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.
108 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
111 JH/02 The smtp transport option "multi_domain" is now expanded.
113 JH/03 The smtp transport now requests PRDR by default, if the server offers
116 JH/04 Certificate name checking on server certificates, when exim is a client,
117 is now done by default. The transport option tls_verify_cert_hostnames
118 can be used to disable this per-host. The build option
119 EXPERIMENTAL_CERTNAMES is withdrawn.
121 JH/05 The value of the tls_verify_certificates smtp transport and main options
122 default to the word "system" to access the system default CA bundle.
123 For GnuTLS, only version 3.0.20 or later.
125 JH/06 Verification of the server certificate for a TLS connection is now tried
126 (but not required) by default. The verification status is now logged by
127 default, for both outbound TLS and client-certificate supplying inbound
130 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
133 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
134 Status Notification (bounce) messages are now MIME format per RFC 3464.
135 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
136 under the control of the dsn_advertise_hosts option, and routers may
137 have a dsn_lasthop option.
139 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
140 default, modifiable by a malware= option. The list separator for
141 the options can now be changed in the usual way. Bug 68.
143 JH/10 The smtp_receive_timeout main option is now expanded before use.
145 JH/11 The incoming_interface log option now also enables logging of the
146 local interface on delivery outgoing connections.
148 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
149 if the interface and destination host and port all match.
151 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
154 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
155 Patch from Andrew Lewis.
157 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
158 now supports optional time-restrictions, weighting, and priority
159 modifiers per server. Patch originally by <rommer@active.by>.
161 JH/16 The spamd_address main option now supports a mixed list of local
162 and remote servers. Remote servers can be IPv6 addresses, and
163 specify a port-range.
165 JH/17 Bug 68: The spamd_address main option now supports an optional
166 timeout value per server.
168 JH/18 Bug 1581: Router and transport options headers_add/remove can
169 now have the list separator specified.
171 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
174 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
177 JH/21 Support for the A6 type of dns record is withdrawn.
179 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
180 rather than the verbs used.
182 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
183 from 255 to 1024 chars.
185 JH/24 Verification callouts now attempt to use TLS by default.
187 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
188 are generic router options now. The defaults didn't change.
190 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
191 Original patch from Alexander Shikoff, worked over by JH.
193 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
194 files if ZCAT_COMMAND is not executable.
196 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
198 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
200 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
201 Normally benign, it bites when the pair was led to by a CNAME;
202 modern usage is to not canoicalize the domain to a CNAME target
203 (and we were inconsistent anyway for A-only vs AAAA+A).
205 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
207 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
208 when evaluating $sender_host_dnssec.
210 JH/31 Check the HELO verification lookup for DNSSEC, adding new
211 $sender_helo_dnssec variable.
213 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
215 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
217 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
219 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
220 documented as working, but never had. Support all but $spam_report.
222 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
223 added for tls authenticator.
228 TL/01 When running the test suite, the README says that variables such as
229 no_msglog_check are global and can be placed anywhere in a specific
230 test's script, however it was observed that placement needed to be near
231 the beginning for it to behave that way. Changed the runtest perl
232 script to read through the entire script once to detect and set these
233 variables, reset to the beginning of the script, and then run through
234 the script parsing/test process like normal.
236 TL/02 The BSD's have an arc4random API. One of the functions to induce
237 adding randomness was arc4random_stir(), but it has been removed in
238 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
239 function when detected.
241 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
242 cause callback expansion.
244 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
245 syntax errors in an expansion can be treated as a string instead of
246 logging or causing an error, due to the internal use of bool_lax
247 instead of bool when processing it.
249 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
250 server certificates when making smtp deliveries.
252 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
254 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
256 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
258 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
259 Merged patch from Sebastian Wiedenroth.
261 JH/05 Fix results-pipe from transport process. Several recipients, combined
262 with certificate use, exposed issues where response data items split
263 over buffer boundaries were not parsed properly. This eventually
264 resulted in duplicates being sent. This issue only became common enough
265 to notice due to the introduction of conection certificate information,
266 the item size being so much larger. Found and fixed by Wolfgang Breyha.
268 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
269 size buffer was used, resulting in syntax errors when an expansion
272 JH/07 Add support for directories of certificates when compiled with a GnuTLS
273 version 3.3.6 or later.
275 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
276 is EXPERIMENTAL_EVENT, the main-configuration and transport options
277 both become "event_action", the variables become $event_name, $event_data
278 and $event_defer_errno. There is a new variable $verify_mode, usable in
279 routers, transports and related events. The tls:cert event is now also
280 raised for inbound connections, if the main configuration event_action
283 TL/06 In test suite, disable OCSP for old versions of openssl which contained
284 early OCSP support, but no stapling (appears to be less than 1.0.0).
286 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
287 server certificate names available under the smtp transport option
288 "tls_verify_cert_hostname" now do not permit multi-component wildcard
291 JH/10 Time-related extraction expansions from certificates now use the main
292 option "timezone" setting for output formatting, and are consistent
293 between OpenSSL and GnuTLS compilations. Bug 1541.
295 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
296 encoded parameter in the incoming message. Bug 1558.
298 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
299 include certificate info, eximon was claiming there were spoolfile
302 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
304 JH/14 Log delivery-related information more consistently, using the sequence
305 "H=<name> [<ip>]" wherever possible.
307 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
308 are problematic for Debian distribution, omit them from the release
311 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
313 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
316 JH/17 Fix a null-indirection in certextract expansions when a nondefault
317 output list separator was used.
322 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
323 checkers that were complaining about end of non-void function with no
326 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
327 This was a regression intruduced in 4.83 by another bugfix.
329 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
331 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
332 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
338 TF/01 Correctly close the server side of TLS when forking for delivery.
340 When a message was received over SMTP with TLS, Exim failed to clear up
341 the incoming connection properly after forking off the child process to
342 deliver the message. In some situations the subsequent outgoing
343 delivery connection happened to have the same fd number as the incoming
344 connection previously had. Exim would try to use TLS and fail, logging
345 a "Bad file descriptor" error.
347 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
348 utilities have not been installed.
350 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
351 temporary space as the ACL may create new global variables.
353 TL/01 LDAP support uses per connection or global context settings, depending
354 upon the detected version of the libraries at build time.
356 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
357 to extract and use the src ip:port in logging and expansions as if it
358 were a direct connection from the outside internet. PPv2 support was
359 updated based on HAProxy spec change in May 2014.
361 JH/02 Add ${listextract {number}{list}{success}{fail}}.
363 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
364 Properly escape header and check for NULL return.
366 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
369 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
371 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
372 characters in header names, implemented as a verify condition.
373 Contributed by Michael Fischer v. Mollard.
375 TL/05 Rename SPF condition results err_perm and err_temp to standardized
376 results permerror and temperror. Previous values are deprecated but
377 still accepted. In a future release, err_perm and err_temp will be
378 completely removed, which will be a backward incompatibility if the
379 ACL tests for either of these two old results. Patch contributed by
380 user bes-internal on the mailing list.
382 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
384 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
385 selectors, in both main and reject logs.
387 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
390 JH/07 Add malware type "sock" for talking to simple daemon.
392 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
394 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
395 routers/transports under cutthrough routing.
397 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
398 numbers. Touch up "bool" conditional to keep the same definition.
400 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
402 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
404 JH/12 Expand items in router/transport headers_add or headers_remove lists
405 individually rather than the list as a whole. Bug 1452.
407 Required for reasonable handling of multiple headers_ options when
408 they may be empty; requires that headers_remove items with embedded
409 colons must have them doubled (or the list-separator changed).
411 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
412 view the policy declared in the DMARC record. Currently, $dmarc_status
413 is a combined value of both the record presence and the result of the
416 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
418 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
419 dnslookup router and the smtp transport (applying to the forward
422 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
423 of ldap servers used for a specific lookup. Patch provided by Heiko
426 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
427 New variable $lookup_dnssec_authenticated for observability.
429 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
430 Patch submitted by Lars Timman.
432 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
434 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
435 Requires trusted mode and valid format message id, aborts otherwise.
436 Patch contributed by Heiko Schlichting.
438 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
439 certextract with support for various fields. Bug 1358.
441 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
442 is requested by default, modifiable by smtp transport option
445 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
446 operate on certificate variables to give certificate fingerprints
447 Also new ${sha256:cert_variable}.
449 JH/23 The PRDR feature is moved from being Experimental into the mainline.
451 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
452 Christian Aistleitner.
454 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
456 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
457 file. Patch from Wolfgang Breyha.
459 JH/25 Expand the coverage of the delivery $host and $host_address to
460 client authenticators run in verify callout. Bug 1476.
462 JH/26 Port service names are now accepted for tls_on_connect_ports, to
463 align with daemon_smtp_ports. Bug 72.
465 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
466 support and error reporting did not work properly.
468 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
469 and is readable. Patch from Andrew Colin Kissa.
471 TL/14 Enhance documentation of ${run expansion and how it parses the
472 commandline after expansion, particularly in the case when an
473 unquoted variable expansion results in an empty value.
475 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
477 PP/02 Fix internal collision of T_APL on systems which support RFC3123
478 by renaming away from it. Addresses GH issue 15, reported by
481 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
483 TL/15 SECURITY: prevent double expansion in math comparison functions
484 (can expand unsanitized data). Not remotely exploitable.
491 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
493 PP/02 Make -n do something, by making it not do something.
494 When combined with -bP, the name of an option is not output.
496 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
499 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
500 $sender_host_name and config options to manage this, and basic check
503 PP/05 DSCP support for outbound connections and control modifier for inbound.
505 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
506 (Only plugin which currently uses this is kerberos4, which nobody should
507 be using, but we should make it available and other future plugins might
508 conceivably use it, even though it would break NAT; stuff *should* be
509 using channel bindings instead).
511 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
512 name; added for Sendmail compatibility; requires admin caller.
513 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
514 just ignore it); requires trusted caller.
515 Also parse but ignore: -Ac -Am -X<logfile>
518 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
520 TL/02 Add +smtp_confirmation as a default logging option.
522 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
523 Patch by Magnus Holmgren from 2007-02-20.
525 TL/04 Bugzilla 1281 - Spec typo.
526 Bugzilla 1283 - Spec typo.
527 Bugzilla 1290 - Spec grammar fixes.
529 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
531 TL/06 Add Experimental DMARC support using libopendmarc libraries.
533 TL/07 Fix an out of order global option causing a segfault. Reported to dev
534 mailing list by by Dmitry Isaikin.
536 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
538 JH/02 Support "G" suffix to numbers in ${if comparisons.
540 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
542 NM/01 Bugzilla 1197 - Spec typo
543 Bugzilla 1196 - Spec examples corrections
545 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
547 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
548 gnutls_enable_pkcs11, but renamed to more accurately indicate its
551 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
552 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
554 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
555 "acl {{name}{arg}...}", and optional args on acl condition
558 JH/05 Permit multiple router/transport headers_add/remove lines.
560 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
562 JH/07 Avoid using a waiting database for a single-message-only transport.
563 Performance patch from Paul Fisher. Bugzilla 1262.
565 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
568 JH/09 Add $headers_added variable, with content from use of ACL modifier
569 add_header (but not yet added to the message). Bugzilla 199.
571 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
572 Pulled from Bugzilla 817 by Wolfgang Breyha.
574 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
576 (nb: this is the same fix as in Exim 4.80.1)
578 JH/11 Add A= logging on delivery lines, and a client_set_id option on
581 JH/12 Add optional authenticated_sender logging to A= and a log_selector
584 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
586 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
587 advertise SMTP AUTH mechanism to us, instead of a generic
588 protocol violation error. Also, make Exim more robust to bad
589 data from the Dovecot auth socket.
591 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
593 When a queue runner is handling a message, Exim first routes the
594 recipient addresses, during which it prunes them based on the retry
595 hints database. After that it attempts to deliver the message to
596 any remaining recipients. It then updates the hints database using
599 So if a recipient address works intermittently, it can get repeatedly
600 deferred at routing time. The retry hints record remains fresh so the
601 address never reaches the final cutoff time.
603 This is a fairly common occurrence when a user is bumping up against
604 their storage quota. Exim had some logic in its local delivery code
605 to deal with this. However it did not apply to per-recipient defers
606 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
608 This change adds a proper retry rule check during routing so that the
609 final cutoff time is checked against the message's age. We only do
610 this check if there is an address retry record and there is not a
611 domain retry record; this implies that previous attempts to handle
612 the address had the retry_use_local_parts option turned on. We use
613 this as an approximation for the destination being like a local
614 delivery, as in LMTP.
616 I suspect this new check makes the old local delivery cutoff check
617 redundant, but I have not verified this so I left the code in place.
619 TF/02 Correct gecos expansion when From: is a prefix of the username.
621 Test 0254 submits a message to Exim with the header
625 When I ran the test suite under the user fanf2, Exim expanded
626 the header to contain my full name, whereas it should have added
627 a Resent-Sender: header. It erroneously treats any prefix of the
628 username as equal to the username.
630 This change corrects that bug.
632 GF/01 DCC debug and logging tidyup
633 Error conditions log to paniclog rather than rejectlog.
634 Debug lines prefixed by "DCC: " to remove any ambiguity.
636 TF/03 Avoid unnecessary rebuilds of lookup-related code.
638 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
639 Bug spotted by Jeremy Harris; was flawed since initial commit.
640 Would have resulted in OCSP responses post-SNI triggering an Exim
641 NULL dereference and crash.
643 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
645 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
646 Bug detection, analysis and fix by Samuel Thibault.
647 Bugzilla 1331, Debian bug #698092.
649 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
651 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
652 Server implementation by Todd Lyons, client by JH.
653 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
654 config variable "prdr_enable" controls whether the server
655 advertises the facility. If the client requests PRDR a new
656 acl_data_smtp_prdr ACL is called once for each recipient, after
657 the body content is received and before the acl_smtp_data ACL.
658 The client is controlled by bolth of: a hosts_try_prdr option
659 on the smtp transport, and the server advertisement.
660 Default client logging of deliveries and rejections involving
661 PRDR are flagged with the string "PRDR".
663 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
664 fclose(). Diagnosis by Todd Lyons.
666 PP/17 Update configure.default to handle IPv6 localhost better.
667 Patch by Alain Williams (plus minor tweaks).
670 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
671 This is now consistent with GnuTLS, and is now documented: the
672 previous undocumented portable approach to treating the option as
673 unset was to force an expansion failure. That still works, and
674 an empty string is now equivalent.
676 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
677 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
678 not performing validation itself.
680 PP/20 Added force_command boolean option to pipe transport.
681 Patch from Nick Koston, of cPanel Inc.
683 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
686 TF/04 Added udpsend ACL modifer and hexquote expansion operator
688 PP/21 Fix eximon continuous updating with timestamped log-files.
689 Broken in a format-string cleanup in 4.80, missed when I repaired the
690 other false fix of the same issue.
691 Report and fix from Heiko Schlichting.
694 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
695 Report from Prashanth Katuri.
697 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
698 It's SecureTransport, so affects any MacOS clients which use the
699 system-integrated TLS libraries, including email clients.
701 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
702 using a MIME ACL for non-SMTP local injection.
703 Report and assistance in diagnosis by Warren Baker.
705 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
707 JH/16 Fix comparisons for 64b. Bugzilla 1385.
709 TL/09 Add expansion variable $authenticated_fail_id to keep track of
710 last id that failed so it may be referenced in subsequent ACL's.
712 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
715 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
716 ldap library initialization, allowing self-signed CA's to be
717 used. Also properly sets require_cert option later in code by
718 using NULL (global ldap config) instead of ldap handle (per
719 session). Bug diagnosis and testing by alxgomz.
721 TL/12 Enhanced documentation in the ratelimit.pl script provided in
722 the src/util/ subdirectory.
724 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
725 renamed to Transport Post Delivery Action by Jeremy Harris, as
728 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
729 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
730 redis_servers = needs to be configured which will be used by the redis
731 lookup. Patch from Warren Baker, of The Packet Hub.
733 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
735 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
736 hostname or reverse DNS when processing a host list. Used suggestions
737 from multiple comments on this bug.
739 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
741 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
742 Missed a few lines, added it to make the runtest require no keyboard
745 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
746 contains upper case chars. Make router use caseful_local_part.
748 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
749 support when GnuTLS has been built with p11-kit.
755 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
757 This, or similar/improved, will also be change PP/11 of 4.82.
763 PP/01 Handle short writes when writing local log-files.
764 In practice, only affects FreeBSD (8 onwards).
765 Bugzilla 1053, with thanks to Dmitry Isaikin.
767 NM/01 Bugzilla 949 - Documentation tweak
769 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
772 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
774 PP/02 Implemented gsasl authenticator.
776 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
778 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
779 `pkg-config foo` for cflags/libs.
781 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
782 with rest of GSASL and with heimdal_gssapi.
784 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
785 `pkg-config foo` for cflags/libs for the TLS implementation.
787 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
788 properties get this fed in as external SSF. A number of robustness
789 and debugging improvements to the cyrus_sasl authenticator.
791 PP/08 cyrus_sasl server now expands the server_realm option.
793 PP/09 Bugzilla 1214 - Log authentication information in reject log.
794 Patch by Jeremy Harris.
796 PP/10 Added dbmjz lookup type.
798 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
800 PP/12 MAIL args handles TAB as well as SP, for better interop with
801 non-compliant senders.
802 Analysis and variant patch by Todd Lyons.
804 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
805 Bug report from Lars Müller <lars@samba.org> (via SUSE),
806 Patch from Dirk Mueller <dmueller@suse.com>
808 PP/13 tls_peerdn now print-escaped for spool files.
809 Observed some $tls_peerdn in wild which contained \n, which resulted
810 in spool file corruption.
812 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
813 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
814 or write after TLS renegotiation, which otherwise led to messages
817 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
818 as a tracking header (ie: a signed header comes before the signature).
819 Patch from Wolfgang Breyha.
821 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
822 comma-sep list; embedded commas doubled.
824 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
826 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
828 Report and patch from Dmitry Banschikov.
830 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
831 Removed SSL_clear() after SSL_new() which led to protocol negotiation
832 failures. We appear to now support TLS1.1+ with Exim.
834 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
835 lets Exim select keys and certificates based upon TLS SNI from client.
836 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
837 before an outbound SMTP session. New log_selector, +tls_sni.
839 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
840 NULL dereference. Report and patch from Alun Jones.
842 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
843 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
844 Not seeing resolver debug output on NetBSD, but suspect this is a
845 resolver implementation change.
847 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
848 Left warnings. Added "eximon gdb" invocation mode.
850 PP/21 Defaulting "accept_8bitmime" to true, not false.
852 PP/22 Added -bw for inetd wait mode support.
854 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
855 locate the relevant includes and libraries. Made this the default.
857 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
858 Bugzilla 1246, report and most of solution from Tomasz Kusy.
860 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
861 This may cause build issues on older platforms.
863 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
864 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
865 gnutls_require_mac & gnutls_require_protocols (no longer supported).
866 Added SNI support via GnuTLS too.
867 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
869 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
871 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
872 Applied second patch from Janne, implementing suggestion to default
873 multiple-strings-in-record handling to match SPF spec.
875 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
877 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
878 read-only, out of scope).
879 Patch from Wolfgang Breyha, report from Stuart Northfield.
881 PP/29 Fix three issues highlighted by clang analyser static analysis.
882 Only crash-plausible issue would require the Cambridge-specific
883 iplookup router and a misconfiguration.
884 Report from Marcin Mirosław.
886 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
888 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
889 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
890 As part of this, removing so much warning spew let me fix some minor
891 real issues in debug logging.
893 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
894 assignment on my part. Fixed.
896 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
897 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
898 Janne Snabb (who went above and beyond: thank you).
900 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
901 string otherwise requires a connection and a bunch more work and it's
902 relatively easy to get wrong. Should also expose TLS library linkage
905 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
906 64-bit ${eval} (JH/03).
908 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
909 GNU libc to support some of the 64-bit stuff, should not lead to
910 conflicts. Defined before os.h is pulled in, so if a given platform
911 needs to override this, it can.
913 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
914 protection layer was required, which is not implemented.
915 Bugzilla 1254, patch from Wolfgang Breyha.
917 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
918 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
919 tls_dhparam take prime identifiers. Also unbreak combination of
920 OpenSSL+DH_params+TLSSNI.
922 PP/39 Disable SSLv2 by default in OpenSSL support.
928 PP/01 Solaris build fix for Oracle's LDAP libraries.
929 Bugzilla 1109, patch from Stephen Usher.
931 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
933 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
936 TF/02 Fix a couple more cases where we did not log the error message
937 when unlink() failed. See also change 4.74-TF/03.
939 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
940 lock up or crash if it happened to be inside a call to libc when it
941 got a SIGUSR1 from exiwhat.
943 The SIGUSR1 handler appends the current process status to the process
944 log which is later printed by exiwhat. It used to use the general
945 purpose logging code to do this, but several functions it calls are
946 not safe for signals.
948 The new output code in the SIGUSR1 handler is specific to the process
949 log, and simple enough that it's easy to inspect for signal safety.
950 Removing some special cases also simplifies the general logging code.
951 Removing the spurious timestamps from the process log simplifies
954 TF/04 Improved ratelimit ACL condition.
956 The /noupdate option has been deprecated in favour of /readonly which
957 has clearer semantics. The /leaky, /strict, and /readonly update modes
958 are mutually exclusive. The update mode is no longer included in the
959 database key; it just determines when the database is updated. (This
960 means that when you upgrde Exim will forget old rate measurements.)
962 Exim now checks that the per_* options are used with an update mode that
963 makes sense for the current ACL. For example, when Exim is processing a
964 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
965 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
966 must specify per_mail/readonly. If you omit the update mode it defaults to
967 /leaky where that makes sense (as before) or /readonly where required.
969 The /noupdate option is now undocumented but still supported for
970 backwards compatibility. It is equivalent to /readonly except that in
971 ACLs where /readonly is required you may specify /leaky/noupdate or
972 /strict/noupdate which are treated the same as /readonly.
974 A useful new feature is the /count= option. This is a generalization
975 of the per_byte option, so that you can measure the throughput of other
976 aggregate values. For example, the per_byte option is now equivalent
977 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
979 The per_rcpt option has been generalized using the /count= mechanism
980 (though it's more complicated than the per_byte equivalence). When it is
981 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
982 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
983 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
984 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
985 Note that using per_rcpt with a non-readonly update mode in more than
986 one ACL will cause the recipients to be double-counted. (The per_mail
987 and per_byte options don't have this problem.)
989 The handling of very low rates has changed slightly. If the computed rate
990 is less than the event's count (usually one) then this event is the first
991 after a long gap. In this case the rate is set to the same as this event's
992 count, so that the first message of a spam run is counted properly.
994 The major new feature is a mechanism for counting the rate of unique
995 events. The new per_addr option counts the number of different
996 recipients that someone has sent messages to in the last time period. It
997 behaves like per_rcpt if all the recipient addresses are different, but
998 duplicate recipient addresses do not increase the measured rate. Like
999 the /count= option this is a general mechanism, so the per_addr option
1000 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1001 example, measure the rate that a client uses different sender addresses
1002 with the options per_mail/unique=$sender_address. There are further
1003 details in the main documentation.
1005 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1007 TF/06 Removed a few PCRE remnants.
1009 TF/07 Automatically extract Exim's version number from tags in the git
1010 repository when doing development or release builds.
1012 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1013 Bugzilla 879. Patch from Paul Fisher.
1015 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1016 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1019 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1021 PP/05 Variable $av_failed, true if the AV scanner deferred.
1022 Bugzilla 1078. Patch from John Horne.
1024 PP/06 Stop make process more reliably on build failure.
1025 Bugzilla 1087. Patch from Heiko Schlittermann.
1027 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1028 Bugzilla 1089. Patch from Heiko Schlittermann.
1030 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1031 Bugzilla 1131. Patch from Holger Weiß.
1033 PP/09 Handle IPv6 addresses with SPF.
1034 Bugzilla 860. Patch from Wolfgang Breyha.
1036 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1038 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1041 PP/11 match_* no longer expand right-hand-side by default.
1042 New compile-time build option, EXPAND_LISTMATCH_RHS.
1043 New expansion conditions, "inlist", "inlisti".
1045 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1047 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1049 PP/14 fix log_write() format string regression from TF/03.
1050 Bugzilla 1152. Patch from Dmitry Isaikin.
1056 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1058 PP/02 Harmonised TLS library version reporting; only show if debugging.
1059 Layout now matches that introduced for other libraries in 4.74 PP/03.
1061 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1063 PP/04 New "dns_use_edns0" global option.
1065 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1068 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1069 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1071 TK/01 Updated PolarSSL code to 0.14.2.
1072 Bugzilla 1097. Patch from Andreas Metzler.
1074 PP/07 Catch divide-by-zero in ${eval:...}.
1075 Fixes bugzilla 1102.
1077 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1080 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1081 format-string attack -- SECURITY: remote arbitrary code execution.
1083 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1084 time unintentionally subject to list matching rules, letting the header
1085 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1086 arbitrary string expansion). This allowed for information disclosure.
1088 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1089 INT_MIN/-1 -- value coerced to INT_MAX.
1095 NM/01 Workround for PCRE version dependency in version reporting
1098 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1099 This fixes portability to compilers other than gcc, notably
1100 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1102 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1103 makefiles for portability to HP-UX and POSIX correctness.
1105 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1106 Also via indented variable definition in the Makefile.
1107 (Debugging by Oliver Heesakkers).
1109 PP/02 Restore caching of spamd results with expanded spamd_address.
1110 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1112 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1113 Improves build reliability. Fix from: Frank Elsner
1115 NM/02 Fix wide character breakage in the rfc2047 coding
1116 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1118 NM/03 Allow underscore in dnslist lookups
1119 Fixes bug 1026. Patch from Graeme Fowler
1121 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1122 Code patches from Adam Ciarcinski of NetBSD.
1124 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1127 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1128 is logged, to avoid truncation. Patch from John Horne.
1130 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1131 Patch from Jakob Hirsch.
1133 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1134 SQL string expansion failure details.
1135 Patch from Andrey Oktyabrski.
1137 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1138 Patch from Simon Arlott.
1140 PP/09 New lookups functionality failed to compile on old gcc which rejects
1141 extern declarations in function scope.
1142 Patch from Oliver Fleischmann
1144 PP/10 Use sig_atomic_t for flags set from signal handlers.
1145 Check getgroups() return and improve debugging.
1146 Fixed developed for diagnosis in bug 927 (which turned out to be
1149 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1150 Patch from Mark Zealey.
1152 PP/12 Bugzilla 1056: Improved spamd server selection.
1153 Patch from Mark Zealey.
1155 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1156 Based on patch from Heiko Schlittermann.
1158 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1159 Patch from Uwe Doering, sign-off by Michael Haardt.
1161 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1162 variable declaration deep within a block. Bug and patch from
1165 PP/15 lookups-Makefile IRIX compatibilty coercion.
1167 PP/16 Make DISABLE_DKIM build knob functional.
1169 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1170 Patch by Simon Arlott
1172 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1173 variable argument macros. Our copy now differs from upstream.
1179 TF/01 Failure to get a lock on a hints database can have serious
1180 consequences so log it to the panic log.
1182 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1183 controlled using the smtp_confirmation log selector.
1185 TF/03 Include the error message when we fail to unlink a spool file.
1187 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1188 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1189 for maintaining out-of-tree patches for some time.
1191 PP/01 Bugzilla 139: Documentation and portability issues.
1192 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1193 Handle per-OS dynamic-module compilation flags.
1195 PP/02 Let /dev/null have normal permissions.
1196 The 4.73 fixes were a little too stringent and complained about the
1197 permissions on /dev/null. Exempt it from some checks.
1198 Reported by Andreas M. Kirchwitz.
1200 PP/03 Report version information for many libraries, including
1201 Exim version information for dynamically loaded libraries. Created
1202 version.h, now support a version extension string for distributors
1203 who patch heavily. Dynamic module ABI change.
1205 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1206 privilege escalation vulnerability whereby the Exim run-time user
1207 can cause root to append content of the attacker's choosing to
1210 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1213 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1214 If dropping privileges for untrusted macros, we disabled normal logging
1215 on the basis that it would fail; for the Exim run-time user, this is not
1216 the case, and it resulted in successful deliveries going unlogged.
1217 Fixed. Reported by Andreas Metzler.
1223 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1224 only prepend for the Resent-* case. Fixes regression introduced in
1225 Exim 4.70 by NM/22 for Bugzilla 607.
1227 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1228 increasing numbers of administrators be bitten by this.
1230 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1232 PP/03 Bugzilla 994: added openssl_options main configuration option.
1234 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1236 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1238 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1240 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1241 without a peer certificate, leading to a segfault because of an
1242 assumption that peers always have certificates. Be a little more
1243 paranoid. Problem reported by Martin Tscholak.
1245 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1246 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1247 NB: ClamAV planning to remove STREAM in "middle of 2010".
1248 CL also introduces -bmalware, various -d+acl logging additions and
1249 more caution in buffer sizes.
1251 PP/09 Implemented reverse_ip expansion operator.
1253 PP/10 Bugzilla 937: provide a "debug" ACL control.
1255 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1257 PP/12 Bugzilla 973: Implement --version.
1259 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1261 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1263 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1265 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1266 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1267 ignore trailing whitespace.
1269 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1271 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1274 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1275 Notification from Dr Andrew Aitchison.
1277 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1278 ExtendedDetectionInfo response format.
1279 Notification from John Horne.
1281 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1284 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1285 XSL and documented dependency on system catalogs, with examples of how
1288 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1291 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1292 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1293 configuration file which is writeable by the Exim user or group.
1295 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1296 of configuration files to cover files specified with the -C option if
1297 they are going to be used with root privileges, not just the default
1300 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1301 option (effectively making it always true).
1303 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1304 files to be used while preserving root privileges.
1306 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1307 that rogue child processes cannot use them.
1309 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1310 run-time user, instead of root.
1312 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1313 Exim run-time user without dropping privileges.
1315 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1316 result string, instead of calling string_vformat() twice with the same
1319 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1320 for other users. Others should always drop root privileges if they use
1321 -C on the command line, even for a whitelisted configure file.
1323 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1325 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1331 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1332 $data_path, and $header_path variables; fixed documentation bugs and
1335 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1336 exipick to access non-standard spools, including the "frozen" queue
1339 NM/01 Bugzilla 965: Support mysql stored procedures.
1340 Patch from Alain Williams
1342 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1344 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1345 Patch from Andreas Metzler
1347 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1348 Patch from Kirill Miazine
1350 NM/05 Bugzilla 671: Added umask to procmail example.
1352 JJ/03 installed exipick 20100323.0, fixing doc bug
1354 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1355 directory. Notification and patch from Dan Rosenberg.
1357 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1359 TK/02 Improve log output when DKIM signing operation fails.
1361 MH/01 Treat the transport option dkim_domain as a colon separated
1362 list, not as a single string, and sign the message with each element,
1363 omitting multiple occurences of the same signer.
1365 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1366 Bugzilla 985, 986. Patch by Simon Arlott
1368 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1369 Patch by Simon Arlott
1371 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1372 MBX locking. Notification from Dan Rosenberg.
1378 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1380 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1382 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1384 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1386 NM/04 Bugzilla 915: Flag broken perl installation during build.
1392 TK/01 Added patch by Johannes Berg that expands the main option
1393 "spamd_address" if it starts with a dollar sign.
1395 TK/02 Write list of recipients to X-Envelope-Sender header when building
1396 the mbox-format spool file for content scanning (suggested by Jakob
1399 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1400 (http://www.dcc-servers.net/) support via dccifd. Activated by
1401 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1403 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1404 by Mark Daniel Reidel <mr@df.eu>.
1406 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1407 When building exim an external PCRE library is now needed -
1408 PCRE is a system library on the majority of modern systems.
1409 See entry on PCRE_LIBS in EDITME file.
1411 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1412 conversation. Added nologin parameter to request.
1413 Patch contributed by Kirill Miazine.
1415 TF/01 Do not log submission mode rewrites if they do not change the address.
1417 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1419 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1420 log files in place. Contributed by Roberto Lima.
1422 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1424 TF/03 Bugzilla 615: When checking the local_parts router precondition
1425 after a local_part_suffix or local_part_prefix option, Exim now
1426 does not use the address's named list lookup cache, since this
1427 contains cached lookups for the whole local part.
1429 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1430 Robert Millan. Documentation is in experimental-spec.txt.
1432 TF/04 Bugzilla 668: Fix parallel build (make -j).
1434 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1436 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1437 Patch provided by Jan Srzednicki.
1439 TF/05 Leading white space used to be stripped from $spam_report which
1440 wrecked the formatting. Now it is preserved.
1442 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1443 that they are available at delivery time.
1445 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1447 TF/08 TLS error reporting now respects the incoming_interface and
1448 incoming_port log selectors.
1450 TF/09 Produce a more useful error message if an SMTP transport's hosts
1451 setting expands to an empty string.
1453 NM/06 Bugzilla 744: EXPN did not work under TLS.
1454 Patch provided by Phil Pennock.
1456 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1457 Patch provided by Richard Godbee.
1459 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1460 acl_smtp_notquit, added index entry.
1462 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1463 Patch provided by Eugene Bujak.
1465 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1466 accept(). Patch provided by Maxim Dounin.
1468 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1469 Patch provided by Phil Pennock.
1471 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1473 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1474 Patch provided by Brad "anomie" Jorsch.
1476 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1477 Patch provided by Dean Brooks.
1479 TK/05 Add native DKIM support (does not depend on external libraries).
1481 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1482 Patch provided by Graeme Fowler.
1484 NM/16 Bugzilla 851: Documentation example syntax fix.
1486 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1488 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1491 NM/19 Bugzilla 745: TLS version reporting.
1492 Patch provided by Phil Pennock.
1494 NM/20 Bugzilla 167: bool: condition support.
1495 Patch provided by Phil Pennock.
1497 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1498 clients. Patch provided by Phil Pennock.
1500 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1501 Patch provided by Brad "anomie" Jorsch.
1503 NM/23 Bugzilla 687: Fix misparses in eximstats.
1504 Patch provided by Heiko Schlittermann.
1506 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1507 Patch provided by Heiko Schlittermann.
1509 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1510 plus update to original patch.
1512 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1514 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1515 Patch provided by David Brownlee.
1517 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1519 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1521 NM/30 Bugzilla 888: TLS documentation bugfixes.
1523 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1525 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1526 Unlike the original bugzilla I have changed all shell scripts in src tree.
1528 NM/33 Bugzilla 898: Transport filter timeout fix.
1529 Patch by Todd Rinaldo.
1531 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1532 Patch by Serge Demonchaux.
1534 NM/35 Bugzilla 39: Base64 decode bug fixes.
1535 Patch by Jakob Hirsch.
1537 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1539 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1541 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1543 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1549 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1550 ALT-N's libdkim that I have put here:
1551 http://duncanthrax.net/exim-experimental/
1553 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1554 were called 'true' and it seems that C99 defines that as a reserved
1555 keyword to be used with 'bool' variable types. That means you could
1556 not include C99-style headers which use bools without triggering
1557 build errors in sieve.c.
1559 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1560 as mailq or other aliases. Changed the --help handling significantly
1561 to do whats expected. exim_usage() emits usage/help information.
1563 SC/01 Added the -bylocaldomain option to eximstats.
1565 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1567 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1569 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1575 PH/01 Another patch from the Sieve maintainer.
1577 PH/02 When an IPv6 address is converted to a string for single-key lookup
1578 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1579 dots are used instead of colons so that keys in lsearch files need not
1580 contain colons. This was done some time before quoting was made available
1581 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1582 (notated using the quote facility) so as to distinguish them from IPv4
1583 keys. This meant that lookups for IP addresses in host lists did not work
1584 for iplsearch lookups.
1586 This has been fixed by arranging for IPv6 addresses to be expressed with
1587 colons if the lookup type is iplsearch. This is not incompatible, because
1588 previously such lookups could never work.
1590 The situation is now rather anomolous, since one *can* have colons in
1591 ordinary lsearch keys. However, making the change in all cases is
1592 incompatible and would probably break a number of configurations.
1594 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1597 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1598 conversion specification without a maximum field width, thereby enabling
1599 a rogue spamd server to cause a buffer overflow. While nobody in their
1600 right mind would setup Exim to query an untrusted spamd server, an
1601 attacker that gains access to a server running spamd could potentially
1602 exploit this vulnerability to run arbitrary code as the Exim user.
1604 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1605 $primary_hostname instead of what libspf2 thinks the hosts name is.
1607 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1608 a directory entry by the name of the lookup key. Previously, if a
1609 symlink pointed to a non-existing file or a file in a directory that
1610 Exim lacked permissions to read, a lookup for a key matching that
1611 symlink would fail. Now it is enough that a matching directory entry
1612 exists, symlink or not. (Bugzilla 503.)
1614 PH/03 The body_linecount and body_zerocount variables are now exported in the
1617 PH/04 Added the $dnslist_matched variable.
1619 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1620 This means they are set thereafter only if the connection becomes
1623 PH/06 Added the client_condition to authenticators so that some can be skipped
1624 by clients under certain conditions.
1626 PH/07 The error message for a badly-placed control=no_multiline_responses left
1627 "_responses" off the end of the name.
1629 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1631 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1632 (without spaces) instead of just copying the configuration text.
1634 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1636 PH/11 Added $max_received_linelength.
1638 PH/12 Added +ignore_defer and +include_defer to host lists.
1640 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1641 way in which PCRE > 7.0 is built.
1643 PH/14 Implemented queue_only_load_latch.
1645 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1646 MAIL command. The effect was to mangle the value on 64-bit systems.
1648 PH/16 Another patch from the Sieve maintainer.
1650 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1652 PH/18 If a system quota error occurred while trying to create the file for
1653 a maildir delivery, the message "Mailbox is full" was not appended to the
1654 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1655 only to a quota excession during the actual writing of the file.
1657 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1658 characters?) which causes problems in log lines. The DN values are now
1659 passed through string_printing() before being added to log lines.
1661 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1662 and InterBase are left for another time.)
1664 PH/21 Added message_body_newlines option.
1666 PH/22 Guard against possible overflow in moan_check_errorcopy().
1668 PH/23 POSIX allows open() to be a macro; guard against that.
1670 PH/24 If the recipient of an error message contained an @ in the local part
1671 (suitably quoted, of course), incorrect values were put in $domain and
1672 $local_part during the evaluation of errors_copy.
1678 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1679 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1682 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1683 issue a MAIL command.
1685 PH/02 In an ACL statement such as
1687 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1689 if a client was not listed at all, or was listed with a value other than
1690 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1691 the condition was not true (as it should be), so access was not denied.
1692 The bug was that the ! inversion was incorrectly passed on to the second
1693 item. This has been fixed.
1695 PH/03 Added additional dnslists conditions == and =& which are different from
1696 = and & when the dns lookup returns more than one IP address.
1698 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1699 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1701 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1702 FSYNC, which compiles an option called disable_fsync that allows for
1703 bypassing fsync(). The documentation is heavily laced with warnings.
1705 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1707 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1708 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1709 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1710 including adding "make clean"; (3) Added -fPIC when compiling the test
1711 dynamically loaded module, to get rid of a warning.
1713 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1714 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1715 = 0s. The bug is otherwise harmless.
1717 PH/07 There was a bug in the dovecot authenticator such that the value of
1718 $auth1 could be overwritten, and so not correctly preserved, after a
1719 successful authentication. This usually meant that the value preserved by
1720 the server_setid option was incorrect.
1722 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1724 PH/09 Installed PCRE release 7.0.
1726 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1727 run for batched SMTP input. It is now run at the start of every message
1728 in the batch. While fixing this I discovered that the process information
1729 (output by running exiwhat) was not always getting set for -bs and -bS
1730 input. This is fixed, and it now also says "batched" for BSMTP.
1732 PH/11 Added control=no_pipelining.
1734 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1735 patch, slightly modified), and move the expansion of helo_data till after
1736 the connection is made in the smtp transport (so it can use these
1739 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1741 PH/14 Added log_selector = +pid.
1743 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1745 PH/16 Add ${if forany and ${if forall.
1747 PH/17 Added dsn_from option to vary the From: line in DSNs.
1749 PH/18 Flush SMTP output before performing a callout, unless control =
1750 no_callout_flush is set.
1752 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1753 was true (the default) a successful delivery failed to delete the retry
1754 item, thus causing premature timeout of the address. The bug is now
1757 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1759 PH/21 Long custom messages for fakedefer and fakereject are now split up
1760 into multiline reponses in the same way that messages for "deny" and
1761 other ACL rejections are.
1763 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1764 with slight modification.
1766 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1767 draft, changing the syntax and factoring some duplicate code".
1769 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1770 for deliveries of the second and subsequent messages over the same SMTP
1773 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1774 ${reduce, with only minor "tidies".
1776 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1778 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1779 expansion side effects.
1781 PH/27 When a message times out after an over-quota error from an Exim-imposed
1782 quota, the bounce message says "mailbox is full". This message was not
1783 being given when it was a system quota that was exceeded. It now should
1786 MH/03 Made $recipients available in local_scan(). local_scan() already has
1787 better access to the recipient list through recipients_list[], but
1788 $recipients can be useful in postmaster-provided expansion strings.
1790 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1791 in the case of a MAIL command with additional options following the
1792 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1793 were accidentally chopped off.
1795 PH/29 SMTP synchronization checks are implemented when a command is read -
1796 there is a check that no more input is waiting when there shouldn't be
1797 any. However, for some commands, a delay in an ACL can mean that it is
1798 some time before the response is written. In this time, more input might
1799 arrive, invalidly. So now there are extra checks after an ACL has run for
1800 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1801 pipelining has not been advertised.
1803 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1805 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1806 returned in the SMTP response when smtp_return_error_details was false.
1807 This has been fixed.
1809 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1810 instead of the C I/O that was originally supplied, because problems were
1811 reported on Solaris.
1813 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1814 Exim which did not show up earlier: it was assuming that a call to
1815 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1816 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1817 was a macro that became an assignment, so it seemed to work. This has
1818 changed to a proper function call with a void return, hence the compile
1819 error. Exim's code has been fixed.
1821 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1824 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1826 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1828 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1829 because it turns out to be too big or there is a timeout) there is no
1830 "Completed" line in the log. When some messages of this type were
1831 selected by exigrep, they were listed as "not completed". Others were
1832 picked up by some special patterns. I have improved the selection
1833 criteria to be more general.
1835 PH/38 The host_find_failed option in the manualroute router can now be set
1836 to "ignore", to completely ignore a host whose IP address cannot be
1837 found. If all hosts are ignored, the behaviour is controlled by the new
1838 host_all_ignored option.
1840 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1841 homing or because of multiple MX records with /mx) generated more than
1842 one IP address, and the following item turned out to be the local host,
1843 all the secondary addresses of the first item were incorrectly removed
1844 from the list, along with the local host and any following hosts (which
1845 is what is supposed to happen).
1847 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1848 whoever called Exim into the -H file. In the case of the daemon it was
1849 behaving confusingly. When first started, it used values for whoever
1850 started the daemon, but after a SIGHUP it used the Exim user (because it
1851 calls itself on a restart). I have changed the code so that it now always
1854 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1855 message are rejected with the same error (e.g. no authentication or bad
1856 sender address), and a DATA command is nevertheless sent (as can happen
1857 with PIPELINING or a stupid MUA), the error message that was given to the
1858 RCPT commands is included in the rejection of the DATA command. This is
1859 intended to be helpful for MUAs that show only the final error to their
1862 PH/42 Another patch from the Sieve maintainer.
1864 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1865 Eximstats - Fixed some broken HTML links and added missing column headers
1867 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1868 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1870 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1872 PH/43 Yet another patch from the Sieve maintainer.
1874 PH/44 I found a way to check for a TCP/IP connection going away before sending
1875 the response to the final '.' that terminates a message, but only in the
1876 case where the client has not sent further data following the '.'
1877 (unfortunately, this is allowed). However, in many cases there won't be
1878 any further data because there won't be any more messages to send. A call
1879 to select() can be used: if it shows that the input is "ready", there is
1880 either input waiting, or the socket has been closed. An attempt to read
1881 the next input character can distinguish the two cases. Previously, Exim
1882 would have sent an OK response which the client would never have see.
1883 This could lead to message repetition. This fix should cure that, at
1884 least in a lot of common cases.
1886 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1887 advertised in response to EHLO.
1893 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1894 fixed by 4.65/MH/01 (is this a record?) are fixed:
1896 (i) An empty string was always treated as zero by the numeric comparison
1897 operators. This behaviour has been restored.
1899 (ii) It is documented that the numeric comparison operators always treat
1900 their arguments as decimal numbers. This was broken in that numbers
1901 starting with 0 were being interpreted as octal.
1903 While fixing these problems I realized that there was another issue that
1904 hadn't been noticed. Values of message_size_limit (both the global option
1905 and the transport option) were treated as octal if they started with 0.
1906 The documentation was vague. These values are now always treated as
1907 decimal, and I will make that clear in the documentation.
1913 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1914 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1917 MH/01 Don't check that the operands of numeric comparison operators are
1918 integers when their expansion is in "skipping" mode (fixes bug
1919 introduced by 4.64-PH/07).
1921 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1922 child addresses, Exim now panics and dies. Previously, because the count
1923 is held in a short int, deliveries were likely to be lost. As such a
1924 large number of recipients for a single message is ridiculous
1925 (performance will be very, very poor), I have chosen to impose a limit
1926 rather than extend the field.
1932 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1933 leftover -K file (the existence of which was triggered by #402).
1934 While we were at it, introduced process PID as part of the -K
1935 filename. This should rule out race conditions when creating
1938 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1939 processing considerably. Previous code took too long for large mails,
1940 triggering a timeout which in turn triggers #401.
1942 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1943 in the DK code in transports.c. sendfile() is not really portable,
1944 hence the _LINUX specificness.
1946 TF/01 In the add_headers option to the mail command in an Exim filter,
1947 there was a bug that Exim would claim a syntax error in any
1948 header after the first one which had an odd number of characters
1951 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1952 callout verification, Exim cached a "reject" for the entire domain. This
1953 is correct for most verifications, but it is not correct for a recipient
1954 verification with use_sender or use_postmaster set, because in that case
1955 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1956 case of MAIL FROM:<> rejection from other early rejections (e.g.
1957 rejection of HELO). When verifying a recipient using a non-null MAIL
1958 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1959 Whatever the result of the callout, the value of the domain cache is
1960 left unchanged (for any other kind of callout, getting as far as trying
1961 RCPT means that the domain itself is ok).
1963 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1966 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1967 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1968 session, the daemon ignores ECONNECT errors and logs others; it now
1969 ignores EPIPE as well.
1971 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1972 (quoted-printable decoding).
1974 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1975 later the small subsequent patch to fix an introduced bug.
1977 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1979 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1981 PH/08 An error is now given if message_size_limit is specified negative.
1983 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1984 to be given (somewhat) arbitrary names.
1986 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1989 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1990 miscellaneous code fixes
1992 PH/10 Added the log_reject_target ACL modifier to specify where to log
1995 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1996 hostname. This is wrong, because it relates to the incoming message (and
1997 probably the interface on which it is arriving) and not to the outgoing
1998 callout (which could be using a different interface). This has been
1999 changed to use the value of the helo_data option from the smtp transport
2000 instead - this is what is used when a message is actually being sent. If
2001 there is no remote transport (possible with a router that sets up host
2002 addresses), $smtp_active_hostname is used.
2004 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2005 tweaks were necessary in order to get it to work (see also 21 below):
2006 (a) The code assumed that strncpy() returns a negative number on buffer
2007 overflow, which isn't the case. Replaced with Exim's string_format()
2009 (b) There were several signed/unsigned issues. I just did the minimum
2010 hacking in of casts. There is scope for a larger refactoring.
2011 (c) The code used strcasecmp() which is not a standard C function.
2012 Replaced with Exim's strcmpic() function.
2013 (d) The code set only $1; it now sets $auth1 as well.
2014 (e) A simple test gave the error "authentication client didn't specify
2015 service in request". It would seem that Dovecot has changed its
2016 interface. Fortunately there's a specification; I followed it and
2017 changed what the client sends and it appears to be working now.
2019 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2022 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2023 address A is aliased to B and C, where B exists and C does not. Without
2024 -v the output is "A verified" because verification stops after a
2025 successful redirection if more than one address is generated. However,
2026 with -v the child addresses are also verified. Exim was outputting "A
2027 failed to verify" and then showing the successful verification for C,
2028 with its parentage. It now outputs "B failed to verify", showing B's
2029 parentage before showing the successful verification of C.
2031 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2032 look up a TXT record in a specific list after matching in a combined
2035 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2036 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2037 they consult the DNS. I had assumed they would set it the way they
2038 wanted; and indeed my experiments on Linux seem to show that in some
2039 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2040 To be on the safe side, however, I have now made the interface to
2041 host_find_byname() similar to host_find_bydns(), with an argument
2042 containing the DNS resolver options. The host_find_byname() function now
2043 sets these options at its start, just as host_find_bydns() does. The smtp
2044 transport options dns_qualify_single and dns_search_parents are passed to
2045 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2046 of host_find_byname() use the default settings of RES_DEFNAMES
2047 (qualify_single) but not RES_DNSRCH (search_parents).
2049 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2050 spool_read_header() do less string testing, by means of a preliminary
2051 switch on the second character of optional "-foo" lines. (This is
2052 overdue, caused by the large number of possibilities that now exist.
2053 Originally there were few.) While I was there, I also converted the
2054 str(n)cmp tests so they don't re-test the leading "-" and the first
2055 character, in the hope this might squeeze out yet more improvement.
2057 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2058 flag allowing group syntax was set by the header_syntax check but not
2059 turned off, possible causing trouble later; (2) The flag was not being
2060 set at all for the header_verify test, causing "group"-style headers to
2061 be rejected. I have now set it in this case, and also caused header_
2062 verify to ignore an empty address taken from a group. While doing this, I
2063 came across some other cases where the code for allowing group syntax
2064 while scanning a header line wasn't quite right (mostly, not resetting
2065 the flag correctly in the right place). These bugs could have caused
2066 trouble for malformed header lines. I hope it is now all correct.
2068 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2069 with the "reply" argument non-NULL. The code, however (which originally
2070 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2071 but it didn't always do it. This confused somebody who was copying the
2072 code for some other use. I have removed all the tests.
2074 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2075 feature that was used to support insecure browsers during the U.S. crypto
2076 embargo. It requires special client support, and Exim is probably the
2077 only MTA that supported it -- and would never use it because real RSA is
2078 always available. This code has been removed, because it had the bad
2079 effect of slowing Exim down by computing (never used) parameters for the
2080 RSA_EXPORT functionality.
2082 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2083 authenticator to fail if there's a tab character in the incoming data
2084 (there should never be unless someone is messing about, as it's supposed
2085 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2086 if the connection is using TLS or if the remote IP is the same as the
2087 local IP, and the "valid-client-cert option" if a client certificate has
2090 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2091 authenticators. This can be used for authorization after authentication
2092 succeeds. (In the case of plaintext, it servers for both authentication
2095 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2096 if any retry times were supplied.
2098 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2099 connection, where there is no client IP address to check. In this
2100 situation, the verify now always succeeds.
2102 PH/25 Applied John Jetmore's -Mset patch.
2104 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2106 PH/27 In a string expansion for a processed (not raw) header when multiple
2107 headers of the same name were present, leading whitespace was being
2108 removed from all of them, but trailing whitespace was being removed only
2109 from the last one. Now trailing whitespace is removed from each header
2110 before concatenation. Completely empty headers in a concatenation (as
2111 before) are ignored.
2113 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2114 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2116 PH/29 [Removed. This was a change that I later backed out, and forgot to
2117 correct the ChangeLog entry (that I had efficiently created) before
2118 committing the later change.]
2120 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2121 address errors (4xx response to RCPT) over the same connection as other
2122 messages routed to the same hosts. Such deliveries are always "forced",
2123 so retry times are not inspected. This resulted in far too many retries
2124 for the affected addresses. The effect occurred only when there were more
2125 hosts than the hosts_max_try setting in the smtp transport when it had
2126 the 4xx errors. Those hosts that it had tried were not added to the list
2127 of hosts for which the message was waiting, so if all were tried, there
2128 was no problem. Two fixes have been applied:
2130 (i) If there are any address or message errors in an SMTP delivery, none
2131 of the hosts (tried or untried) are now added to the list of hosts
2132 for which the message is waiting, so the message should not be a
2133 candidate for sending over the same connection that was used for a
2134 successful delivery of some other message. This seems entirely
2135 reasonable: after all the message is NOT "waiting for some host".
2136 This is so "obvious" that I'm not sure why it wasn't done
2137 previously. Hope I haven't missed anything, but it can't do any
2138 harm, as the worst effect is to miss an optimization.
2140 (ii) If, despite (i), such a delivery is accidentally attempted, the
2141 routing retry time is respected, so at least it doesn't keep
2142 hammering the server.
2144 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2145 in ${readsocket because some servers need this prod.
2147 PH/32 Added some extra debug output when updating a wait-xxx database.
2149 PH/33 The hint "could be header name not terminated by colon", which has been
2150 given for certain expansion errors for a long time, was not being given
2151 for the ${if def:h_colon_omitted{... case.
2153 PH/34 The spec says: "With one important exception, whenever a domain list is
2154 being scanned, $domain contains the subject domain." There was at least
2155 one case where this was not true.
2157 PH/35 The error "getsockname() failed: connection reset by peer" was being
2158 written to the panic log as well as the main log, but it isn't really
2159 panic-worthy as it just means the connection died rather early on. I have
2160 removed the panic log writing for the ECONNRESET error when getsockname()
2163 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2164 runs only) independently of the message's sender address. This meant
2165 that, if the 4xx error was in fact related to the sender, a different
2166 message to the same recipient with a different sender could confuse
2167 things. In particualar, this can happen when sending to a greylisting
2168 server, but other circumstances could also provoke similar problems.
2169 I have changed the default so that the retry time for these errors is now
2170 based a combination of the sender and recipient addresses. This change
2171 can be overridden by setting address_retry_include_sender=false in the
2174 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2175 remote server are returned as part of bounce messages. This was not
2176 happening for LMTP over a pipe (the lmtp transport), but now it is the
2177 same for both kinds of LMTP.
2179 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2180 in header lines that were in fact CNAMEs. This is no longer the case.
2182 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2183 and queue runs started by the daemon processed all messages. This has
2184 been fixed so that -R and -S can now usefully be given with -q<time>.
2186 PH/40 Import PCRE release 6.7 (fixes some bugs).
2188 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2190 PH/42 Give an error if -q is specified more than once.
2192 PH/43 Renamed the variables $interface_address and $interface_port as
2193 $received_ip_address and $received_port, to make it clear that these
2194 values apply to message reception, and not to the outgoing interface when
2195 a message is delivered. (The old names remain recognized, of course.)
2197 PH/44 There was no timeout on the connect() call when using a Unix domain
2198 socket in the ${readsocket expansion. There now is.
2200 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2201 be meaningful with "accept".
2203 SC/01 Eximstats V1.43
2204 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2206 SC/02 Eximstats V1.44
2207 Use a glob alias rather than an array ref in the generated
2208 parser. This improves both readability and performance.
2210 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2211 Collect SpamAssassin and rejection statistics.
2212 Don't display local sender or destination tables unless
2213 there is data to show.
2214 Added average volumes into the top table text output.
2216 SC/04 Eximstats V1.46
2217 Collect data on the number of addresses (recipients)
2218 as well as the number of messages.
2220 SC/05 Eximstats V1.47
2221 Added 'Message too big' to the list of mail rejection
2222 reasons (thanks to Marco Gaiarin).
2224 SC/06 Eximstats V1.48
2225 Mainlog lines which have GMT offsets and are too short to
2226 have a flag are now skipped.
2228 SC/07 Eximstats V1.49 (Alain Williams)
2229 Added the -emptyok flag.
2231 SC/08 Eximstats V1.50
2232 Fixes for obtaining the IP address from reject messages.
2234 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2235 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2236 whitesspace changes from 4.64-PH/27
2238 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2241 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2242 are found, allow negative numbers in numeric criteria)
2244 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2246 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2247 to match changes made in 4.64-PH/43
2249 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2251 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2252 "domain" is actually an IP address, and force a failure. This locks out
2253 those revolvers/nameservers that support "A-for-A" lookups, in
2254 contravention of the specifications.
2256 PH/48 When a host name was looked up from an IP address, and the subsequent
2257 forward lookup of the name timed out, the host name was left in
2258 $sender_host_name, contrary to the specification.
2260 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2261 restricted to single-key lookups, Exim was not diagnosing an error if
2262 * or *@ was used with a query-style lookup.
2264 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2266 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2267 long ago, but noone interested enough thought of it. Let's just say that
2268 the "1.1" means that there are some new functions that weren't there at
2269 some point in the past.
2271 PH/51 Error processing for expansion failure of helo_data from an smtp
2272 transport during callout processing was broken.
2274 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2275 tested/used via the -bh/-bhc/-bs options.
2277 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2278 bug, fixed in subsequent PCRE releases).
2280 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2281 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2283 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2289 SC/01 Use a glob alias rather than an array ref in eximstats generated
2290 parser. This improves both readability and performance.
2292 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2293 Don't display local sender or destination tables in eximstats unless
2294 there is data to show.
2295 Added average volumes into the eximstats top table text output.
2297 SC/03 Collect data on the number of addresses (recipients) as well
2298 as the number of messages in eximstats.
2300 TF/01 Correct an error in the documentation for the redirect router. Exim
2301 does (usually) call initgroups() when daemonizing.
2303 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2304 with consistent privilege compared to when running as a daemon.
2306 TF/03 Note in the spec that $authenticated_id is not set for local
2307 submissions from trusted users.
2309 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2310 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2312 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2313 by adding some example configuration directives to the default
2314 configuration file. A little bit of work is required to uncomment the
2315 directives and define how usernames and passwords are checked, but
2316 there is now a framework to start from.
2318 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2319 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2320 without this. I don't know how relevant this is to other LDAP libraries.
2322 PH/02 Add the verb name to the "unknown ACL verb" error.
2324 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2326 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2328 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2329 directory not expanded when it should be if an expanded home directory
2330 was set for the address (which is overridden by the transport).
2332 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2335 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2336 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2337 because it is too late at that time, and has no effect.
2339 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2340 security issue with \' (bugzilla #107). I could not use the
2341 PQescapeStringConn() function, because it needs a PGconn value as one of
2344 PH/08 When testing addresses using -bt, indicate those final addresses that
2345 are duplicates that would not cause an additional delivery. At least one
2346 person was confused, thinking that -bt output corresponded to deliveries.
2347 (Suppressing duplicates isn't a good idea as you lose the information
2348 about possibly different redirections that led to the duplicates.)
2350 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2351 systems where poll() doesn't work, in particular OS X.
2353 PH/10 Added more information to debugging output for retry time not reached.
2355 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2356 operations in malware.c.
2358 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2361 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2362 syslog_duplication set false, log lines that would normally be written
2363 both the the main log and to the reject log were not written to syslog at
2366 PH/14 In the default configuration, change the use of "message" in ACL warn
2367 statements to "add_header".
2369 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2370 not followed by a command (e.g. "seen endif").
2372 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2373 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2376 PH/17 Added extra conditions to the default value of delay_warning_condition
2380 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2381 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2382 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2385 The Auto-Submitted: and various List- headers are standardised, whereas I
2386 don't think Precedence: ever was.
2388 PH/18 Refactored debugging code in route_finduser() to show more information,
2389 in particular, the error code if getpwnam() issues one.
2391 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2392 This is apparently needed in addition to the PH/07 change above to avoid
2393 any possible encoding problems.
2395 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2396 but not after initializing Perl.
2398 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2399 output them only if debugging. By default they are written stderr,
2400 apparently, which is not desirable.
2402 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2405 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2408 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2410 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2411 authenticated or an ident call has been made. Suppress the default
2412 values for $authenticated_id and $authenticated_sender (but permit -oMai
2413 and -oMas) when testing with -bh.
2415 PH/24 Re-jigged the order of the tests in the default configuration so that the
2416 tests for valid domains and recipients precede the DNS black list and CSA
2417 tests, on the grounds that those ones are more expensive.
2419 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2420 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2421 This bug exists in every version of Exim that I still have, right back to
2424 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2425 However, an attempt to turn on case-sensitivity in a regex key by
2426 including (?-i) didn't work because the subject string was already
2427 lowercased, and the effects were non-intuitive. It turns out that a
2428 one-line patch can be used to allow (?-i) to work as expected.
2434 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2435 other effects) broke the use of negated acl sub-conditions.
2437 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2440 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2441 "Deny" causes Exim to reject the incoming connection with a 554 error.
2442 Unfortunately, if there is a major crisis, such as a disk failure,
2443 tcp-wrappers gives "deny", whereas what one would like would be some
2444 kind of temporary error. A kludge has been added to help with this.
2445 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2446 554 error is used if errno is still zero or contains ENOENT (which occurs
2447 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2450 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2452 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2453 errors. Otherwise a message that provokes a temporary error (when other
2454 messages do not) can cause a whole host to time out.
2456 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2457 addresses were routed directly to files or pipes from a redirect router.
2458 File deliveries just didn't batch; pipe deliveries might have suffered
2461 PH/06 A failure to get a lock for a hints database would erroneously always say
2462 "Failed to get write lock", even when it was really a read lock.
2464 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2465 of 0600. This has been changed to use the value of the lockfile_mode
2466 option (which defaults to 0600).
2468 PH/08 Applied small patch from the Sieve maintainer.
2470 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2471 folder from quota calculations, a direct delivery into this folder messed
2472 up the contents of the maildirsize file. This was because the regex was
2473 used only to exclude .Trash (or whatever) when the size of the mailbox
2474 was calculated. There was no check that a delivery was happening into an
2475 excluded directory. This bug has been fixed by ignoring all quota
2476 processing for deliveries into excluded directories.
2478 PH/10 Added the maildirfolder_create_regex option to appendfile.
2484 PH/01 The code for finding all the local interface addresses on a FreeBSD
2485 system running IPv6 was broken. This may well have applied to all BSD
2486 systems, as well as to others that have similar system calls. The broken
2487 code found IPv4 interfaces correctly, but gave incorrect values for the
2488 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2489 that it would not match correctly against @[] and not recognize the IPv6
2492 PH/02 The ipliteral router was not recognizing addresses of the form user@
2493 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2495 PH/03 Added disable_ipv6.
2497 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2498 decoded form, because it is most often used to construct To: headers
2499 lines in autoreplies, and the decoded form may well be syntactically
2500 invalid. However, $reply_address has leading white space removed, and all
2501 newlines turned into spaces so that the autoreply transport does not
2504 PH/05 If group was specified without a user on a router, and no group or user
2505 was specified on a transport, the group from the router was ignored.
2507 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2508 for visible compile-time settings that can be used to change these
2509 numbers, for those that want even more. Backwards compatibility with old
2510 spool files has been maintained. However, going back to a previous Exim
2511 release will lost any variables that are in spool files.
2513 PH/07 Two small changes when running in the test harness: increase delay when
2514 passing a TCP/IP connection to a new process, in case the original
2515 process has to generate a bounce, and remove special handling of
2516 127.0.0.2 (sic), which is no longer necessary.
2518 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2519 be the same on different OS.
2521 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2524 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2525 whether --show-vars was specified or not
2527 JJ/02 exipick: Added support for new ACL variable spool format introduced
2530 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2531 syntactically invalid From: or Reply-to: line, and a filter used this to
2532 generate an autoreply, and therefore failed to obtain an address for the
2533 autoreply, Exim could try to deliver to a non-existent relative file
2534 name, causing unrelated and misleading errors. What now happens is that
2535 it logs this as a hard delivery error, but does not attempt to create a
2538 PH/11 The exinext utility has a -C option for testing purposes, but although
2539 the given file was scanned by exinext itself; it wasn't being passed on
2540 when Exim was called.
2542 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2543 an end-of-file indication when reading a command response.
2545 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2546 compiled. In many other places in Exim, IPv6 addresses are always
2547 recognized, so I have changed this. It also means that IPv4 domain
2548 literals of the form [IPV4:n.n.n.n] are now always recognized.
2550 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2551 used if the router is not running as root, for example, when verifying at
2552 ACL time, or when using -bh. The debugging output from this situation was
2553 non-existent - all you got was a failure to exec. I have made two
2556 (a) Failures to set uid/gid, the current directory, or a process leader
2557 in a subprocess such as that created by queryprogram now generate
2558 suitable debugging ouput when -d is set.
2560 (b) The queryprogram router detects when it is not running as root,
2561 outputs suitable debugging information if -d is set, and then runs
2562 the subprocess without attempting to change uid/gid.
2564 PH/15 Minor change to Makefile for building test_host (undocumented testing
2567 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2568 additional section of a DNS packet that returns MX or SRV records.
2569 Instead, it always explicitly searches for A/AAAA records. This avoids
2570 major problems that occur when a DNS server includes only records of one
2571 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2572 fixed another bug: if SRV records were looked up and the corresponding
2573 address records were *not* found in the additional section, the port
2574 values from the SRV records were lost.
2576 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2577 using the correct key (the original address) when searching the retry
2578 rules in order to find which one to use for generating the retry hint.
2580 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2581 adding the default one. Similarly, if it contains a Reply-To: header, the
2582 errors_reply_to option, if set, is not used.
2584 PH/19 When calculating a retry time, Exim used to measure the "time since
2585 failure" by looking at the "first failed" field in the retry record. Now
2586 it does not use this if it is later than than the arrival time of the
2587 message. Instead it uses the arrival time. This makes for better
2588 behaviour in cases where some deliveries succeed, thus re-setting the
2589 "first failed" field. An example is a quota failure for a huge message
2590 when small messages continue to be delivered. Without this change, the
2591 "time since failure" will always be short, possible causing more frequent
2592 delivery attempts for the huge message than are intended.
2593 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2595 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2596 $1, $2, $3) because the numerical variables can be reset during some
2597 expansion items (e.g. "match"), thereby losing the authentication data.
2599 PH/21 Make -bV show the size of off_t variables so that the test suite can
2600 decide whether to run tests for quotas > 2G.
2602 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2603 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2604 filecount value is greater than 2G or if a quota value is greater than 2G
2605 on a system where the size of off_t is not greater than 4, a panic error
2608 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2609 never match. The debug and -bh output now contains an explicit error
2610 message indicating a malformed IPv4 address or mask.
2612 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2613 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2614 PH/23 above applies.
2616 PH/25 Do not write to syslog when running in the test harness. The only
2617 occasion when this arises is a failure to open the main or panic logs
2618 (for which there is an explicit test).
2620 PH/26 Added the /no_tell option to "control=freeze".
2622 PH/27 If a host name lookup failed very early in a connection, for example, if
2623 the IP address matched host_lookup and the reverse lookup yielded a name
2624 that did not have a forward lookup, an error message of the form "no IP
2625 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2626 could be logged. Now it outputs the IP address instead of "NULL".
2628 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2629 allows the sender and the authenticated sender to be set when
2630 submitting a message from within Exim. Since child_open_exim() is
2631 documented for local_scan(), the new function should be too.
2633 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2634 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2635 results in an empty string is now treated as unset.
2637 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2639 PH/31 Added sender_verify_fail logging option.
2641 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2642 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2643 tidied the source and removed it altogether.
2645 PH/33 When a queue run was abandoned because the load average was too high, a
2646 log line was always written; now it is written only if the queue_run log
2647 selector is set. In addition, the log line for abandonment now contains
2648 information about the queue run such as the pid. This is always present
2649 in "start" and "stop" lines but was omitted from the "abandon" line.
2651 PH/34 Omit spaces between a header name and the colon in the error message that
2652 is given when verify = headers_syntax fails (if there are lots of them,
2653 the message gets confusing).
2655 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2656 names, as there are now some PTR records that contain slashes. This check
2657 is only to protect against broken name servers that fall over on strange
2658 characters, so the fact that it applies to all lookups doesn't matter.
2660 PH/36 Now that the new test suite is complete, we can remove some of the
2661 special code in Exim that was needed for the old test suite. For example,
2662 sorting DNS records because real resolvers return them in an arbitrary
2663 order. The new test suite's fake resolver always returns records in the
2666 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2667 bounces) except when queue_only is set, to avoid logging races between
2668 the different processes.
2670 PH/38 Panic-die if .include specifies a non-absolute path.
2672 PH/39 A tweak to the "H" retry rule from its user.
2674 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2675 a label. They prevented compilation on older perls.
2677 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2678 a warning to be raised on newish perls.
2680 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2681 on queue. Changes to match documented behaviour of showing count of
2682 messages matching specified criteria.
2684 PH/40 Changed the default ident timeout from 30s to 5s.
2686 PH/41 Added support for the use of login_cap features, on those BSD systems
2687 that have them, for controlling the resources used by pipe deliveries.
2689 PH/42 The content-scanning code uses fopen() to create files in which to put
2690 message data. Previously it was not paying any attention to the mode of
2691 the files. Exim runs with umask(0) because the rest of the code creates
2692 files with open(), and sets the required mode explicitly. Thus, these
2693 files were ending up world-writeable. This was not a big issue, because,
2694 being within the spool directory, they were not world-accessible. I have
2695 created a function called modefopen, which takes an additional mode
2696 argument. It sets umask(777), creates the file, chmods it to the required
2697 mode, then resets the umask. All the relevant calls to fopen() in the
2698 content scanning code have been changed to use this function.
2700 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2701 to 24 hours. This avoids potential overflow problems when processing G
2702 and H retry rules. I suspect nobody ever tinkers with this value.
2704 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2706 PH/45 When the plaintext authenticator is running as a client, the server's
2707 challenges are checked to ensure they are valid base64 strings. By
2708 default, the authentication attempt is cancelled if an invalid string is
2709 received. Setting client_ignore_invalid_base64 true ignores these errors.
2710 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2711 they are received. Thus, the responses can be made to depend on the
2712 challenges. If an invalid string is ignored, an empty string is placed in
2715 PH/46 Messages that are created by the autoreply transport now contains a
2716 References: header, in accordance with RFCs 2822 and 3834.
2718 PH/47 Added authenticated_sender_force to the smtp transport.
2720 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2722 PH/49 Installed latest patch from the Sieve maintainer.
2724 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2725 was also set, the appendfile transport was unnecessarily scanning a
2726 directory of message files (e.g. for maildir delivery) to find the count
2727 of files (along with the size), even though it did not need this
2728 information. It now does the scan only if it needs to find either the
2729 size of the count of files.
2731 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2733 PH/52 Two bugs concerned with error handling when the smtp transport is
2736 (i) Exim was not creating retry information for temporary errors given
2737 for individual recipients after the DATA command when the smtp transport
2738 was used in LMTP mode. This meant that they could be retried too
2739 frequently, and not timed out correctly.
2741 (ii) Exim was setting the flag that allows error details to be returned
2742 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2743 recipients that were returned after the DATA command.
2745 PH/53 This is related to PH/52, but is more general: for any failing address,
2746 when detailed error information was permitted to be returned to the
2747 sender, but the error was temporary, then after the final timeout, only
2748 "retry timeout exceeded" was returned. Now it returns the full error as
2749 well as "retry timeout exceeded".
2751 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2752 do this, and (what is worse) MTAs that accept it.
2754 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2755 will now be deprecated.
2757 PH/56 New os.c-cygwin from the Cygwin maintainer.
2759 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2760 formats (previously only available in exim formats via -bpr, -bpru,
2761 and -bpra. Now also available in native and exiqgrep formats)
2763 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2764 with very large, slow to parse queues
2766 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2768 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2770 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2771 responses to authentication challenges, though it was showing the
2772 challenges; (ii) I've removed the CR characters from the debug output for
2775 PH/58 Allow for the insertion of a newline as well as a space when a string
2776 is turned into more than one encoded-word during RFC 2047 encoding. The
2777 Sieve code now uses this.
2779 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2780 data_4xx, lost_connection, tls_required.
2782 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2783 message was being sent as an SMTP response.
2785 PH/61 Add -l and -k options to exicyclog.
2787 PH/62 When verifying, if an address was redirected to one new address, so that
2788 verification continued, and the new address failed or deferred after
2789 having set something in $address_data, the value of $address_data was not
2790 passed back to the ACL. This was different to the case when no
2791 redirection occurred. The value is now passed back in both cases.
2793 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2794 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2795 use login_cap.h, so on its own it isn't the distinguishing feature. The
2796 new name refers directly to the setclassresources() function.
2798 PH/65 Added configuration files for NetBSD3.
2800 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2802 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2803 is preferred over IPv4.
2805 PH/68 The bounce_return_message and bounce_return_body options were not being
2806 honoured for bounces generated during the reception of non-SMTP messages.
2807 In particular, this applied to messages rejected by the ACL. This bug has
2808 been fixed. However, if bounce_return_message is true and bounce_return_
2809 body is false, the headers that are returned for a non-SMTP message
2810 include only those that have been read before the error was detected.
2811 (In the case of an ACL rejection, they have all been read.)
2813 PH/69 The HTML version of the specification is now built in a directory called
2814 spec_html instead of spec.html, because the latter looks like a path with
2815 a MIME-type, and this confuses some software.
2817 PH/70 Catch two compiler warnings in sieve.c.
2819 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2820 function verify_get_ident() calls ip_connect() to connect a socket, but
2821 if the "connect()" function timed out, ip_connect() used to close the
2822 socket. However, verify_get_ident() also closes the socket later, and in
2823 between Exim writes to the log, which may get opened at this point. When
2824 the socket was closed in ip_connect(), the log could get the same file
2825 descriptor number as the socket. This naturally causes chaos. The fix is
2826 not to close the socket in ip_connect(); the socket should be closed by
2827 the function that creates it. There was only one place in the code where
2828 this was missing, in the iplookup router, which I don't think anybody now
2829 uses, but I've fixed it anyway.
2831 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2832 well as to direct DNS lookups. Otherwise the handling of names in host
2833 lists is inconsistent and therefore confusing.
2839 PH/01 Two changes to the default runtime configuration:
2841 (1) Move the checks for relay_from_hosts and authenticated clients from
2842 after to before the (commented out) DNS black list checks.
2844 (2) Add control=submission to the relay_from_hosts and authenticated
2845 clients checks, on the grounds that messages accepted by these
2846 statements are most likely to be submissions.
2848 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2850 (1) Generate an error if the third argument for the ${prvs expansion is
2853 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2856 (3) Reset the variables that are obtained from the first argument of
2857 ${prvscheck and used in the second argument before leaving the code,
2858 because their memory is reclaimed, so using them afterwards may do
2861 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2862 one (it's much easier than Tom thought :-).
2864 (5) Because of (4), we can now allow for the use of $prvscheck_result
2865 inside the third argument.
2867 PH/03 For some reason, the default setting of PATH when running a command from
2868 a pipe transport was just "/usr/bin". I have changed it to
2871 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2872 anything to be listed in the output from -bV.
2874 PH/05 When a filter generated an autoreply, the entire To: header line was
2875 quoted in the delivery log line, like this:
2877 => >A.N.Other <ano@some.domain> <original@ddress> ...
2879 This has been changed so that it extracts the operative address. There
2880 may be more than one such address. If so, they are comma-separated, like
2883 => >ano@some.domain,ona@other.domain <original@ddress> ...
2885 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2886 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2887 not being looked up in the rDNS to get a host name, Exim was showing the
2888 IP address twice in Received: lines, even though the IP addresses were
2889 identical. For example:
2891 Received: from [1.2.3.4] (helo=[1.2.3.4])
2893 However, if the real host name was known, it was omitting the HELO data
2894 if it matched the actual IP address. This has been tidied up so that it
2895 doesn't show the same IP address twice.
2897 PH/07 When both +timestamp and +memory debugging was on, the value given by
2898 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2899 called by the debug printing, thereby overwriting the timestamp buffer.
2900 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2902 PH/08 When the original message was included in an autoreply transport, it
2903 always said "this is a copy of the message, including all the headers",
2904 even if body_only or headers_only was set. It now gives an appropriate
2907 PH/09 Applied a patch from the Sieve maintainer which:
2909 o fixes some comments
2910 o adds the (disabled) notify extension core
2911 o adds some debug output for the result of if/elsif tests
2912 o points to the current vacation draft in the documentation
2913 and documents the missing references header update
2917 o fixes a bug in processing the envelope test (when testing
2918 multiple envelope elements, the last element determinted the
2921 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2922 Electronic Mail") by including:
2924 Auto-submitted: auto-generated
2926 in the messages that it generates (bounce messages and others, such as
2927 warnings). In the case of bounce messages for non-SMTP mesages, there was
2928 also a typo: it was using "Auto_submitted" (underscore instead of
2929 hyphen). Since every message generated by Exim is necessarily in response
2930 to another message, thes have all been changed to:
2932 Auto-Submitted: auto-replied
2934 in accordance with these statements in the RFC:
2936 The auto-replied keyword:
2938 - SHOULD be used on messages sent in direct response to another
2939 message by an automatic process,
2941 - MUST NOT be used on manually-generated messages,
2943 - MAY be used on Delivery Status Notifications (DSNs) and Message
2944 Disposition Notifications (MDNs),
2946 - MUST NOT be used on messages generated by automatic or periodic
2947 processes, except for messages which are automatic responses to
2950 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2951 to the default Received: header definition.
2953 PH/12 Added log selector acl_warn_skipped (default on).
2955 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2956 variables because (a) they are in the wrong storage pool and (b) even if
2957 they were copied, it wouldn't work properly because of the caching.
2959 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2960 checking when decoding. Apparently there are clients that generate
2961 overlong encoded strings. Why am I not surprised?
2963 PH/15 If the first argument of "${if match_address" was not empty, but did not
2964 contain an "@" character, Exim crashed. Now it writes a panic log message
2965 and treats the condition as false.
2967 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2969 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2970 extension "envelope-auth". The code is finished and in agreement with
2971 other implementations, but there is no documentation so far and in fact,
2972 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2973 not changing the active code.
2975 Print executed "if" and "elsif" statements when debugging is used. This
2976 helps a great deal to understand what a filter does.
2978 Document more things not specified clearly in RFC3028. I had all this
2979 sorted out, when out of a sudden new issues came to my mind. Oops."
2981 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2984 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2985 canonical form (as documented). However, after a host name lookup from
2986 the IP address, check_host() was doing a simple string comparison with
2987 addresses acquired from the DNS when checking that the found name did
2988 have the original IP as one of its addresses. Since any found IPv6
2989 addresses are likely to be in abbreviated form, the comparison could
2990 fail. Luckily, there already exists a function for doing the comparison
2991 by converting both addresses to binary, so now that is used instead of
2992 the text comparison.
2994 PH/20 There was another similar case to PH/19, when a complete host name was
2995 given in a host list; looking up its IP address could give an abbreviated
2996 form, whereas the current host's name might or might not be abbreviated.
2997 The same fix has been applied.
3003 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3004 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3007 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3008 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3010 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3012 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3013 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3014 Also, for "Auto-Submitted", treat anything other than "no" as
3015 non-personal, in accordance with RFC 3834. (Previously it treated
3016 anything starting "auto-" as non-personal.)
3018 TF/01 The control=submission/name=... option had a problem with syntax
3019 errors if the name included a slash character. The /name= option
3020 now slurps the rest of the string, so it can include any characters
3021 but it must come last in the list of options (after /sender_retain
3024 PH/05 Some modifications to the interface to the fake nameserver for the new
3032 TK/01 Added the "success_on_redirect" address verification option. See
3033 NewStuff for rationale and an example.
3035 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3037 PH/02 Patch to exigrep to allow it to work on syslog lines.
3039 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3040 fread() to skip over the body file's header line, because in Cygwin the
3041 header line is locked and is inaccessible.
3043 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3044 co-exist for some time) to make it clear that it is the Exim ID that is
3045 referenced, not the Message-ID: header line.
3047 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3048 string_format() function, because snprintf() does not exist on all
3051 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3052 ${stat: expansion item.
3054 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3055 protocol synchronization error", to keep the pedants happy.
3057 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3058 well as for IRIX systems, when gcc is being used. See the host.c source
3061 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3063 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3066 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3067 transport and to the smtp transport in LMTP mode.
3069 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3071 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3072 run a filter in a subprocess. This could lead to confusion in subsequent
3073 lookups in the parent process. There should also be a search_tidyup() at
3074 the end of the subprocess.
3076 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3077 only if the host matched helo_try_verify_hosts, which caused the
3078 verification to occur when the EHLO/HELO command was issued. The ACL just
3079 tested the remembered result. Now, if a previous verification attempt has
3080 not happened, "verify = helo" does it there and then.
3082 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3084 TK/03 Fix log output including CR from clamd.
3086 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3087 exist provoked a memory error which could cause a segfault.
3089 PH/15 Installed PCRE 6.2
3091 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3093 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3094 of the problem. Specifically, suggested +O2 rather than +O1 for the
3097 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3099 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3100 record, the port was not correctly logged when the outgoing_port log
3101 selector was set (it logged the transort's default port).
3103 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3104 fallback_hosts, and "hosts" in the smtp transport.
3106 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3107 host errors such as "Connection refused".
3109 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3110 authentication with radiusclient 0.4.9:
3112 - Error returned from rc_read_config was caught wrongly
3113 - Username/password not passed on to radius server due to wrong length.
3115 The presumption is that some radiusclient API changes for 4.51/PH/17
3116 were not taken care of correctly. The code is still untested by me (my
3117 Linux distribution still has 0.3.2 of radiusclient), but it was
3118 contributed by a Radius user.
3120 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3121 expanding the "port" option of the smtp transport.
3123 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3124 while reading a MIME header. Thanks to Tom Hughes for a patch.
3126 PH/24 Include config.h inside local_scan.h so that configuration settings are
3129 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3130 that in an ACL for RCPT (for example), you can examine exactly what was
3133 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3134 commands, but it was not correctly comparing the address with the actual
3135 client host address. Thus, it would show the EHLO address in Received:
3136 header lines when this was not necessary.
3138 PH/27 Added the % operator to ${eval:}.
3140 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3141 it should be ignoring failures (because with -C, for example, it has lost
3142 privilege). It wasn't ignoring creation failures other than "already
3145 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3148 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3149 because an input line was too long, either on its own, or by virtue of
3150 too many continuations, the temporary file was not being removed, and the
3151 return code was incorrect.
3153 PH/31 Missing "BOOL" in function definition in filtertest.c.
3155 PH/32 Applied Sieve patches from the maintainer.
3157 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3159 PH/33 Added "verify = not_blind".
3161 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3162 Local/Makefile (with some defaults set). These are used in built scripts
3163 such as exicyclog, but they have never been used in the exim_install
3164 script (though there are many overriding facilities there). I have
3165 arranged that the exim_install script now takes note of these two
3168 PH/35 Installed configuration files for Dragonfly.
3170 PH/36 When a locally submitted message by a trusted user did not contain a
3171 From: header, and the sender address was obtained from -f or from an SMTP
3172 MAIL command, and the trusted user did not use -F to supply a sender
3173 name, $originator_name was incorrectly used when constructing a From:
3174 header. Furthermore, $originator_name was used for submission mode
3175 messages from external hosts without From: headers in a similar way,
3176 which is clearly wrong.
3178 PH/37 Added control=suppress_local_fixups.
3180 PH/38 When log_selector = +received_sender was set, and the addition of the
3181 sender made the log line's construction buffer exactly full, or one byte
3182 less than full, an overflow happened when the terminating "\n" was
3185 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3186 when the result of a list match is failure because a DNS lookup failed.
3188 PH/40 RM_COMMAND is now used in the building process.
3190 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3191 the "build-* directories that it finds.
3193 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3194 address in a domain literal was a prefix of an interface address.
3196 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3197 when verifying a sender address, unless rewrite_headers is false.
3199 PH/44 Wrote a long comment about why errors_to addresses are verified as
3200 recipients, not senders.
3202 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3203 the ratelimit ACL was added.
3205 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3207 PH/46 Added extra information about PostgreSQL errors to the error string.
3209 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3210 avoiding the need to install special zones in a real server. This is
3211 backwards compatible; if it can't find the fake resolver, it drops back.
3212 Thus, both old and new test suites can be run.
3214 TF/02 Added util/ratelimit.pl
3216 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3217 clock is set back in time.
3219 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3220 Candler <B.Candler@pobox.com>.
3222 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3223 OK for addresses that are the result of redirections.
3225 PH/48 A number of further additions for the benefit of the new test suite,
3226 including a fake gethostbyname() that interfaces to the fake DNS resolver
3229 TF/06 The fix for widen_domains has also been applied to qualify_single and
3230 search_parents which are the other dnslookup options that can cause
3233 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3236 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3238 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3239 service" if it encountered an MX record with an empty target hostname.
3240 The message is now "an MX or SRV record indicated no SMTP service".
3242 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3243 if the DNS of the sending site is misconfigured. This is quite a
3244 common situation. This change restores the behaviour of treating a
3245 helo verification defer as a failure.
3247 PH/51 If self=fail was set on a router, the bounce message did not include the
3248 actual error message.
3254 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3256 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3257 command itself ended in error, the underlying message about the transport
3258 filter timeout was being overwritten with the pipe command error. Now the
3259 underlying error message should be appended to the second error message.
3261 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3263 PH/02 Reduce the amount of output that "make" produces by default. Full output
3264 can still be requested.
3266 PH/03 The warning log line about a condition test deferring for a "warn" verb
3267 was being output only once per connection, rather than after each
3268 occurrence (because it was using the same function as for successful
3269 "warn" verbs). This seems wrong, so I have changed it.
3271 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3272 it should not have, which might have caused a crash in the right
3273 circumstances, but probably never did.
3275 PH/04 Installed a modified version of Tony Finch's patch to make submission
3276 mode fix the return path as well as the Sender: header line, and to
3277 add a /name= option so that you can make the user's friendly name appear
3280 TF/03 Added the control = fakedefer ACL modifier.
3282 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3283 Mark Lowes for thorough testing.
3285 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3287 TK/03 Merged latest SRS patch from Miles Wilton.
3289 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3290 with the definition in sysexits.h (which is #included earlier).
3291 Fortunately, Exim does not actually use EX_OK. The code used to try to
3292 preserve the sysexits.h value, by assumimg that macro definitions were
3293 scanned for macro replacements. I have been disabused of this notion,
3294 so now the code just undefines EX_OK before #including unistd.h.
3296 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3297 in the smtp transport. When a block could not be written in a single
3298 write() function, the timeout was being re-applied to each part-write.
3299 This seems wrong - if the receiver was accepting one byte at a time it
3300 would take for ever. The timeout is now adjusted when this happens. It
3301 doesn't have to be particularly precise.
3303 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3304 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3306 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3307 without a domain if the check to <postmaster@domain> fails.
3309 SC/01 Eximstats: added -xls and the ability to specify output files
3310 (patch written by Frank Heydlauf).
3312 SC/02 Eximstats: use FileHandles for outputing results.
3314 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3316 SC/04 Eximstats: fixed display of large numbers with -nvr option
3318 SC/05 Eximstats: fixed merging of reports with empty tables.
3320 SC/06 Eximstats: added the -include_original_destination flag
3322 SC/07 Eximstats: removed tabs and trailing whitespace.
3324 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3326 TK/06 MBOX spool code: Add real "From " MBOX separator line
3327 so the .eml file is really in mbox format (even though
3328 most programs do not really care). Patch from Alex Miller.
3330 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3331 The latter is generated from $received_to and is only set if the
3332 message has one envelope recipient. SA can use these headers,
3333 obviously out-of-the-box. Patch from Alex Miller.
3335 PH/08 The ${def test on a variable was returning false if the variable's
3336 value was "0", contrary to what the specification has always said!
3337 The result should be true unless the variable is empty.
3339 PH/09 The syntax error of a character other than { following "${if
3340 def:variable_name" (after optional whitespace) was not being diagnosed.
3341 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3342 accidental colon was present, for example, could give incorrect results.
3344 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3345 result is used (not including appendfile, where other changes are about
3348 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3349 This involved changing a lot of size variables from int to off_t. It
3350 should work with maildirs and everything.
3352 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3353 spamd dying while we are connected to it.
3355 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3358 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3359 function for BDB 4.3.
3361 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3363 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3364 the sources for PCRE have been re-arranged and more files are now
3367 PH/15 The code I had for printing potentially long long variables in PH/11
3368 above was not the best (it lost precision). The length of off_t variables
3369 is now inspected at build time, and an appropriate printing format (%ld
3370 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3371 to be "long long int" or "long int". This is needed for the internal
3372 formatting function string_vformat().
3374 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3375 the configuration file to be ":syslog", then the script "guesses" where
3376 the logs files are, rather than using the compiled in default. In our
3377 case the guess is not the same as the compiled default, so the script
3378 suddenly stopped working when I started to use syslog. The patch checks
3379 to see if log_file_path is "". If so, it attempts to read it from exim
3380 with no configuration file to get the compiled in version, before it
3381 falls back to the previous guessing code."
3383 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3384 implementing BATV in an Exim configuration. See NewStuff for the gory
3387 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3388 Makefile that are specific to HP-UX.
3390 PH/18 If the "use_postmaster" option was set for a recipient callout together
3391 with the "random" option, the postmaster address was used as the MAIL
3392 FROM address for the random test, but not for the subsequent recipient
3393 test. It is now used for both.
3395 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3396 patch removes a few documentation additions to RFC 3028, because the
3397 latest draft now contains them. It adds the new en;ascii-case comparator
3398 and a new error check for 8bit text in MIME parts. Comparator and
3399 require names are now matched exactly. I enabled the subaddress
3400 extension, but it is not well tested yet (read: it works for me)."
3402 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3403 rework some of the code of TK/09 above to avoid the hardwired use of
3404 "%lld" and "long long". Replaced the call to snprintf() with a call to
3407 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3408 records point to non-existent hosts", "retry timeout exceeded", and
3409 "retry time not reached for any host after a long failure period".
3411 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3412 experimental DomainKeys support:
3414 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3415 (2) On an error such as an illegally used "control", the wrong name for
3416 the control was given.
3418 These problems did NOT occur unless DomainKeys support was compiled.
3420 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3422 PH/24 Added ${if match_ip condition.
3424 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3425 databases so that it will be absolutely obvious if a crash occurs in the
3426 DB library. This is a regular occurrence (often caused by mis-matched
3429 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3430 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3431 picked up on a user's system that detects such things. There doesn't seem
3432 to be a gcc warning option for this - only an attribute that has to be
3433 put on the function's prototype. It seems that in Fedora Core 4 they have
3434 set this on a number of new functions. No doubt there will be more in due
3437 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3438 specify a transport. However, if an address that was verified by such a
3439 router was the subject of a callout, Exim crashed because it tried to
3440 read the rcpt_include_affixes from the non-existent transport. Now it
3441 just assumes that the setting of that option is false. This bug was
3442 introduced by 4.51/PH/31.
3444 PH/28 Changed -d+all to exclude +memory, because that information is very
3445 rarely of interest, but it makes the output a lot bigger. People tend to
3446 do -d+all out of habit.
3448 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3449 code in os-type was giving problems when libc.so lives in lib64, like on
3452 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3453 aren't the modern standard, and it seems that some systems' include files
3454 don't always have them. Exim was already checking for some of the newer
3455 ones like T_AAAA, and defining it itself. I've added checks for all the
3456 record types that Exim uses.
3458 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3459 not automatically generating a new one, as it is supposed to. This
3460 prevented TLS from working. If the file did exist, but contained invalid
3461 data, a new version was generated, as expected. It was only the case of a
3462 non-existent file that was broken.
3464 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3465 with a change in libdomainkeys > 0.64.
3467 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3468 from DNS. If the selector record carries the flag, it now has
3469 precedence over the domain-wide flag.
3471 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3473 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3474 the use of a port name that isn't defined in /etc/services) occurred, the
3475 message was deferred as in a normal delivery, and thus remained on the
3476 spool, instead of being failed because of the mua_wrapper setting. This
3477 is now fixed, and I tidied up some of the mua_wrapper messages at the
3480 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3481 the messages in a hash of arrays rather than using individual hashes.
3482 This is a bit cleaner and results in dramatic memory savings, albeit
3483 at a slight CPU cost.
3485 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3486 as requested by Marc Sherman.
3488 SC/10 Eximstats: added histograms for user specified patterns as requested
3491 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3493 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3494 fopen() in the content-scanning modules that did not already have it.
3500 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3501 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3503 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3505 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3507 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3508 to test Sieve filters that use "vacation".
3510 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3511 that changes the way the GnuTLS parameters are stored in the cache file.
3512 The new format can be generated externally. For backward compatibility,
3513 if the data in the cache doesn't make sense, Exim assumes it has read an
3514 old-format file, and it generates new data and writes a new file. This
3515 means that you can't go back to an older release without removing the
3518 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3519 work if there are any delivery delays because "one_time" forces the
3520 parent to be marked "delivered", so its unseen clone is never tried
3521 again. For this reason, Exim now forbids the simultaneous setting of
3524 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3525 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3526 entry if you want to know the details. The fix, however, neglected to
3527 consider the case when local delivery batching is involved. The test for
3528 "previously delivered" was not happening when checking to see if an
3529 address could be batched with a previous (undelivered) one; under
3530 certain circumstances this could lead to multiple deliveries to the same
3533 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3534 in its include files, and this causes problems building Exim.
3536 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3537 header_syntax) but Exim was just ignoring anything given after a slash.
3538 In particular, this caused confusion with an attempt to use "verify =
3539 reverse_host_lookup/defer_ok". An error is now given when options are
3540 supplied for verify items that do not have them. (Maybe reverse_host_
3541 lookup should have a defer_ok option, but that's a different point.)
3543 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3544 defined by RFC 821) to 2048, because there were problems with some AUTH
3545 commands, and RFC 1869 says the size should be increased for extended
3546 SMTP commands that take arguments.
3548 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3551 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3552 "unknown" error; now it says that the functionality isn't in the binary.
3554 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3555 an address' error message when a string expansion fails (syntax or
3556 whatever). Otherwise the password may appear in the log. Following change
3557 PH/42 below, there is no longer a chance of it appearing in a bounce
3560 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3562 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3563 of its addresses was ever tried. (Bugzilla bug #2.)
3565 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3566 the result incorrectly in the debug output. (It correctly added a newline
3567 to what was transported.)
3569 TF/01 Added $received_time.
3571 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3572 commented out examples of how to interface to a virus scanner and to
3573 SpamAssassin. Also added commented examples of av_scanner and
3574 spamd_address settings.
3576 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3577 and controls are allowed in which ACLs. There were a couple of minor
3578 errors. Some of the entries in the conditions table (which is a table of
3579 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3580 negation of where the condition IS allowed.
3582 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3584 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3585 header file does not have a version number, so I've had to invent a new
3586 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3587 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3588 radiusclient), but it was contributed by a Radius user.
3590 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3591 files or directories, for OpenSSL.
3593 PH/19 When an Exim process that is running as root has to create an Exim log
3594 file, it does so in a subprocess that runs as exim:exim so as to get the
3595 ownership right at creation (otherwise, other Exim processes might see
3596 the file with the wrong ownership). There was no test for failure of this
3597 fork() call, which would lead to the process getting stuck as it waited
3598 for a non-existent subprocess. Forks do occasionally fail when resources
3599 run out. I reviewed all the other calls to fork(); they all seem to check
3602 PH/20 When checking for unexpected SMTP input at connect time (before writing
3603 the banner), Exim was not dealing correctly with a non-positive return
3604 from the read() function. If the client had disconnected by this time,
3605 the result was a log entry for a synchronization error with an empty
3606 string after "input=" when read() returned zero. If read() returned -1
3607 (an event I could not check), uninitialized data bytes were printed.
3608 There were reports of junk text (parts of files, etc) appearing after
3611 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3613 PH/22 Added support for macro redefinition, and (re)definition in between
3614 driver and ACL definitions.
3616 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3617 forgetting to use the resulting value; it was using the unexpanded value.
3619 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3620 hadn't been configured. The fix is from Juergen Kreileder, who
3621 understands it better than I do:
3623 "Here's what I see happening with three configured cyrus_sasl
3624 authenticators configured (plain, login, cram-md5):
3626 On startup auth_cyrus_sasl_init() gets called for each of these.
3627 This means three calls to sasl_listmech() without a specified mech_list.
3628 => SASL tests which mechs of all available mechs actually work
3629 => three warnings about OTP not working
3630 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3632 With the patch, sasl_listmech() also gets called three times. But now
3633 SASL's mech_list option is set to the server_mech specified in the the
3634 authenticator. Or in other words, the answer from sasl_listmech()
3635 gets limited to just the mech you're testing for (which is different
3637 => the return list contains just 'plain' or 'login', 'cram-md5' or
3638 nothing depending on the value of ob->server_mech.
3640 I've just tested the patch: Authentication still works fine,
3641 unavailable mechs specified in the exim configuration are still
3642 caught, and the auth.log warnings about OTP are gone."
3644 PH/25 When debugging is enabled, the contents of the command line are added
3645 to the debugging output, even when log_selector=+arguments is not
3648 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3649 answer is "GNU", and only if the return is "GNU/something" is the answer
3652 PH/27 $acl_verify_message is now set immediately after the failure of a
3653 verification in an ACL, and so is available in subsequent modifiers. In
3654 particular, the message can be preserved by coding like this:
3656 warn !verify = sender
3657 set acl_m0 = $acl_verify_message
3659 Previously, $acl_verify_message was set only while expanding "message"
3660 and "log_message" when a very denied access.
3662 PH/28 Modified OS/os.c-Linux with
3664 -#ifndef OS_LOAD_AVERAGE
3665 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3667 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3668 nomenclature these days.)
3670 PH/29 Installed patch from the Sieve maintainer that adds the options
3671 sieve_useraddress and sieve_subaddress to the redirect router.
3673 PH/30 In these circumstances:
3674 . Two addresses routed to the same list of hosts;
3675 . First host does not offer TLS;
3676 . First host accepts first address;
3677 . First host gives temporary error to second address;
3678 . Second host offers TLS and a TLS session is established;
3679 . Second host accepts second address.
3680 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3681 and peerdn, if requested) that were in fact used only for the second
3684 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3685 attention to any local part prefix or suffix that was matched by the
3686 router that accepted the address. It now behaves in the same way as it
3687 does for delivery: the affixes are removed from the local part unless
3688 rcpt_include_affixes is set on the transport.
3690 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3691 timeout during the DATA phase of an incoming message.
3693 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3694 applied a patch sanctioned by the Sieve maintainer.
3696 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3697 the uid or gid is negative. A case of a negative gid caused this to be
3698 noticed. The fix allows for either to be negative.
3700 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3701 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3704 PH/36 The change PH/12 above was broken. Fixed it.
3706 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3707 the grounds that routing the same address twice would always produce the
3708 same answer. This might have been true once, but it is certainly no
3709 longer true now. Routing a child address may depend on the previous
3710 routing that produced that child. Some complicated redirection strategies
3711 went wrong when messages had multiple recipients, and made Exim's
3712 behaviour dependent on the order in which the addresses were given.
3714 I have moved the duplicate checking until after the routing is complete.
3715 Exim scans the addresses that are assigned to local and remote
3716 transports, and removes any duplicates. This means that more work will be
3717 done, as duplicates will always all be routed, but duplicates are
3718 presumably rare, so I don't expect this is of any significance.
3720 For deliveries to pipes, files, and autoreplies, the duplicate checking
3721 still happens during the routing process, since they are not going to be
3724 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3725 It corrects a timeout issue with spamd. This is Ian's comment: "The
3726 background is that sometimes spamd either never reads data from a
3727 connection it has accepted, or it never writes response data. The exiscan
3728 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3729 blindly assumes that writes won't block so it may never time out."
3731 PH/39 Allow G after quota size as well as K and M.
3733 PH/40 The value set for $authenticated_id in an authenticator may not contain
3734 binary zeroes or newlines because the value is written to log lines and
3735 to spool files. There was no check on this. Now the value is run through
3736 the string_printing() function so that such characters are converted to
3737 printable escape sequences.
3739 PH/41 $message_linecount is a new variable that contains the total number of
3740 lines in the message. Compare $body_linecount, which is the count for the
3743 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3744 bounce and delay warning messages, except in certain special cases, which
3747 (a) An SMTP error message from a remote host;
3748 (b) A message specified in a :fail: redirection;
3749 (c) A message specified in a "fail" command in a system filter;
3750 (d) A message specified in a FAIL return from the queryprogram router;
3751 (e) A message specified by the cannot_route_message router option.
3753 In these cases only, Exim does include the error details in bounce and
3754 warning messages. There are also a few cases where bland messages such
3755 as "unrouteable address" or "local delivery error" are given.
3757 PH/43 $value is now also set for the "else" part of a ${run expansion.
3759 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3760 being worked on, but at least Exim now implements the latest version to
3763 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3764 process to complete was treated as a delivery failure, a timeout while
3765 writing the message to the pipe was logged, but erroneously treated as a
3766 successful delivery. Such timeouts include transport filter timeouts. For
3767 consistency with the overall process timeout, these timeouts are now
3768 treated as errors, giving rise to delivery failures by default. However,
3769 there is now a new Boolean option for the pipe transport called
3770 timeout_defer, which, if set TRUE, converts the failures into defers for
3771 both kinds of timeout. A transport filter timeout is now identified in
3774 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3775 systems where "make" and "gmake" are different, calling "gmake" at top
3776 level broke things. I've arranged for the value of $(MAKE) to be passed
3777 from the Makefile to this script so that it can call the same version of
3781 A note about Exim versions 4.44 and 4.50
3782 ----------------------------------------
3784 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3785 changes of various kinds. As a consequence, a big documentation update was
3786 needed. This delayed the release for rather longer than seemed good, especially
3787 in the light of a couple of (minor) security issues. Therefore, the changes
3788 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3789 release. So 4.44 and 4.50 are in effect two different branches that both start
3792 I have left the 4.50 change log unchanged; it contains all the changes since
3793 4.43. The change log for 4.44 is below; many of its items are identical to
3794 those for 4.50. This seems to be the most sensible way to preserve the
3795 historical information.
3801 1. Minor wording change to the doc/README.SIEVE file.
3803 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3804 computation of the current number of files was incorrect.
3806 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3807 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3810 4. Give more explanation in the error message when the command for a transport
3811 filter fails to execute.
3813 5. There are several places where Exim runs a non-Exim command in a
3814 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3815 was being done only for the command run by the queryprogram router. It is
3816 now done for all such subprocesses. The other cases are: ${run, transport
3817 filters, and the commands run by the lmtp and pipe transports.
3819 6. Added CONFIGURE_GROUP build-time option.
3821 7. Some older OS have a limit of 256 on the maximum number of file
3822 descriptors. Exim was using setrlimit() to set 1000 as a large value
3823 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3824 systems. I've change it so that if it can't get 1000, it tries for 256.
3826 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3827 was an oversight, and furthermore, ever since the addition of extra
3828 controls (e.g. 4.43/32), the checks on when to allow different forms of
3829 "control" were broken. There should now be diagnostics for all cases when a
3830 control that does not make sense is encountered.
3832 9. Added the /retain_sender option to "control=submission".
3834 10. $recipients is now available in the predata ACL (oversight).
3836 11. Tidy the search cache before the fork to do a delivery from a message
3837 received from the command line. Otherwise the child will trigger a lookup
3838 failure and thereby defer the delivery if it tries to use (for example) a
3839 cached ldap connection that the parent has called unbind on.
3841 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3842 of $address_data from the recipient verification was clobbered by the
3843 sender verification.
3845 13. The value of address_data from a sender verification is now available in
3846 $sender_address_data in subsequent conditions in the ACL statement.
3848 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3850 15. Added a new option "connect=<time>" to callout options, to set a different
3853 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3854 was its contents. (It was OK if the option was not defined at all.)
3856 17. A "Completed" log line is now written for messages that are removed from
3857 the spool by the -Mrm option.
3859 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3860 information about exactly what failed.
3862 19. Added -dd to debug only the daemon process.
3864 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3865 handles timeouts, both on the server side and network timeouts. Renamed the
3866 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3868 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3869 It is now set to "smtps".
3871 22. $host_address is now set to the target address during the checking of
3872 ignore_target_hosts.
3874 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3875 being passed; this would have caused $sender_host_name to have been used if
3876 matching the list had actually called for a host name (not very likely,
3877 since this list is usually IP addresses). A host name is now passed as
3880 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3881 code that specifies a non-restarting handler (typically sigaction() in
3882 modern systems) in an attempt to fix a rare and obscure crash bug.
3884 25. Narrowed the window for a race in the daemon that could cause it to ignore
3885 SIGCHLD signals. This is not a major problem, because they are used only to
3886 wake it up if nothing else does.
3888 26. A malformed maildirsize file could cause Exim to calculate negative values
3889 for the mailbox size or file count. Odd effects could occur as a result.
3890 The maildirsize information is now recalculated if the size or filecount
3893 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3894 support for a long time. Removed HAVE_SYS_VFS_H.
3896 28. Installed the latest version of exipick from John Jetmore.
3898 29. In an address list, if the pattern was not a regular expression, an empty
3899 subject address (from a bounce message) matched only if the pattern was an
3900 empty string. Non-empty patterns were not even tested. This was the wrong
3901 because it is perfectly reasonable to use an empty address as part of a
3902 database query. An empty address is now tested by patterns that are
3903 lookups. However, all the other forms of pattern expect the subject to
3904 contain a local part and a domain, and therefore, for them, an empty
3905 address still always fails if the pattern is not itself empty.
3907 30. Exim went into a mad DNS loop when attempting to do a callout where the
3908 host was specified on an smtp transport, and looking it up yielded more
3909 than one IP address.
3911 31. Re-factored the code for checking spool and log partition space into a
3912 function that finds that data and another that does the check. The former
3913 is then used to implement four new variables: $spool_space, $log_space,
3914 $spool_inodes, and $log_inodes.
3916 32. The RFC2047 encoding function was originally intended for short strings
3917 such as real names; it was not keeping to the 75-character limit for
3918 encoded words that the RFC imposes. It now respects the limit, and
3919 generates multiple encoded words if necessary. To be on the safe side, I
3920 have increased the buffer size for the ${rfc2047: expansion operator from
3923 33. It is now permitted to omit both strings after an "if" condition; if the
3924 condition is true, the result is "true". As before, when the second string
3925 is omitted, a false condition yields an empty string. This makes it less
3926 cumbersome to write custom ACL and router conditions.
3928 34. Failure to deliver a bounce message always caused it to be frozen, even if
3929 there was an errors_to setting on the router. The errors_to setting is now
3932 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3933 canonical form (fully expanded) before being placed in
3934 $sender_host_address.
3936 36. The table in the code that translates DNS record types into text (T_A to
3937 "A" for instance) was missing entries for NS and CNAME. It is just possible
3938 that this could have caused confusion if both these types were looked up
3939 for the same domain, because the text type is used as part of Exim's
3940 per-process caching. But the chance of anyone hitting this buglet seems
3943 37. The dnsdb lookup has been extended in a number of ways.
3945 (1) There is a new type, "zns", which walks up the domain tree until it
3946 finds some nameserver records. It should be used with care.
3948 (2) There is a new type, "mxh", which is like "mx" except that it returns
3949 just the host names, not the priorities.
3951 (3) It is now possible to give a list of domains (or IP addresses) to be
3952 looked up. The behaviour when one of the lookups defers can be
3953 controlled by a keyword.
3955 (4) It is now possible to specify the separator character for use when
3956 multiple records are returned.
3958 38. The dnslists ACL condition has been extended: it is now possible to supply
3959 a list of IP addresses and/or domains to be looked up in a particular DNS
3962 39. Added log_selector=+queue_time_overall.
3964 40. When running the queue in the test harness, wait just a tad after forking a
3965 delivery process, to get repeatability of debugging output.
3967 41. Include certificate and key file names in error message when GnuTLS fails
3968 to set them up, because the GnuTLS error message doesn't include the name
3969 of the failing file when there is a problem reading it.
3971 42. Allow both -bf and -bF in the same test run.
3973 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3975 44. The "Exiscan patch" is now merged into the mainline Exim source.
3977 45. Sometimes the final signoff response after QUIT could fail to get
3978 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3979 < 0 before doing a fflush(). This bug looks as though it goes back to the
3980 introduction of TLS in release 3.20, but "sometimes" must have been rare
3981 because the tests only now provoked it.
3983 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3984 (this can affect the format of dates).
3986 47. exim_tidydb, when checking for the continued existence of a message for
3987 which it has found a message-specific retry record, was not finding
3988 messages that were in split spool directories. Consequently, it was
3989 deleting retry records that should have stayed in existence.
3991 48. Steve fixed some bugs in eximstats.
3993 49. The SPA authentication driver was not abandoning authentication and moving
3994 on to the next authenticator when an expansion was forced to fail,
3995 contradicting the general specification for all authenticators. Instead it
3996 was generating a temporary error. It now behaves as specified.
3998 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3999 (the order specifies the preference for clients). The order is now AES256,
4000 AES128, 3DES, ARCFOUR128.
4002 51. Small patch to Sieve code - explicitly set From: when generating an
4005 52. Exim crashed if a remote delivery caused a very long error message to be
4006 recorded - for instance if somebody sent an entire SpamAssassin report back
4007 as a large number of 550 error lines. This bug was coincidentally fixed by
4008 increasing the size of one of Exim's internal buffers (big_buffer) that
4009 happened as part of the Exiscan merge. However, to be on the safe side, I
4010 have made the code more robust (and fixed the comments that describe what
4013 53. Now that there can be additional text after "Completed" in log lines (if
4014 the queue_time_overall log selector is set), a one-byte patch to exigrep
4015 was needed to allow it to recognize "Completed" as not the last thing in
4018 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4019 patch that reportedly fixes this has been added. I am not expert enough to
4020 create a test for it. This is what the patch creator wrote:
4022 "I found a little strange behaviour of ldap code when working with
4023 Windows 2003 AD Domain, where users was placed in more than one
4024 Organization Units. When I tried to give exim partial DN, the exit code
4025 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4026 But simultaneously result of request was absolutely normal ldap result,
4027 so I produce this patch..."
4029 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4030 REFERENCE, so I have modified the code to exclude the patch when that macro
4033 55. Some experimental protocols are using DNS PTR records for new purposes. The
4034 keys for these records are domain names, not reversed IP addresses. The
4035 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4036 leaves it alone. Component reversal etc. now happens only for IP addresses.
4039 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4041 57. Double the size of the debug message buffer (to 2048) so that more of very
4042 long debug lines gets shown.
4044 58. The exicyclog utility now does better if the number of log files to keep
4045 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4047 59. Two changes related to the smtp_active_hostname option:
4049 (1) $smtp_active_hostname is now available as a variable.
4050 (2) The default for smtp_banner uses $smtp_active_hostname instead
4051 of $primary_hostname.
4053 60. The host_aton() function is supposed to be passed a string that is known
4054 to be a valid IP address. However, in the case of IPv6 addresses, it was
4055 not checking this. This is a hostage to fortune. Exim now panics and dies
4056 if the condition is not met. A case was found where this could be provoked
4057 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4058 components; fortuitously, this particular loophole had already been fixed
4059 by change 4.50/55 above.
4061 If there are any other similar loopholes, the new check in host_aton()
4062 itself should stop them being exploited. The report I received stated that
4063 data on the command line could provoke the exploit when Exim was running as
4064 exim, but did not say which command line option was involved. All I could
4065 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4066 running as the user.
4069 61. There was a buffer overflow vulnerability in the SPA authentication code
4070 (which came originally from the Samba project). I have added a test to the
4071 spa_base64_to_bits() function which I hope fixes it.
4074 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4075 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4077 63. The daemon start-up calls getloadavg() while still root for those OS that
4078 need the first call to be done as root, but it missed one case: when
4079 deliver_queue_load_max is set with deliver_drop_privilege. This is
4080 necessary for the benefit of the queue runner, because there is no re-exec
4081 when deliver_drop_privilege is set.
4083 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4084 This has been fixed.
4086 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4087 was in use, was not putting the data itself into the right store pool;
4088 consequently, it could be overwritten for a subsequent message in the same
4089 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4092 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4094 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4095 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4096 the function were treating the return as a boolean value, which happened to
4097 work because 0=false and not-0=true, but is not correct code.
4099 68. The host_aton() function was not handling scoped IPv6 addresses (those
4100 with, for example, "%eth0" on the end) correctly.
4102 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4103 negated items (that is, ~something) in unsigned ints. Some compilers
4104 apparently mutter when there is no cast.
4106 70. If an address verification called from an ACL failed, and did not produce a
4107 user-specific message (i.e. there was only a "system" message), nothing was
4108 put in $acl_verify_message. In this situation, it now puts the system
4111 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4112 session; change 4.31/43 added the unwanted input to the log line - except
4113 that it did not do this in the start of session case. It now does.
4115 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4116 This is wrong and can cause the other end to generate a synchronization
4117 error if it is another Exim or anything else that does the synchronization
4118 check. A QUIT command is no longer sent after a timeout.
4120 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4121 during host lookups.
4123 74. The defer_ok option of callout verification was not working if it was used
4124 when verifying addresses in header lines, that is, for this case:
4126 verify = header_sender/callout=defer_ok
4128 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4129 those file descriptors could be used for SMTP connections. If anything
4130 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4131 could be sent to the SMTP client, causing chaos. The daemon now opens
4132 stdin, stdout, and stderr to /dev/null when it puts itself into the
4135 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4136 log by default. The user can override this with suitable Perl magic.
4138 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4139 the log message when discard triggers, was not working for the DATA ACL or
4140 for the non-SMTP ACL.
4142 78. Error message wording change in sieve.c.
4144 79. If smtp_accept_max_per_host was set, the number of connections could be
4145 restricted to fewer than expected, because the daemon was trying to set up
4146 a new connection before checking whether the processes handling previous
4147 connections had finished. The check for completed processes is now done
4148 earlier. On busy systems, this bug wouldn't be noticed because something
4149 else would have woken the daemon, and it would have reaped the completed
4152 80. If a message was submitted locally by a user whose login name contained one
4153 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4154 It caused a spool format error. I have fixed the spool reading code. A
4155 related problem was that the "from" clause in the Received: line became
4156 illegal because of the space(s). It is now covered by ${quote_local_part.
4158 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4161 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4163 83. Patch from Sieve maintainer to fix unterminated string problem in
4164 "vacation" handling.
4166 84. Some minor changes to the Linux configuration files to help with other
4167 OS variants using glibc.
4169 85. One more patch for Sieve to update vacation handling to latest spec.
4172 ----------------------------------------------------
4173 See the note above about the 4.44 and 4.50 releases.
4174 ----------------------------------------------------
4180 1. Change 4.43/35 introduced a bug that caused file counts to be
4181 incorrectly computed when quota_filecount was set in an appendfile
4184 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4185 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4188 3. Give more explanation in the error message when the command for a transport
4189 filter fails to execute.
4191 4. There are several places where Exim runs a non-Exim command in a
4192 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4193 was being done only for the command run by the queryprogram router. It is
4194 now done for all such subprocesses. The other cases are: ${run, transport
4195 filters, and the commands run by the lmtp and pipe transports.
4197 5. Some older OS have a limit of 256 on the maximum number of file
4198 descriptors. Exim was using setrlimit() to set 1000 as a large value
4199 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4200 systems. I've change it so that if it can't get 1000, it tries for 256.
4202 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4203 was an oversight, and furthermore, ever since the addition of extra
4204 controls (e.g. 4.43/32), the checks on when to allow different forms of
4205 "control" were broken. There should now be diagnostics for all cases when a
4206 control that does not make sense is encountered.
4208 7. $recipients is now available in the predata ACL (oversight).
4210 8. Tidy the search cache before the fork to do a delivery from a message
4211 received from the command line. Otherwise the child will trigger a lookup
4212 failure and thereby defer the delivery if it tries to use (for example) a
4213 cached ldap connection that the parent has called unbind on.
4215 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4216 of $address_data from the recipient verification was clobbered by the
4217 sender verification.
4219 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4220 was its contents. (It was OK if the option was not defined at all.)
4222 11. A "Completed" log line is now written for messages that are removed from
4223 the spool by the -Mrm option.
4225 12. $host_address is now set to the target address during the checking of
4226 ignore_target_hosts.
4228 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4229 being passed; this would have caused $sender_host_name to have been used if
4230 matching the list had actually called for a host name (not very likely,
4231 since this list is usually IP addresses). A host name is now passed as
4234 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4235 code that specifies a non-restarting handler (typically sigaction() in
4236 modern systems) in an attempt to fix a rare and obscure crash bug.
4238 15. Narrowed the window for a race in the daemon that could cause it to ignore
4239 SIGCHLD signals. This is not a major problem, because they are used only to
4240 wake it up if nothing else does.
4242 16. A malformed maildirsize file could cause Exim to calculate negative values
4243 for the mailbox size or file count. Odd effects could occur as a result.
4244 The maildirsize information is now recalculated if the size or filecount
4247 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4248 support for a long time. Removed HAVE_SYS_VFS_H.
4250 18. Updated exipick to current release from John Jetmore.
4252 19. Allow an empty sender to be matched against a lookup in an address list.
4253 Previously the only cases considered were a regular expression, or an
4256 20. Exim went into a mad DNS lookup loop when doing a callout where the
4257 host was specified on the transport, if the DNS lookup yielded more than
4260 21. The RFC2047 encoding function was originally intended for short strings
4261 such as real names; it was not keeping to the 75-character limit for
4262 encoded words that the RFC imposes. It now respects the limit, and
4263 generates multiple encoded words if necessary. To be on the safe side, I
4264 have increased the buffer size for the ${rfc2047: expansion operator from
4267 22. Failure to deliver a bounce message always caused it to be frozen, even if
4268 there was an errors_to setting on the router. The errors_to setting is now
4271 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4272 canonical form (fully expanded) before being placed in
4273 $sender_host_address.
4275 24. Updated eximstats to version 1.33
4277 25. Include certificate and key file names in error message when GnuTLS fails
4278 to set them up, because the GnuTLS error message doesn't include the name
4279 of the failing file when there is a problem reading it.
4281 26. Expand error message when OpenSSL has problems setting up cert/key files.
4284 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4285 (this can affect the format of dates).
4287 28. exim_tidydb, when checking for the continued existence of a message for
4288 which it has found a message-specific retry record, was not finding
4289 messages that were in split spool directories. Consequently, it was
4290 deleting retry records that should have stayed in existence.
4292 29. eximstats updated to version 1.35
4293 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4294 1.35 - bugfix such that pie charts by volume are generated correctly
4296 30. The SPA authentication driver was not abandoning authentication and moving
4297 on to the next authenticator when an expansion was forced to fail,
4298 contradicting the general specification for all authenticators. Instead it
4299 was generating a temporary error. It now behaves as specified.
4301 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4302 (the order specifies the preference for clients). The order is now AES256,
4303 AES128, 3DES, ARCFOUR128.
4305 31. Small patch to Sieve code - explicitly set From: when generating an
4308 32. Exim crashed if a remote delivery caused a very long error message to be
4309 recorded - for instance if somebody sent an entire SpamAssassin report back
4310 as a large number of 550 error lines. This bug was coincidentally fixed by
4311 increasing the size of one of Exim's internal buffers (big_buffer) that
4312 happened as part of the Exiscan merge. However, to be on the safe side, I
4313 have made the code more robust (and fixed the comments that describe what
4316 33. Some experimental protocols are using DNS PTR records for new purposes. The
4317 keys for these records are domain names, not reversed IP addresses. The
4318 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4319 leaves it alone. Component reversal etc. now happens only for IP addresses.
4322 34. The host_aton() function is supposed to be passed a string that is known
4323 to be a valid IP address. However, in the case of IPv6 addresses, it was
4324 not checking this. This is a hostage to fortune. Exim now panics and dies
4325 if the condition is not met. A case was found where this could be provoked
4326 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4327 components; fortuitously, this particular loophole had already been fixed
4328 by change 4.50/55 or 4.44/33 above.
4330 If there are any other similar loopholes, the new check in host_aton()
4331 itself should stop them being exploited. The report I received stated that
4332 data on the command line could provoke the exploit when Exim was running as
4333 exim, but did not say which command line option was involved. All I could
4334 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4335 running as the user.
4338 35. There was a buffer overflow vulnerability in the SPA authentication code
4339 (which came originally from the Samba project). I have added a test to the
4340 spa_base64_to_bits() function which I hope fixes it.
4343 36. The daemon start-up calls getloadavg() while still root for those OS that
4344 need the first call to be done as root, but it missed one case: when
4345 deliver_queue_load_max is set with deliver_drop_privilege. This is
4346 necessary for the benefit of the queue runner, because there is no re-exec
4347 when deliver_drop_privilege is set.
4349 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4350 was in use, was not putting the data itself into the right store pool;
4351 consequently, it could be overwritten for a subsequent message in the same
4352 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4355 38. Sometimes the final signoff response after QUIT could fail to get
4356 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4357 < 0 before doing a fflush(). This bug looks as though it goes back to the
4358 introduction of TLS in release 3.20, but "sometimes" must have been rare
4359 because the tests only now provoked it.
4365 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4366 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4367 Now it returns an integer. A number of calls were still expecting a T/F
4368 return. Fortuitously, in all cases, the tests worked in OK situations,
4369 which is the norm. However, things would have gone wrong on any write
4370 failures on the smtp file descriptor. This function is used when sending
4371 messages over SMTP and also when doing verify callouts.
4373 2. When Exim is called to do synchronous delivery of a locally submitted
4374 message (the -odf or -odi options), it no longer closes stderr before doing
4377 3. Implemented the mua_wrapper option.
4379 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4381 5. Implemented the functions header_remove(), header_testname(),
4382 header_add_at_position(), and receive_remove_recipient(), and exported them
4385 6. If an ACL "warn" statement specified the addition of headers, Exim already
4386 inserted X-ACL-Warn: at the start if there was no header name. However, it
4387 was not making this test for the second and subsequent header lines if
4388 there were newlines in the string. This meant that an invalid header could
4389 be inserted if Exim was badly configured.
4391 7. Allow an ACL "warn" statement to add header lines at the start or after all
4392 the Received: headers, as well as at the end.
4394 8. Added the rcpt_4xx retry error code.
4396 9. Added postmaster_mailfrom=xxx to callout verification option.
4398 10. Added mailfrom=xxxx to the callout verification option, for verify=
4401 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4402 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4404 12. Inserted some casts to stop certain compilers warning when using pointer
4405 differences as field lengths or precisions in printf-type calls (mostly
4406 affecting debugging statements).
4408 13. Added optional readline() support for -be (dynamically loaded).
4410 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4411 same clock tick as a message's arrival, so that its received time was the
4412 same as the "first fail" time on the retry record, and that message
4413 remained on the queue past the ultimate address timeout, every queue runner
4414 would try a delivery (because it was past the ultimate address timeout) but
4415 after another failure, the ultimate address timeout, which should have then
4416 bounced the address, did not kick in. This was a "< instead of <=" error;
4417 in most cases the first failure would have been in the next clock tick
4418 after the received time, and all would be well.
4420 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4421 being recognized when the domain list was tested by the match_domain
4422 condition in an expansion string.
4424 16. Added the ${str2b64: operator.
4426 17. Exim was always calling setrlimit() to set a large limit for the number of
4427 processes, without checking whether the existing limit was already
4428 adequate. (It did check for the limit on file descriptors.) Furthermore,
4429 errors from getrlimit() and setrlimit() were being ignored. Now they are
4430 logged to the main and panic logs, but Exim does carry on, to try to do its
4431 job under whatever limits there are.
4433 18. Imported PCRE 5.0.
4435 19. Trivial typo in log message " temporarily refused connection" (the leading
4438 20. If the log selector return_path_on_delivery was set and an address was
4439 redirected to /dev/null, the delivery process crashed because it assumed
4440 that a return path would always be set for a "successful" delivery. In this
4441 case, the whole delivery is bypassed as an optimization, and therefore no
4444 21. Internal re-arrangement: the function for sending a challenge and reading
4445 a response while authentication was assuming a zero-terminated challenge
4446 string. It's now changed to take a pointer and a length, to allow for
4447 binary data in such strings.
4449 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4451 23. Exim was not respecting finduser_retries when seeking the login of the
4452 uid under which it was called; it was always trying 10 times. (The default
4453 setting of finduser_retries is zero.) Also, it was sleeping after the final
4454 failure, which is pointless.
4456 24. Implemented tls_on_connect_ports.
4458 25. Implemented acl_smtp_predata.
4460 26. If the domain in control=submission is set empty, Exim assumes that the
4461 authenticated id is a complete email address when it generates From: or
4462 Sender: header lines.
4464 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4465 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4466 chown and chgrp in /bin and hostname in /usr/bin.
4468 28. Exim was keeping the "process log" file open after each use, just as it
4469 does for the main log. This opens the possibility of it remaining open for
4470 long periods when the USR1 signal hits a daemon. Occasional processlog
4471 errors were reported, that could have been caused by this. Anyway, it seems
4472 much more sensible not to leave this file open at all, so that is what now
4475 29. The long-running daemon process does not normally write to the log once it
4476 has entered its main loop, and it closes the log before doing so. This is
4477 so that log files can straightforwardly be renamed and moved. However,
4478 there are a couple of unusual error situations where the daemon does write
4479 log entries, and I had neglected to close the log afterwards.
4481 30. The text of an SMTP error response that was received during a remote
4482 delivery was being truncated at 512 bytes. This is too short for some of
4483 the long messages that one sometimes sees. I've increased the limit to
4486 31. It is now possible to make retry rules that apply only when a message has a
4487 specific sender, in particular, an empty sender.
4489 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4490 it possible to be selective about when SMTP synchronization is enforced.
4492 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4494 32. Implemented hosts_connection_nolog.
4496 33. Added an ACL for QUIT.
4498 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4501 35. Added mailbox_size and mailbox_filecount to appendfile.
4503 36. Added control = no_multiline_responses to ACLs.
4505 37. There was a bug in the logic of the code that waits for the clock to tick
4506 in the case where the clock went backwards by a substantial amount such
4507 that the microsecond fraction of "now" was more than the microsecond
4508 fraction of "then" (but the whole seconds number was less).
4510 38. Added support for the libradius Radius client library this is found on
4511 FreeBSD (previously only the radiusclient library was supported).
4517 1. When certain lookups returned multiple values in the form name=value, the
4518 quoting of the values was not always being done properly. Specifically:
4519 (a) If the value started with a double quote, but contained no whitespace,
4521 (b) If the value contained whitespace other than a space character (i.e.
4522 tabs or newlines or carriage returns) it was not quoted.
4523 This fix has been applied to the mysql and pgsql lookups by writing a
4524 separate quoting function and calling it from the lookup code. The fix
4525 should probably also be applied to nisplus, ibase and oracle lookups, but
4526 since I cannot test any of those, I have not disturbed their existing code.
4528 2. A hit in the callout cache for a specific address caused a log line with no
4529 reason for rejecting RCPT. Now it says "Previous (cached) callout
4530 verification failure".
4532 3. There was an off-by-one bug in the queryprogram router. An over-long
4533 return line was truncated at 256 instead of 255 characters, thereby
4534 overflowing its buffer with the terminating zero. As well as fixing this, I
4535 have increased the buffer size to 1024 (and made a note to document this).
4537 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4538 when Exim is waiting for an SMTP response from a remote server, Exim
4539 restarts its select() call on the socket, thereby resetting its timeout.
4540 This is not a problem when such interrupts are rare. Somebody set up a cron
4541 job to run exiwhat every 2 minutes, which is less than the normal select()
4542 timeout (5 or 10 minutes). This meant that the select() timeout never
4543 kicked in because it was always reset. I have fixed this by comparing the
4544 time when an interrupt arrives with the time at the start of the first call
4545 to select(). If more time than the timeout has elapsed, the interrupt is
4546 treated as a timeout.
4548 5. Some internal re-factoring in preparation for the addition of Sieve
4549 extensions (by MH). In particular, the "personal" test is moved to a
4550 separate function, and given an option for scanning Cc: and Bcc: (which is
4551 not set for Exim filters).
4553 6. When Exim created an email address using the login of the caller as the
4554 local part (e.g. when creating a From: or Sender: header line), it was not
4555 quoting the local part when it contained special characters such as @.
4557 7. Installed new OpenBSD configuration files.
4559 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4560 try to make them clearer.
4562 9. Callout options, other than the timeout value, were being ignored when
4563 verifying sender addresses in header lines. For example, when using
4565 verify = header_sender/callout=no_cache
4567 the cache was (incorrectly) being used.
4569 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4570 only the Cygwin environment.
4572 11. When return_path_on_delivery was set as a log selector, if different remote
4573 addresses in the same message used different return paths and parallel
4574 remote delivery occurred, the wrong values would sometimes be logged.
4575 (Whenever a remote delivery process finished, the return path value from
4576 the most recently started remote delivery process was logged.)
4578 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4579 lines when AUTH and/or TLS are in use. This is the "received protocol"
4580 field. Exim used to use "asmtp" for authenticated SMTP, without any
4581 indication (in the protocol name) for TLS use. Now it follows the RFC and
4582 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4583 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4584 names appear in log lines as well as in Received: header lines.
4586 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4587 extensions, and comparison tests, and to fix some bugs.
4589 14. Changes to the "personal" filter test:
4591 (1) The test was buggy in that it was just doing the equivalent of
4592 "contains" tests on header lines. For example, if a user's address was
4593 anne@some.where, the "personal" test would incorrectly be true for
4595 To: susanne@some.where
4597 This test is now done by extracting each address from the header in turn,
4598 and checking the entire address. Other tests that are part of "personal"
4599 are now done using regular expressions (for example, to check local parts
4600 of addresses in From: header lines).
4602 (2) The list of non-personal local parts in From: addresses has been
4603 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4604 taken from the Sieve specification recommendations.
4606 (3) If the message contains any header line starting with "List-" it is
4607 treated as non-personal.
4609 (4) The test for "circular" in the Subject: header line has been removed
4610 because it now seems ill-conceived.
4612 15. Minor typos in src/EDITME comments corrected.
4614 16. Installed latest exipick from John Jetmore.
4616 17. If headers_add on a router specified a text string that was too long for
4617 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4618 of string_sprintf() is now avoided.
4620 18. $message_body_size was not set (it was always zero) when running the DATA
4621 ACL and the local_scan() function.
4623 19. For the "mail" command in an Exim filter, no default was being set for
4624 the once_repeat time, causing a random time value to be used if "once" was
4625 specified. (If the value happened to be <= 0, no repeat happened.) The
4626 default is now 0s, meaning "never repeat". The "vacation" command was OK
4627 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4628 (I found it when inspecting the code).
4630 20. There is now an overall timeout for performing a callout verification. It
4631 defaults to 4 times the callout timeout, which applies to individual SMTP
4632 commands during the callout. The overall timeout applies when there is more
4633 than one host that can be tried. The timeout is checked before trying the
4634 next host. This prevents very long delays if there are a large number of
4635 hosts and all are timing out (e.g. when the network connections are timing
4636 out). The value of the overall timeout can be changed by specifying an
4637 additional sub-option for "callout", called "maxwait". For example:
4639 verify = sender/callout=5s,maxwait=20s
4641 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4642 to the end before writing, but this should make it even safer).
4644 22. Exim was forgetting that it had advertised PIPELINING for the second and
4645 subsequent messages on an SMTP connection. It was also not resetting its
4646 memory on STARTTLS and an internal HELO.
4648 23. When Exim logs an SMTP synchronization error within a session, it now
4649 records whether PIPELINING has been advertised or not.
4651 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4652 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4653 rather than long int.
4655 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4657 26. Added the never_mail option to autoreply.
4663 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4664 crash if the getsockname() call failed; this can happen if a connection is
4665 closed very soon after it is established. The problem was simply in the
4666 order in which certain operations were done, causing Exim to try to write
4667 to the SMTP stream before it had set up the file descriptor. The bug has
4668 been fixed by making things happen in the correct order.
4674 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4675 before the connection was closed, thus losing the rejection response.
4677 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4678 some early Solaris releases, but causes trouble in current releases where
4679 socklen_t is defined.
4681 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4684 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4687 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4688 that releasing the top of it at the end releases what was used for sub-
4689 expansions (unless the block got too big). However, discard this block if
4690 the first thing is a variable or header, so that we can use its block when
4691 it is dynamic (useful for very large $message_headers, for example).
4693 6. Lookups now cache *every* query, not just the most recent. A new, separate
4694 store pool is used for this. It can be recovered when all lookup caches are
4695 flushed. Lookups now release memory at the end of their result strings.
4696 This has involved some general refactoring of the lookup sources.
4698 7. Some code has been added to the store_xxx() functions to reduce the amount
4699 of flapping under certain conditions.
4701 8. log_incoming_interface used to affect only the <= reception log lines. Now
4702 it causes the local interface and port to be added to several more SMTP log
4703 lines, for example "SMTP connection from", and rejection lines.
4705 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4707 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4709 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4710 could be overwritten at the end of the current message (or the start of a
4711 new message if it was set in a HELO ACL). The value is now preserved for
4712 the duration of the SMTP connection.
4714 12. If a transport had a headers_rewrite setting, and a matching header line
4715 contained an unqualified address, that address was qualified, even if it
4716 did not match any rewriting rules. The underlying bug was that the values
4717 of the flags that permit the existence of unqualified sender and recipient
4718 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4719 non-local messages, and by -bnq for local messages) were not being
4720 preserved with the message after it was received.
4722 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4723 "next input=" as part of the text comprising the host identity instead of
4724 the correct text. The code was using the same buffer for two different
4725 strings. However, depending on which order the printing function evaluated
4726 its arguments, the bug did not always show up. Under Linux, for example, my
4727 test suite worked just fine.
4729 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4730 doesn't work with some older versions of Perl. It has been changed to "my",
4731 which in any case is probably the better facility to use.
4733 15. A really picky compiler found some instances of statements for creating
4734 error messages that either had too many or two few arguments for the format
4737 16. The size of the buffer for calls to the DNS resolver has been increased
4738 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4739 for addresses that have a lot of PTR records. This alleviates a problem; it
4740 does not fully solve it.
4742 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4743 buffer now truncates the list and logs the incident, which is the same
4744 action as happens when Exim is looking up a host name and its aliases.
4745 Previously in this situation something unpredictable would happen;
4746 sometimes it was "internal error: store_reset failed".
4748 18. If a server dropped the connection unexpectedly when an Exim client was
4749 using GnuTLS and trying to read a response, the client delivery process
4750 crashed while trying to generate an error log message.
4752 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4753 string, for example:
4755 warn message = H1: something\nH2: something
4757 the text was added as a single header line from Exim's point of view
4758 though it ended up OK in the delivered message. However, searching for the
4759 second and subsequent header lines using $h_h2: did not work. This has been
4760 fixed. Similarly, if a system filter added multiple headers in this way,
4761 the routers could not see them.
4763 20. Expanded the error message when iplsearch is called with an invalid key to
4764 suggest using net-iplsearch in a host list.
4766 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4767 ACLs are no longer actually imposed (and a message to that effect is
4770 22. If a "gecos" field in a passwd entry contained escaped characters, in
4771 particular, if it contained a \" sequence, Exim got it wrong when building
4772 a From: or a Sender: header from that name. A second bug also caused
4773 incorrect handling when an unquoted " was present following a character
4774 that needed quoting.
4776 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4777 was not being matched caselessly.
4779 24. Arranged for all hyphens in the exim.8 source to be escaped with
4782 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4783 properly. Recipient callout cache records were still being keyed to include
4784 the sender, even when use_sender was set false. This led to far more
4785 callouts that were necessary. The sender is no longer included in the key
4786 when use_sender is false.
4788 26. Added "control = submission" modifier to ACLs.
4790 27. Added the ${base62d: operator to decode base 62 numbers.
4792 28. dnsdb lookups can now access SRV records.
4794 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4795 the configuration file.
4797 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4798 (-v) mode. This makes the output for a verbose queue run more intelligible.
4800 31. Added a use_postmaster feature to recipient callouts.
4802 32. Added the $body_zerocount variable, containing the number of binary zero
4803 bytes in the message body.
4805 33. The time of last modification of the "new" subdirectory is now used as the
4806 "mailbox time last read" when there is a quota error for a maildir
4809 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4811 35. Added +ignore_unknown as a special item in host lists.
4813 36. Code for decoding IPv6 addresses in host lists is now included, even if
4814 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4815 address was recognized as an IP address, but was then not correctly decoded
4816 into binary, causing unexpected and incorrect effects when compared with
4823 1. Very minor rewording of debugging text in manualroute to say "list of
4824 hosts" instead of "hostlist".
4826 2. If verify=header_syntax was set, and a header line with an unqualified
4827 address (no domain) and a large number of spaces between the end of the
4828 name and the colon was received, the reception process suffered a buffer
4829 overflow, and (when I tested it) crashed. This was caused by some obsolete
4830 code that should have been removed. The fix is to remove it!
4832 3. When running in the test harness, delay a bit after writing a bounce
4833 message to get a bit more predictability in the log output.
4835 4. Added a call to search_tidyup() just before forking a reception process. In
4836 theory, someone could use a lookup in the expansion of smtp_accept_max_
4837 per_host which, without the tidyup, could leave open a database connection.
4839 5. Added the variables $recipient_data and $sender_data which get set from a
4840 lookup success in an ACL "recipients" or "senders" condition, or a router
4841 "senders" option, similar to $domain_data and $local_part_data.
4843 6. Moved the writing of debug_print from before to after the "senders" test
4846 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4847 problems for message scanning, either using a data ACL, or using
4848 local_scan() because the Received: header was not generated till after they
4849 were called (in order to set the time as the time of reception completion).
4850 I have revised the way this works. The header is now generated after the
4851 body is received, but before the ACL or local_scan() are called. After they
4852 are run, the timestamp in the header is updated.
4858 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4859 before starting a queue runner without re-exec. This happened only when
4860 deliver_drop_privilege was set or when the Exim user was set to root. The
4861 effect of the bug was that timeouts during subsequent deliveries caused
4862 crashes instead of being properly handled. The handler is now left at its
4863 default (and expected) setting.
4865 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4866 message, again when deliver_drop_privilege is set or Exim is run as root.
4867 The bug described in (1) was not present in this case, but the tidying up
4868 of the other signals was missing. I have made the two cases consistent.
4870 3. The ignore_target_hosts setting on a manualroute router was being ignored
4871 for hosts that were looked up using the /MX notation.
4873 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4876 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4877 operated on the sender address. After changing the $sender_address to <>
4878 for the sender address verify, Exim was re-instated it as the original
4879 (before rewriting) address, but remembering that it had rewritten it, so it
4880 wasn't rewriting it again. This bug also had the effect of breaking the
4881 sender address verification caching when the sender address was rewritten.
4883 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4884 This has been changed so that if the ip literal address matches
4885 ignore_target_hosts, the router declines.
4887 7. Added expansion conditions match_domain, match_address, and match_local_
4888 part (NOT match_host).
4890 8. The placeholder for the Received: header didn't have a length field set.
4892 9. Added code to Exim itself and to exim_lock to test for a specific race
4893 condition that could lead to file corruption when using MBX delivery. The
4894 issue is with the lockfile that is created in /tmp. If this file is removed
4895 after a process has opened it but before that process has acquired a lock,
4896 there is the potential for a second process to recreate the file and also
4897 acquire a lock. This could lead to two Exim processes writing to the file
4898 at the same time. The added code performs the same test as UW imapd; it
4899 checks after acquiring the lock that its file descriptor still refers to
4900 the same named file.
4902 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4903 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4906 11. Added the smtp_active_hostname option. If used, this will typically be made
4907 to depend on the incoming interface address. Because $interface_address is
4908 not set up until the daemon has forked a reception process, error responses
4909 that can happen earlier (such as "too many connections") no longer contain
4912 12. If an expansion in a condition on a "warn" statement fails because a lookup
4913 defers, the "warn" statement is abandoned, and the next ACL statement is
4914 processed. Previously this caused the whole ACL to be aborted.
4916 13. Added the iplsearch lookup type.
4918 14. Added ident_timeout as a log selector.
4920 15. Added tls_certificate_verified as a log selector.
4922 16. Added a global option tls_require_ciphers (compare the smtp transport
4923 option of the same name). This controls incoming TLS connections.
4925 17. I finally figured out how to make tls_require_ciphers do a similar thing
4926 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4927 before starting the TLS session.
4929 18. Tabs are now shown as \t in -bP output.
4931 19. If the log selector return_path_on_delivery was set, Exim crashed when
4932 bouncing a message because it had too many Received: header lines.
4934 20. If two routers both had headers_remove settings, and the first one included
4935 a superfluous trailing colon, the final name in the first list and the
4936 first name in the second list were incorrectly joined into one item (with a
4937 colon in the middle).
4943 1. Added -C and -D options to the exinext utility, mainly to make it easier
4944 to include in the automated testing, but these could be helpful when
4945 multiple configurations are in use.
4947 2. The exinext utility was not formatting the output nicely when there was
4948 an alternate port involved in the retry record key, nor when there was a
4949 message id as well (for retries that were specific to a specific message
4950 and a specific host). It was also confused by IPv6 addresses, because of
4951 the additional colons they contain. I have fixed the IPv4 problem, and
4952 patched it up to do a reasonable job for IPv6.
4954 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4955 delivery, the log line now contains "pipelined" if PIPELINING was used.
4957 4. An SMTP transport process used to panic and die if the bind() call to set
4958 an explicit outgoing interface failed. This has been changed; it is now
4959 treated in the same way as a connect() failure.
4961 5. A reference to $sender_host_name in the part of a conditional expansion
4962 that was being skipped was still causing a DNS lookup. This no longer
4965 6. The def: expansion condition was not recognizing references to header lines
4966 that used bh_ and bheader_.
4968 7. Added the _cache feature to named lists.
4970 8. The code for checking quota_filecount in the appendfile transport was
4971 allowing one more file than it should have been.
4973 9. For compatibility with Sendmail, the command line option
4981 and sets the incoming protocol and host name (for trusted callers). The
4982 host name and its colon can be omitted when only the protocol is to be set.
4983 Note the Exim already has two private options, -pd and -ps, that refer to
4984 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4985 "s", but I don't think that's a major issue.
4987 10. A number of refactoring changes to the code, none of which should affect
4990 (a) The number of logging options was getting close to filling up the
4991 32-bit word that was used as a bit map. I have split them into two classes:
4992 those that are passed in the argument to log_write(), and those that are
4993 only ever tested independently outside of that function. These are now in
4994 separate 32-bit words, so there is plenty of room for expansion again.
4995 There is no change in the user interface or the logging behaviour.
4997 (b) When building, for example, log lines, the code previously used a
4998 macro that called string_cat() twice, in order to add two strings. This is
4999 not really sufficiently general. Furthermore, there was one instance where
5000 it was actually wrong because one of the argument was used twice, and in
5001 one call a function was used. (As it happened, calling the function twice
5002 did not affect the overall behaviour.) The macro has been replaced by a
5003 function that can join an arbitrary number of extra strings onto a growing
5006 (c) The code for expansion conditions now uses a table and a binary chop
5007 instead of a serial search (which was left over from when there were very
5008 few conditions). Also, it now recognizes conditions like "pam" even when
5009 the relevant support is not compiled in: a suitably worded error message is
5010 given if an attempt is made to use such a condition.
5012 11. Added ${time_interval:xxxxx}.
5014 12. A bug was causing one of the ddress fields not to be passed back correctly
5015 from remote delivery subprocesses. The field in question was not being
5016 subsequently used, so this caused to problems in practice.
5018 13. Added new log selectors queue_time and deliver_time.
5020 14. Might have fixed a bug in maildirsizefile handling that threw up
5021 "unexpected character" debug warnings, and recalculated the data
5022 unnecessarily. In any case, I expanded the warning message to give more
5025 15. Added the message "Restricted characters in address" to the statements in
5026 the default ACL that block characters like @ and % in local parts.
5028 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5029 Three changes have been made:
5031 (a) There was a serious bug; a negative response to MAIL caused the whole
5032 recipient domain to be cached as invalid, thereby blocking all messages
5033 to all local parts at the same domain, from all senders. This bug has
5034 been fixed. The domain is no longer cached after a negative response to
5035 MAIL if the sender used is not empty.
5037 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5040 (c) A new callout option, "use_sender" has been added for people who want
5041 the modified behaviour.
5047 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5050 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5051 indeed breaks things for older releases.
5053 3. Added additional logging to the case where there is a problem reading data
5054 from a filter that is running in a subprocess using a pipe, in order to
5055 try to track down a specific problem.
5057 4. Testing facility fudge: when running in the test harness and attempting
5058 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5059 getting "No route to host". Convert this to a timeout.
5061 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5064 6. Some OS don't have socklen_t but use size_t instead. This affects the
5065 fifth argument of getsockopt() amongst other things. This is now
5066 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5067 can be set for individual OS. I have set it for SunOS5, OSF1, and
5068 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5069 some earlier ones do not.
5071 7. Change 4.30/15 was not doing the test caselessly.
5073 8. The standard form for an IPv6 address literal was being rejected by address
5074 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5075 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5076 this, as well as the form without the "IPv6" on the front (but only when
5077 address literals are enabled, of course).
5079 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5081 10. Exim crashed if a message with an empty sender address specified by -f
5082 encountered a router with an errors_to setting. This could be provoked only
5083 by a command such as
5087 where an empty string was supplied; "<>" did not hit this bug.
5089 11. Installed PCRE release 4.5.
5091 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5092 remained set. It is now erased.
5094 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5095 times from message ids (which are base 36 rather than the normal 62).
5097 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5098 were being counted as actual protocol errors, and logged if the log
5099 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5100 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5101 and DATA following a set of rejected RCPTs do not count as protocol errors.
5102 In other words, Exim assumes they were pipelined, though this may not
5103 actually be the case. Of course, in all cases the client gets an
5104 appropriate error code.
5106 15. If a lookup fails in an ACL condition, a message about the failure may
5107 be available; it is used if testing the ACL cannot continue, because most
5108 such messages specify what the cause of the deferral is. However, some
5109 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5110 that caused an old message to be retained and used if a later statement
5111 caused a defer, replacing the real cause of the deferral.
5113 16. If an IP address had so many PTR records that the DNS lookup buffer
5114 was not large enough to hold them, Exim could crash while trying to process
5115 the truncated data. It now detects and logs this case.
5117 17. Further to 4.21/58, another change has been made: if (and only if) the
5118 first line of a message (the first header line) ends with CRLF, a bare LF
5119 in a subsequent header line has a space inserted after it, so as not to
5120 terminate the header.
5122 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5123 unnecessarily, used atoi() instead of strtol(), and didn't check the
5124 termination when getting file sizes from file names by regex.
5126 19. Completely re-implemented the support for maildirsize files, in the light
5127 of a number of problems with the previous contributed implementation
5128 (4.30/29). In particular:
5130 . If the quota is zero, the maildirsize file is maintained, but no quota is
5133 . If the maildir directory does not exist, it is created before any attempt
5134 to write a maildirsize file.
5136 . The quota value in the file is just a cache; if the quota is changed in
5137 the transport, the new value overrides.
5139 . A regular expression is available for excluding directories from the
5142 20. The autoreply transport checks the characters in options that define the
5143 message's headers; it allows continued headers, but it was checking with
5144 isspace() after an embedded newline instead of explicitly looking for a
5147 21. If all the "regular" hosts to which an address was routed had passed their
5148 expiry times, and had not reached their retry times, the address was
5149 bounced, even if fallback hosts were defined. Now Exim should go on to try
5152 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5153 equivalent code in the SMTP transport. Some hosts send humungous responses
5154 to HELO/EHLO, more than 1024 it seems.
5156 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5157 gives compiler warnings in some environments. I've now done it "properly",
5160 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5161 (because of problems with the built-in one) was declared to return uschar *
5162 instead of char *, causing compiler failure.
5164 25. Fixed a file descriptor leak when processing alias/forward files.
5166 26. Fixed a minor format string issue in dbfn.c.
5168 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5170 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5171 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5174 29. When a sender address is verified, it is cached, to save repeating the test
5175 when there is more than one recipient in a message. However, when the
5176 verification involves a callout, it is possible for different callout
5177 options to be set for different recipients. It is too complicated to keep
5178 track of this in the cache, so now Exim always runs a verification when a
5179 callout is required, relying on the callout cache for the optimization.
5180 The overhead is duplication of the address routing, but this should not be
5183 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5184 to be verified with callout=postmaster, and the main callout worked but the
5185 postmaster check failed, the verification correctly failed. However, if a
5186 subsequent RCPT command asked for sender verification *without* the
5187 postmaster check, incorrect caching caused this verification also to fail,
5190 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5191 it was not caching the DNS options (qualify_single, search_parents) that
5192 were used when the lookup failed. A subsequent lookup with different
5193 options therefore always gave the same answer, though there were cases
5194 where it should not have. (Example: a "domains = !$mx_any" option on a
5195 dnslookup router: the "domains" option is always processed without any
5196 widening, but the router might have qualify_single set.) Now Exim uses the
5197 cached value only when the same options are set.
5199 32. Added John Jetmore's "exipick" utility to the distribution.
5201 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5202 than a timeout (e.g. a certificate is required, and is not provided), an
5203 Exim server now closes the connection immediately. Previously it waited for
5204 the client to close - but if the client is SSL, it seems that they each
5205 wait for each other, leading to a delay before one of them times out.
5207 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5208 maintained 0.8.x compatibility because I don't think many are using it, and
5209 it is clearly obsolete.
5211 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5214 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5215 client certificate was expired. A simple patch fixes this, though I don't
5216 understand the full logic of why the verify callback is called multiple
5219 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5220 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5221 which causes problems with some clients (such as the Certicom SSL Plus
5222 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5223 disables the coutermeasure allowing Eudora to connect."
5225 38. Exim was not checking that a write() to a log file succeeded. This could
5226 lead to Bad Things if a log got too big, in particular if it hit a file
5227 size limit. Exim now panics and dies if it cannot write to a log file, just
5228 as it does if it cannot open a log file.
5230 39. Modified OS/Makefile-Linux so that it now contains
5232 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5234 The two -D definitions ensure that Exim is compiled with large file
5235 support, which makes it possible to handle log files that are bigger than
5238 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5239 instance) a domain was checked against a named list that involved a lookup,
5240 causing $domain_data to be set, then another domain was checked against the
5241 same list, then the first domain was re-checked, the value of $domain_data
5242 after the final check could be wrong. In particular, if the second check
5243 failed, it could be set empty. This bug probably also applied to
5246 41. The strip_trailing_dot option was not being applied to the address given
5247 with the -f command-line option.
5249 42. The code for reading a message's header from the spool was incrementing
5250 $received_count, but never initializing it. This meant that the value was
5251 incorrect (doubled) while delivering a message in the same process in which
5252 it was received. In the most common configuration of Exim, this never
5253 happens - a fresh exec is done - but it can happen when
5254 deliver_drop_privilege is set.
5256 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5257 it now includes up to 150 characters of the unexpected data in the log
5260 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5261 and building data strings. The size of both of these buffers was 10 000
5262 bytes - far larger than anybody would *ever* want, thought I. Needless to
5263 say, somebody hit the limit. I have increased the maximum line length to
5264 20 000 and the maximum data length of concatenated lines to 100 000. I have
5265 also fixed two bugs, because there was no checking on these buffers. Tsk,
5266 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5267 buffer is too small.
5269 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5270 lsearch lookups. Now it does.
5272 46. When parsing a route_list item in a manualroute router, a fixed-length
5273 buffer was used for the list of hosts. I made this 1024 bytes long,
5274 thinking that nobody would ever have a list of hosts that long. Wrong.
5275 Somebody had a whole pile of complicated expansion conditions, and the
5276 string was silently truncated, leading to an expansion error. It turns out
5277 that it is easier to change to an unlimited length (owing to other changes
5278 that have happened since this code was originally written) than to build
5279 structure for giving a limitation error. The length of the item that
5280 expands into the list of hosts is now unlimited.
5282 47. The lsearch lookup could not handle data where the length of text line was
5283 more than 4095 characters. Such lines were truncated, leading to shortened
5284 data being returned. It should now handle lines of any length.
5286 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5287 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5290 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5291 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5292 no longer happens in comment lines. A list of replacements is now placed
5293 at the head of all of the source files, except those whose only change is
5294 to replace PERL_COMMAND in the very first #! line.
5296 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5297 messages on the queue, with a bottom-up merge sort, using code contributed
5298 by Michael Haardt. This should make operations like -bp somewhat faster on
5299 large queues. It won't affect queue runners, except when queue_run_in_order
5302 51. Installed eximstats 1.31 in the distribution.
5304 52. Added support for SRV lookups to the dnslookup router.
5306 53. If an ACL referred to $message_body or $message_body_end, the value was not
5307 reset for any messages that followed in the same SMTP session.
5309 54. The store-handling optimization for building very long strings was not
5310 differentiating between the different store pools. I don't think this
5311 actually made any difference in practice, but I've tidied it.
5313 55. While running the routers to verify a sender address, $sender_address
5314 was still set to the sender address. This is wrong, because when routing to
5315 send a bounce to the sender, it would be empty. Therefore, I have changed
5316 it so that, while verifying a sender address, $sender_address is set to <>.
5317 (There is no change to what happens when verifying a recipient address.)
5319 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5320 target A or AAAA records (if not already returned) without resetting the
5321 qualify_single or search_parents options of the DNS resolver. These are
5322 inappropriate in this case because the targets of MX and SRV records must
5323 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5324 target that, when qualified, matched something in the local domain. These
5325 two options are now turned off when doing these lookups.
5327 57. It seems that at least some releases of Reiserfs (which does not have the
5328 concept of a fixed number of inodes) returns zero and not -1 for the
5329 number of available inodes. This interacted badly with check_spool_inodes,
5330 which assumed that -1 was the "no such thing" setting. What I have done is
5331 to check that the total number of inodes is greater than zero before doing
5332 the test of how many are available.
5334 58. When a "warn" ACL statement has a log_message modifier, the message is
5335 remembered, and not repeated. This is to avoid a lot of repetition when a
5336 message has many recipients that cause the same warning to be written.
5337 Howewer, Exim was preserving the list of already written lines for an
5338 entire SMTP session, which doesn't seem right. The memory is now reset if a
5339 new message is started.
5341 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5342 debugging output unless log_rewrite was also set.
5344 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5345 of (int)(handle) when we know that handle contains (void *)(-1).
5347 61. The Exim daemon panic-logs an error return when it closes the incoming
5348 connection. However "connection reset by peer" seems to be common, and
5349 isn't really an error worthy of noting specially, so that particular error
5352 62. When Exim is trying to find all the local interfaces, it used to panic and
5353 die if the ioctl to get the interface flags failed. However, it seems that
5354 on at least one OS (Solaris 9) it is possible to have an interface that is
5355 included in the list of interfaces, but for which you get a failure error
5356 for this call. This happens when the interface is not "plumbed" into a
5357 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5358 failure of the "get flags" call assumes that the interface is down.
5360 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5361 makes life easier for people who are doing arithmetic on fields extracted
5362 from dates, where you often get leading zeros that should not be
5363 interpreted as octal.
5365 64. Added qualify_domain to the redirect router, to override the global
5368 65. If a pathologically long header line contained very many addresses (the
5369 report of this problem mentioned 10 000) and each of them was rewritten,
5370 Exim could use up a very large amount of memory. (It kept on making new
5371 copies of the header line as it rewrote, and never released the old ones.)
5372 At the expense of a bit more processing, the header rewriting function has
5373 been changed so that it no longer eats memory in this way.
5375 66. The generation of the Received: header has been moved from the time that a
5376 message starts to be received, to the time that it finishes. The timestamp
5377 in the Received: header should now be very close to that of the <= log
5378 line. There are two side-effects of this change:
5380 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5381 logged header lines no longer include the local Received: line, because
5382 it has not yet been created. The same applies to a copy of the message
5383 that is returned to a non-SMTP sender when a message is rejected.
5385 (b) When a filter file is tested using -bf, no additional Received: header
5386 is added to the test message. After some thought, I decided that this
5389 This change does not affect the value of $received_for. It is still set
5390 after address rewriting, but before local_scan() is called.
5392 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5394 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5395 gave an unhelpful panic error message, and a defer error. I have managed to
5396 change this behaviour so that it now rejects any supplied certificate,
5397 which seems right, as the list of acceptable certificates is empty.
5399 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5400 gave an unhelpful defer error. I have not managed to make this reject any
5401 supplied certificates, but the error message it gives is "no certificate
5402 supplied", which is not helpful.
5404 70. exigrep's output now also includes lines that are not associated with any
5405 message, but which match the given pattern. Implemented by a patch from
5406 Martin Sluka, which also tidied up the Perl a bit.
5408 71. Recipient callout verification, like sender verification, was using <> in
5409 the MAIL FROM command. This isn't really the right thing, since the actual
5410 sender may affect whether the remote host accepts the recipient or not. I
5411 have changed it to use the actual sender in the callout; this means that
5412 the cache record is now keyed on a recipient/sender pair, not just the
5413 recipient address. There doesn't seem to be a real danger of callout loops,
5414 since a callout by the remote host to check the sender would use <>.
5415 [SEE ABOVE: changed after hitting problems.]
5417 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5418 temporary errors. However, in the case of such a code being given after
5419 the end of a data transmission (i.e. after ".") Exim was failing to write
5420 a retry record for the message. (Yes, there was some broken host that was
5421 actually sending 8xx at this point.)
5423 73. An unknown lookup type in a host list could cause Exim to panic-die when
5424 the list was checked. (An example that provoked this was putting <; in the
5425 middle of a list instead of at the start.) If this happened during a DATA
5426 ACL check, a -D file could be left lying around. This kind of configuration
5427 error no longer causes Exim to die; instead it causes a defer errror. The
5428 incident is still logged to the main and panic logs.
5430 74. Buglet left over from Exim 3 conversion. The message "too many messages
5431 in one connection" was written to the rejectlog but not the mainlog, except
5432 when address rewriting (yes!) was being logged.
5434 75. Added write_rejectlog option.
5436 76. When a system filter was run not as root (that is, when system_filter_user
5437 was set), the values of the $n variables were not being returned to the
5438 main process; thus, they were not subsequently available in the $sn
5441 77. Added +return_path_on_delivery log selector.
5443 78. A connection timeout was being treated differently from recipients deferred
5444 when testing hosts_max_try with a message that was older than the host's
5445 retry timeout. (The host should not be counted, thus allowing all hosts to
5446 be tried at least once before bouncing.) This may have been the cause of an
5447 occasionally reported bug whereby a message would remain on the queue
5448 longer than the retry timeout, but would be bounced if a delivery was
5449 forced. I say "may" because I never totally pinned down the problem;
5450 setting up timeout/retry tests is difficult. See also the next item.
5452 79. The ultimate address timeout was not being applied to errors that involved
5453 a combination of host plus message (for example, a timeout on a MAIL
5454 command). When an address resolved to a number of possible hosts, and they
5455 were not all tried for each delivery (e.g. because of hosts_max_try), a
5456 message could remain on the queue longer than the retry timeout.
5458 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5461 81. Fixed an obscure SMTP outgoing bug which required at least the following
5462 conditions: (a) there was another message waiting for the same server;
5463 (b) the server returned 5xx to all RCPT commands in the first message so
5464 that the message was not completed; (c) the server dropped the connection
5465 or gave a negative response to the RSET that Exim sends to abort the
5466 transaction. The observed case was a dropped connection after DATA that had
5467 been sent in pipelining mode. That is, the server had advertised PIPELINING
5468 but was not implementing it correctly. The effect of the bug was incorrect
5469 behaviour, such as trying another host, and this could lead to a crash.
5475 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5476 and daemon.c were passed as pointers to ints; they should have been
5477 pointers to socklen_t variables (which are typically unsigned ints).
5479 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5482 3. Fixed a really odd bug that affected only the testing scheme; patching a
5483 certain fixed string in the binary changed the value of another string that
5484 happened to be identical to the end of the original first string.
5486 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5487 name", it returns that address as the IP address. On some operating
5488 systems (e.g. Solaris), it also passes back the IP address string as the
5489 "host name". However, on others (e.g. Linux), it passes back an empty
5490 string. Exim wasn't checking for this, and was changing the host name to an
5491 empty string, assuming it had been canonicized.
5493 5. Although rare, it is permitted to have more than one PTR record for a given
5494 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5495 all the names associated with an address, because they do in Solaris.
5496 However, it seems that they do not in Linux for data that comes from the
5497 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5498 I found this out when I moved to a new Linux workstation and tried to run
5499 the Exim test suite.
5501 To get round this problem I have changed the code so that it now does its
5502 own call to the DNS to look up PTR records when searching for a host name.
5503 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5504 addresses that are only in /etc/hosts are still found.
5506 This behaviour is, however, controlled by an option called host_lookup_
5507 order, which defaults to "bydns:byaddr". If people want to use the other
5508 order, or indeed, just use one or the other means of lookup, they can
5509 specify it in this variable.
5511 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5512 some operating systems, this comes back from gethostbyaddr() as an empty
5513 string, and this is what Exim used to test for. However, it seems that in
5514 other systems, "." is yielded. Exim now tests for this case too.
5516 7. The values of check_spool_space and check_log_space are now held internally
5517 as a number of kilobytes instead of an absolute number of bytes. If a
5518 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5519 kilobyte. This means that much larger values can be stored.
5521 8. Exim monitor: an attempt to get the action menu when not actually pointing
5522 at a message produces an empty menu entitled "No message selected". This
5523 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5524 no entries in it ("Shell widget menu has zero width and/or height"). So I
5525 have added a single, blank menu entry in this case.
5527 9. Added ${quote_local_part.
5529 10. MIME decoding is now applied to the contents of Subject: header lines when
5532 11. Now that a reference to $sender_host_address automatically causes a reverse
5533 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5534 host lookup before query-style lookups in lists that might use this
5535 variable. This has therefore been abolished, and the "net-" prefix is no
5536 longer necessary for query-style lookups.
5538 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5539 have been a typo for LFLAGS, so it has been changed.
5541 13. The install script calls Exim with "-C /dev/null" in order to find the
5542 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5543 to be output. Howeve, since Exim outputs its version number before the
5544 error, it didn't break the script. It just looked ugly. I fixed this by
5545 always allowing "-C /dev/null" if the caller is root.
5547 14. Ignore overlarge ACL variable number when reading spool file - insurance
5548 against a later release with more variables having written the file.
5550 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5551 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5552 this, as well as the form without the "IPv6" on the front.
5554 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5555 OS/Makefile-Darwin file.
5557 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5558 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5560 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5561 variable. However, some OS use ints for inodes. I've added cast to long int
5562 to get rid of the compiler warning.
5564 19. I had forgotten to lock out "/../" in configuration file names when
5565 ALT_CONFIG_PREFIX was set.
5567 20. Routers used for verification do not need to specify transports. However,
5568 if such a router generated a host list, and callout was configured, Exim
5569 crashed, because it could not find a port number from the (non-existent)
5570 transport. It now assumes port 25 in this circumstance.
5572 21. Added the -t option to exigrep.
5574 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5575 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5576 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5577 them is set, LOOKUP_LSEARCH is forced.
5579 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5581 24. Added sender and host information to the "rejected by local_scan()" log
5582 line; previously there was no indication of these.
5584 25. Added .include_if_exists.
5586 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5587 Linux. It turns out that not all file systems support this. Apparently some
5588 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5589 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5590 not supported on the file descriptor, is now ignored when Exim is trying to
5591 sync a directory. This applies only to Linux.
5593 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5595 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5596 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5599 29. Added support for maildirsize files from supplied patch (modified a bit).
5601 30. The use of :fail: followed by an empty string could lead Exim to respond to
5602 sender verification failures with (e.g.):
5604 550 Verification failed for <xxx>
5605 550 Sender verify failed
5607 where the first response line was missing the '-' that indicates it is not
5608 the final line of the response.
5610 31. The loop for finding the name of the user that called Exim had a hardwired
5611 limit of 10; it now uses the value of finduser_retries, which is used for
5612 all other user lookups.
5614 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5617 33. Exim was neglecting to zero errno before one call of strtol() when
5618 expanding a string and expecting an integer value. On some systems this
5619 resulted in spurious "integer overflow" errors. Also, it was casting the
5620 result into an int without checking.
5622 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5623 did not work. The code looks as if it has *never* worked, though it appears
5624 to have been documented since at least releast 1.62. I have made it work.
5626 35. The "timeout_DNS" error in retry rules, also documented since at least
5627 1.62, also never worked. As it isn't clear exactly what this means, and
5628 clearly it isn't a major issue, I have abolished the feature by treating it
5629 as "timeout", and writing a warning to the main and panic logs.
5631 36. The display of retry rules for -brt wasn't always showing the error code
5634 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5635 timeout_connect_A, timeout_connect_MX.
5637 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5638 to the empty sender.
5640 39. The daemon was not analysing the content of -oX till after it had closed
5641 stderr and disconnected from the controlling terminal. This meant that any
5642 syntax errors were only noted on the panic log, and the return code from
5643 the command was 0. By re-arranging the code a little, I've made the
5644 decoding happen first, so such errors now appear on stderr, and the return
5645 code is 1. However, the actual setting up of the sockets still happens in
5646 the disconnected process, so errors there are still only recorded on the
5649 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5650 connections (as happens on some IP stacks) was logged at start up time as
5651 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5652 it from "IPv6 and IPv4", which means that two separate sockets are being
5655 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5656 says whether AF_INET or AF_INET6 was passed as an argument.
5658 42. Exiwhat output was messed up when time zones were included in log
5661 43. Exiwhat now gives more information about the daemon's listening ports,
5662 and whether -tls-on-connect was used.
5664 44. The "port" option of the smtp transport is now expanded.
5666 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5667 silently ignored. Now an error message is written to the main and panic
5670 46. There's a new ACL modifier called "logwrite" which writes to a log file
5671 as soon as it is encountered.
5673 47. Added $local_user_uid and $local_user_gid at routing time.
5675 48. Exim crashed when trying to verify a sender address that was being
5678 49. Exim was recognizing only a space character after ".include". It now also
5679 recognizes a tab character.
5681 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5682 extracting the relevant information from the specification. The man page no
5683 longer contains scrambled data for the -d option, and I've added a section
5684 at the front about calling Exim under different names.
5686 51. Added "extra_headers" argument to the "mail" command in filter files.
5688 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5691 53. Installed eximstats 1.29.
5693 54. Added transport_filter_timeout as a generic transport option.
5695 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5696 Cc: header lines. This was required by RFC 822, but it not required by RFC
5699 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5700 incoming messages that did not have them. Now it does so only if the
5701 message originates locally, that is, if there is no associated remote host
5702 address. When Resent- header lines are present, this applies to the Resent-
5703 lines rather than the non-Resent- lines.
5705 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5706 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5708 58. Messages for configuration errors now include the name of the main
5709 configuration file - useful now that there may be more than one file in a
5710 list (.included file names were always shown).
5712 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5713 for those rare installations that do not start the daemon as root or run it
5714 setuid root. I've cut out the call to initgroups() if the daemon is not
5717 60. The Exim user and group can now be bound into the binary as text strings
5718 that are looked up at the start of Exim's processing.
5720 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5722 62. Added $mailstore_basename variable.
5724 63. Installed patch to sieve.c from Michael Haardt.
5726 64. When Exim failed to open the panic log after failing to open the main log,
5727 the original message it was trying to log was written to stderr and debug
5728 output, but if they were not available (the usual case in production), it
5729 was lost. Now it is written to syslog before the two lines that record the
5730 failures to open the logs.
5732 65. Users' Exim filters run in subprocesses under the user's uid. It is
5733 possible for a "deliver" command or an alias in a "personal" command to
5734 provoke an address rewrite. If logging of address rewriting is configured,
5735 this fails because the process is not running as root or exim. There may be
5736 a better way of dealing with this, but for the moment (because 4.30 needs
5737 to be released), I have disabled address rewrite logging when running a
5738 filter in a non-root, non-exim process.
5744 1. The buildconfig auxiliary program wasn't quoting the value set for
5745 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5746 not defined. This bug was masked in 4.22 by the effect that was fixed in
5749 2. Some messages that were rejected after a message id was allocated were
5750 shown as "incomplete" by exigrep. It no longer does this for messages that
5751 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5753 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5754 have allow_domain_literals set, the ID did not get logged in the <= line.
5755 Domain literals are now always recognized in Message-ID: header lines.
5757 4. The first argument for a ${extract expansion item is the key name or field
5758 number. Leading and trailing spaces in this item were not being ignored,
5759 causing some misleading effects.
5761 5. When deliver_drop_privilege was set, single queue runner processes started
5762 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5763 same command in the process it spins off) were not dropping privilege.
5765 6. When the daemon running as "exim" started a queue runner, it always
5766 re-executed Exim in the spun-off process. This is a waste of effort when
5767 deliver_drop_privilege is set. The new process now just calls the
5768 queue-runner function directly.
5774 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5777 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5778 ignored. Though the use of -oP was forcing the writing of a pid file, it
5779 was always written to the default place.
5781 3. If the message "no IP address found for host xxxx" is generated during
5782 incoming verification, it is now followed by identification of the incoming
5783 connection (so you can more easily find what provoked it).
5785 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5787 5. Added some features to "harden" Exim a bit more against certain attacks:
5789 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5790 be put in Local/Makefile. This is like the never_users runtime option,
5791 but it cannot be overridden. The default setting is "root".
5793 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5794 prefix string with which any file named in a -C command line option
5797 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5798 is retained for -C and -D only if the caller of Exim is root. Without
5799 it, the exim user may also use -C and -D and retain privilege.
5801 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5802 command line option is disabled.
5804 6. Macro names set by the -D option must start with an upper case letter, just
5805 like macro names defined in the configuration file.
5807 7. Added "dereference=" facility to LDAP.
5809 8. Two instances of the typo "uknown" in the source files are fixed.
5811 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5812 the Configure-Makefile script screwed up while processing it.
5814 10. Incorporated PCRE 4.4.
5816 11. The SMTP synchronization check was not operating right at the start of an
5817 SMTP session. For example, it could not catch a HELO sent before the client
5818 waited for the greeting. There is now a check for outstanding input at the
5819 point when the greeting is written. Because of the duplex, asynchronous
5820 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5821 way, but not yet received, when the check is performed.
5823 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5824 on TCP/IP sockets, because this apparently causes some broken clients to
5827 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5828 unchanged) from the Cygwin maintainer.
5830 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5831 was supported instead of testing for mbx. Effectively a typo.
5833 15. The spa authenticator server code was not checking that the input it
5834 received was valid base64.
5836 16. The debug output line for the "set" modifier in ACLs was not showing the
5837 name of the variable that was being set.
5839 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5841 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5842 lookup on its own. Something else was needed to trigger the lookup. For
5843 example, a match in host_lookup or the need for a host name in a host list.
5844 Now, if $sender_host_name is referenced and the host name has not yet been
5845 looked up, a lookup is performed. If the lookup fails, the variable remains
5846 empty, and $host_lookup_failed is set to "1".
5848 19. Added "eqi" as a case-independent comparison operator.
5850 20. The saslauthd authentication condition could segfault if neither service
5851 nor realm was specified.
5853 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5854 error message that was logged was misleading, and incoming SMTP
5855 connections were dropped. The message is now more accurate, and temporary
5856 errors are given to SMTP connections.
5858 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5859 (except RSET) with a 421 error, until QUIT is received. However, it was
5860 failing to send a response to QUIT.
5862 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5863 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5864 ACL is now run after the helo_try_verify_hosts code.
5866 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5867 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5868 case-independent, so other case variants are also recognized). Apparently
5869 some systems use these upper case variants.
5871 25. If more than two messages were waiting for the same host, and a transport
5872 filter was specified for the transport, Exim sent two messages over the
5873 same TCP/IP connection, and then failed with "socket operation on non-
5874 socket" when it tried to send the third.
5876 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5878 27. The extern definition of crypt16() in expand.c was not being excluded when
5879 the OS had its own crypt16() function.
5881 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5882 as a preferred synonym for return_size_limit, both as an option and as an
5885 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5887 30. Changed the default configuration ACL to relax the local part checking rule
5888 for addresses that are not in any local domains. For these addresses,
5889 slashes and pipe symbols are allowed within local parts, but the sequence
5890 /../ is explicitly forbidden.
5892 31. SPA server authentication was not clearing the challenge buffer before
5895 32. log_message in a "warn" ACL statement was writing to the reject log as
5896 well as to the main log, which contradicts the documentation and doesn't
5897 seem right (because no rejection is happening). So I have stopped it.
5899 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5900 However, I am unable to do any testing of this.
5902 34. Fixed an infelicity in the appendfile transport. When checking directories
5903 for a mailbox, to see if any needed to be created, it was accidentally
5904 using path names with one or more superfluous leading slashes; tracing
5905 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5907 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5908 contents are added to the log line that is written for every discarded
5909 recipient. (Previously a log_message setting was ignored.)
5911 36. The ${quote: operator now quotes the string if it is empty.
5913 37. The install script runs exim in order to find its version number. If for
5914 some reason other than non-existence or emptiness, which it checks, it
5915 could not run './exim', it was installing it with an empty version number,
5916 i.e. as "exim-". This error state is now caught, and the installation is
5919 38. An argument was missing from the function that creates an error message
5920 when Exim fails to connect to the socket for saslauthd authentication.
5921 This could cause Exim to crash, or give a corrupted message.
5923 39. Added isip, isip4, and isip6 to ${if conditions.
5925 40. The ACL variables $acl_xx are now saved with the message, and can be
5926 accessed later in routers, transports, and filters.
5928 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5929 strings in the file are not string-expanded.
5931 42. If a MAIL command specified a SIZE value that was too large to fit into an
5932 int variable, the check against message_size_limit failed. Such values are
5933 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5934 day this will have to be increased, but I don't think I want to be around
5935 when emails are that large.
5942 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5943 iconv() is not standard in FreeBSD.
5945 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5946 IPv6 enabled. The observed symptom was a segmentation fault on return from
5947 the function os_common_find_running_interfaces() in src/os.c.
5949 3. In the check_special_case() function in daemon.c I had used "errno" as an
5950 argument name, which causes warnings on some systems. This was basically a
5951 typo, since it was named "eno" in the comments!
5953 4. The code that waits for the clock to tick (at a resolution of some fraction
5954 of a second) so as to ensure message-id uniqueness was always waiting for
5955 at least one whole tick, when it could have waited for less. [This is
5956 almost certainly not relevant at current processor speeds, where it is
5957 unlikely to ever wait at all. But we try to future-proof.]
5959 5. The function that sleeps for a time interval that includes fractions of a
5960 second contained a race. It did not block SIGALRM between setting the
5961 timer, and suspending (a couple of lines later). If the interval was short
5962 and the sigsuspend() was delayed until after it had expired, the suspension
5963 never ended. On busy systems this could lead to processes getting stuck for
5966 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5967 process, before it forks any delivery processes. The open lookup caching
5968 mechanism meant that the open file or database connection was passed into
5969 the delivery process. The problem was that delivery processes always tidy
5970 up cached lookup data. This could cause a problem for the next delivery
5971 process started by the queue runner, because the external queue runner
5972 process does not know about the closure. So the next delivery process
5973 still has data in the lookup cache. In the case of a file lookup, there was
5974 no problem because closing a file descriptor in a subprocess doesn't affect
5975 the parent. However, if the lookup was caching a connection to a database,
5976 the connection was closed, and the second delivery process was likely to
5977 see errors such as "PGSQL: query failed: server closed the connection
5978 unexpectedly". The problem has been fixed by closing all cached lookups
5979 in a queue runner before running a delivery process.
5981 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5982 seem to have the "const" qualifier which it has on other OS. I've
5985 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5986 *of the same type* (client or server) with the same public name that an
5987 error should be diagnosed.
5989 9. When Exim looked up a host name for an IP address, but failed to find the
5990 original IP address when looking up the host name (a safety check), it
5991 output the message "<ip address> does not match any IP for NULL", which was
5992 confusing, to say the least. The bug was that the host name should have
5993 appeared instead of "NULL".
5995 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5996 user that is built into the binary, and the -C or -D options is used, root
5997 privilege is dropped before the configuration file is read. In addition,
5998 logging is switched to stderr instead of the normal log files. If the
5999 configuration then re-defines the Exim user, the unprivileged environment
6000 is probably not what is expected, so Exim logs a panic warning message (but
6003 However, if deliver_drop_privilege is set, the unprivileged state may well
6004 be exactly what is intended, so the warning has been cut out in that case,
6005 and Exim is allowed to try to write to its normal log files.
6011 1. smtp_return_error_details was not giving details for temporary sender
6012 or receiver verification errors.
6014 2. Diagnose a configuration error if two authenticators have the same public
6017 3. Exim used not to create the message log file for a message until the first
6018 delivery attempt. This could be confusing when incoming messages were held
6019 for policy or load reasons. The message log file is now created at the time
6020 the message is received, and an initial "Received" line is written to it.
6022 4. The automatically generated man page for command line options had a minor
6023 bug that caused no ill effects; however, a more serious problem was that
6024 the procedure for building the man page automatically didn't always
6025 operate. Consequently, release 4.20 contains an out-of-date version. This
6026 shouldn't happen again.
6028 5. When building Exim with embedded Perl support, the script that builds the
6029 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6030 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6032 6. The freeze_tell option was not being used for messages that were frozen on
6033 arrival, either by an ACL or by local_scan().
6035 7. Added the smtp_incomplete_transaction log selector.
6037 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6038 was accepting AUTH without a new EHLO.
6040 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6041 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6044 10. Logging of TCP/IP connections (when configured) now happens in the main
6045 daemon process instead of the child process, so that the TCP/IP connection
6046 count is more accurate (but it can never be perfect).
6048 11. The use of "drop" in a nested ACL was not being handled correctly in the
6049 outer ACL. Now, if condition failure induced by the nested "drop" causes
6050 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6051 or "require"), the connection is dropped.
6053 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6054 that yield "discard" can now be used with an "accept" or a "discard" verb,
6055 but an error is generated for any others (because I can't see a useful way
6056 to define what should happen).
6058 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6059 are now processed in the same way as lines in the Exim configuration file.
6060 In particular, continuation lines are supported.
6062 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6064 15. Added -ti meaning -t -i.
6066 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6067 domains, and warn by logging if others are found.
6069 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6070 returned from GIFCONF when Exim is trying to find the list of interfaces on
6071 a host. The code in os.c has been modified to copy each ifreq to an aligned
6072 structure in all cases.
6074 Also, in some cases, the returned ifreq's were being copied to a 'struct
6075 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6076 means the last couple of bytes of an IPv6 address could be chopped if the
6077 ifreq contained only a normal sockaddr (14 bytes storage).
6079 18. Named domain lists were not supported in the hosts_treat_as_local option.
6080 An entry such as +xxxx was not recognized, and was treated as a literal
6083 19. Ensure that header lines added by a DATA ACL are included in the reject log
6084 if the ACL subsequently rejects the message.
6086 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6087 MD5 (which is deprecated).
6089 21. When testing a filter file using -bf, Exim was writing a message when it
6090 took the sender from a "From " line in the message, but it was not doing so
6091 when it took $return_path from a Return-Path: header line. It now does.
6093 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6094 with a valid header line field name (a series of printing characters
6095 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6097 23. Changed "disc" in the source to "disk" to conform to the documentation and
6098 the book and for uniformity.
6100 24. Ignore Sendmail's -Ooption=value command line item.
6102 25. When execve() failed while trying to run a command in a pipe transport,
6103 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6104 could be confused with a return value of 69 from the command itself. This
6105 has been changed to 127, the value the shell returns if it is asked to run
6106 a non-existent command. The wording for the related log line suggests a
6107 non-existent command as the problem.
6109 26. If received_header_text expands to an empty string, do not add a Received:
6110 header line to the message. (Well, it adds a token one on the spool, but
6111 marks it "old" so that it doesn't get used or transmitted.)
6113 27. Installed eximstats 1.28 (addition of -nt option).
6115 28. There was no check for failure on the call to getsockname() in the daemon
6116 code. This can fail if there is a shortage of resources on the system, with
6117 ENOMEM, for example. A temporary error is now given on failure.
6119 29. Contrary to the C standard, it seems that in some environments, the
6120 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6121 program. Exim now does this explicitly; it affects the formatting of
6122 timestamps using strftime().
6124 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6125 complaints. I've now initialized all the variables, to avoid this.
6127 32. Header lines added by a system filter were not being "seen" during
6128 transport-time rewrites.
6130 33. The info_callback() function passed to OpenSSL is set up with type void
6131 (*)(SSL *, int, int), as described somewhere. However, when calling the
6132 function (actually a macro) that sets it up, the type void(*)() is
6133 expected. I've put in a cast to prevent warnings from picky compilers.
6135 34. If a DNS black list lookup found a CNAME record, but there were no A
6136 records associated with the domain it pointed at, Exim crashed.
6138 35. If a DNS black list lookup returned more than one A record, Exim ignored
6139 all but the first. It now scans all returned addresses if a particular IP
6140 value is being sought. In this situation, the contents of the
6141 $dnslist_value variable are a list of all the addresses, separated by a
6144 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6145 to accept a host name and all its aliases if the forward lookup for any of
6146 them yielded the IP address of the incoming connection. Now it accepts only
6147 those names whose forward lookup yields the correct IP address. Any other
6148 names are discarded. This closes a loophole whereby a rogue DNS
6149 administrator could create reverse DNS records to break through a
6150 wildcarded host restriction in an ACL.
6152 37. If a user filter or a system filter that ran in a subprocess used any of
6153 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6154 the wrong values were passed to the pipe command ($thisaddress had the
6155 value of $0, $0 had the value of $1, etc). This bug was introduced by
6156 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6158 38. Improved the line breaking for long SMTP error messages from ACLs.
6159 Previously, if there was no break point between 40 and 75 characters, Exim
6160 left the rest of the message alone. Two changes have been made: (a) I've
6161 reduced the minimum length to 35 characters; (b) if it can't find a break
6162 point between 35 and 75 characters, it looks ahead and uses the first one
6163 that it finds. This may give the occasional overlong line, but at least the
6164 remaining text gets split now.
6166 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6167 file descriptors might be low, and that setting 1000 would always raise it.
6168 It turns out that in some environments, the limit is already over 1000 and
6169 that lowering it causes trouble. So now Exim takes care not to decrease it.
6171 40. When delivering a message, the value of $return_path is set to $sender_
6172 address at the start of routing (routers may change the value). By an
6173 oversight, this default was not being set up when an address was tested by
6174 -bt or -bv, which affected the outcome if any router or filter referred to
6177 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6178 log only when "message" or "log_message" are set. However, if one of the
6179 conditions was an address verification, or a call to a nested ACL, the
6180 messages generated by the underlying test were being passed through. This
6181 no longer happens. The underlying message is available in $acl_verify_
6182 message for both "message" and "log_message" expansions, so it can be
6183 passed through if needed.
6185 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6186 new expansion $bh_ to give the encoded byte string without charset
6187 translation. Translation happens only if iconv() is available; HAVE_ICONV
6188 indicates this at build time. HEADERS_CHARSET gives the charset to
6189 translate to; headers_charset can change it in the configuration, and
6190 "headers charset" can change it in an individual filter file.
6192 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6193 that creates RFC 2047 encoded "words" labels them as that charset instead
6194 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6195 expansion operator; (ii) when Exim creates a From: line for a local
6196 message; (iii) when a header line is rewritten to include a "phrase" part.
6198 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6199 buggy, causing it to skip the first lines of messages whose message ID
6200 ended in 'D'. This would not have bitten before Exim release 4.14, because
6201 message IDs were unlikely to end in 'D' before then. The effect was to have
6202 incorrect size information for certain domains.
6204 45. #include "config.h" was missing at the start of the crypt16.c module. This
6205 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6208 46. If there was a timeout during a "random" callout check, Exim treated it as
6209 a failure of the random address, and carried on sending RSET and the real
6210 address. If the delay was just some slowness somewhere, the response to the
6211 original RCPT would be taken as a response to RSET and so on, causing
6212 mayhem of various kinds.
6214 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6215 when I implemented it. It didn't allow for the fact that some option values
6216 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6217 the right test for positive values.
6219 48. Domain names in DNS records are case-independent. Exim always looks them up
6220 in lower case. Some resolvers return domain names in exactly the case they
6221 appear in the zone file, that is, they may contain uppercase letters. Not
6222 all resolvers do this - some return always lower case. Exim was treating a
6223 change of case by a resolver as a change of domain, similar to a widening
6224 of a domain abbreviation. This triggered its re-routing code and so it was
6225 trying to route what was effectively the same domain again. This normally
6226 caused routing to fail (because the router wouldn't handle the domain
6227 twice). Now Exim checks for this case specially, and just changes the
6228 casing of the domain that it ultimately uses when it transmits the message
6231 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6234 50. If a filter generated a file delivery with a non-absolute name (possible if
6235 no home directory exists for the router), the forbid_file option was not
6238 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6239 the existing equality matching.
6241 52. Exim was using ints instead of ino_t variables in some places where it was
6242 dealing with inode numbers.
6244 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6245 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6246 called TMPDIR, and if it finds it is different, it changes its value.
6248 54. The smtp_printf() function is now made available to local_scan() so
6249 additional output lines can be written before returning. There is also an
6250 smtp_fflush() function to enable the detection of a dropped connection.
6251 The variables smtp_input and smtp_batched_input are exported to
6254 55. Changed the default runtime configuration: the message "Unknown user"
6255 has been removed from the ACL, and instead placed on the localuser router,
6256 using the cannot_route_message feature. This means that any verification
6257 failures that generate their own messages won't get overridden. Similarly,
6258 the "Unrouteable address" message that was in the ACL for unverifiable
6259 relay addresses has also been removed.
6261 56. Added hosts_avoid_esmtp to the smtp transport.
6263 57. The exicyclog script was not checking for the esoteric option
6264 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6265 will work only if exicyclog is run under the appropriate euid.
6267 58. Following a discussion on the list, the rules by which Exim recognises line
6268 endings on incoming messages have been changed. The -dropcr and drop_cr
6269 options are now no-ops, retained only for backwards compatibility. The
6270 following line terminators are recognized: LF CRLF CR. However, special
6271 processing applies to CR:
6273 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6274 nor a local message in the state where . is a terminator.
6276 (ii) If a bare CR is encountered in a header line, an extra space is added
6277 after the line terminator so as not to end the header. The reasoning
6278 behind this is that bare CRs in header lines are most likely either
6279 to be mistakes, or people trying to play silly games.
6281 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6282 was being incorrectly given as 18 bytes larger than it should have been.
6283 This is a VOB (very old bug).
6285 60. This may never have affected anything current, but just in case it has:
6286 When the local host is found other than at the start of a list of hosts,
6287 the local host, those with the same MX, and any that follow, are discarded.
6288 When the list in question was part of a longer list of hosts, the following
6289 hosts (not currently being processed) were also being discarded. This no
6290 longer happens. I'm not sure if this situation could ever has previously
6293 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6296 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6297 header. This is something that is recommended in a new Internet Draft, and
6298 is something that is documented as being done by Sendmail. There are two
6299 possible values. For messages generated by the autoreply transport, Exim
6302 Auto-Submitted: auto-replied
6304 whereas for all other generated messages (e.g. bounces) it adds
6306 Auto-Submitted: auto-generated
6308 63. The "personal" condition in filters now includes a test for the
6309 Auto-Submitted: header. If it contains the string "auto-" the message it
6310 not considered personal.
6312 64. Added rcpt_include_affixes as a generic transport option.
6314 65. Added queue_only_override (default true).
6316 66. Added the syslog_duplication option.
6318 67. If what should have been the first header line of a message consisted of
6319 a space followed by a colon, Exim was mis-interpreting it as a header line.
6320 It isn't of course - it is syntactically invalid and should therefore be
6321 treated as the start of the message body. The misbehaviour could have
6322 caused a number of strange effects, including loss of data in subsequent
6323 header lines, and spool format errors.
6325 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6326 client host had authenticated. This control can now be exercised by an ACL
6327 for more flexibility.
6329 69. By default, callouts do not happen when testing with -bh. There is now a
6330 variant, -bhc, which does actually run the callout code, including
6331 consulting and updating the callout cache.
6333 70. Added support for saslauthd authentication, courtesy of Alexander
6336 71. If statvfs() failed on the spool or log directories while checking their
6337 size for availability, Exim confusingly gave the error "space shortage".
6338 Furthermore, in debugging mode it crashed with a floating point exception.
6339 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6340 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6341 serious problem, Exim now writes to the main and panic logs when this
6342 happens, with details of the failure. It then refuses to accept the
6343 incoming message, giving the message "spool directory problem" or "log
6344 directory problem" with a 421 code for SMTP messages.
6346 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6347 0, 1, and 2 exist, because some OS complain for execs without them (see
6348 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6349 descriptors. However, the code omitted to check that the open succeeded,
6350 causing mysterious errors if for some reason the permissions on /dev/null
6351 got screwed. Now Exim writes a message to the main and panic logs, and
6352 bombs out if it can't open /dev/null.
6354 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6355 interact so that it is all more flexible. It is supposed to remain
6356 backwards compatible. Also added extra_local_interfaces.
6358 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6359 to bomb out with an assertion failure - to the client this appears as a
6360 connection drop. This problem occurs in the part of the code that was taken
6361 from the Samba project. Fortunately, the assertion is in a very simple
6362 function, so I have fixed this by reproducing the function inline in the
6363 one place where it is called, and arranging for authentication to fail
6364 instead of killing the process with assert().
6366 75. The SPA client code was not working when the server requested OEM rather
6367 than Unicode encoding.
6369 76. Added code to make require_files with a specific uid setting more usable in
6370 the case where statting the file as root fails - usually a non-root-mounted
6371 NFS file system. When this happens and the failure is EACCES, Exim now
6372 forks a subprocess and does the per-uid checking as the relevant uid.
6374 77. Added process_log_path.
6376 78. If log_file_path was not explicitly set, a setting of check_log_space or
6377 check_log_inodes was ignored.
6379 79. If a space check for the spool or log partitions fails, the incident is now
6380 logged. Of course, in the latter case the data may get lost...
6382 80. Added the %p formatting code to string_format() so that it can be used to
6383 print addresses in debug_print(). Adjusted all the address printing in the
6384 debugging in store.c to use %p rather than %d.
6386 81. There was a concern that a line of code in smtp_in.c could overflow a
6387 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6388 initially expressed, the concern was not well-founded, because trailing
6389 spaces are removed early. However, if the trailing spaces were followed by
6390 a NULL, they did not get removed, so the overflow was possible. Two fixes
6393 (a) I re-wrote the offending code in a cleaner fashion.
6394 (b) If an incoming SMTP command contains a NULL character, it is rejected
6397 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6398 runs initgroups(), so that if the Exim user is in any additional groups,
6399 they will be used during message reception.
6405 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob