: (a) I've
6212 reduced the minimum length to 35 characters; (b) if it can't find a break
6213 point between 35 and 75 characters, it looks ahead and uses the first one
6214 that it finds. This may give the occasional overlong line, but at least the
6215 remaining text gets split now.
6217 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6218 file descriptors might be low, and that setting 1000 would always raise it.
6219 It turns out that in some environments, the limit is already over 1000 and
6220 that lowering it causes trouble. So now Exim takes care not to decrease it.
6222 40. When delivering a message, the value of $return_path is set to $sender_
6223 address at the start of routing (routers may change the value). By an
6224 oversight, this default was not being set up when an address was tested by
6225 -bt or -bv, which affected the outcome if any router or filter referred to
6228 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6229 log only when "message" or "log_message" are set. However, if one of the
6230 conditions was an address verification, or a call to a nested ACL, the
6231 messages generated by the underlying test were being passed through. This
6232 no longer happens. The underlying message is available in $acl_verify_
6233 message for both "message" and "log_message" expansions, so it can be
6234 passed through if needed.
6236 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6237 new expansion $bh_ to give the encoded byte string without charset
6238 translation. Translation happens only if iconv() is available; HAVE_ICONV
6239 indicates this at build time. HEADERS_CHARSET gives the charset to
6240 translate to; headers_charset can change it in the configuration, and
6241 "headers charset" can change it in an individual filter file.
6243 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6244 that creates RFC 2047 encoded "words" labels them as that charset instead
6245 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6246 expansion operator; (ii) when Exim creates a From: line for a local
6247 message; (iii) when a header line is rewritten to include a "phrase" part.
6249 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6250 buggy, causing it to skip the first lines of messages whose message ID
6251 ended in 'D'. This would not have bitten before Exim release 4.14, because
6252 message IDs were unlikely to end in 'D' before then. The effect was to have
6253 incorrect size information for certain domains.
6255 45. #include "config.h" was missing at the start of the crypt16.c module. This
6256 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6259 46. If there was a timeout during a "random" callout check, Exim treated it as
6260 a failure of the random address, and carried on sending RSET and the real
6261 address. If the delay was just some slowness somewhere, the response to the
6262 original RCPT would be taken as a response to RSET and so on, causing
6263 mayhem of various kinds.
6265 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6266 when I implemented it. It didn't allow for the fact that some option values
6267 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6268 the right test for positive values.
6270 48. Domain names in DNS records are case-independent. Exim always looks them up
6271 in lower case. Some resolvers return domain names in exactly the case they
6272 appear in the zone file, that is, they may contain uppercase letters. Not
6273 all resolvers do this - some return always lower case. Exim was treating a
6274 change of case by a resolver as a change of domain, similar to a widening
6275 of a domain abbreviation. This triggered its re-routing code and so it was
6276 trying to route what was effectively the same domain again. This normally
6277 caused routing to fail (because the router wouldn't handle the domain
6278 twice). Now Exim checks for this case specially, and just changes the
6279 casing of the domain that it ultimately uses when it transmits the message
6282 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6285 50. If a filter generated a file delivery with a non-absolute name (possible if
6286 no home directory exists for the router), the forbid_file option was not
6289 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6290 the existing equality matching.
6292 52. Exim was using ints instead of ino_t variables in some places where it was
6293 dealing with inode numbers.
6295 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6296 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6297 called TMPDIR, and if it finds it is different, it changes its value.
6299 54. The smtp_printf() function is now made available to local_scan() so
6300 additional output lines can be written before returning. There is also an
6301 smtp_fflush() function to enable the detection of a dropped connection.
6302 The variables smtp_input and smtp_batched_input are exported to
6305 55. Changed the default runtime configuration: the message "Unknown user"
6306 has been removed from the ACL, and instead placed on the localuser router,
6307 using the cannot_route_message feature. This means that any verification
6308 failures that generate their own messages won't get overridden. Similarly,
6309 the "Unrouteable address" message that was in the ACL for unverifiable
6310 relay addresses has also been removed.
6312 56. Added hosts_avoid_esmtp to the smtp transport.
6314 57. The exicyclog script was not checking for the esoteric option
6315 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6316 will work only if exicyclog is run under the appropriate euid.
6318 58. Following a discussion on the list, the rules by which Exim recognises line
6319 endings on incoming messages have been changed. The -dropcr and drop_cr
6320 options are now no-ops, retained only for backwards compatibility. The
6321 following line terminators are recognized: LF CRLF CR. However, special
6322 processing applies to CR:
6324 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6325 nor a local message in the state where . is a terminator.
6327 (ii) If a bare CR is encountered in a header line, an extra space is added
6328 after the line terminator so as not to end the header. The reasoning
6329 behind this is that bare CRs in header lines are most likely either
6330 to be mistakes, or people trying to play silly games.
6332 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6333 was being incorrectly given as 18 bytes larger than it should have been.
6334 This is a VOB (very old bug).
6336 60. This may never have affected anything current, but just in case it has:
6337 When the local host is found other than at the start of a list of hosts,
6338 the local host, those with the same MX, and any that follow, are discarded.
6339 When the list in question was part of a longer list of hosts, the following
6340 hosts (not currently being processed) were also being discarded. This no
6341 longer happens. I'm not sure if this situation could ever has previously
6344 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6347 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6348 header. This is something that is recommended in a new Internet Draft, and
6349 is something that is documented as being done by Sendmail. There are two
6350 possible values. For messages generated by the autoreply transport, Exim
6353 Auto-Submitted: auto-replied
6355 whereas for all other generated messages (e.g. bounces) it adds
6357 Auto-Submitted: auto-generated
6359 63. The "personal" condition in filters now includes a test for the
6360 Auto-Submitted: header. If it contains the string "auto-" the message it
6361 not considered personal.
6363 64. Added rcpt_include_affixes as a generic transport option.
6365 65. Added queue_only_override (default true).
6367 66. Added the syslog_duplication option.
6369 67. If what should have been the first header line of a message consisted of
6370 a space followed by a colon, Exim was mis-interpreting it as a header line.
6371 It isn't of course - it is syntactically invalid and should therefore be
6372 treated as the start of the message body. The misbehaviour could have
6373 caused a number of strange effects, including loss of data in subsequent
6374 header lines, and spool format errors.
6376 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6377 client host had authenticated. This control can now be exercised by an ACL
6378 for more flexibility.
6380 69. By default, callouts do not happen when testing with -bh. There is now a
6381 variant, -bhc, which does actually run the callout code, including
6382 consulting and updating the callout cache.
6384 70. Added support for saslauthd authentication, courtesy of Alexander
6387 71. If statvfs() failed on the spool or log directories while checking their
6388 size for availability, Exim confusingly gave the error "space shortage".
6389 Furthermore, in debugging mode it crashed with a floating point exception.
6390 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6391 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6392 serious problem, Exim now writes to the main and panic logs when this
6393 happens, with details of the failure. It then refuses to accept the
6394 incoming message, giving the message "spool directory problem" or "log
6395 directory problem" with a 421 code for SMTP messages.
6397 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6398 0, 1, and 2 exist, because some OS complain for execs without them (see
6399 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6400 descriptors. However, the code omitted to check that the open succeeded,
6401 causing mysterious errors if for some reason the permissions on /dev/null
6402 got screwed. Now Exim writes a message to the main and panic logs, and
6403 bombs out if it can't open /dev/null.
6405 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6406 interact so that it is all more flexible. It is supposed to remain
6407 backwards compatible. Also added extra_local_interfaces.
6409 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6410 to bomb out with an assertion failure - to the client this appears as a
6411 connection drop. This problem occurs in the part of the code that was taken
6412 from the Samba project. Fortunately, the assertion is in a very simple
6413 function, so I have fixed this by reproducing the function inline in the
6414 one place where it is called, and arranging for authentication to fail
6415 instead of killing the process with assert().
6417 75. The SPA client code was not working when the server requested OEM rather
6418 than Unicode encoding.
6420 76. Added code to make require_files with a specific uid setting more usable in
6421 the case where statting the file as root fails - usually a non-root-mounted
6422 NFS file system. When this happens and the failure is EACCES, Exim now
6423 forks a subprocess and does the per-uid checking as the relevant uid.
6425 77. Added process_log_path.
6427 78. If log_file_path was not explicitly set, a setting of check_log_space or
6428 check_log_inodes was ignored.
6430 79. If a space check for the spool or log partitions fails, the incident is now
6431 logged. Of course, in the latter case the data may get lost...
6433 80. Added the %p formatting code to string_format() so that it can be used to
6434 print addresses in debug_print(). Adjusted all the address printing in the
6435 debugging in store.c to use %p rather than %d.
6437 81. There was a concern that a line of code in smtp_in.c could overflow a
6438 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6439 initially expressed, the concern was not well-founded, because trailing
6440 spaces are removed early. However, if the trailing spaces were followed by
6441 a NULL, they did not get removed, so the overflow was possible. Two fixes
6444 (a) I re-wrote the offending code in a cleaner fashion.
6445 (b) If an incoming SMTP command contains a NULL character, it is rejected
6448 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6449 runs initgroups(), so that if the Exim user is in any additional groups,
6450 they will be used during message reception.
6456 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob