1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
8 and 3.4.4 - once the server is enabled to respond to an OCSP request
9 it does even when not requested, resulting in a stapling non-aware
10 client dropping the TLS connection.
12 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
13 support variable-length bit vectors. No functional change.
15 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
16 The I= interface field on outgoing lines is now after the H= remote
17 host field, same as incoming lines. There is a separate
18 outgoing_interface log selector which allows you to disable the
21 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
22 If not running log_selector +smtp_connection the mainlog would be held
23 open indefinitely after a "too many connections" event, including to a
24 deleted file after a log rotate. Leave the per net connection logging
25 leaving it open for efficiency as that will be quickly detected by the
26 check on the next write.
28 HS/01 Bug 1671: Fix post transport crash.
29 Processing the wait-<transport> messages could crash the delivery
30 process if the message IDs didn't exist for some reason. When
31 using 'split_spool_directory=yes' the construction of the spool
32 file name failed already, exposing the same netto behaviour.
34 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
35 mime_regex ACL conditions.
37 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
38 to DSN fail messages (bounces): remote IP, remote greeting, remote response
39 to HELO, local diagnostic string.
41 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
42 log line to debug. Even when configured with a tls authenticator many
43 client connections are expected to not authenticate in this way, so
44 an authenticate fail is not an error.
46 HS/02 Add the Exim version string to the process info. This way exiwhat
47 gives some more detail about the running daemon.
49 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
50 matter for fast-change records such as DNSBLs.
52 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
53 retry record, even if constant. There may be multiple transports with
54 different interface settings and the retry behaviour needs to be kept
57 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
59 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
61 JH/10 Bug 840: fix log_defer_output option of pipe transport
63 JH/11 Bug 830: use same host for all RCPTS of a message, even under
64 hosts_randomize. This matters a lot when combined with mua_wrapper.
66 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
67 ${quote_pgsql:<string>} operator.
69 JH/13 Bug 1708: avoid misaligned access in cached lookup.
71 JH/14 Change header file name for freeradius-client. Relevant if compiling
72 with Radius support; from the Gentoo tree and checked under Fedora.
74 JH/15 Bug 1712: Introduce $prdr_requested flag variable
76 JH/16 Bug 1714: Permit an empty string as expansion result for transport
77 option transport_filter, meaning no filtering.
79 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
81 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
82 defaults to "*" (all hosts). The variable is now available when not built
83 with TLS, default unset, mainly to enable keeping the testuite sane.
84 If a server certificate is not supplied (via tls_certificate) an error is
85 logged, and clients will find TLS connections fail on startup. Presumably
86 they will retry in-clear.
87 Packagers of Exim are strongly encouraged to create a server certificate
90 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
91 with the $config_file variable.
93 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
94 in transport context, after the attempt, and per-recipient. The latter type
95 is per host attempted. The event data is the error message, and the errno
96 information encodes the lookup type (A vs. MX) used for the (first) host,
97 and the trailing two digits of the smtp 4xx reponse.
99 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
100 to write to mainlog (or rejectlog, paniclog) in the window between file
101 creation and permissions/ownership being changed. Particularly affects
102 installations where exicyclog is run as root, rather than exim user;
103 result is that the running daemon panics and dies.
105 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
107 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
108 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
109 "pri" and "weight". Note that the previous implicit priority given by the
110 list order is no longer honoured.
112 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalisation
118 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
121 JH/02 The smtp transport option "multi_domain" is now expanded.
123 JH/03 The smtp transport now requests PRDR by default, if the server offers
126 JH/04 Certificate name checking on server certificates, when exim is a client,
127 is now done by default. The transport option tls_verify_cert_hostnames
128 can be used to disable this per-host. The build option
129 EXPERIMENTAL_CERTNAMES is withdrawn.
131 JH/05 The value of the tls_verify_certificates smtp transport and main options
132 default to the word "system" to access the system default CA bundle.
133 For GnuTLS, only version 3.0.20 or later.
135 JH/06 Verification of the server certificate for a TLS connection is now tried
136 (but not required) by default. The verification status is now logged by
137 default, for both outbound TLS and client-certificate supplying inbound
140 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
143 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
144 Status Notification (bounce) messages are now MIME format per RFC 3464.
145 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
146 under the control of the dsn_advertise_hosts option, and routers may
147 have a dsn_lasthop option.
149 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
150 default, modifiable by a malware= option. The list separator for
151 the options can now be changed in the usual way. Bug 68.
153 JH/10 The smtp_receive_timeout main option is now expanded before use.
155 JH/11 The incoming_interface log option now also enables logging of the
156 local interface on delivery outgoing connections.
158 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
159 if the interface and destination host and port all match.
161 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
164 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
165 Patch from Andrew Lewis.
167 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
168 now supports optional time-restrictions, weighting, and priority
169 modifiers per server. Patch originally by <rommer@active.by>.
171 JH/16 The spamd_address main option now supports a mixed list of local
172 and remote servers. Remote servers can be IPv6 addresses, and
173 specify a port-range.
175 JH/17 Bug 68: The spamd_address main option now supports an optional
176 timeout value per server.
178 JH/18 Bug 1581: Router and transport options headers_add/remove can
179 now have the list separator specified.
181 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
184 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
187 JH/21 Support for the A6 type of dns record is withdrawn.
189 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
190 rather than the verbs used.
192 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
193 from 255 to 1024 chars.
195 JH/24 Verification callouts now attempt to use TLS by default.
197 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
198 are generic router options now. The defaults didn't change.
200 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
201 Original patch from Alexander Shikoff, worked over by JH.
203 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
204 files if ZCAT_COMMAND is not executable.
206 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
208 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
210 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
211 Normally benign, it bites when the pair was led to by a CNAME;
212 modern usage is to not canoicalize the domain to a CNAME target
213 (and we were inconsistent anyway for A-only vs AAAA+A).
215 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
217 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
218 when evaluating $sender_host_dnssec.
220 JH/31 Check the HELO verification lookup for DNSSEC, adding new
221 $sender_helo_dnssec variable.
223 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
225 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
227 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
229 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
230 documented as working, but never had. Support all but $spam_report.
232 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
233 added for tls authenticator.
238 TL/01 When running the test suite, the README says that variables such as
239 no_msglog_check are global and can be placed anywhere in a specific
240 test's script, however it was observed that placement needed to be near
241 the beginning for it to behave that way. Changed the runtest perl
242 script to read through the entire script once to detect and set these
243 variables, reset to the beginning of the script, and then run through
244 the script parsing/test process like normal.
246 TL/02 The BSD's have an arc4random API. One of the functions to induce
247 adding randomness was arc4random_stir(), but it has been removed in
248 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
249 function when detected.
251 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
252 cause callback expansion.
254 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
255 syntax errors in an expansion can be treated as a string instead of
256 logging or causing an error, due to the internal use of bool_lax
257 instead of bool when processing it.
259 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
260 server certificates when making smtp deliveries.
262 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
264 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
266 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
268 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
269 Merged patch from Sebastian Wiedenroth.
271 JH/05 Fix results-pipe from transport process. Several recipients, combined
272 with certificate use, exposed issues where response data items split
273 over buffer boundaries were not parsed properly. This eventually
274 resulted in duplicates being sent. This issue only became common enough
275 to notice due to the introduction of conection certificate information,
276 the item size being so much larger. Found and fixed by Wolfgang Breyha.
278 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
279 size buffer was used, resulting in syntax errors when an expansion
282 JH/07 Add support for directories of certificates when compiled with a GnuTLS
283 version 3.3.6 or later.
285 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
286 is EXPERIMENTAL_EVENT, the main-configuration and transport options
287 both become "event_action", the variables become $event_name, $event_data
288 and $event_defer_errno. There is a new variable $verify_mode, usable in
289 routers, transports and related events. The tls:cert event is now also
290 raised for inbound connections, if the main configuration event_action
293 TL/06 In test suite, disable OCSP for old versions of openssl which contained
294 early OCSP support, but no stapling (appears to be less than 1.0.0).
296 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
297 server certificate names available under the smtp transport option
298 "tls_verify_cert_hostname" now do not permit multi-component wildcard
301 JH/10 Time-related extraction expansions from certificates now use the main
302 option "timezone" setting for output formatting, and are consistent
303 between OpenSSL and GnuTLS compilations. Bug 1541.
305 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
306 encoded parameter in the incoming message. Bug 1558.
308 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
309 include certificate info, eximon was claiming there were spoolfile
312 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
314 JH/14 Log delivery-related information more consistently, using the sequence
315 "H=<name> [<ip>]" wherever possible.
317 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
318 are problematic for Debian distribution, omit them from the release
321 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
323 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
326 JH/17 Fix a null-indirection in certextract expansions when a nondefault
327 output list separator was used.
332 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
333 checkers that were complaining about end of non-void function with no
336 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
337 This was a regression intruduced in 4.83 by another bugfix.
339 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
341 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
342 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
348 TF/01 Correctly close the server side of TLS when forking for delivery.
350 When a message was received over SMTP with TLS, Exim failed to clear up
351 the incoming connection properly after forking off the child process to
352 deliver the message. In some situations the subsequent outgoing
353 delivery connection happened to have the same fd number as the incoming
354 connection previously had. Exim would try to use TLS and fail, logging
355 a "Bad file descriptor" error.
357 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
358 utilities have not been installed.
360 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
361 temporary space as the ACL may create new global variables.
363 TL/01 LDAP support uses per connection or global context settings, depending
364 upon the detected version of the libraries at build time.
366 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
367 to extract and use the src ip:port in logging and expansions as if it
368 were a direct connection from the outside internet. PPv2 support was
369 updated based on HAProxy spec change in May 2014.
371 JH/02 Add ${listextract {number}{list}{success}{fail}}.
373 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
374 Properly escape header and check for NULL return.
376 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
379 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
381 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
382 characters in header names, implemented as a verify condition.
383 Contributed by Michael Fischer v. Mollard.
385 TL/05 Rename SPF condition results err_perm and err_temp to standardized
386 results permerror and temperror. Previous values are deprecated but
387 still accepted. In a future release, err_perm and err_temp will be
388 completely removed, which will be a backward incompatibility if the
389 ACL tests for either of these two old results. Patch contributed by
390 user bes-internal on the mailing list.
392 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
394 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
395 selectors, in both main and reject logs.
397 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
400 JH/07 Add malware type "sock" for talking to simple daemon.
402 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
404 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
405 routers/transports under cutthrough routing.
407 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
408 numbers. Touch up "bool" conditional to keep the same definition.
410 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
412 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
414 JH/12 Expand items in router/transport headers_add or headers_remove lists
415 individually rather than the list as a whole. Bug 1452.
417 Required for reasonable handling of multiple headers_ options when
418 they may be empty; requires that headers_remove items with embedded
419 colons must have them doubled (or the list-separator changed).
421 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
422 view the policy declared in the DMARC record. Currently, $dmarc_status
423 is a combined value of both the record presence and the result of the
426 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
428 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
429 dnslookup router and the smtp transport (applying to the forward
432 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
433 of ldap servers used for a specific lookup. Patch provided by Heiko
436 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
437 New variable $lookup_dnssec_authenticated for observability.
439 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
440 Patch submitted by Lars Timman.
442 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
444 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
445 Requires trusted mode and valid format message id, aborts otherwise.
446 Patch contributed by Heiko Schlichting.
448 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
449 certextract with support for various fields. Bug 1358.
451 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
452 is requested by default, modifiable by smtp transport option
455 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
456 operate on certificate variables to give certificate fingerprints
457 Also new ${sha256:cert_variable}.
459 JH/23 The PRDR feature is moved from being Experimental into the mainline.
461 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
462 Christian Aistleitner.
464 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
466 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
467 file. Patch from Wolfgang Breyha.
469 JH/25 Expand the coverage of the delivery $host and $host_address to
470 client authenticators run in verify callout. Bug 1476.
472 JH/26 Port service names are now accepted for tls_on_connect_ports, to
473 align with daemon_smtp_ports. Bug 72.
475 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
476 support and error reporting did not work properly.
478 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
479 and is readable. Patch from Andrew Colin Kissa.
481 TL/14 Enhance documentation of ${run expansion and how it parses the
482 commandline after expansion, particularly in the case when an
483 unquoted variable expansion results in an empty value.
485 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
487 PP/02 Fix internal collision of T_APL on systems which support RFC3123
488 by renaming away from it. Addresses GH issue 15, reported by
491 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
493 TL/15 SECURITY: prevent double expansion in math comparison functions
494 (can expand unsanitized data). Not remotely exploitable.
501 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
503 PP/02 Make -n do something, by making it not do something.
504 When combined with -bP, the name of an option is not output.
506 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
509 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
510 $sender_host_name and config options to manage this, and basic check
513 PP/05 DSCP support for outbound connections and control modifier for inbound.
515 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
516 (Only plugin which currently uses this is kerberos4, which nobody should
517 be using, but we should make it available and other future plugins might
518 conceivably use it, even though it would break NAT; stuff *should* be
519 using channel bindings instead).
521 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
522 name; added for Sendmail compatibility; requires admin caller.
523 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
524 just ignore it); requires trusted caller.
525 Also parse but ignore: -Ac -Am -X<logfile>
528 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
530 TL/02 Add +smtp_confirmation as a default logging option.
532 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
533 Patch by Magnus Holmgren from 2007-02-20.
535 TL/04 Bugzilla 1281 - Spec typo.
536 Bugzilla 1283 - Spec typo.
537 Bugzilla 1290 - Spec grammar fixes.
539 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
541 TL/06 Add Experimental DMARC support using libopendmarc libraries.
543 TL/07 Fix an out of order global option causing a segfault. Reported to dev
544 mailing list by by Dmitry Isaikin.
546 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
548 JH/02 Support "G" suffix to numbers in ${if comparisons.
550 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
552 NM/01 Bugzilla 1197 - Spec typo
553 Bugzilla 1196 - Spec examples corrections
555 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
557 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
558 gnutls_enable_pkcs11, but renamed to more accurately indicate its
561 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
562 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
564 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
565 "acl {{name}{arg}...}", and optional args on acl condition
568 JH/05 Permit multiple router/transport headers_add/remove lines.
570 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
572 JH/07 Avoid using a waiting database for a single-message-only transport.
573 Performance patch from Paul Fisher. Bugzilla 1262.
575 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
578 JH/09 Add $headers_added variable, with content from use of ACL modifier
579 add_header (but not yet added to the message). Bugzilla 199.
581 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
582 Pulled from Bugzilla 817 by Wolfgang Breyha.
584 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
586 (nb: this is the same fix as in Exim 4.80.1)
588 JH/11 Add A= logging on delivery lines, and a client_set_id option on
591 JH/12 Add optional authenticated_sender logging to A= and a log_selector
594 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
596 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
597 advertise SMTP AUTH mechanism to us, instead of a generic
598 protocol violation error. Also, make Exim more robust to bad
599 data from the Dovecot auth socket.
601 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
603 When a queue runner is handling a message, Exim first routes the
604 recipient addresses, during which it prunes them based on the retry
605 hints database. After that it attempts to deliver the message to
606 any remaining recipients. It then updates the hints database using
609 So if a recipient address works intermittently, it can get repeatedly
610 deferred at routing time. The retry hints record remains fresh so the
611 address never reaches the final cutoff time.
613 This is a fairly common occurrence when a user is bumping up against
614 their storage quota. Exim had some logic in its local delivery code
615 to deal with this. However it did not apply to per-recipient defers
616 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
618 This change adds a proper retry rule check during routing so that the
619 final cutoff time is checked against the message's age. We only do
620 this check if there is an address retry record and there is not a
621 domain retry record; this implies that previous attempts to handle
622 the address had the retry_use_local_parts option turned on. We use
623 this as an approximation for the destination being like a local
624 delivery, as in LMTP.
626 I suspect this new check makes the old local delivery cutoff check
627 redundant, but I have not verified this so I left the code in place.
629 TF/02 Correct gecos expansion when From: is a prefix of the username.
631 Test 0254 submits a message to Exim with the header
635 When I ran the test suite under the user fanf2, Exim expanded
636 the header to contain my full name, whereas it should have added
637 a Resent-Sender: header. It erroneously treats any prefix of the
638 username as equal to the username.
640 This change corrects that bug.
642 GF/01 DCC debug and logging tidyup
643 Error conditions log to paniclog rather than rejectlog.
644 Debug lines prefixed by "DCC: " to remove any ambiguity.
646 TF/03 Avoid unnecessary rebuilds of lookup-related code.
648 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
649 Bug spotted by Jeremy Harris; was flawed since initial commit.
650 Would have resulted in OCSP responses post-SNI triggering an Exim
651 NULL dereference and crash.
653 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
655 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
656 Bug detection, analysis and fix by Samuel Thibault.
657 Bugzilla 1331, Debian bug #698092.
659 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
661 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
662 Server implementation by Todd Lyons, client by JH.
663 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
664 config variable "prdr_enable" controls whether the server
665 advertises the facility. If the client requests PRDR a new
666 acl_data_smtp_prdr ACL is called once for each recipient, after
667 the body content is received and before the acl_smtp_data ACL.
668 The client is controlled by bolth of: a hosts_try_prdr option
669 on the smtp transport, and the server advertisement.
670 Default client logging of deliveries and rejections involving
671 PRDR are flagged with the string "PRDR".
673 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
674 fclose(). Diagnosis by Todd Lyons.
676 PP/17 Update configure.default to handle IPv6 localhost better.
677 Patch by Alain Williams (plus minor tweaks).
680 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
681 This is now consistent with GnuTLS, and is now documented: the
682 previous undocumented portable approach to treating the option as
683 unset was to force an expansion failure. That still works, and
684 an empty string is now equivalent.
686 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
687 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
688 not performing validation itself.
690 PP/20 Added force_command boolean option to pipe transport.
691 Patch from Nick Koston, of cPanel Inc.
693 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
696 TF/04 Added udpsend ACL modifer and hexquote expansion operator
698 PP/21 Fix eximon continuous updating with timestamped log-files.
699 Broken in a format-string cleanup in 4.80, missed when I repaired the
700 other false fix of the same issue.
701 Report and fix from Heiko Schlichting.
704 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
705 Report from Prashanth Katuri.
707 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
708 It's SecureTransport, so affects any MacOS clients which use the
709 system-integrated TLS libraries, including email clients.
711 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
712 using a MIME ACL for non-SMTP local injection.
713 Report and assistance in diagnosis by Warren Baker.
715 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
717 JH/16 Fix comparisons for 64b. Bugzilla 1385.
719 TL/09 Add expansion variable $authenticated_fail_id to keep track of
720 last id that failed so it may be referenced in subsequent ACL's.
722 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
725 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
726 ldap library initialization, allowing self-signed CA's to be
727 used. Also properly sets require_cert option later in code by
728 using NULL (global ldap config) instead of ldap handle (per
729 session). Bug diagnosis and testing by alxgomz.
731 TL/12 Enhanced documentation in the ratelimit.pl script provided in
732 the src/util/ subdirectory.
734 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
735 renamed to Transport Post Delivery Action by Jeremy Harris, as
738 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
739 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
740 redis_servers = needs to be configured which will be used by the redis
741 lookup. Patch from Warren Baker, of The Packet Hub.
743 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
745 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
746 hostname or reverse DNS when processing a host list. Used suggestions
747 from multiple comments on this bug.
749 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
751 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
752 Missed a few lines, added it to make the runtest require no keyboard
755 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
756 contains upper case chars. Make router use caseful_local_part.
758 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
759 support when GnuTLS has been built with p11-kit.
765 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
767 This, or similar/improved, will also be change PP/11 of 4.82.
773 PP/01 Handle short writes when writing local log-files.
774 In practice, only affects FreeBSD (8 onwards).
775 Bugzilla 1053, with thanks to Dmitry Isaikin.
777 NM/01 Bugzilla 949 - Documentation tweak
779 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
782 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
784 PP/02 Implemented gsasl authenticator.
786 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
788 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
789 `pkg-config foo` for cflags/libs.
791 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
792 with rest of GSASL and with heimdal_gssapi.
794 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
795 `pkg-config foo` for cflags/libs for the TLS implementation.
797 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
798 properties get this fed in as external SSF. A number of robustness
799 and debugging improvements to the cyrus_sasl authenticator.
801 PP/08 cyrus_sasl server now expands the server_realm option.
803 PP/09 Bugzilla 1214 - Log authentication information in reject log.
804 Patch by Jeremy Harris.
806 PP/10 Added dbmjz lookup type.
808 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
810 PP/12 MAIL args handles TAB as well as SP, for better interop with
811 non-compliant senders.
812 Analysis and variant patch by Todd Lyons.
814 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
815 Bug report from Lars Müller <lars@samba.org> (via SUSE),
816 Patch from Dirk Mueller <dmueller@suse.com>
818 PP/13 tls_peerdn now print-escaped for spool files.
819 Observed some $tls_peerdn in wild which contained \n, which resulted
820 in spool file corruption.
822 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
823 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
824 or write after TLS renegotiation, which otherwise led to messages
827 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
828 as a tracking header (ie: a signed header comes before the signature).
829 Patch from Wolfgang Breyha.
831 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
832 comma-sep list; embedded commas doubled.
834 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
836 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
838 Report and patch from Dmitry Banschikov.
840 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
841 Removed SSL_clear() after SSL_new() which led to protocol negotiation
842 failures. We appear to now support TLS1.1+ with Exim.
844 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
845 lets Exim select keys and certificates based upon TLS SNI from client.
846 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
847 before an outbound SMTP session. New log_selector, +tls_sni.
849 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
850 NULL dereference. Report and patch from Alun Jones.
852 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
853 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
854 Not seeing resolver debug output on NetBSD, but suspect this is a
855 resolver implementation change.
857 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
858 Left warnings. Added "eximon gdb" invocation mode.
860 PP/21 Defaulting "accept_8bitmime" to true, not false.
862 PP/22 Added -bw for inetd wait mode support.
864 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
865 locate the relevant includes and libraries. Made this the default.
867 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
868 Bugzilla 1246, report and most of solution from Tomasz Kusy.
870 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
871 This may cause build issues on older platforms.
873 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
874 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
875 gnutls_require_mac & gnutls_require_protocols (no longer supported).
876 Added SNI support via GnuTLS too.
877 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
879 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
881 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
882 Applied second patch from Janne, implementing suggestion to default
883 multiple-strings-in-record handling to match SPF spec.
885 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
887 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
888 read-only, out of scope).
889 Patch from Wolfgang Breyha, report from Stuart Northfield.
891 PP/29 Fix three issues highlighted by clang analyser static analysis.
892 Only crash-plausible issue would require the Cambridge-specific
893 iplookup router and a misconfiguration.
894 Report from Marcin Mirosław.
896 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
898 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
899 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
900 As part of this, removing so much warning spew let me fix some minor
901 real issues in debug logging.
903 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
904 assignment on my part. Fixed.
906 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
907 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
908 Janne Snabb (who went above and beyond: thank you).
910 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
911 string otherwise requires a connection and a bunch more work and it's
912 relatively easy to get wrong. Should also expose TLS library linkage
915 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
916 64-bit ${eval} (JH/03).
918 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
919 GNU libc to support some of the 64-bit stuff, should not lead to
920 conflicts. Defined before os.h is pulled in, so if a given platform
921 needs to override this, it can.
923 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
924 protection layer was required, which is not implemented.
925 Bugzilla 1254, patch from Wolfgang Breyha.
927 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
928 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
929 tls_dhparam take prime identifiers. Also unbreak combination of
930 OpenSSL+DH_params+TLSSNI.
932 PP/39 Disable SSLv2 by default in OpenSSL support.
938 PP/01 Solaris build fix for Oracle's LDAP libraries.
939 Bugzilla 1109, patch from Stephen Usher.
941 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
943 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
946 TF/02 Fix a couple more cases where we did not log the error message
947 when unlink() failed. See also change 4.74-TF/03.
949 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
950 lock up or crash if it happened to be inside a call to libc when it
951 got a SIGUSR1 from exiwhat.
953 The SIGUSR1 handler appends the current process status to the process
954 log which is later printed by exiwhat. It used to use the general
955 purpose logging code to do this, but several functions it calls are
956 not safe for signals.
958 The new output code in the SIGUSR1 handler is specific to the process
959 log, and simple enough that it's easy to inspect for signal safety.
960 Removing some special cases also simplifies the general logging code.
961 Removing the spurious timestamps from the process log simplifies
964 TF/04 Improved ratelimit ACL condition.
966 The /noupdate option has been deprecated in favour of /readonly which
967 has clearer semantics. The /leaky, /strict, and /readonly update modes
968 are mutually exclusive. The update mode is no longer included in the
969 database key; it just determines when the database is updated. (This
970 means that when you upgrde Exim will forget old rate measurements.)
972 Exim now checks that the per_* options are used with an update mode that
973 makes sense for the current ACL. For example, when Exim is processing a
974 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
975 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
976 must specify per_mail/readonly. If you omit the update mode it defaults to
977 /leaky where that makes sense (as before) or /readonly where required.
979 The /noupdate option is now undocumented but still supported for
980 backwards compatibility. It is equivalent to /readonly except that in
981 ACLs where /readonly is required you may specify /leaky/noupdate or
982 /strict/noupdate which are treated the same as /readonly.
984 A useful new feature is the /count= option. This is a generalization
985 of the per_byte option, so that you can measure the throughput of other
986 aggregate values. For example, the per_byte option is now equivalent
987 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
989 The per_rcpt option has been generalized using the /count= mechanism
990 (though it's more complicated than the per_byte equivalence). When it is
991 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
992 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
993 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
994 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
995 Note that using per_rcpt with a non-readonly update mode in more than
996 one ACL will cause the recipients to be double-counted. (The per_mail
997 and per_byte options don't have this problem.)
999 The handling of very low rates has changed slightly. If the computed rate
1000 is less than the event's count (usually one) then this event is the first
1001 after a long gap. In this case the rate is set to the same as this event's
1002 count, so that the first message of a spam run is counted properly.
1004 The major new feature is a mechanism for counting the rate of unique
1005 events. The new per_addr option counts the number of different
1006 recipients that someone has sent messages to in the last time period. It
1007 behaves like per_rcpt if all the recipient addresses are different, but
1008 duplicate recipient addresses do not increase the measured rate. Like
1009 the /count= option this is a general mechanism, so the per_addr option
1010 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1011 example, measure the rate that a client uses different sender addresses
1012 with the options per_mail/unique=$sender_address. There are further
1013 details in the main documentation.
1015 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1017 TF/06 Removed a few PCRE remnants.
1019 TF/07 Automatically extract Exim's version number from tags in the git
1020 repository when doing development or release builds.
1022 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1023 Bugzilla 879. Patch from Paul Fisher.
1025 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1026 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1029 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1031 PP/05 Variable $av_failed, true if the AV scanner deferred.
1032 Bugzilla 1078. Patch from John Horne.
1034 PP/06 Stop make process more reliably on build failure.
1035 Bugzilla 1087. Patch from Heiko Schlittermann.
1037 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1038 Bugzilla 1089. Patch from Heiko Schlittermann.
1040 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1041 Bugzilla 1131. Patch from Holger Weiß.
1043 PP/09 Handle IPv6 addresses with SPF.
1044 Bugzilla 860. Patch from Wolfgang Breyha.
1046 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1048 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1051 PP/11 match_* no longer expand right-hand-side by default.
1052 New compile-time build option, EXPAND_LISTMATCH_RHS.
1053 New expansion conditions, "inlist", "inlisti".
1055 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1057 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1059 PP/14 fix log_write() format string regression from TF/03.
1060 Bugzilla 1152. Patch from Dmitry Isaikin.
1066 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1068 PP/02 Harmonised TLS library version reporting; only show if debugging.
1069 Layout now matches that introduced for other libraries in 4.74 PP/03.
1071 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1073 PP/04 New "dns_use_edns0" global option.
1075 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1078 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1079 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1081 TK/01 Updated PolarSSL code to 0.14.2.
1082 Bugzilla 1097. Patch from Andreas Metzler.
1084 PP/07 Catch divide-by-zero in ${eval:...}.
1085 Fixes bugzilla 1102.
1087 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1090 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1091 format-string attack -- SECURITY: remote arbitrary code execution.
1093 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1094 time unintentionally subject to list matching rules, letting the header
1095 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1096 arbitrary string expansion). This allowed for information disclosure.
1098 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1099 INT_MIN/-1 -- value coerced to INT_MAX.
1105 NM/01 Workround for PCRE version dependency in version reporting
1108 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1109 This fixes portability to compilers other than gcc, notably
1110 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1112 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1113 makefiles for portability to HP-UX and POSIX correctness.
1115 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1116 Also via indented variable definition in the Makefile.
1117 (Debugging by Oliver Heesakkers).
1119 PP/02 Restore caching of spamd results with expanded spamd_address.
1120 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1122 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1123 Improves build reliability. Fix from: Frank Elsner
1125 NM/02 Fix wide character breakage in the rfc2047 coding
1126 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1128 NM/03 Allow underscore in dnslist lookups
1129 Fixes bug 1026. Patch from Graeme Fowler
1131 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1132 Code patches from Adam Ciarcinski of NetBSD.
1134 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1137 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1138 is logged, to avoid truncation. Patch from John Horne.
1140 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1141 Patch from Jakob Hirsch.
1143 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1144 SQL string expansion failure details.
1145 Patch from Andrey Oktyabrski.
1147 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1148 Patch from Simon Arlott.
1150 PP/09 New lookups functionality failed to compile on old gcc which rejects
1151 extern declarations in function scope.
1152 Patch from Oliver Fleischmann
1154 PP/10 Use sig_atomic_t for flags set from signal handlers.
1155 Check getgroups() return and improve debugging.
1156 Fixed developed for diagnosis in bug 927 (which turned out to be
1159 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1160 Patch from Mark Zealey.
1162 PP/12 Bugzilla 1056: Improved spamd server selection.
1163 Patch from Mark Zealey.
1165 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1166 Based on patch from Heiko Schlittermann.
1168 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1169 Patch from Uwe Doering, sign-off by Michael Haardt.
1171 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1172 variable declaration deep within a block. Bug and patch from
1175 PP/15 lookups-Makefile IRIX compatibilty coercion.
1177 PP/16 Make DISABLE_DKIM build knob functional.
1179 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1180 Patch by Simon Arlott
1182 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1183 variable argument macros. Our copy now differs from upstream.
1189 TF/01 Failure to get a lock on a hints database can have serious
1190 consequences so log it to the panic log.
1192 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1193 controlled using the smtp_confirmation log selector.
1195 TF/03 Include the error message when we fail to unlink a spool file.
1197 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1198 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1199 for maintaining out-of-tree patches for some time.
1201 PP/01 Bugzilla 139: Documentation and portability issues.
1202 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1203 Handle per-OS dynamic-module compilation flags.
1205 PP/02 Let /dev/null have normal permissions.
1206 The 4.73 fixes were a little too stringent and complained about the
1207 permissions on /dev/null. Exempt it from some checks.
1208 Reported by Andreas M. Kirchwitz.
1210 PP/03 Report version information for many libraries, including
1211 Exim version information for dynamically loaded libraries. Created
1212 version.h, now support a version extension string for distributors
1213 who patch heavily. Dynamic module ABI change.
1215 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1216 privilege escalation vulnerability whereby the Exim run-time user
1217 can cause root to append content of the attacker's choosing to
1220 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1223 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1224 If dropping privileges for untrusted macros, we disabled normal logging
1225 on the basis that it would fail; for the Exim run-time user, this is not
1226 the case, and it resulted in successful deliveries going unlogged.
1227 Fixed. Reported by Andreas Metzler.
1233 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1234 only prepend for the Resent-* case. Fixes regression introduced in
1235 Exim 4.70 by NM/22 for Bugzilla 607.
1237 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1238 increasing numbers of administrators be bitten by this.
1240 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1242 PP/03 Bugzilla 994: added openssl_options main configuration option.
1244 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1246 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1248 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1250 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1251 without a peer certificate, leading to a segfault because of an
1252 assumption that peers always have certificates. Be a little more
1253 paranoid. Problem reported by Martin Tscholak.
1255 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1256 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1257 NB: ClamAV planning to remove STREAM in "middle of 2010".
1258 CL also introduces -bmalware, various -d+acl logging additions and
1259 more caution in buffer sizes.
1261 PP/09 Implemented reverse_ip expansion operator.
1263 PP/10 Bugzilla 937: provide a "debug" ACL control.
1265 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1267 PP/12 Bugzilla 973: Implement --version.
1269 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1271 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1273 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1275 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1276 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1277 ignore trailing whitespace.
1279 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1281 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1284 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1285 Notification from Dr Andrew Aitchison.
1287 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1288 ExtendedDetectionInfo response format.
1289 Notification from John Horne.
1291 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1294 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1295 XSL and documented dependency on system catalogs, with examples of how
1298 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1301 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1302 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1303 configuration file which is writeable by the Exim user or group.
1305 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1306 of configuration files to cover files specified with the -C option if
1307 they are going to be used with root privileges, not just the default
1310 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1311 option (effectively making it always true).
1313 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1314 files to be used while preserving root privileges.
1316 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1317 that rogue child processes cannot use them.
1319 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1320 run-time user, instead of root.
1322 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1323 Exim run-time user without dropping privileges.
1325 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1326 result string, instead of calling string_vformat() twice with the same
1329 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1330 for other users. Others should always drop root privileges if they use
1331 -C on the command line, even for a whitelisted configure file.
1333 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1335 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1341 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1342 $data_path, and $header_path variables; fixed documentation bugs and
1345 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1346 exipick to access non-standard spools, including the "frozen" queue
1349 NM/01 Bugzilla 965: Support mysql stored procedures.
1350 Patch from Alain Williams
1352 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1354 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1355 Patch from Andreas Metzler
1357 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1358 Patch from Kirill Miazine
1360 NM/05 Bugzilla 671: Added umask to procmail example.
1362 JJ/03 installed exipick 20100323.0, fixing doc bug
1364 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1365 directory. Notification and patch from Dan Rosenberg.
1367 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1369 TK/02 Improve log output when DKIM signing operation fails.
1371 MH/01 Treat the transport option dkim_domain as a colon separated
1372 list, not as a single string, and sign the message with each element,
1373 omitting multiple occurences of the same signer.
1375 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1376 Bugzilla 985, 986. Patch by Simon Arlott
1378 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1379 Patch by Simon Arlott
1381 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1382 MBX locking. Notification from Dan Rosenberg.
1388 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1390 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1392 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1394 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1396 NM/04 Bugzilla 915: Flag broken perl installation during build.
1402 TK/01 Added patch by Johannes Berg that expands the main option
1403 "spamd_address" if it starts with a dollar sign.
1405 TK/02 Write list of recipients to X-Envelope-Sender header when building
1406 the mbox-format spool file for content scanning (suggested by Jakob
1409 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1410 (http://www.dcc-servers.net/) support via dccifd. Activated by
1411 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1413 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1414 by Mark Daniel Reidel <mr@df.eu>.
1416 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1417 When building exim an external PCRE library is now needed -
1418 PCRE is a system library on the majority of modern systems.
1419 See entry on PCRE_LIBS in EDITME file.
1421 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1422 conversation. Added nologin parameter to request.
1423 Patch contributed by Kirill Miazine.
1425 TF/01 Do not log submission mode rewrites if they do not change the address.
1427 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1429 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1430 log files in place. Contributed by Roberto Lima.
1432 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1434 TF/03 Bugzilla 615: When checking the local_parts router precondition
1435 after a local_part_suffix or local_part_prefix option, Exim now
1436 does not use the address's named list lookup cache, since this
1437 contains cached lookups for the whole local part.
1439 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1440 Robert Millan. Documentation is in experimental-spec.txt.
1442 TF/04 Bugzilla 668: Fix parallel build (make -j).
1444 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1446 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1447 Patch provided by Jan Srzednicki.
1449 TF/05 Leading white space used to be stripped from $spam_report which
1450 wrecked the formatting. Now it is preserved.
1452 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1453 that they are available at delivery time.
1455 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1457 TF/08 TLS error reporting now respects the incoming_interface and
1458 incoming_port log selectors.
1460 TF/09 Produce a more useful error message if an SMTP transport's hosts
1461 setting expands to an empty string.
1463 NM/06 Bugzilla 744: EXPN did not work under TLS.
1464 Patch provided by Phil Pennock.
1466 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1467 Patch provided by Richard Godbee.
1469 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1470 acl_smtp_notquit, added index entry.
1472 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1473 Patch provided by Eugene Bujak.
1475 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1476 accept(). Patch provided by Maxim Dounin.
1478 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1479 Patch provided by Phil Pennock.
1481 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1483 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1484 Patch provided by Brad "anomie" Jorsch.
1486 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1487 Patch provided by Dean Brooks.
1489 TK/05 Add native DKIM support (does not depend on external libraries).
1491 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1492 Patch provided by Graeme Fowler.
1494 NM/16 Bugzilla 851: Documentation example syntax fix.
1496 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1498 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1501 NM/19 Bugzilla 745: TLS version reporting.
1502 Patch provided by Phil Pennock.
1504 NM/20 Bugzilla 167: bool: condition support.
1505 Patch provided by Phil Pennock.
1507 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1508 clients. Patch provided by Phil Pennock.
1510 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1511 Patch provided by Brad "anomie" Jorsch.
1513 NM/23 Bugzilla 687: Fix misparses in eximstats.
1514 Patch provided by Heiko Schlittermann.
1516 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1517 Patch provided by Heiko Schlittermann.
1519 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1520 plus update to original patch.
1522 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1524 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1525 Patch provided by David Brownlee.
1527 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1529 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1531 NM/30 Bugzilla 888: TLS documentation bugfixes.
1533 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1535 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1536 Unlike the original bugzilla I have changed all shell scripts in src tree.
1538 NM/33 Bugzilla 898: Transport filter timeout fix.
1539 Patch by Todd Rinaldo.
1541 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1542 Patch by Serge Demonchaux.
1544 NM/35 Bugzilla 39: Base64 decode bug fixes.
1545 Patch by Jakob Hirsch.
1547 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1549 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1551 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1553 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1559 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1560 ALT-N's libdkim that I have put here:
1561 http://duncanthrax.net/exim-experimental/
1563 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1564 were called 'true' and it seems that C99 defines that as a reserved
1565 keyword to be used with 'bool' variable types. That means you could
1566 not include C99-style headers which use bools without triggering
1567 build errors in sieve.c.
1569 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1570 as mailq or other aliases. Changed the --help handling significantly
1571 to do whats expected. exim_usage() emits usage/help information.
1573 SC/01 Added the -bylocaldomain option to eximstats.
1575 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1577 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1579 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1585 PH/01 Another patch from the Sieve maintainer.
1587 PH/02 When an IPv6 address is converted to a string for single-key lookup
1588 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1589 dots are used instead of colons so that keys in lsearch files need not
1590 contain colons. This was done some time before quoting was made available
1591 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1592 (notated using the quote facility) so as to distinguish them from IPv4
1593 keys. This meant that lookups for IP addresses in host lists did not work
1594 for iplsearch lookups.
1596 This has been fixed by arranging for IPv6 addresses to be expressed with
1597 colons if the lookup type is iplsearch. This is not incompatible, because
1598 previously such lookups could never work.
1600 The situation is now rather anomolous, since one *can* have colons in
1601 ordinary lsearch keys. However, making the change in all cases is
1602 incompatible and would probably break a number of configurations.
1604 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1607 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1608 conversion specification without a maximum field width, thereby enabling
1609 a rogue spamd server to cause a buffer overflow. While nobody in their
1610 right mind would setup Exim to query an untrusted spamd server, an
1611 attacker that gains access to a server running spamd could potentially
1612 exploit this vulnerability to run arbitrary code as the Exim user.
1614 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1615 $primary_hostname instead of what libspf2 thinks the hosts name is.
1617 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1618 a directory entry by the name of the lookup key. Previously, if a
1619 symlink pointed to a non-existing file or a file in a directory that
1620 Exim lacked permissions to read, a lookup for a key matching that
1621 symlink would fail. Now it is enough that a matching directory entry
1622 exists, symlink or not. (Bugzilla 503.)
1624 PH/03 The body_linecount and body_zerocount variables are now exported in the
1627 PH/04 Added the $dnslist_matched variable.
1629 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1630 This means they are set thereafter only if the connection becomes
1633 PH/06 Added the client_condition to authenticators so that some can be skipped
1634 by clients under certain conditions.
1636 PH/07 The error message for a badly-placed control=no_multiline_responses left
1637 "_responses" off the end of the name.
1639 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1641 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1642 (without spaces) instead of just copying the configuration text.
1644 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1646 PH/11 Added $max_received_linelength.
1648 PH/12 Added +ignore_defer and +include_defer to host lists.
1650 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1651 way in which PCRE > 7.0 is built.
1653 PH/14 Implemented queue_only_load_latch.
1655 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1656 MAIL command. The effect was to mangle the value on 64-bit systems.
1658 PH/16 Another patch from the Sieve maintainer.
1660 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1662 PH/18 If a system quota error occurred while trying to create the file for
1663 a maildir delivery, the message "Mailbox is full" was not appended to the
1664 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1665 only to a quota excession during the actual writing of the file.
1667 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1668 characters?) which causes problems in log lines. The DN values are now
1669 passed through string_printing() before being added to log lines.
1671 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1672 and InterBase are left for another time.)
1674 PH/21 Added message_body_newlines option.
1676 PH/22 Guard against possible overflow in moan_check_errorcopy().
1678 PH/23 POSIX allows open() to be a macro; guard against that.
1680 PH/24 If the recipient of an error message contained an @ in the local part
1681 (suitably quoted, of course), incorrect values were put in $domain and
1682 $local_part during the evaluation of errors_copy.
1688 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1689 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1692 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1693 issue a MAIL command.
1695 PH/02 In an ACL statement such as
1697 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1699 if a client was not listed at all, or was listed with a value other than
1700 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1701 the condition was not true (as it should be), so access was not denied.
1702 The bug was that the ! inversion was incorrectly passed on to the second
1703 item. This has been fixed.
1705 PH/03 Added additional dnslists conditions == and =& which are different from
1706 = and & when the dns lookup returns more than one IP address.
1708 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1709 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1711 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1712 FSYNC, which compiles an option called disable_fsync that allows for
1713 bypassing fsync(). The documentation is heavily laced with warnings.
1715 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1717 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1718 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1719 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1720 including adding "make clean"; (3) Added -fPIC when compiling the test
1721 dynamically loaded module, to get rid of a warning.
1723 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1724 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1725 = 0s. The bug is otherwise harmless.
1727 PH/07 There was a bug in the dovecot authenticator such that the value of
1728 $auth1 could be overwritten, and so not correctly preserved, after a
1729 successful authentication. This usually meant that the value preserved by
1730 the server_setid option was incorrect.
1732 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1734 PH/09 Installed PCRE release 7.0.
1736 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1737 run for batched SMTP input. It is now run at the start of every message
1738 in the batch. While fixing this I discovered that the process information
1739 (output by running exiwhat) was not always getting set for -bs and -bS
1740 input. This is fixed, and it now also says "batched" for BSMTP.
1742 PH/11 Added control=no_pipelining.
1744 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1745 patch, slightly modified), and move the expansion of helo_data till after
1746 the connection is made in the smtp transport (so it can use these
1749 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1751 PH/14 Added log_selector = +pid.
1753 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1755 PH/16 Add ${if forany and ${if forall.
1757 PH/17 Added dsn_from option to vary the From: line in DSNs.
1759 PH/18 Flush SMTP output before performing a callout, unless control =
1760 no_callout_flush is set.
1762 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1763 was true (the default) a successful delivery failed to delete the retry
1764 item, thus causing premature timeout of the address. The bug is now
1767 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1769 PH/21 Long custom messages for fakedefer and fakereject are now split up
1770 into multiline reponses in the same way that messages for "deny" and
1771 other ACL rejections are.
1773 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1774 with slight modification.
1776 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1777 draft, changing the syntax and factoring some duplicate code".
1779 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1780 for deliveries of the second and subsequent messages over the same SMTP
1783 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1784 ${reduce, with only minor "tidies".
1786 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1788 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1789 expansion side effects.
1791 PH/27 When a message times out after an over-quota error from an Exim-imposed
1792 quota, the bounce message says "mailbox is full". This message was not
1793 being given when it was a system quota that was exceeded. It now should
1796 MH/03 Made $recipients available in local_scan(). local_scan() already has
1797 better access to the recipient list through recipients_list[], but
1798 $recipients can be useful in postmaster-provided expansion strings.
1800 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1801 in the case of a MAIL command with additional options following the
1802 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1803 were accidentally chopped off.
1805 PH/29 SMTP synchronization checks are implemented when a command is read -
1806 there is a check that no more input is waiting when there shouldn't be
1807 any. However, for some commands, a delay in an ACL can mean that it is
1808 some time before the response is written. In this time, more input might
1809 arrive, invalidly. So now there are extra checks after an ACL has run for
1810 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1811 pipelining has not been advertised.
1813 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1815 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1816 returned in the SMTP response when smtp_return_error_details was false.
1817 This has been fixed.
1819 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1820 instead of the C I/O that was originally supplied, because problems were
1821 reported on Solaris.
1823 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1824 Exim which did not show up earlier: it was assuming that a call to
1825 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1826 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1827 was a macro that became an assignment, so it seemed to work. This has
1828 changed to a proper function call with a void return, hence the compile
1829 error. Exim's code has been fixed.
1831 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1834 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1836 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1838 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1839 because it turns out to be too big or there is a timeout) there is no
1840 "Completed" line in the log. When some messages of this type were
1841 selected by exigrep, they were listed as "not completed". Others were
1842 picked up by some special patterns. I have improved the selection
1843 criteria to be more general.
1845 PH/38 The host_find_failed option in the manualroute router can now be set
1846 to "ignore", to completely ignore a host whose IP address cannot be
1847 found. If all hosts are ignored, the behaviour is controlled by the new
1848 host_all_ignored option.
1850 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1851 homing or because of multiple MX records with /mx) generated more than
1852 one IP address, and the following item turned out to be the local host,
1853 all the secondary addresses of the first item were incorrectly removed
1854 from the list, along with the local host and any following hosts (which
1855 is what is supposed to happen).
1857 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1858 whoever called Exim into the -H file. In the case of the daemon it was
1859 behaving confusingly. When first started, it used values for whoever
1860 started the daemon, but after a SIGHUP it used the Exim user (because it
1861 calls itself on a restart). I have changed the code so that it now always
1864 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1865 message are rejected with the same error (e.g. no authentication or bad
1866 sender address), and a DATA command is nevertheless sent (as can happen
1867 with PIPELINING or a stupid MUA), the error message that was given to the
1868 RCPT commands is included in the rejection of the DATA command. This is
1869 intended to be helpful for MUAs that show only the final error to their
1872 PH/42 Another patch from the Sieve maintainer.
1874 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1875 Eximstats - Fixed some broken HTML links and added missing column headers
1877 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1878 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1880 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1882 PH/43 Yet another patch from the Sieve maintainer.
1884 PH/44 I found a way to check for a TCP/IP connection going away before sending
1885 the response to the final '.' that terminates a message, but only in the
1886 case where the client has not sent further data following the '.'
1887 (unfortunately, this is allowed). However, in many cases there won't be
1888 any further data because there won't be any more messages to send. A call
1889 to select() can be used: if it shows that the input is "ready", there is
1890 either input waiting, or the socket has been closed. An attempt to read
1891 the next input character can distinguish the two cases. Previously, Exim
1892 would have sent an OK response which the client would never have see.
1893 This could lead to message repetition. This fix should cure that, at
1894 least in a lot of common cases.
1896 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1897 advertised in response to EHLO.
1903 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1904 fixed by 4.65/MH/01 (is this a record?) are fixed:
1906 (i) An empty string was always treated as zero by the numeric comparison
1907 operators. This behaviour has been restored.
1909 (ii) It is documented that the numeric comparison operators always treat
1910 their arguments as decimal numbers. This was broken in that numbers
1911 starting with 0 were being interpreted as octal.
1913 While fixing these problems I realized that there was another issue that
1914 hadn't been noticed. Values of message_size_limit (both the global option
1915 and the transport option) were treated as octal if they started with 0.
1916 The documentation was vague. These values are now always treated as
1917 decimal, and I will make that clear in the documentation.
1923 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1924 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1927 MH/01 Don't check that the operands of numeric comparison operators are
1928 integers when their expansion is in "skipping" mode (fixes bug
1929 introduced by 4.64-PH/07).
1931 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1932 child addresses, Exim now panics and dies. Previously, because the count
1933 is held in a short int, deliveries were likely to be lost. As such a
1934 large number of recipients for a single message is ridiculous
1935 (performance will be very, very poor), I have chosen to impose a limit
1936 rather than extend the field.
1942 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1943 leftover -K file (the existence of which was triggered by #402).
1944 While we were at it, introduced process PID as part of the -K
1945 filename. This should rule out race conditions when creating
1948 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1949 processing considerably. Previous code took too long for large mails,
1950 triggering a timeout which in turn triggers #401.
1952 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1953 in the DK code in transports.c. sendfile() is not really portable,
1954 hence the _LINUX specificness.
1956 TF/01 In the add_headers option to the mail command in an Exim filter,
1957 there was a bug that Exim would claim a syntax error in any
1958 header after the first one which had an odd number of characters
1961 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1962 callout verification, Exim cached a "reject" for the entire domain. This
1963 is correct for most verifications, but it is not correct for a recipient
1964 verification with use_sender or use_postmaster set, because in that case
1965 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1966 case of MAIL FROM:<> rejection from other early rejections (e.g.
1967 rejection of HELO). When verifying a recipient using a non-null MAIL
1968 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1969 Whatever the result of the callout, the value of the domain cache is
1970 left unchanged (for any other kind of callout, getting as far as trying
1971 RCPT means that the domain itself is ok).
1973 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1976 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1977 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1978 session, the daemon ignores ECONNECT errors and logs others; it now
1979 ignores EPIPE as well.
1981 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1982 (quoted-printable decoding).
1984 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1985 later the small subsequent patch to fix an introduced bug.
1987 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1989 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1991 PH/08 An error is now given if message_size_limit is specified negative.
1993 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1994 to be given (somewhat) arbitrary names.
1996 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1999 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2000 miscellaneous code fixes
2002 PH/10 Added the log_reject_target ACL modifier to specify where to log
2005 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2006 hostname. This is wrong, because it relates to the incoming message (and
2007 probably the interface on which it is arriving) and not to the outgoing
2008 callout (which could be using a different interface). This has been
2009 changed to use the value of the helo_data option from the smtp transport
2010 instead - this is what is used when a message is actually being sent. If
2011 there is no remote transport (possible with a router that sets up host
2012 addresses), $smtp_active_hostname is used.
2014 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2015 tweaks were necessary in order to get it to work (see also 21 below):
2016 (a) The code assumed that strncpy() returns a negative number on buffer
2017 overflow, which isn't the case. Replaced with Exim's string_format()
2019 (b) There were several signed/unsigned issues. I just did the minimum
2020 hacking in of casts. There is scope for a larger refactoring.
2021 (c) The code used strcasecmp() which is not a standard C function.
2022 Replaced with Exim's strcmpic() function.
2023 (d) The code set only $1; it now sets $auth1 as well.
2024 (e) A simple test gave the error "authentication client didn't specify
2025 service in request". It would seem that Dovecot has changed its
2026 interface. Fortunately there's a specification; I followed it and
2027 changed what the client sends and it appears to be working now.
2029 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2032 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2033 address A is aliased to B and C, where B exists and C does not. Without
2034 -v the output is "A verified" because verification stops after a
2035 successful redirection if more than one address is generated. However,
2036 with -v the child addresses are also verified. Exim was outputting "A
2037 failed to verify" and then showing the successful verification for C,
2038 with its parentage. It now outputs "B failed to verify", showing B's
2039 parentage before showing the successful verification of C.
2041 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2042 look up a TXT record in a specific list after matching in a combined
2045 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2046 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2047 they consult the DNS. I had assumed they would set it the way they
2048 wanted; and indeed my experiments on Linux seem to show that in some
2049 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2050 To be on the safe side, however, I have now made the interface to
2051 host_find_byname() similar to host_find_bydns(), with an argument
2052 containing the DNS resolver options. The host_find_byname() function now
2053 sets these options at its start, just as host_find_bydns() does. The smtp
2054 transport options dns_qualify_single and dns_search_parents are passed to
2055 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2056 of host_find_byname() use the default settings of RES_DEFNAMES
2057 (qualify_single) but not RES_DNSRCH (search_parents).
2059 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2060 spool_read_header() do less string testing, by means of a preliminary
2061 switch on the second character of optional "-foo" lines. (This is
2062 overdue, caused by the large number of possibilities that now exist.
2063 Originally there were few.) While I was there, I also converted the
2064 str(n)cmp tests so they don't re-test the leading "-" and the first
2065 character, in the hope this might squeeze out yet more improvement.
2067 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2068 flag allowing group syntax was set by the header_syntax check but not
2069 turned off, possible causing trouble later; (2) The flag was not being
2070 set at all for the header_verify test, causing "group"-style headers to
2071 be rejected. I have now set it in this case, and also caused header_
2072 verify to ignore an empty address taken from a group. While doing this, I
2073 came across some other cases where the code for allowing group syntax
2074 while scanning a header line wasn't quite right (mostly, not resetting
2075 the flag correctly in the right place). These bugs could have caused
2076 trouble for malformed header lines. I hope it is now all correct.
2078 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2079 with the "reply" argument non-NULL. The code, however (which originally
2080 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2081 but it didn't always do it. This confused somebody who was copying the
2082 code for some other use. I have removed all the tests.
2084 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2085 feature that was used to support insecure browsers during the U.S. crypto
2086 embargo. It requires special client support, and Exim is probably the
2087 only MTA that supported it -- and would never use it because real RSA is
2088 always available. This code has been removed, because it had the bad
2089 effect of slowing Exim down by computing (never used) parameters for the
2090 RSA_EXPORT functionality.
2092 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2093 authenticator to fail if there's a tab character in the incoming data
2094 (there should never be unless someone is messing about, as it's supposed
2095 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2096 if the connection is using TLS or if the remote IP is the same as the
2097 local IP, and the "valid-client-cert option" if a client certificate has
2100 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2101 authenticators. This can be used for authorization after authentication
2102 succeeds. (In the case of plaintext, it servers for both authentication
2105 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2106 if any retry times were supplied.
2108 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2109 connection, where there is no client IP address to check. In this
2110 situation, the verify now always succeeds.
2112 PH/25 Applied John Jetmore's -Mset patch.
2114 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2116 PH/27 In a string expansion for a processed (not raw) header when multiple
2117 headers of the same name were present, leading whitespace was being
2118 removed from all of them, but trailing whitespace was being removed only
2119 from the last one. Now trailing whitespace is removed from each header
2120 before concatenation. Completely empty headers in a concatenation (as
2121 before) are ignored.
2123 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2124 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2126 PH/29 [Removed. This was a change that I later backed out, and forgot to
2127 correct the ChangeLog entry (that I had efficiently created) before
2128 committing the later change.]
2130 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2131 address errors (4xx response to RCPT) over the same connection as other
2132 messages routed to the same hosts. Such deliveries are always "forced",
2133 so retry times are not inspected. This resulted in far too many retries
2134 for the affected addresses. The effect occurred only when there were more
2135 hosts than the hosts_max_try setting in the smtp transport when it had
2136 the 4xx errors. Those hosts that it had tried were not added to the list
2137 of hosts for which the message was waiting, so if all were tried, there
2138 was no problem. Two fixes have been applied:
2140 (i) If there are any address or message errors in an SMTP delivery, none
2141 of the hosts (tried or untried) are now added to the list of hosts
2142 for which the message is waiting, so the message should not be a
2143 candidate for sending over the same connection that was used for a
2144 successful delivery of some other message. This seems entirely
2145 reasonable: after all the message is NOT "waiting for some host".
2146 This is so "obvious" that I'm not sure why it wasn't done
2147 previously. Hope I haven't missed anything, but it can't do any
2148 harm, as the worst effect is to miss an optimization.
2150 (ii) If, despite (i), such a delivery is accidentally attempted, the
2151 routing retry time is respected, so at least it doesn't keep
2152 hammering the server.
2154 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2155 in ${readsocket because some servers need this prod.
2157 PH/32 Added some extra debug output when updating a wait-xxx database.
2159 PH/33 The hint "could be header name not terminated by colon", which has been
2160 given for certain expansion errors for a long time, was not being given
2161 for the ${if def:h_colon_omitted{... case.
2163 PH/34 The spec says: "With one important exception, whenever a domain list is
2164 being scanned, $domain contains the subject domain." There was at least
2165 one case where this was not true.
2167 PH/35 The error "getsockname() failed: connection reset by peer" was being
2168 written to the panic log as well as the main log, but it isn't really
2169 panic-worthy as it just means the connection died rather early on. I have
2170 removed the panic log writing for the ECONNRESET error when getsockname()
2173 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2174 runs only) independently of the message's sender address. This meant
2175 that, if the 4xx error was in fact related to the sender, a different
2176 message to the same recipient with a different sender could confuse
2177 things. In particualar, this can happen when sending to a greylisting
2178 server, but other circumstances could also provoke similar problems.
2179 I have changed the default so that the retry time for these errors is now
2180 based a combination of the sender and recipient addresses. This change
2181 can be overridden by setting address_retry_include_sender=false in the
2184 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2185 remote server are returned as part of bounce messages. This was not
2186 happening for LMTP over a pipe (the lmtp transport), but now it is the
2187 same for both kinds of LMTP.
2189 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2190 in header lines that were in fact CNAMEs. This is no longer the case.
2192 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2193 and queue runs started by the daemon processed all messages. This has
2194 been fixed so that -R and -S can now usefully be given with -q<time>.
2196 PH/40 Import PCRE release 6.7 (fixes some bugs).
2198 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2200 PH/42 Give an error if -q is specified more than once.
2202 PH/43 Renamed the variables $interface_address and $interface_port as
2203 $received_ip_address and $received_port, to make it clear that these
2204 values apply to message reception, and not to the outgoing interface when
2205 a message is delivered. (The old names remain recognized, of course.)
2207 PH/44 There was no timeout on the connect() call when using a Unix domain
2208 socket in the ${readsocket expansion. There now is.
2210 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2211 be meaningful with "accept".
2213 SC/01 Eximstats V1.43
2214 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2216 SC/02 Eximstats V1.44
2217 Use a glob alias rather than an array ref in the generated
2218 parser. This improves both readability and performance.
2220 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2221 Collect SpamAssassin and rejection statistics.
2222 Don't display local sender or destination tables unless
2223 there is data to show.
2224 Added average volumes into the top table text output.
2226 SC/04 Eximstats V1.46
2227 Collect data on the number of addresses (recipients)
2228 as well as the number of messages.
2230 SC/05 Eximstats V1.47
2231 Added 'Message too big' to the list of mail rejection
2232 reasons (thanks to Marco Gaiarin).
2234 SC/06 Eximstats V1.48
2235 Mainlog lines which have GMT offsets and are too short to
2236 have a flag are now skipped.
2238 SC/07 Eximstats V1.49 (Alain Williams)
2239 Added the -emptyok flag.
2241 SC/08 Eximstats V1.50
2242 Fixes for obtaining the IP address from reject messages.
2244 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2245 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2246 whitesspace changes from 4.64-PH/27
2248 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2251 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2252 are found, allow negative numbers in numeric criteria)
2254 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2256 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2257 to match changes made in 4.64-PH/43
2259 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2261 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2262 "domain" is actually an IP address, and force a failure. This locks out
2263 those revolvers/nameservers that support "A-for-A" lookups, in
2264 contravention of the specifications.
2266 PH/48 When a host name was looked up from an IP address, and the subsequent
2267 forward lookup of the name timed out, the host name was left in
2268 $sender_host_name, contrary to the specification.
2270 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2271 restricted to single-key lookups, Exim was not diagnosing an error if
2272 * or *@ was used with a query-style lookup.
2274 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2276 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2277 long ago, but noone interested enough thought of it. Let's just say that
2278 the "1.1" means that there are some new functions that weren't there at
2279 some point in the past.
2281 PH/51 Error processing for expansion failure of helo_data from an smtp
2282 transport during callout processing was broken.
2284 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2285 tested/used via the -bh/-bhc/-bs options.
2287 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2288 bug, fixed in subsequent PCRE releases).
2290 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2291 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2293 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2299 SC/01 Use a glob alias rather than an array ref in eximstats generated
2300 parser. This improves both readability and performance.
2302 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2303 Don't display local sender or destination tables in eximstats unless
2304 there is data to show.
2305 Added average volumes into the eximstats top table text output.
2307 SC/03 Collect data on the number of addresses (recipients) as well
2308 as the number of messages in eximstats.
2310 TF/01 Correct an error in the documentation for the redirect router. Exim
2311 does (usually) call initgroups() when daemonizing.
2313 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2314 with consistent privilege compared to when running as a daemon.
2316 TF/03 Note in the spec that $authenticated_id is not set for local
2317 submissions from trusted users.
2319 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2320 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2322 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2323 by adding some example configuration directives to the default
2324 configuration file. A little bit of work is required to uncomment the
2325 directives and define how usernames and passwords are checked, but
2326 there is now a framework to start from.
2328 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2329 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2330 without this. I don't know how relevant this is to other LDAP libraries.
2332 PH/02 Add the verb name to the "unknown ACL verb" error.
2334 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2336 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2338 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2339 directory not expanded when it should be if an expanded home directory
2340 was set for the address (which is overridden by the transport).
2342 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2345 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2346 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2347 because it is too late at that time, and has no effect.
2349 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2350 security issue with \' (bugzilla #107). I could not use the
2351 PQescapeStringConn() function, because it needs a PGconn value as one of
2354 PH/08 When testing addresses using -bt, indicate those final addresses that
2355 are duplicates that would not cause an additional delivery. At least one
2356 person was confused, thinking that -bt output corresponded to deliveries.
2357 (Suppressing duplicates isn't a good idea as you lose the information
2358 about possibly different redirections that led to the duplicates.)
2360 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2361 systems where poll() doesn't work, in particular OS X.
2363 PH/10 Added more information to debugging output for retry time not reached.
2365 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2366 operations in malware.c.
2368 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2371 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2372 syslog_duplication set false, log lines that would normally be written
2373 both the the main log and to the reject log were not written to syslog at
2376 PH/14 In the default configuration, change the use of "message" in ACL warn
2377 statements to "add_header".
2379 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2380 not followed by a command (e.g. "seen endif").
2382 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2383 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2386 PH/17 Added extra conditions to the default value of delay_warning_condition
2390 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2391 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2392 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2395 The Auto-Submitted: and various List- headers are standardised, whereas I
2396 don't think Precedence: ever was.
2398 PH/18 Refactored debugging code in route_finduser() to show more information,
2399 in particular, the error code if getpwnam() issues one.
2401 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2402 This is apparently needed in addition to the PH/07 change above to avoid
2403 any possible encoding problems.
2405 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2406 but not after initializing Perl.
2408 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2409 output them only if debugging. By default they are written stderr,
2410 apparently, which is not desirable.
2412 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2415 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2418 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2420 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2421 authenticated or an ident call has been made. Suppress the default
2422 values for $authenticated_id and $authenticated_sender (but permit -oMai
2423 and -oMas) when testing with -bh.
2425 PH/24 Re-jigged the order of the tests in the default configuration so that the
2426 tests for valid domains and recipients precede the DNS black list and CSA
2427 tests, on the grounds that those ones are more expensive.
2429 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2430 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2431 This bug exists in every version of Exim that I still have, right back to
2434 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2435 However, an attempt to turn on case-sensitivity in a regex key by
2436 including (?-i) didn't work because the subject string was already
2437 lowercased, and the effects were non-intuitive. It turns out that a
2438 one-line patch can be used to allow (?-i) to work as expected.
2444 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2445 other effects) broke the use of negated acl sub-conditions.
2447 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2450 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2451 "Deny" causes Exim to reject the incoming connection with a 554 error.
2452 Unfortunately, if there is a major crisis, such as a disk failure,
2453 tcp-wrappers gives "deny", whereas what one would like would be some
2454 kind of temporary error. A kludge has been added to help with this.
2455 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2456 554 error is used if errno is still zero or contains ENOENT (which occurs
2457 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2460 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2462 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2463 errors. Otherwise a message that provokes a temporary error (when other
2464 messages do not) can cause a whole host to time out.
2466 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2467 addresses were routed directly to files or pipes from a redirect router.
2468 File deliveries just didn't batch; pipe deliveries might have suffered
2471 PH/06 A failure to get a lock for a hints database would erroneously always say
2472 "Failed to get write lock", even when it was really a read lock.
2474 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2475 of 0600. This has been changed to use the value of the lockfile_mode
2476 option (which defaults to 0600).
2478 PH/08 Applied small patch from the Sieve maintainer.
2480 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2481 folder from quota calculations, a direct delivery into this folder messed
2482 up the contents of the maildirsize file. This was because the regex was
2483 used only to exclude .Trash (or whatever) when the size of the mailbox
2484 was calculated. There was no check that a delivery was happening into an
2485 excluded directory. This bug has been fixed by ignoring all quota
2486 processing for deliveries into excluded directories.
2488 PH/10 Added the maildirfolder_create_regex option to appendfile.
2494 PH/01 The code for finding all the local interface addresses on a FreeBSD
2495 system running IPv6 was broken. This may well have applied to all BSD
2496 systems, as well as to others that have similar system calls. The broken
2497 code found IPv4 interfaces correctly, but gave incorrect values for the
2498 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2499 that it would not match correctly against @[] and not recognize the IPv6
2502 PH/02 The ipliteral router was not recognizing addresses of the form user@
2503 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2505 PH/03 Added disable_ipv6.
2507 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2508 decoded form, because it is most often used to construct To: headers
2509 lines in autoreplies, and the decoded form may well be syntactically
2510 invalid. However, $reply_address has leading white space removed, and all
2511 newlines turned into spaces so that the autoreply transport does not
2514 PH/05 If group was specified without a user on a router, and no group or user
2515 was specified on a transport, the group from the router was ignored.
2517 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2518 for visible compile-time settings that can be used to change these
2519 numbers, for those that want even more. Backwards compatibility with old
2520 spool files has been maintained. However, going back to a previous Exim
2521 release will lost any variables that are in spool files.
2523 PH/07 Two small changes when running in the test harness: increase delay when
2524 passing a TCP/IP connection to a new process, in case the original
2525 process has to generate a bounce, and remove special handling of
2526 127.0.0.2 (sic), which is no longer necessary.
2528 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2529 be the same on different OS.
2531 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2534 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2535 whether --show-vars was specified or not
2537 JJ/02 exipick: Added support for new ACL variable spool format introduced
2540 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2541 syntactically invalid From: or Reply-to: line, and a filter used this to
2542 generate an autoreply, and therefore failed to obtain an address for the
2543 autoreply, Exim could try to deliver to a non-existent relative file
2544 name, causing unrelated and misleading errors. What now happens is that
2545 it logs this as a hard delivery error, but does not attempt to create a
2548 PH/11 The exinext utility has a -C option for testing purposes, but although
2549 the given file was scanned by exinext itself; it wasn't being passed on
2550 when Exim was called.
2552 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2553 an end-of-file indication when reading a command response.
2555 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2556 compiled. In many other places in Exim, IPv6 addresses are always
2557 recognized, so I have changed this. It also means that IPv4 domain
2558 literals of the form [IPV4:n.n.n.n] are now always recognized.
2560 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2561 used if the router is not running as root, for example, when verifying at
2562 ACL time, or when using -bh. The debugging output from this situation was
2563 non-existent - all you got was a failure to exec. I have made two
2566 (a) Failures to set uid/gid, the current directory, or a process leader
2567 in a subprocess such as that created by queryprogram now generate
2568 suitable debugging ouput when -d is set.
2570 (b) The queryprogram router detects when it is not running as root,
2571 outputs suitable debugging information if -d is set, and then runs
2572 the subprocess without attempting to change uid/gid.
2574 PH/15 Minor change to Makefile for building test_host (undocumented testing
2577 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2578 additional section of a DNS packet that returns MX or SRV records.
2579 Instead, it always explicitly searches for A/AAAA records. This avoids
2580 major problems that occur when a DNS server includes only records of one
2581 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2582 fixed another bug: if SRV records were looked up and the corresponding
2583 address records were *not* found in the additional section, the port
2584 values from the SRV records were lost.
2586 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2587 using the correct key (the original address) when searching the retry
2588 rules in order to find which one to use for generating the retry hint.
2590 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2591 adding the default one. Similarly, if it contains a Reply-To: header, the
2592 errors_reply_to option, if set, is not used.
2594 PH/19 When calculating a retry time, Exim used to measure the "time since
2595 failure" by looking at the "first failed" field in the retry record. Now
2596 it does not use this if it is later than than the arrival time of the
2597 message. Instead it uses the arrival time. This makes for better
2598 behaviour in cases where some deliveries succeed, thus re-setting the
2599 "first failed" field. An example is a quota failure for a huge message
2600 when small messages continue to be delivered. Without this change, the
2601 "time since failure" will always be short, possible causing more frequent
2602 delivery attempts for the huge message than are intended.
2603 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2605 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2606 $1, $2, $3) because the numerical variables can be reset during some
2607 expansion items (e.g. "match"), thereby losing the authentication data.
2609 PH/21 Make -bV show the size of off_t variables so that the test suite can
2610 decide whether to run tests for quotas > 2G.
2612 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2613 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2614 filecount value is greater than 2G or if a quota value is greater than 2G
2615 on a system where the size of off_t is not greater than 4, a panic error
2618 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2619 never match. The debug and -bh output now contains an explicit error
2620 message indicating a malformed IPv4 address or mask.
2622 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2623 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2624 PH/23 above applies.
2626 PH/25 Do not write to syslog when running in the test harness. The only
2627 occasion when this arises is a failure to open the main or panic logs
2628 (for which there is an explicit test).
2630 PH/26 Added the /no_tell option to "control=freeze".
2632 PH/27 If a host name lookup failed very early in a connection, for example, if
2633 the IP address matched host_lookup and the reverse lookup yielded a name
2634 that did not have a forward lookup, an error message of the form "no IP
2635 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2636 could be logged. Now it outputs the IP address instead of "NULL".
2638 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2639 allows the sender and the authenticated sender to be set when
2640 submitting a message from within Exim. Since child_open_exim() is
2641 documented for local_scan(), the new function should be too.
2643 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2644 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2645 results in an empty string is now treated as unset.
2647 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2649 PH/31 Added sender_verify_fail logging option.
2651 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2652 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2653 tidied the source and removed it altogether.
2655 PH/33 When a queue run was abandoned because the load average was too high, a
2656 log line was always written; now it is written only if the queue_run log
2657 selector is set. In addition, the log line for abandonment now contains
2658 information about the queue run such as the pid. This is always present
2659 in "start" and "stop" lines but was omitted from the "abandon" line.
2661 PH/34 Omit spaces between a header name and the colon in the error message that
2662 is given when verify = headers_syntax fails (if there are lots of them,
2663 the message gets confusing).
2665 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2666 names, as there are now some PTR records that contain slashes. This check
2667 is only to protect against broken name servers that fall over on strange
2668 characters, so the fact that it applies to all lookups doesn't matter.
2670 PH/36 Now that the new test suite is complete, we can remove some of the
2671 special code in Exim that was needed for the old test suite. For example,
2672 sorting DNS records because real resolvers return them in an arbitrary
2673 order. The new test suite's fake resolver always returns records in the
2676 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2677 bounces) except when queue_only is set, to avoid logging races between
2678 the different processes.
2680 PH/38 Panic-die if .include specifies a non-absolute path.
2682 PH/39 A tweak to the "H" retry rule from its user.
2684 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2685 a label. They prevented compilation on older perls.
2687 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2688 a warning to be raised on newish perls.
2690 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2691 on queue. Changes to match documented behaviour of showing count of
2692 messages matching specified criteria.
2694 PH/40 Changed the default ident timeout from 30s to 5s.
2696 PH/41 Added support for the use of login_cap features, on those BSD systems
2697 that have them, for controlling the resources used by pipe deliveries.
2699 PH/42 The content-scanning code uses fopen() to create files in which to put
2700 message data. Previously it was not paying any attention to the mode of
2701 the files. Exim runs with umask(0) because the rest of the code creates
2702 files with open(), and sets the required mode explicitly. Thus, these
2703 files were ending up world-writeable. This was not a big issue, because,
2704 being within the spool directory, they were not world-accessible. I have
2705 created a function called modefopen, which takes an additional mode
2706 argument. It sets umask(777), creates the file, chmods it to the required
2707 mode, then resets the umask. All the relevant calls to fopen() in the
2708 content scanning code have been changed to use this function.
2710 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2711 to 24 hours. This avoids potential overflow problems when processing G
2712 and H retry rules. I suspect nobody ever tinkers with this value.
2714 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2716 PH/45 When the plaintext authenticator is running as a client, the server's
2717 challenges are checked to ensure they are valid base64 strings. By
2718 default, the authentication attempt is cancelled if an invalid string is
2719 received. Setting client_ignore_invalid_base64 true ignores these errors.
2720 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2721 they are received. Thus, the responses can be made to depend on the
2722 challenges. If an invalid string is ignored, an empty string is placed in
2725 PH/46 Messages that are created by the autoreply transport now contains a
2726 References: header, in accordance with RFCs 2822 and 3834.
2728 PH/47 Added authenticated_sender_force to the smtp transport.
2730 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2732 PH/49 Installed latest patch from the Sieve maintainer.
2734 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2735 was also set, the appendfile transport was unnecessarily scanning a
2736 directory of message files (e.g. for maildir delivery) to find the count
2737 of files (along with the size), even though it did not need this
2738 information. It now does the scan only if it needs to find either the
2739 size of the count of files.
2741 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2743 PH/52 Two bugs concerned with error handling when the smtp transport is
2746 (i) Exim was not creating retry information for temporary errors given
2747 for individual recipients after the DATA command when the smtp transport
2748 was used in LMTP mode. This meant that they could be retried too
2749 frequently, and not timed out correctly.
2751 (ii) Exim was setting the flag that allows error details to be returned
2752 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2753 recipients that were returned after the DATA command.
2755 PH/53 This is related to PH/52, but is more general: for any failing address,
2756 when detailed error information was permitted to be returned to the
2757 sender, but the error was temporary, then after the final timeout, only
2758 "retry timeout exceeded" was returned. Now it returns the full error as
2759 well as "retry timeout exceeded".
2761 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2762 do this, and (what is worse) MTAs that accept it.
2764 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2765 will now be deprecated.
2767 PH/56 New os.c-cygwin from the Cygwin maintainer.
2769 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2770 formats (previously only available in exim formats via -bpr, -bpru,
2771 and -bpra. Now also available in native and exiqgrep formats)
2773 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2774 with very large, slow to parse queues
2776 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2778 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2780 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2781 responses to authentication challenges, though it was showing the
2782 challenges; (ii) I've removed the CR characters from the debug output for
2785 PH/58 Allow for the insertion of a newline as well as a space when a string
2786 is turned into more than one encoded-word during RFC 2047 encoding. The
2787 Sieve code now uses this.
2789 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2790 data_4xx, lost_connection, tls_required.
2792 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2793 message was being sent as an SMTP response.
2795 PH/61 Add -l and -k options to exicyclog.
2797 PH/62 When verifying, if an address was redirected to one new address, so that
2798 verification continued, and the new address failed or deferred after
2799 having set something in $address_data, the value of $address_data was not
2800 passed back to the ACL. This was different to the case when no
2801 redirection occurred. The value is now passed back in both cases.
2803 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2804 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2805 use login_cap.h, so on its own it isn't the distinguishing feature. The
2806 new name refers directly to the setclassresources() function.
2808 PH/65 Added configuration files for NetBSD3.
2810 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2812 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2813 is preferred over IPv4.
2815 PH/68 The bounce_return_message and bounce_return_body options were not being
2816 honoured for bounces generated during the reception of non-SMTP messages.
2817 In particular, this applied to messages rejected by the ACL. This bug has
2818 been fixed. However, if bounce_return_message is true and bounce_return_
2819 body is false, the headers that are returned for a non-SMTP message
2820 include only those that have been read before the error was detected.
2821 (In the case of an ACL rejection, they have all been read.)
2823 PH/69 The HTML version of the specification is now built in a directory called
2824 spec_html instead of spec.html, because the latter looks like a path with
2825 a MIME-type, and this confuses some software.
2827 PH/70 Catch two compiler warnings in sieve.c.
2829 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2830 function verify_get_ident() calls ip_connect() to connect a socket, but
2831 if the "connect()" function timed out, ip_connect() used to close the
2832 socket. However, verify_get_ident() also closes the socket later, and in
2833 between Exim writes to the log, which may get opened at this point. When
2834 the socket was closed in ip_connect(), the log could get the same file
2835 descriptor number as the socket. This naturally causes chaos. The fix is
2836 not to close the socket in ip_connect(); the socket should be closed by
2837 the function that creates it. There was only one place in the code where
2838 this was missing, in the iplookup router, which I don't think anybody now
2839 uses, but I've fixed it anyway.
2841 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2842 well as to direct DNS lookups. Otherwise the handling of names in host
2843 lists is inconsistent and therefore confusing.
2849 PH/01 Two changes to the default runtime configuration:
2851 (1) Move the checks for relay_from_hosts and authenticated clients from
2852 after to before the (commented out) DNS black list checks.
2854 (2) Add control=submission to the relay_from_hosts and authenticated
2855 clients checks, on the grounds that messages accepted by these
2856 statements are most likely to be submissions.
2858 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2860 (1) Generate an error if the third argument for the ${prvs expansion is
2863 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2866 (3) Reset the variables that are obtained from the first argument of
2867 ${prvscheck and used in the second argument before leaving the code,
2868 because their memory is reclaimed, so using them afterwards may do
2871 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2872 one (it's much easier than Tom thought :-).
2874 (5) Because of (4), we can now allow for the use of $prvscheck_result
2875 inside the third argument.
2877 PH/03 For some reason, the default setting of PATH when running a command from
2878 a pipe transport was just "/usr/bin". I have changed it to
2881 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2882 anything to be listed in the output from -bV.
2884 PH/05 When a filter generated an autoreply, the entire To: header line was
2885 quoted in the delivery log line, like this:
2887 => >A.N.Other <ano@some.domain> <original@ddress> ...
2889 This has been changed so that it extracts the operative address. There
2890 may be more than one such address. If so, they are comma-separated, like
2893 => >ano@some.domain,ona@other.domain <original@ddress> ...
2895 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2896 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2897 not being looked up in the rDNS to get a host name, Exim was showing the
2898 IP address twice in Received: lines, even though the IP addresses were
2899 identical. For example:
2901 Received: from [1.2.3.4] (helo=[1.2.3.4])
2903 However, if the real host name was known, it was omitting the HELO data
2904 if it matched the actual IP address. This has been tidied up so that it
2905 doesn't show the same IP address twice.
2907 PH/07 When both +timestamp and +memory debugging was on, the value given by
2908 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2909 called by the debug printing, thereby overwriting the timestamp buffer.
2910 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2912 PH/08 When the original message was included in an autoreply transport, it
2913 always said "this is a copy of the message, including all the headers",
2914 even if body_only or headers_only was set. It now gives an appropriate
2917 PH/09 Applied a patch from the Sieve maintainer which:
2919 o fixes some comments
2920 o adds the (disabled) notify extension core
2921 o adds some debug output for the result of if/elsif tests
2922 o points to the current vacation draft in the documentation
2923 and documents the missing references header update
2927 o fixes a bug in processing the envelope test (when testing
2928 multiple envelope elements, the last element determinted the
2931 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2932 Electronic Mail") by including:
2934 Auto-submitted: auto-generated
2936 in the messages that it generates (bounce messages and others, such as
2937 warnings). In the case of bounce messages for non-SMTP mesages, there was
2938 also a typo: it was using "Auto_submitted" (underscore instead of
2939 hyphen). Since every message generated by Exim is necessarily in response
2940 to another message, thes have all been changed to:
2942 Auto-Submitted: auto-replied
2944 in accordance with these statements in the RFC:
2946 The auto-replied keyword:
2948 - SHOULD be used on messages sent in direct response to another
2949 message by an automatic process,
2951 - MUST NOT be used on manually-generated messages,
2953 - MAY be used on Delivery Status Notifications (DSNs) and Message
2954 Disposition Notifications (MDNs),
2956 - MUST NOT be used on messages generated by automatic or periodic
2957 processes, except for messages which are automatic responses to
2960 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2961 to the default Received: header definition.
2963 PH/12 Added log selector acl_warn_skipped (default on).
2965 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2966 variables because (a) they are in the wrong storage pool and (b) even if
2967 they were copied, it wouldn't work properly because of the caching.
2969 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2970 checking when decoding. Apparently there are clients that generate
2971 overlong encoded strings. Why am I not surprised?
2973 PH/15 If the first argument of "${if match_address" was not empty, but did not
2974 contain an "@" character, Exim crashed. Now it writes a panic log message
2975 and treats the condition as false.
2977 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2979 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2980 extension "envelope-auth". The code is finished and in agreement with
2981 other implementations, but there is no documentation so far and in fact,
2982 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2983 not changing the active code.
2985 Print executed "if" and "elsif" statements when debugging is used. This
2986 helps a great deal to understand what a filter does.
2988 Document more things not specified clearly in RFC3028. I had all this
2989 sorted out, when out of a sudden new issues came to my mind. Oops."
2991 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2994 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2995 canonical form (as documented). However, after a host name lookup from
2996 the IP address, check_host() was doing a simple string comparison with
2997 addresses acquired from the DNS when checking that the found name did
2998 have the original IP as one of its addresses. Since any found IPv6
2999 addresses are likely to be in abbreviated form, the comparison could
3000 fail. Luckily, there already exists a function for doing the comparison
3001 by converting both addresses to binary, so now that is used instead of
3002 the text comparison.
3004 PH/20 There was another similar case to PH/19, when a complete host name was
3005 given in a host list; looking up its IP address could give an abbreviated
3006 form, whereas the current host's name might or might not be abbreviated.
3007 The same fix has been applied.
3013 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3014 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3017 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3018 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3020 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3022 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3023 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3024 Also, for "Auto-Submitted", treat anything other than "no" as
3025 non-personal, in accordance with RFC 3834. (Previously it treated
3026 anything starting "auto-" as non-personal.)
3028 TF/01 The control=submission/name=... option had a problem with syntax
3029 errors if the name included a slash character. The /name= option
3030 now slurps the rest of the string, so it can include any characters
3031 but it must come last in the list of options (after /sender_retain
3034 PH/05 Some modifications to the interface to the fake nameserver for the new
3042 TK/01 Added the "success_on_redirect" address verification option. See
3043 NewStuff for rationale and an example.
3045 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3047 PH/02 Patch to exigrep to allow it to work on syslog lines.
3049 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3050 fread() to skip over the body file's header line, because in Cygwin the
3051 header line is locked and is inaccessible.
3053 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3054 co-exist for some time) to make it clear that it is the Exim ID that is
3055 referenced, not the Message-ID: header line.
3057 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3058 string_format() function, because snprintf() does not exist on all
3061 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3062 ${stat: expansion item.
3064 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3065 protocol synchronization error", to keep the pedants happy.
3067 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3068 well as for IRIX systems, when gcc is being used. See the host.c source
3071 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3073 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3076 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3077 transport and to the smtp transport in LMTP mode.
3079 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3081 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3082 run a filter in a subprocess. This could lead to confusion in subsequent
3083 lookups in the parent process. There should also be a search_tidyup() at
3084 the end of the subprocess.
3086 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3087 only if the host matched helo_try_verify_hosts, which caused the
3088 verification to occur when the EHLO/HELO command was issued. The ACL just
3089 tested the remembered result. Now, if a previous verification attempt has
3090 not happened, "verify = helo" does it there and then.
3092 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3094 TK/03 Fix log output including CR from clamd.
3096 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3097 exist provoked a memory error which could cause a segfault.
3099 PH/15 Installed PCRE 6.2
3101 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3103 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3104 of the problem. Specifically, suggested +O2 rather than +O1 for the
3107 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3109 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3110 record, the port was not correctly logged when the outgoing_port log
3111 selector was set (it logged the transort's default port).
3113 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3114 fallback_hosts, and "hosts" in the smtp transport.
3116 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3117 host errors such as "Connection refused".
3119 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3120 authentication with radiusclient 0.4.9:
3122 - Error returned from rc_read_config was caught wrongly
3123 - Username/password not passed on to radius server due to wrong length.
3125 The presumption is that some radiusclient API changes for 4.51/PH/17
3126 were not taken care of correctly. The code is still untested by me (my
3127 Linux distribution still has 0.3.2 of radiusclient), but it was
3128 contributed by a Radius user.
3130 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3131 expanding the "port" option of the smtp transport.
3133 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3134 while reading a MIME header. Thanks to Tom Hughes for a patch.
3136 PH/24 Include config.h inside local_scan.h so that configuration settings are
3139 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3140 that in an ACL for RCPT (for example), you can examine exactly what was
3143 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3144 commands, but it was not correctly comparing the address with the actual
3145 client host address. Thus, it would show the EHLO address in Received:
3146 header lines when this was not necessary.
3148 PH/27 Added the % operator to ${eval:}.
3150 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3151 it should be ignoring failures (because with -C, for example, it has lost
3152 privilege). It wasn't ignoring creation failures other than "already
3155 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3158 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3159 because an input line was too long, either on its own, or by virtue of
3160 too many continuations, the temporary file was not being removed, and the
3161 return code was incorrect.
3163 PH/31 Missing "BOOL" in function definition in filtertest.c.
3165 PH/32 Applied Sieve patches from the maintainer.
3167 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3169 PH/33 Added "verify = not_blind".
3171 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3172 Local/Makefile (with some defaults set). These are used in built scripts
3173 such as exicyclog, but they have never been used in the exim_install
3174 script (though there are many overriding facilities there). I have
3175 arranged that the exim_install script now takes note of these two
3178 PH/35 Installed configuration files for Dragonfly.
3180 PH/36 When a locally submitted message by a trusted user did not contain a
3181 From: header, and the sender address was obtained from -f or from an SMTP
3182 MAIL command, and the trusted user did not use -F to supply a sender
3183 name, $originator_name was incorrectly used when constructing a From:
3184 header. Furthermore, $originator_name was used for submission mode
3185 messages from external hosts without From: headers in a similar way,
3186 which is clearly wrong.
3188 PH/37 Added control=suppress_local_fixups.
3190 PH/38 When log_selector = +received_sender was set, and the addition of the
3191 sender made the log line's construction buffer exactly full, or one byte
3192 less than full, an overflow happened when the terminating "\n" was
3195 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3196 when the result of a list match is failure because a DNS lookup failed.
3198 PH/40 RM_COMMAND is now used in the building process.
3200 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3201 the "build-* directories that it finds.
3203 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3204 address in a domain literal was a prefix of an interface address.
3206 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3207 when verifying a sender address, unless rewrite_headers is false.
3209 PH/44 Wrote a long comment about why errors_to addresses are verified as
3210 recipients, not senders.
3212 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3213 the ratelimit ACL was added.
3215 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3217 PH/46 Added extra information about PostgreSQL errors to the error string.
3219 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3220 avoiding the need to install special zones in a real server. This is
3221 backwards compatible; if it can't find the fake resolver, it drops back.
3222 Thus, both old and new test suites can be run.
3224 TF/02 Added util/ratelimit.pl
3226 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3227 clock is set back in time.
3229 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3230 Candler <B.Candler@pobox.com>.
3232 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3233 OK for addresses that are the result of redirections.
3235 PH/48 A number of further additions for the benefit of the new test suite,
3236 including a fake gethostbyname() that interfaces to the fake DNS resolver
3239 TF/06 The fix for widen_domains has also been applied to qualify_single and
3240 search_parents which are the other dnslookup options that can cause
3243 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3246 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3248 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3249 service" if it encountered an MX record with an empty target hostname.
3250 The message is now "an MX or SRV record indicated no SMTP service".
3252 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3253 if the DNS of the sending site is misconfigured. This is quite a
3254 common situation. This change restores the behaviour of treating a
3255 helo verification defer as a failure.
3257 PH/51 If self=fail was set on a router, the bounce message did not include the
3258 actual error message.
3264 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3266 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3267 command itself ended in error, the underlying message about the transport
3268 filter timeout was being overwritten with the pipe command error. Now the
3269 underlying error message should be appended to the second error message.
3271 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3273 PH/02 Reduce the amount of output that "make" produces by default. Full output
3274 can still be requested.
3276 PH/03 The warning log line about a condition test deferring for a "warn" verb
3277 was being output only once per connection, rather than after each
3278 occurrence (because it was using the same function as for successful
3279 "warn" verbs). This seems wrong, so I have changed it.
3281 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3282 it should not have, which might have caused a crash in the right
3283 circumstances, but probably never did.
3285 PH/04 Installed a modified version of Tony Finch's patch to make submission
3286 mode fix the return path as well as the Sender: header line, and to
3287 add a /name= option so that you can make the user's friendly name appear
3290 TF/03 Added the control = fakedefer ACL modifier.
3292 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3293 Mark Lowes for thorough testing.
3295 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3297 TK/03 Merged latest SRS patch from Miles Wilton.
3299 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3300 with the definition in sysexits.h (which is #included earlier).
3301 Fortunately, Exim does not actually use EX_OK. The code used to try to
3302 preserve the sysexits.h value, by assumimg that macro definitions were
3303 scanned for macro replacements. I have been disabused of this notion,
3304 so now the code just undefines EX_OK before #including unistd.h.
3306 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3307 in the smtp transport. When a block could not be written in a single
3308 write() function, the timeout was being re-applied to each part-write.
3309 This seems wrong - if the receiver was accepting one byte at a time it
3310 would take for ever. The timeout is now adjusted when this happens. It
3311 doesn't have to be particularly precise.
3313 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3314 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3316 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3317 without a domain if the check to <postmaster@domain> fails.
3319 SC/01 Eximstats: added -xls and the ability to specify output files
3320 (patch written by Frank Heydlauf).
3322 SC/02 Eximstats: use FileHandles for outputing results.
3324 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3326 SC/04 Eximstats: fixed display of large numbers with -nvr option
3328 SC/05 Eximstats: fixed merging of reports with empty tables.
3330 SC/06 Eximstats: added the -include_original_destination flag
3332 SC/07 Eximstats: removed tabs and trailing whitespace.
3334 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3336 TK/06 MBOX spool code: Add real "From " MBOX separator line
3337 so the .eml file is really in mbox format (even though
3338 most programs do not really care). Patch from Alex Miller.
3340 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3341 The latter is generated from $received_to and is only set if the
3342 message has one envelope recipient. SA can use these headers,
3343 obviously out-of-the-box. Patch from Alex Miller.
3345 PH/08 The ${def test on a variable was returning false if the variable's
3346 value was "0", contrary to what the specification has always said!
3347 The result should be true unless the variable is empty.
3349 PH/09 The syntax error of a character other than { following "${if
3350 def:variable_name" (after optional whitespace) was not being diagnosed.
3351 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3352 accidental colon was present, for example, could give incorrect results.
3354 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3355 result is used (not including appendfile, where other changes are about
3358 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3359 This involved changing a lot of size variables from int to off_t. It
3360 should work with maildirs and everything.
3362 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3363 spamd dying while we are connected to it.
3365 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3368 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3369 function for BDB 4.3.
3371 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3373 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3374 the sources for PCRE have been re-arranged and more files are now
3377 PH/15 The code I had for printing potentially long long variables in PH/11
3378 above was not the best (it lost precision). The length of off_t variables
3379 is now inspected at build time, and an appropriate printing format (%ld
3380 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3381 to be "long long int" or "long int". This is needed for the internal
3382 formatting function string_vformat().
3384 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3385 the configuration file to be ":syslog", then the script "guesses" where
3386 the logs files are, rather than using the compiled in default. In our
3387 case the guess is not the same as the compiled default, so the script
3388 suddenly stopped working when I started to use syslog. The patch checks
3389 to see if log_file_path is "". If so, it attempts to read it from exim
3390 with no configuration file to get the compiled in version, before it
3391 falls back to the previous guessing code."
3393 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3394 implementing BATV in an Exim configuration. See NewStuff for the gory
3397 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3398 Makefile that are specific to HP-UX.
3400 PH/18 If the "use_postmaster" option was set for a recipient callout together
3401 with the "random" option, the postmaster address was used as the MAIL
3402 FROM address for the random test, but not for the subsequent recipient
3403 test. It is now used for both.
3405 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3406 patch removes a few documentation additions to RFC 3028, because the
3407 latest draft now contains them. It adds the new en;ascii-case comparator
3408 and a new error check for 8bit text in MIME parts. Comparator and
3409 require names are now matched exactly. I enabled the subaddress
3410 extension, but it is not well tested yet (read: it works for me)."
3412 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3413 rework some of the code of TK/09 above to avoid the hardwired use of
3414 "%lld" and "long long". Replaced the call to snprintf() with a call to
3417 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3418 records point to non-existent hosts", "retry timeout exceeded", and
3419 "retry time not reached for any host after a long failure period".
3421 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3422 experimental DomainKeys support:
3424 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3425 (2) On an error such as an illegally used "control", the wrong name for
3426 the control was given.
3428 These problems did NOT occur unless DomainKeys support was compiled.
3430 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3432 PH/24 Added ${if match_ip condition.
3434 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3435 databases so that it will be absolutely obvious if a crash occurs in the
3436 DB library. This is a regular occurrence (often caused by mis-matched
3439 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3440 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3441 picked up on a user's system that detects such things. There doesn't seem
3442 to be a gcc warning option for this - only an attribute that has to be
3443 put on the function's prototype. It seems that in Fedora Core 4 they have
3444 set this on a number of new functions. No doubt there will be more in due
3447 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3448 specify a transport. However, if an address that was verified by such a
3449 router was the subject of a callout, Exim crashed because it tried to
3450 read the rcpt_include_affixes from the non-existent transport. Now it
3451 just assumes that the setting of that option is false. This bug was
3452 introduced by 4.51/PH/31.
3454 PH/28 Changed -d+all to exclude +memory, because that information is very
3455 rarely of interest, but it makes the output a lot bigger. People tend to
3456 do -d+all out of habit.
3458 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3459 code in os-type was giving problems when libc.so lives in lib64, like on
3462 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3463 aren't the modern standard, and it seems that some systems' include files
3464 don't always have them. Exim was already checking for some of the newer
3465 ones like T_AAAA, and defining it itself. I've added checks for all the
3466 record types that Exim uses.
3468 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3469 not automatically generating a new one, as it is supposed to. This
3470 prevented TLS from working. If the file did exist, but contained invalid
3471 data, a new version was generated, as expected. It was only the case of a
3472 non-existent file that was broken.
3474 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3475 with a change in libdomainkeys > 0.64.
3477 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3478 from DNS. If the selector record carries the flag, it now has
3479 precedence over the domain-wide flag.
3481 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3483 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3484 the use of a port name that isn't defined in /etc/services) occurred, the
3485 message was deferred as in a normal delivery, and thus remained on the
3486 spool, instead of being failed because of the mua_wrapper setting. This
3487 is now fixed, and I tidied up some of the mua_wrapper messages at the
3490 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3491 the messages in a hash of arrays rather than using individual hashes.
3492 This is a bit cleaner and results in dramatic memory savings, albeit
3493 at a slight CPU cost.
3495 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3496 as requested by Marc Sherman.
3498 SC/10 Eximstats: added histograms for user specified patterns as requested
3501 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3503 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3504 fopen() in the content-scanning modules that did not already have it.
3510 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3511 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3513 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3515 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3517 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3518 to test Sieve filters that use "vacation".
3520 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3521 that changes the way the GnuTLS parameters are stored in the cache file.
3522 The new format can be generated externally. For backward compatibility,
3523 if the data in the cache doesn't make sense, Exim assumes it has read an
3524 old-format file, and it generates new data and writes a new file. This
3525 means that you can't go back to an older release without removing the
3528 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3529 work if there are any delivery delays because "one_time" forces the
3530 parent to be marked "delivered", so its unseen clone is never tried
3531 again. For this reason, Exim now forbids the simultaneous setting of
3534 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3535 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3536 entry if you want to know the details. The fix, however, neglected to
3537 consider the case when local delivery batching is involved. The test for
3538 "previously delivered" was not happening when checking to see if an
3539 address could be batched with a previous (undelivered) one; under
3540 certain circumstances this could lead to multiple deliveries to the same
3543 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3544 in its include files, and this causes problems building Exim.
3546 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3547 header_syntax) but Exim was just ignoring anything given after a slash.
3548 In particular, this caused confusion with an attempt to use "verify =
3549 reverse_host_lookup/defer_ok". An error is now given when options are
3550 supplied for verify items that do not have them. (Maybe reverse_host_
3551 lookup should have a defer_ok option, but that's a different point.)
3553 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3554 defined by RFC 821) to 2048, because there were problems with some AUTH
3555 commands, and RFC 1869 says the size should be increased for extended
3556 SMTP commands that take arguments.
3558 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3561 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3562 "unknown" error; now it says that the functionality isn't in the binary.
3564 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3565 an address' error message when a string expansion fails (syntax or
3566 whatever). Otherwise the password may appear in the log. Following change
3567 PH/42 below, there is no longer a chance of it appearing in a bounce
3570 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3572 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3573 of its addresses was ever tried. (Bugzilla bug #2.)
3575 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3576 the result incorrectly in the debug output. (It correctly added a newline
3577 to what was transported.)
3579 TF/01 Added $received_time.
3581 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3582 commented out examples of how to interface to a virus scanner and to
3583 SpamAssassin. Also added commented examples of av_scanner and
3584 spamd_address settings.
3586 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3587 and controls are allowed in which ACLs. There were a couple of minor
3588 errors. Some of the entries in the conditions table (which is a table of
3589 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3590 negation of where the condition IS allowed.
3592 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3594 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3595 header file does not have a version number, so I've had to invent a new
3596 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3597 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3598 radiusclient), but it was contributed by a Radius user.
3600 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3601 files or directories, for OpenSSL.
3603 PH/19 When an Exim process that is running as root has to create an Exim log
3604 file, it does so in a subprocess that runs as exim:exim so as to get the
3605 ownership right at creation (otherwise, other Exim processes might see
3606 the file with the wrong ownership). There was no test for failure of this
3607 fork() call, which would lead to the process getting stuck as it waited
3608 for a non-existent subprocess. Forks do occasionally fail when resources
3609 run out. I reviewed all the other calls to fork(); they all seem to check
3612 PH/20 When checking for unexpected SMTP input at connect time (before writing
3613 the banner), Exim was not dealing correctly with a non-positive return
3614 from the read() function. If the client had disconnected by this time,
3615 the result was a log entry for a synchronization error with an empty
3616 string after "input=" when read() returned zero. If read() returned -1
3617 (an event I could not check), uninitialized data bytes were printed.
3618 There were reports of junk text (parts of files, etc) appearing after
3621 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3623 PH/22 Added support for macro redefinition, and (re)definition in between
3624 driver and ACL definitions.
3626 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3627 forgetting to use the resulting value; it was using the unexpanded value.
3629 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3630 hadn't been configured. The fix is from Juergen Kreileder, who
3631 understands it better than I do:
3633 "Here's what I see happening with three configured cyrus_sasl
3634 authenticators configured (plain, login, cram-md5):
3636 On startup auth_cyrus_sasl_init() gets called for each of these.
3637 This means three calls to sasl_listmech() without a specified mech_list.
3638 => SASL tests which mechs of all available mechs actually work
3639 => three warnings about OTP not working
3640 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3642 With the patch, sasl_listmech() also gets called three times. But now
3643 SASL's mech_list option is set to the server_mech specified in the the
3644 authenticator. Or in other words, the answer from sasl_listmech()
3645 gets limited to just the mech you're testing for (which is different
3647 => the return list contains just 'plain' or 'login', 'cram-md5' or
3648 nothing depending on the value of ob->server_mech.
3650 I've just tested the patch: Authentication still works fine,
3651 unavailable mechs specified in the exim configuration are still
3652 caught, and the auth.log warnings about OTP are gone."
3654 PH/25 When debugging is enabled, the contents of the command line are added
3655 to the debugging output, even when log_selector=+arguments is not
3658 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3659 answer is "GNU", and only if the return is "GNU/something" is the answer
3662 PH/27 $acl_verify_message is now set immediately after the failure of a
3663 verification in an ACL, and so is available in subsequent modifiers. In
3664 particular, the message can be preserved by coding like this:
3666 warn !verify = sender
3667 set acl_m0 = $acl_verify_message
3669 Previously, $acl_verify_message was set only while expanding "message"
3670 and "log_message" when a very denied access.
3672 PH/28 Modified OS/os.c-Linux with
3674 -#ifndef OS_LOAD_AVERAGE
3675 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3677 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3678 nomenclature these days.)
3680 PH/29 Installed patch from the Sieve maintainer that adds the options
3681 sieve_useraddress and sieve_subaddress to the redirect router.
3683 PH/30 In these circumstances:
3684 . Two addresses routed to the same list of hosts;
3685 . First host does not offer TLS;
3686 . First host accepts first address;
3687 . First host gives temporary error to second address;
3688 . Second host offers TLS and a TLS session is established;
3689 . Second host accepts second address.
3690 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3691 and peerdn, if requested) that were in fact used only for the second
3694 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3695 attention to any local part prefix or suffix that was matched by the
3696 router that accepted the address. It now behaves in the same way as it
3697 does for delivery: the affixes are removed from the local part unless
3698 rcpt_include_affixes is set on the transport.
3700 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3701 timeout during the DATA phase of an incoming message.
3703 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3704 applied a patch sanctioned by the Sieve maintainer.
3706 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3707 the uid or gid is negative. A case of a negative gid caused this to be
3708 noticed. The fix allows for either to be negative.
3710 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3711 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3714 PH/36 The change PH/12 above was broken. Fixed it.
3716 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3717 the grounds that routing the same address twice would always produce the
3718 same answer. This might have been true once, but it is certainly no
3719 longer true now. Routing a child address may depend on the previous
3720 routing that produced that child. Some complicated redirection strategies
3721 went wrong when messages had multiple recipients, and made Exim's
3722 behaviour dependent on the order in which the addresses were given.
3724 I have moved the duplicate checking until after the routing is complete.
3725 Exim scans the addresses that are assigned to local and remote
3726 transports, and removes any duplicates. This means that more work will be
3727 done, as duplicates will always all be routed, but duplicates are
3728 presumably rare, so I don't expect this is of any significance.
3730 For deliveries to pipes, files, and autoreplies, the duplicate checking
3731 still happens during the routing process, since they are not going to be
3734 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3735 It corrects a timeout issue with spamd. This is Ian's comment: "The
3736 background is that sometimes spamd either never reads data from a
3737 connection it has accepted, or it never writes response data. The exiscan
3738 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3739 blindly assumes that writes won't block so it may never time out."
3741 PH/39 Allow G after quota size as well as K and M.
3743 PH/40 The value set for $authenticated_id in an authenticator may not contain
3744 binary zeroes or newlines because the value is written to log lines and
3745 to spool files. There was no check on this. Now the value is run through
3746 the string_printing() function so that such characters are converted to
3747 printable escape sequences.
3749 PH/41 $message_linecount is a new variable that contains the total number of
3750 lines in the message. Compare $body_linecount, which is the count for the
3753 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3754 bounce and delay warning messages, except in certain special cases, which
3757 (a) An SMTP error message from a remote host;
3758 (b) A message specified in a :fail: redirection;
3759 (c) A message specified in a "fail" command in a system filter;
3760 (d) A message specified in a FAIL return from the queryprogram router;
3761 (e) A message specified by the cannot_route_message router option.
3763 In these cases only, Exim does include the error details in bounce and
3764 warning messages. There are also a few cases where bland messages such
3765 as "unrouteable address" or "local delivery error" are given.
3767 PH/43 $value is now also set for the "else" part of a ${run expansion.
3769 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3770 being worked on, but at least Exim now implements the latest version to
3773 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3774 process to complete was treated as a delivery failure, a timeout while
3775 writing the message to the pipe was logged, but erroneously treated as a
3776 successful delivery. Such timeouts include transport filter timeouts. For
3777 consistency with the overall process timeout, these timeouts are now
3778 treated as errors, giving rise to delivery failures by default. However,
3779 there is now a new Boolean option for the pipe transport called
3780 timeout_defer, which, if set TRUE, converts the failures into defers for
3781 both kinds of timeout. A transport filter timeout is now identified in
3784 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3785 systems where "make" and "gmake" are different, calling "gmake" at top
3786 level broke things. I've arranged for the value of $(MAKE) to be passed
3787 from the Makefile to this script so that it can call the same version of
3791 A note about Exim versions 4.44 and 4.50
3792 ----------------------------------------
3794 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3795 changes of various kinds. As a consequence, a big documentation update was
3796 needed. This delayed the release for rather longer than seemed good, especially
3797 in the light of a couple of (minor) security issues. Therefore, the changes
3798 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3799 release. So 4.44 and 4.50 are in effect two different branches that both start
3802 I have left the 4.50 change log unchanged; it contains all the changes since
3803 4.43. The change log for 4.44 is below; many of its items are identical to
3804 those for 4.50. This seems to be the most sensible way to preserve the
3805 historical information.
3811 1. Minor wording change to the doc/README.SIEVE file.
3813 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3814 computation of the current number of files was incorrect.
3816 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3817 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3820 4. Give more explanation in the error message when the command for a transport
3821 filter fails to execute.
3823 5. There are several places where Exim runs a non-Exim command in a
3824 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3825 was being done only for the command run by the queryprogram router. It is
3826 now done for all such subprocesses. The other cases are: ${run, transport
3827 filters, and the commands run by the lmtp and pipe transports.
3829 6. Added CONFIGURE_GROUP build-time option.
3831 7. Some older OS have a limit of 256 on the maximum number of file
3832 descriptors. Exim was using setrlimit() to set 1000 as a large value
3833 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3834 systems. I've change it so that if it can't get 1000, it tries for 256.
3836 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3837 was an oversight, and furthermore, ever since the addition of extra
3838 controls (e.g. 4.43/32), the checks on when to allow different forms of
3839 "control" were broken. There should now be diagnostics for all cases when a
3840 control that does not make sense is encountered.
3842 9. Added the /retain_sender option to "control=submission".
3844 10. $recipients is now available in the predata ACL (oversight).
3846 11. Tidy the search cache before the fork to do a delivery from a message
3847 received from the command line. Otherwise the child will trigger a lookup
3848 failure and thereby defer the delivery if it tries to use (for example) a
3849 cached ldap connection that the parent has called unbind on.
3851 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3852 of $address_data from the recipient verification was clobbered by the
3853 sender verification.
3855 13. The value of address_data from a sender verification is now available in
3856 $sender_address_data in subsequent conditions in the ACL statement.
3858 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3860 15. Added a new option "connect=<time>" to callout options, to set a different
3863 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3864 was its contents. (It was OK if the option was not defined at all.)
3866 17. A "Completed" log line is now written for messages that are removed from
3867 the spool by the -Mrm option.
3869 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3870 information about exactly what failed.
3872 19. Added -dd to debug only the daemon process.
3874 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3875 handles timeouts, both on the server side and network timeouts. Renamed the
3876 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3878 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3879 It is now set to "smtps".
3881 22. $host_address is now set to the target address during the checking of
3882 ignore_target_hosts.
3884 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3885 being passed; this would have caused $sender_host_name to have been used if
3886 matching the list had actually called for a host name (not very likely,
3887 since this list is usually IP addresses). A host name is now passed as
3890 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3891 code that specifies a non-restarting handler (typically sigaction() in
3892 modern systems) in an attempt to fix a rare and obscure crash bug.
3894 25. Narrowed the window for a race in the daemon that could cause it to ignore
3895 SIGCHLD signals. This is not a major problem, because they are used only to
3896 wake it up if nothing else does.
3898 26. A malformed maildirsize file could cause Exim to calculate negative values
3899 for the mailbox size or file count. Odd effects could occur as a result.
3900 The maildirsize information is now recalculated if the size or filecount
3903 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3904 support for a long time. Removed HAVE_SYS_VFS_H.
3906 28. Installed the latest version of exipick from John Jetmore.
3908 29. In an address list, if the pattern was not a regular expression, an empty
3909 subject address (from a bounce message) matched only if the pattern was an
3910 empty string. Non-empty patterns were not even tested. This was the wrong
3911 because it is perfectly reasonable to use an empty address as part of a
3912 database query. An empty address is now tested by patterns that are
3913 lookups. However, all the other forms of pattern expect the subject to
3914 contain a local part and a domain, and therefore, for them, an empty
3915 address still always fails if the pattern is not itself empty.
3917 30. Exim went into a mad DNS loop when attempting to do a callout where the
3918 host was specified on an smtp transport, and looking it up yielded more
3919 than one IP address.
3921 31. Re-factored the code for checking spool and log partition space into a
3922 function that finds that data and another that does the check. The former
3923 is then used to implement four new variables: $spool_space, $log_space,
3924 $spool_inodes, and $log_inodes.
3926 32. The RFC2047 encoding function was originally intended for short strings
3927 such as real names; it was not keeping to the 75-character limit for
3928 encoded words that the RFC imposes. It now respects the limit, and
3929 generates multiple encoded words if necessary. To be on the safe side, I
3930 have increased the buffer size for the ${rfc2047: expansion operator from
3933 33. It is now permitted to omit both strings after an "if" condition; if the
3934 condition is true, the result is "true". As before, when the second string
3935 is omitted, a false condition yields an empty string. This makes it less
3936 cumbersome to write custom ACL and router conditions.
3938 34. Failure to deliver a bounce message always caused it to be frozen, even if
3939 there was an errors_to setting on the router. The errors_to setting is now
3942 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3943 canonical form (fully expanded) before being placed in
3944 $sender_host_address.
3946 36. The table in the code that translates DNS record types into text (T_A to
3947 "A" for instance) was missing entries for NS and CNAME. It is just possible
3948 that this could have caused confusion if both these types were looked up
3949 for the same domain, because the text type is used as part of Exim's
3950 per-process caching. But the chance of anyone hitting this buglet seems
3953 37. The dnsdb lookup has been extended in a number of ways.
3955 (1) There is a new type, "zns", which walks up the domain tree until it
3956 finds some nameserver records. It should be used with care.
3958 (2) There is a new type, "mxh", which is like "mx" except that it returns
3959 just the host names, not the priorities.
3961 (3) It is now possible to give a list of domains (or IP addresses) to be
3962 looked up. The behaviour when one of the lookups defers can be
3963 controlled by a keyword.
3965 (4) It is now possible to specify the separator character for use when
3966 multiple records are returned.
3968 38. The dnslists ACL condition has been extended: it is now possible to supply
3969 a list of IP addresses and/or domains to be looked up in a particular DNS
3972 39. Added log_selector=+queue_time_overall.
3974 40. When running the queue in the test harness, wait just a tad after forking a
3975 delivery process, to get repeatability of debugging output.
3977 41. Include certificate and key file names in error message when GnuTLS fails
3978 to set them up, because the GnuTLS error message doesn't include the name
3979 of the failing file when there is a problem reading it.
3981 42. Allow both -bf and -bF in the same test run.
3983 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3985 44. The "Exiscan patch" is now merged into the mainline Exim source.
3987 45. Sometimes the final signoff response after QUIT could fail to get
3988 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3989 < 0 before doing a fflush(). This bug looks as though it goes back to the
3990 introduction of TLS in release 3.20, but "sometimes" must have been rare
3991 because the tests only now provoked it.
3993 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3994 (this can affect the format of dates).
3996 47. exim_tidydb, when checking for the continued existence of a message for
3997 which it has found a message-specific retry record, was not finding
3998 messages that were in split spool directories. Consequently, it was
3999 deleting retry records that should have stayed in existence.
4001 48. Steve fixed some bugs in eximstats.
4003 49. The SPA authentication driver was not abandoning authentication and moving
4004 on to the next authenticator when an expansion was forced to fail,
4005 contradicting the general specification for all authenticators. Instead it
4006 was generating a temporary error. It now behaves as specified.
4008 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4009 (the order specifies the preference for clients). The order is now AES256,
4010 AES128, 3DES, ARCFOUR128.
4012 51. Small patch to Sieve code - explicitly set From: when generating an
4015 52. Exim crashed if a remote delivery caused a very long error message to be
4016 recorded - for instance if somebody sent an entire SpamAssassin report back
4017 as a large number of 550 error lines. This bug was coincidentally fixed by
4018 increasing the size of one of Exim's internal buffers (big_buffer) that
4019 happened as part of the Exiscan merge. However, to be on the safe side, I
4020 have made the code more robust (and fixed the comments that describe what
4023 53. Now that there can be additional text after "Completed" in log lines (if
4024 the queue_time_overall log selector is set), a one-byte patch to exigrep
4025 was needed to allow it to recognize "Completed" as not the last thing in
4028 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4029 patch that reportedly fixes this has been added. I am not expert enough to
4030 create a test for it. This is what the patch creator wrote:
4032 "I found a little strange behaviour of ldap code when working with
4033 Windows 2003 AD Domain, where users was placed in more than one
4034 Organization Units. When I tried to give exim partial DN, the exit code
4035 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4036 But simultaneously result of request was absolutely normal ldap result,
4037 so I produce this patch..."
4039 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4040 REFERENCE, so I have modified the code to exclude the patch when that macro
4043 55. Some experimental protocols are using DNS PTR records for new purposes. The
4044 keys for these records are domain names, not reversed IP addresses. The
4045 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4046 leaves it alone. Component reversal etc. now happens only for IP addresses.
4049 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4051 57. Double the size of the debug message buffer (to 2048) so that more of very
4052 long debug lines gets shown.
4054 58. The exicyclog utility now does better if the number of log files to keep
4055 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4057 59. Two changes related to the smtp_active_hostname option:
4059 (1) $smtp_active_hostname is now available as a variable.
4060 (2) The default for smtp_banner uses $smtp_active_hostname instead
4061 of $primary_hostname.
4063 60. The host_aton() function is supposed to be passed a string that is known
4064 to be a valid IP address. However, in the case of IPv6 addresses, it was
4065 not checking this. This is a hostage to fortune. Exim now panics and dies
4066 if the condition is not met. A case was found where this could be provoked
4067 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4068 components; fortuitously, this particular loophole had already been fixed
4069 by change 4.50/55 above.
4071 If there are any other similar loopholes, the new check in host_aton()
4072 itself should stop them being exploited. The report I received stated that
4073 data on the command line could provoke the exploit when Exim was running as
4074 exim, but did not say which command line option was involved. All I could
4075 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4076 running as the user.
4079 61. There was a buffer overflow vulnerability in the SPA authentication code
4080 (which came originally from the Samba project). I have added a test to the
4081 spa_base64_to_bits() function which I hope fixes it.
4084 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4085 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4087 63. The daemon start-up calls getloadavg() while still root for those OS that
4088 need the first call to be done as root, but it missed one case: when
4089 deliver_queue_load_max is set with deliver_drop_privilege. This is
4090 necessary for the benefit of the queue runner, because there is no re-exec
4091 when deliver_drop_privilege is set.
4093 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4094 This has been fixed.
4096 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4097 was in use, was not putting the data itself into the right store pool;
4098 consequently, it could be overwritten for a subsequent message in the same
4099 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4102 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4104 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4105 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4106 the function were treating the return as a boolean value, which happened to
4107 work because 0=false and not-0=true, but is not correct code.
4109 68. The host_aton() function was not handling scoped IPv6 addresses (those
4110 with, for example, "%eth0" on the end) correctly.
4112 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4113 negated items (that is, ~something) in unsigned ints. Some compilers
4114 apparently mutter when there is no cast.
4116 70. If an address verification called from an ACL failed, and did not produce a
4117 user-specific message (i.e. there was only a "system" message), nothing was
4118 put in $acl_verify_message. In this situation, it now puts the system
4121 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4122 session; change 4.31/43 added the unwanted input to the log line - except
4123 that it did not do this in the start of session case. It now does.
4125 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4126 This is wrong and can cause the other end to generate a synchronization
4127 error if it is another Exim or anything else that does the synchronization
4128 check. A QUIT command is no longer sent after a timeout.
4130 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4131 during host lookups.
4133 74. The defer_ok option of callout verification was not working if it was used
4134 when verifying addresses in header lines, that is, for this case:
4136 verify = header_sender/callout=defer_ok
4138 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4139 those file descriptors could be used for SMTP connections. If anything
4140 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4141 could be sent to the SMTP client, causing chaos. The daemon now opens
4142 stdin, stdout, and stderr to /dev/null when it puts itself into the
4145 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4146 log by default. The user can override this with suitable Perl magic.
4148 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4149 the log message when discard triggers, was not working for the DATA ACL or
4150 for the non-SMTP ACL.
4152 78. Error message wording change in sieve.c.
4154 79. If smtp_accept_max_per_host was set, the number of connections could be
4155 restricted to fewer than expected, because the daemon was trying to set up
4156 a new connection before checking whether the processes handling previous
4157 connections had finished. The check for completed processes is now done
4158 earlier. On busy systems, this bug wouldn't be noticed because something
4159 else would have woken the daemon, and it would have reaped the completed
4162 80. If a message was submitted locally by a user whose login name contained one
4163 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4164 It caused a spool format error. I have fixed the spool reading code. A
4165 related problem was that the "from" clause in the Received: line became
4166 illegal because of the space(s). It is now covered by ${quote_local_part.
4168 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4171 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4173 83. Patch from Sieve maintainer to fix unterminated string problem in
4174 "vacation" handling.
4176 84. Some minor changes to the Linux configuration files to help with other
4177 OS variants using glibc.
4179 85. One more patch for Sieve to update vacation handling to latest spec.
4182 ----------------------------------------------------
4183 See the note above about the 4.44 and 4.50 releases.
4184 ----------------------------------------------------
4190 1. Change 4.43/35 introduced a bug that caused file counts to be
4191 incorrectly computed when quota_filecount was set in an appendfile
4194 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4195 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4198 3. Give more explanation in the error message when the command for a transport
4199 filter fails to execute.
4201 4. There are several places where Exim runs a non-Exim command in a
4202 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4203 was being done only for the command run by the queryprogram router. It is
4204 now done for all such subprocesses. The other cases are: ${run, transport
4205 filters, and the commands run by the lmtp and pipe transports.
4207 5. Some older OS have a limit of 256 on the maximum number of file
4208 descriptors. Exim was using setrlimit() to set 1000 as a large value
4209 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4210 systems. I've change it so that if it can't get 1000, it tries for 256.
4212 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4213 was an oversight, and furthermore, ever since the addition of extra
4214 controls (e.g. 4.43/32), the checks on when to allow different forms of
4215 "control" were broken. There should now be diagnostics for all cases when a
4216 control that does not make sense is encountered.
4218 7. $recipients is now available in the predata ACL (oversight).
4220 8. Tidy the search cache before the fork to do a delivery from a message
4221 received from the command line. Otherwise the child will trigger a lookup
4222 failure and thereby defer the delivery if it tries to use (for example) a
4223 cached ldap connection that the parent has called unbind on.
4225 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4226 of $address_data from the recipient verification was clobbered by the
4227 sender verification.
4229 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4230 was its contents. (It was OK if the option was not defined at all.)
4232 11. A "Completed" log line is now written for messages that are removed from
4233 the spool by the -Mrm option.
4235 12. $host_address is now set to the target address during the checking of
4236 ignore_target_hosts.
4238 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4239 being passed; this would have caused $sender_host_name to have been used if
4240 matching the list had actually called for a host name (not very likely,
4241 since this list is usually IP addresses). A host name is now passed as
4244 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4245 code that specifies a non-restarting handler (typically sigaction() in
4246 modern systems) in an attempt to fix a rare and obscure crash bug.
4248 15. Narrowed the window for a race in the daemon that could cause it to ignore
4249 SIGCHLD signals. This is not a major problem, because they are used only to
4250 wake it up if nothing else does.
4252 16. A malformed maildirsize file could cause Exim to calculate negative values
4253 for the mailbox size or file count. Odd effects could occur as a result.
4254 The maildirsize information is now recalculated if the size or filecount
4257 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4258 support for a long time. Removed HAVE_SYS_VFS_H.
4260 18. Updated exipick to current release from John Jetmore.
4262 19. Allow an empty sender to be matched against a lookup in an address list.
4263 Previously the only cases considered were a regular expression, or an
4266 20. Exim went into a mad DNS lookup loop when doing a callout where the
4267 host was specified on the transport, if the DNS lookup yielded more than
4270 21. The RFC2047 encoding function was originally intended for short strings
4271 such as real names; it was not keeping to the 75-character limit for
4272 encoded words that the RFC imposes. It now respects the limit, and
4273 generates multiple encoded words if necessary. To be on the safe side, I
4274 have increased the buffer size for the ${rfc2047: expansion operator from
4277 22. Failure to deliver a bounce message always caused it to be frozen, even if
4278 there was an errors_to setting on the router. The errors_to setting is now
4281 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4282 canonical form (fully expanded) before being placed in
4283 $sender_host_address.
4285 24. Updated eximstats to version 1.33
4287 25. Include certificate and key file names in error message when GnuTLS fails
4288 to set them up, because the GnuTLS error message doesn't include the name
4289 of the failing file when there is a problem reading it.
4291 26. Expand error message when OpenSSL has problems setting up cert/key files.
4294 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4295 (this can affect the format of dates).
4297 28. exim_tidydb, when checking for the continued existence of a message for
4298 which it has found a message-specific retry record, was not finding
4299 messages that were in split spool directories. Consequently, it was
4300 deleting retry records that should have stayed in existence.
4302 29. eximstats updated to version 1.35
4303 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4304 1.35 - bugfix such that pie charts by volume are generated correctly
4306 30. The SPA authentication driver was not abandoning authentication and moving
4307 on to the next authenticator when an expansion was forced to fail,
4308 contradicting the general specification for all authenticators. Instead it
4309 was generating a temporary error. It now behaves as specified.
4311 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4312 (the order specifies the preference for clients). The order is now AES256,
4313 AES128, 3DES, ARCFOUR128.
4315 31. Small patch to Sieve code - explicitly set From: when generating an
4318 32. Exim crashed if a remote delivery caused a very long error message to be
4319 recorded - for instance if somebody sent an entire SpamAssassin report back
4320 as a large number of 550 error lines. This bug was coincidentally fixed by
4321 increasing the size of one of Exim's internal buffers (big_buffer) that
4322 happened as part of the Exiscan merge. However, to be on the safe side, I
4323 have made the code more robust (and fixed the comments that describe what
4326 33. Some experimental protocols are using DNS PTR records for new purposes. The
4327 keys for these records are domain names, not reversed IP addresses. The
4328 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4329 leaves it alone. Component reversal etc. now happens only for IP addresses.
4332 34. The host_aton() function is supposed to be passed a string that is known
4333 to be a valid IP address. However, in the case of IPv6 addresses, it was
4334 not checking this. This is a hostage to fortune. Exim now panics and dies
4335 if the condition is not met. A case was found where this could be provoked
4336 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4337 components; fortuitously, this particular loophole had already been fixed
4338 by change 4.50/55 or 4.44/33 above.
4340 If there are any other similar loopholes, the new check in host_aton()
4341 itself should stop them being exploited. The report I received stated that
4342 data on the command line could provoke the exploit when Exim was running as
4343 exim, but did not say which command line option was involved. All I could
4344 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4345 running as the user.
4348 35. There was a buffer overflow vulnerability in the SPA authentication code
4349 (which came originally from the Samba project). I have added a test to the
4350 spa_base64_to_bits() function which I hope fixes it.
4353 36. The daemon start-up calls getloadavg() while still root for those OS that
4354 need the first call to be done as root, but it missed one case: when
4355 deliver_queue_load_max is set with deliver_drop_privilege. This is
4356 necessary for the benefit of the queue runner, because there is no re-exec
4357 when deliver_drop_privilege is set.
4359 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4360 was in use, was not putting the data itself into the right store pool;
4361 consequently, it could be overwritten for a subsequent message in the same
4362 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4365 38. Sometimes the final signoff response after QUIT could fail to get
4366 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4367 < 0 before doing a fflush(). This bug looks as though it goes back to the
4368 introduction of TLS in release 3.20, but "sometimes" must have been rare
4369 because the tests only now provoked it.
4375 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4376 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4377 Now it returns an integer. A number of calls were still expecting a T/F
4378 return. Fortuitously, in all cases, the tests worked in OK situations,
4379 which is the norm. However, things would have gone wrong on any write
4380 failures on the smtp file descriptor. This function is used when sending
4381 messages over SMTP and also when doing verify callouts.
4383 2. When Exim is called to do synchronous delivery of a locally submitted
4384 message (the -odf or -odi options), it no longer closes stderr before doing
4387 3. Implemented the mua_wrapper option.
4389 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4391 5. Implemented the functions header_remove(), header_testname(),
4392 header_add_at_position(), and receive_remove_recipient(), and exported them
4395 6. If an ACL "warn" statement specified the addition of headers, Exim already
4396 inserted X-ACL-Warn: at the start if there was no header name. However, it
4397 was not making this test for the second and subsequent header lines if
4398 there were newlines in the string. This meant that an invalid header could
4399 be inserted if Exim was badly configured.
4401 7. Allow an ACL "warn" statement to add header lines at the start or after all
4402 the Received: headers, as well as at the end.
4404 8. Added the rcpt_4xx retry error code.
4406 9. Added postmaster_mailfrom=xxx to callout verification option.
4408 10. Added mailfrom=xxxx to the callout verification option, for verify=
4411 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4412 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4414 12. Inserted some casts to stop certain compilers warning when using pointer
4415 differences as field lengths or precisions in printf-type calls (mostly
4416 affecting debugging statements).
4418 13. Added optional readline() support for -be (dynamically loaded).
4420 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4421 same clock tick as a message's arrival, so that its received time was the
4422 same as the "first fail" time on the retry record, and that message
4423 remained on the queue past the ultimate address timeout, every queue runner
4424 would try a delivery (because it was past the ultimate address timeout) but
4425 after another failure, the ultimate address timeout, which should have then
4426 bounced the address, did not kick in. This was a "< instead of <=" error;
4427 in most cases the first failure would have been in the next clock tick
4428 after the received time, and all would be well.
4430 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4431 being recognized when the domain list was tested by the match_domain
4432 condition in an expansion string.
4434 16. Added the ${str2b64: operator.
4436 17. Exim was always calling setrlimit() to set a large limit for the number of
4437 processes, without checking whether the existing limit was already
4438 adequate. (It did check for the limit on file descriptors.) Furthermore,
4439 errors from getrlimit() and setrlimit() were being ignored. Now they are
4440 logged to the main and panic logs, but Exim does carry on, to try to do its
4441 job under whatever limits there are.
4443 18. Imported PCRE 5.0.
4445 19. Trivial typo in log message " temporarily refused connection" (the leading
4448 20. If the log selector return_path_on_delivery was set and an address was
4449 redirected to /dev/null, the delivery process crashed because it assumed
4450 that a return path would always be set for a "successful" delivery. In this
4451 case, the whole delivery is bypassed as an optimization, and therefore no
4454 21. Internal re-arrangement: the function for sending a challenge and reading
4455 a response while authentication was assuming a zero-terminated challenge
4456 string. It's now changed to take a pointer and a length, to allow for
4457 binary data in such strings.
4459 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4461 23. Exim was not respecting finduser_retries when seeking the login of the
4462 uid under which it was called; it was always trying 10 times. (The default
4463 setting of finduser_retries is zero.) Also, it was sleeping after the final
4464 failure, which is pointless.
4466 24. Implemented tls_on_connect_ports.
4468 25. Implemented acl_smtp_predata.
4470 26. If the domain in control=submission is set empty, Exim assumes that the
4471 authenticated id is a complete email address when it generates From: or
4472 Sender: header lines.
4474 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4475 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4476 chown and chgrp in /bin and hostname in /usr/bin.
4478 28. Exim was keeping the "process log" file open after each use, just as it
4479 does for the main log. This opens the possibility of it remaining open for
4480 long periods when the USR1 signal hits a daemon. Occasional processlog
4481 errors were reported, that could have been caused by this. Anyway, it seems
4482 much more sensible not to leave this file open at all, so that is what now
4485 29. The long-running daemon process does not normally write to the log once it
4486 has entered its main loop, and it closes the log before doing so. This is
4487 so that log files can straightforwardly be renamed and moved. However,
4488 there are a couple of unusual error situations where the daemon does write
4489 log entries, and I had neglected to close the log afterwards.
4491 30. The text of an SMTP error response that was received during a remote
4492 delivery was being truncated at 512 bytes. This is too short for some of
4493 the long messages that one sometimes sees. I've increased the limit to
4496 31. It is now possible to make retry rules that apply only when a message has a
4497 specific sender, in particular, an empty sender.
4499 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4500 it possible to be selective about when SMTP synchronization is enforced.
4502 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4504 32. Implemented hosts_connection_nolog.
4506 33. Added an ACL for QUIT.
4508 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4511 35. Added mailbox_size and mailbox_filecount to appendfile.
4513 36. Added control = no_multiline_responses to ACLs.
4515 37. There was a bug in the logic of the code that waits for the clock to tick
4516 in the case where the clock went backwards by a substantial amount such
4517 that the microsecond fraction of "now" was more than the microsecond
4518 fraction of "then" (but the whole seconds number was less).
4520 38. Added support for the libradius Radius client library this is found on
4521 FreeBSD (previously only the radiusclient library was supported).
4527 1. When certain lookups returned multiple values in the form name=value, the
4528 quoting of the values was not always being done properly. Specifically:
4529 (a) If the value started with a double quote, but contained no whitespace,
4531 (b) If the value contained whitespace other than a space character (i.e.
4532 tabs or newlines or carriage returns) it was not quoted.
4533 This fix has been applied to the mysql and pgsql lookups by writing a
4534 separate quoting function and calling it from the lookup code. The fix
4535 should probably also be applied to nisplus, ibase and oracle lookups, but
4536 since I cannot test any of those, I have not disturbed their existing code.
4538 2. A hit in the callout cache for a specific address caused a log line with no
4539 reason for rejecting RCPT. Now it says "Previous (cached) callout
4540 verification failure".
4542 3. There was an off-by-one bug in the queryprogram router. An over-long
4543 return line was truncated at 256 instead of 255 characters, thereby
4544 overflowing its buffer with the terminating zero. As well as fixing this, I
4545 have increased the buffer size to 1024 (and made a note to document this).
4547 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4548 when Exim is waiting for an SMTP response from a remote server, Exim
4549 restarts its select() call on the socket, thereby resetting its timeout.
4550 This is not a problem when such interrupts are rare. Somebody set up a cron
4551 job to run exiwhat every 2 minutes, which is less than the normal select()
4552 timeout (5 or 10 minutes). This meant that the select() timeout never
4553 kicked in because it was always reset. I have fixed this by comparing the
4554 time when an interrupt arrives with the time at the start of the first call
4555 to select(). If more time than the timeout has elapsed, the interrupt is
4556 treated as a timeout.
4558 5. Some internal re-factoring in preparation for the addition of Sieve
4559 extensions (by MH). In particular, the "personal" test is moved to a
4560 separate function, and given an option for scanning Cc: and Bcc: (which is
4561 not set for Exim filters).
4563 6. When Exim created an email address using the login of the caller as the
4564 local part (e.g. when creating a From: or Sender: header line), it was not
4565 quoting the local part when it contained special characters such as @.
4567 7. Installed new OpenBSD configuration files.
4569 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4570 try to make them clearer.
4572 9. Callout options, other than the timeout value, were being ignored when
4573 verifying sender addresses in header lines. For example, when using
4575 verify = header_sender/callout=no_cache
4577 the cache was (incorrectly) being used.
4579 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4580 only the Cygwin environment.
4582 11. When return_path_on_delivery was set as a log selector, if different remote
4583 addresses in the same message used different return paths and parallel
4584 remote delivery occurred, the wrong values would sometimes be logged.
4585 (Whenever a remote delivery process finished, the return path value from
4586 the most recently started remote delivery process was logged.)
4588 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4589 lines when AUTH and/or TLS are in use. This is the "received protocol"
4590 field. Exim used to use "asmtp" for authenticated SMTP, without any
4591 indication (in the protocol name) for TLS use. Now it follows the RFC and
4592 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4593 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4594 names appear in log lines as well as in Received: header lines.
4596 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4597 extensions, and comparison tests, and to fix some bugs.
4599 14. Changes to the "personal" filter test:
4601 (1) The test was buggy in that it was just doing the equivalent of
4602 "contains" tests on header lines. For example, if a user's address was
4603 anne@some.where, the "personal" test would incorrectly be true for
4605 To: susanne@some.where
4607 This test is now done by extracting each address from the header in turn,
4608 and checking the entire address. Other tests that are part of "personal"
4609 are now done using regular expressions (for example, to check local parts
4610 of addresses in From: header lines).
4612 (2) The list of non-personal local parts in From: addresses has been
4613 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4614 taken from the Sieve specification recommendations.
4616 (3) If the message contains any header line starting with "List-" it is
4617 treated as non-personal.
4619 (4) The test for "circular" in the Subject: header line has been removed
4620 because it now seems ill-conceived.
4622 15. Minor typos in src/EDITME comments corrected.
4624 16. Installed latest exipick from John Jetmore.
4626 17. If headers_add on a router specified a text string that was too long for
4627 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4628 of string_sprintf() is now avoided.
4630 18. $message_body_size was not set (it was always zero) when running the DATA
4631 ACL and the local_scan() function.
4633 19. For the "mail" command in an Exim filter, no default was being set for
4634 the once_repeat time, causing a random time value to be used if "once" was
4635 specified. (If the value happened to be <= 0, no repeat happened.) The
4636 default is now 0s, meaning "never repeat". The "vacation" command was OK
4637 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4638 (I found it when inspecting the code).
4640 20. There is now an overall timeout for performing a callout verification. It
4641 defaults to 4 times the callout timeout, which applies to individual SMTP
4642 commands during the callout. The overall timeout applies when there is more
4643 than one host that can be tried. The timeout is checked before trying the
4644 next host. This prevents very long delays if there are a large number of
4645 hosts and all are timing out (e.g. when the network connections are timing
4646 out). The value of the overall timeout can be changed by specifying an
4647 additional sub-option for "callout", called "maxwait". For example:
4649 verify = sender/callout=5s,maxwait=20s
4651 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4652 to the end before writing, but this should make it even safer).
4654 22. Exim was forgetting that it had advertised PIPELINING for the second and
4655 subsequent messages on an SMTP connection. It was also not resetting its
4656 memory on STARTTLS and an internal HELO.
4658 23. When Exim logs an SMTP synchronization error within a session, it now
4659 records whether PIPELINING has been advertised or not.
4661 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4662 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4663 rather than long int.
4665 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4667 26. Added the never_mail option to autoreply.
4673 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4674 crash if the getsockname() call failed; this can happen if a connection is
4675 closed very soon after it is established. The problem was simply in the
4676 order in which certain operations were done, causing Exim to try to write
4677 to the SMTP stream before it had set up the file descriptor. The bug has
4678 been fixed by making things happen in the correct order.
4684 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4685 before the connection was closed, thus losing the rejection response.
4687 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4688 some early Solaris releases, but causes trouble in current releases where
4689 socklen_t is defined.
4691 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4694 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4697 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4698 that releasing the top of it at the end releases what was used for sub-
4699 expansions (unless the block got too big). However, discard this block if
4700 the first thing is a variable or header, so that we can use its block when
4701 it is dynamic (useful for very large $message_headers, for example).
4703 6. Lookups now cache *every* query, not just the most recent. A new, separate
4704 store pool is used for this. It can be recovered when all lookup caches are
4705 flushed. Lookups now release memory at the end of their result strings.
4706 This has involved some general refactoring of the lookup sources.
4708 7. Some code has been added to the store_xxx() functions to reduce the amount
4709 of flapping under certain conditions.
4711 8. log_incoming_interface used to affect only the <= reception log lines. Now
4712 it causes the local interface and port to be added to several more SMTP log
4713 lines, for example "SMTP connection from", and rejection lines.
4715 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4717 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4719 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4720 could be overwritten at the end of the current message (or the start of a
4721 new message if it was set in a HELO ACL). The value is now preserved for
4722 the duration of the SMTP connection.
4724 12. If a transport had a headers_rewrite setting, and a matching header line
4725 contained an unqualified address, that address was qualified, even if it
4726 did not match any rewriting rules. The underlying bug was that the values
4727 of the flags that permit the existence of unqualified sender and recipient
4728 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4729 non-local messages, and by -bnq for local messages) were not being
4730 preserved with the message after it was received.
4732 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4733 "next input=" as part of the text comprising the host identity instead of
4734 the correct text. The code was using the same buffer for two different
4735 strings. However, depending on which order the printing function evaluated
4736 its arguments, the bug did not always show up. Under Linux, for example, my
4737 test suite worked just fine.
4739 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4740 doesn't work with some older versions of Perl. It has been changed to "my",
4741 which in any case is probably the better facility to use.
4743 15. A really picky compiler found some instances of statements for creating
4744 error messages that either had too many or two few arguments for the format
4747 16. The size of the buffer for calls to the DNS resolver has been increased
4748 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4749 for addresses that have a lot of PTR records. This alleviates a problem; it
4750 does not fully solve it.
4752 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4753 buffer now truncates the list and logs the incident, which is the same
4754 action as happens when Exim is looking up a host name and its aliases.
4755 Previously in this situation something unpredictable would happen;
4756 sometimes it was "internal error: store_reset failed".
4758 18. If a server dropped the connection unexpectedly when an Exim client was
4759 using GnuTLS and trying to read a response, the client delivery process
4760 crashed while trying to generate an error log message.
4762 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4763 string, for example:
4765 warn message = H1: something\nH2: something
4767 the text was added as a single header line from Exim's point of view
4768 though it ended up OK in the delivered message. However, searching for the
4769 second and subsequent header lines using $h_h2: did not work. This has been
4770 fixed. Similarly, if a system filter added multiple headers in this way,
4771 the routers could not see them.
4773 20. Expanded the error message when iplsearch is called with an invalid key to
4774 suggest using net-iplsearch in a host list.
4776 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4777 ACLs are no longer actually imposed (and a message to that effect is
4780 22. If a "gecos" field in a passwd entry contained escaped characters, in
4781 particular, if it contained a \" sequence, Exim got it wrong when building
4782 a From: or a Sender: header from that name. A second bug also caused
4783 incorrect handling when an unquoted " was present following a character
4784 that needed quoting.
4786 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4787 was not being matched caselessly.
4789 24. Arranged for all hyphens in the exim.8 source to be escaped with
4792 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4793 properly. Recipient callout cache records were still being keyed to include
4794 the sender, even when use_sender was set false. This led to far more
4795 callouts that were necessary. The sender is no longer included in the key
4796 when use_sender is false.
4798 26. Added "control = submission" modifier to ACLs.
4800 27. Added the ${base62d: operator to decode base 62 numbers.
4802 28. dnsdb lookups can now access SRV records.
4804 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4805 the configuration file.
4807 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4808 (-v) mode. This makes the output for a verbose queue run more intelligible.
4810 31. Added a use_postmaster feature to recipient callouts.
4812 32. Added the $body_zerocount variable, containing the number of binary zero
4813 bytes in the message body.
4815 33. The time of last modification of the "new" subdirectory is now used as the
4816 "mailbox time last read" when there is a quota error for a maildir
4819 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4821 35. Added +ignore_unknown as a special item in host lists.
4823 36. Code for decoding IPv6 addresses in host lists is now included, even if
4824 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4825 address was recognized as an IP address, but was then not correctly decoded
4826 into binary, causing unexpected and incorrect effects when compared with
4833 1. Very minor rewording of debugging text in manualroute to say "list of
4834 hosts" instead of "hostlist".
4836 2. If verify=header_syntax was set, and a header line with an unqualified
4837 address (no domain) and a large number of spaces between the end of the
4838 name and the colon was received, the reception process suffered a buffer
4839 overflow, and (when I tested it) crashed. This was caused by some obsolete
4840 code that should have been removed. The fix is to remove it!
4842 3. When running in the test harness, delay a bit after writing a bounce
4843 message to get a bit more predictability in the log output.
4845 4. Added a call to search_tidyup() just before forking a reception process. In
4846 theory, someone could use a lookup in the expansion of smtp_accept_max_
4847 per_host which, without the tidyup, could leave open a database connection.
4849 5. Added the variables $recipient_data and $sender_data which get set from a
4850 lookup success in an ACL "recipients" or "senders" condition, or a router
4851 "senders" option, similar to $domain_data and $local_part_data.
4853 6. Moved the writing of debug_print from before to after the "senders" test
4856 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4857 problems for message scanning, either using a data ACL, or using
4858 local_scan() because the Received: header was not generated till after they
4859 were called (in order to set the time as the time of reception completion).
4860 I have revised the way this works. The header is now generated after the
4861 body is received, but before the ACL or local_scan() are called. After they
4862 are run, the timestamp in the header is updated.
4868 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4869 before starting a queue runner without re-exec. This happened only when
4870 deliver_drop_privilege was set or when the Exim user was set to root. The
4871 effect of the bug was that timeouts during subsequent deliveries caused
4872 crashes instead of being properly handled. The handler is now left at its
4873 default (and expected) setting.
4875 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4876 message, again when deliver_drop_privilege is set or Exim is run as root.
4877 The bug described in (1) was not present in this case, but the tidying up
4878 of the other signals was missing. I have made the two cases consistent.
4880 3. The ignore_target_hosts setting on a manualroute router was being ignored
4881 for hosts that were looked up using the /MX notation.
4883 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4886 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4887 operated on the sender address. After changing the $sender_address to <>
4888 for the sender address verify, Exim was re-instated it as the original
4889 (before rewriting) address, but remembering that it had rewritten it, so it
4890 wasn't rewriting it again. This bug also had the effect of breaking the
4891 sender address verification caching when the sender address was rewritten.
4893 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4894 This has been changed so that if the ip literal address matches
4895 ignore_target_hosts, the router declines.
4897 7. Added expansion conditions match_domain, match_address, and match_local_
4898 part (NOT match_host).
4900 8. The placeholder for the Received: header didn't have a length field set.
4902 9. Added code to Exim itself and to exim_lock to test for a specific race
4903 condition that could lead to file corruption when using MBX delivery. The
4904 issue is with the lockfile that is created in /tmp. If this file is removed
4905 after a process has opened it but before that process has acquired a lock,
4906 there is the potential for a second process to recreate the file and also
4907 acquire a lock. This could lead to two Exim processes writing to the file
4908 at the same time. The added code performs the same test as UW imapd; it
4909 checks after acquiring the lock that its file descriptor still refers to
4910 the same named file.
4912 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4913 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4916 11. Added the smtp_active_hostname option. If used, this will typically be made
4917 to depend on the incoming interface address. Because $interface_address is
4918 not set up until the daemon has forked a reception process, error responses
4919 that can happen earlier (such as "too many connections") no longer contain
4922 12. If an expansion in a condition on a "warn" statement fails because a lookup
4923 defers, the "warn" statement is abandoned, and the next ACL statement is
4924 processed. Previously this caused the whole ACL to be aborted.
4926 13. Added the iplsearch lookup type.
4928 14. Added ident_timeout as a log selector.
4930 15. Added tls_certificate_verified as a log selector.
4932 16. Added a global option tls_require_ciphers (compare the smtp transport
4933 option of the same name). This controls incoming TLS connections.
4935 17. I finally figured out how to make tls_require_ciphers do a similar thing
4936 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4937 before starting the TLS session.
4939 18. Tabs are now shown as \t in -bP output.
4941 19. If the log selector return_path_on_delivery was set, Exim crashed when
4942 bouncing a message because it had too many Received: header lines.
4944 20. If two routers both had headers_remove settings, and the first one included
4945 a superfluous trailing colon, the final name in the first list and the
4946 first name in the second list were incorrectly joined into one item (with a
4947 colon in the middle).
4953 1. Added -C and -D options to the exinext utility, mainly to make it easier
4954 to include in the automated testing, but these could be helpful when
4955 multiple configurations are in use.
4957 2. The exinext utility was not formatting the output nicely when there was
4958 an alternate port involved in the retry record key, nor when there was a
4959 message id as well (for retries that were specific to a specific message
4960 and a specific host). It was also confused by IPv6 addresses, because of
4961 the additional colons they contain. I have fixed the IPv4 problem, and
4962 patched it up to do a reasonable job for IPv6.
4964 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4965 delivery, the log line now contains "pipelined" if PIPELINING was used.
4967 4. An SMTP transport process used to panic and die if the bind() call to set
4968 an explicit outgoing interface failed. This has been changed; it is now
4969 treated in the same way as a connect() failure.
4971 5. A reference to $sender_host_name in the part of a conditional expansion
4972 that was being skipped was still causing a DNS lookup. This no longer
4975 6. The def: expansion condition was not recognizing references to header lines
4976 that used bh_ and bheader_.
4978 7. Added the _cache feature to named lists.
4980 8. The code for checking quota_filecount in the appendfile transport was
4981 allowing one more file than it should have been.
4983 9. For compatibility with Sendmail, the command line option
4991 and sets the incoming protocol and host name (for trusted callers). The
4992 host name and its colon can be omitted when only the protocol is to be set.
4993 Note the Exim already has two private options, -pd and -ps, that refer to
4994 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4995 "s", but I don't think that's a major issue.
4997 10. A number of refactoring changes to the code, none of which should affect
5000 (a) The number of logging options was getting close to filling up the
5001 32-bit word that was used as a bit map. I have split them into two classes:
5002 those that are passed in the argument to log_write(), and those that are
5003 only ever tested independently outside of that function. These are now in
5004 separate 32-bit words, so there is plenty of room for expansion again.
5005 There is no change in the user interface or the logging behaviour.
5007 (b) When building, for example, log lines, the code previously used a
5008 macro that called string_cat() twice, in order to add two strings. This is
5009 not really sufficiently general. Furthermore, there was one instance where
5010 it was actually wrong because one of the argument was used twice, and in
5011 one call a function was used. (As it happened, calling the function twice
5012 did not affect the overall behaviour.) The macro has been replaced by a
5013 function that can join an arbitrary number of extra strings onto a growing
5016 (c) The code for expansion conditions now uses a table and a binary chop
5017 instead of a serial search (which was left over from when there were very
5018 few conditions). Also, it now recognizes conditions like "pam" even when
5019 the relevant support is not compiled in: a suitably worded error message is
5020 given if an attempt is made to use such a condition.
5022 11. Added ${time_interval:xxxxx}.
5024 12. A bug was causing one of the ddress fields not to be passed back correctly
5025 from remote delivery subprocesses. The field in question was not being
5026 subsequently used, so this caused to problems in practice.
5028 13. Added new log selectors queue_time and deliver_time.
5030 14. Might have fixed a bug in maildirsizefile handling that threw up
5031 "unexpected character" debug warnings, and recalculated the data
5032 unnecessarily. In any case, I expanded the warning message to give more
5035 15. Added the message "Restricted characters in address" to the statements in
5036 the default ACL that block characters like @ and % in local parts.
5038 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5039 Three changes have been made:
5041 (a) There was a serious bug; a negative response to MAIL caused the whole
5042 recipient domain to be cached as invalid, thereby blocking all messages
5043 to all local parts at the same domain, from all senders. This bug has
5044 been fixed. The domain is no longer cached after a negative response to
5045 MAIL if the sender used is not empty.
5047 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5050 (c) A new callout option, "use_sender" has been added for people who want
5051 the modified behaviour.
5057 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5060 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5061 indeed breaks things for older releases.
5063 3. Added additional logging to the case where there is a problem reading data
5064 from a filter that is running in a subprocess using a pipe, in order to
5065 try to track down a specific problem.
5067 4. Testing facility fudge: when running in the test harness and attempting
5068 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5069 getting "No route to host". Convert this to a timeout.
5071 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5074 6. Some OS don't have socklen_t but use size_t instead. This affects the
5075 fifth argument of getsockopt() amongst other things. This is now
5076 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5077 can be set for individual OS. I have set it for SunOS5, OSF1, and
5078 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5079 some earlier ones do not.
5081 7. Change 4.30/15 was not doing the test caselessly.
5083 8. The standard form for an IPv6 address literal was being rejected by address
5084 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5085 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5086 this, as well as the form without the "IPv6" on the front (but only when
5087 address literals are enabled, of course).
5089 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5091 10. Exim crashed if a message with an empty sender address specified by -f
5092 encountered a router with an errors_to setting. This could be provoked only
5093 by a command such as
5097 where an empty string was supplied; "<>" did not hit this bug.
5099 11. Installed PCRE release 4.5.
5101 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5102 remained set. It is now erased.
5104 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5105 times from message ids (which are base 36 rather than the normal 62).
5107 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5108 were being counted as actual protocol errors, and logged if the log
5109 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5110 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5111 and DATA following a set of rejected RCPTs do not count as protocol errors.
5112 In other words, Exim assumes they were pipelined, though this may not
5113 actually be the case. Of course, in all cases the client gets an
5114 appropriate error code.
5116 15. If a lookup fails in an ACL condition, a message about the failure may
5117 be available; it is used if testing the ACL cannot continue, because most
5118 such messages specify what the cause of the deferral is. However, some
5119 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5120 that caused an old message to be retained and used if a later statement
5121 caused a defer, replacing the real cause of the deferral.
5123 16. If an IP address had so many PTR records that the DNS lookup buffer
5124 was not large enough to hold them, Exim could crash while trying to process
5125 the truncated data. It now detects and logs this case.
5127 17. Further to 4.21/58, another change has been made: if (and only if) the
5128 first line of a message (the first header line) ends with CRLF, a bare LF
5129 in a subsequent header line has a space inserted after it, so as not to
5130 terminate the header.
5132 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5133 unnecessarily, used atoi() instead of strtol(), and didn't check the
5134 termination when getting file sizes from file names by regex.
5136 19. Completely re-implemented the support for maildirsize files, in the light
5137 of a number of problems with the previous contributed implementation
5138 (4.30/29). In particular:
5140 . If the quota is zero, the maildirsize file is maintained, but no quota is
5143 . If the maildir directory does not exist, it is created before any attempt
5144 to write a maildirsize file.
5146 . The quota value in the file is just a cache; if the quota is changed in
5147 the transport, the new value overrides.
5149 . A regular expression is available for excluding directories from the
5152 20. The autoreply transport checks the characters in options that define the
5153 message's headers; it allows continued headers, but it was checking with
5154 isspace() after an embedded newline instead of explicitly looking for a
5157 21. If all the "regular" hosts to which an address was routed had passed their
5158 expiry times, and had not reached their retry times, the address was
5159 bounced, even if fallback hosts were defined. Now Exim should go on to try
5162 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5163 equivalent code in the SMTP transport. Some hosts send humungous responses
5164 to HELO/EHLO, more than 1024 it seems.
5166 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5167 gives compiler warnings in some environments. I've now done it "properly",
5170 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5171 (because of problems with the built-in one) was declared to return uschar *
5172 instead of char *, causing compiler failure.
5174 25. Fixed a file descriptor leak when processing alias/forward files.
5176 26. Fixed a minor format string issue in dbfn.c.
5178 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5180 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5181 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5184 29. When a sender address is verified, it is cached, to save repeating the test
5185 when there is more than one recipient in a message. However, when the
5186 verification involves a callout, it is possible for different callout
5187 options to be set for different recipients. It is too complicated to keep
5188 track of this in the cache, so now Exim always runs a verification when a
5189 callout is required, relying on the callout cache for the optimization.
5190 The overhead is duplication of the address routing, but this should not be
5193 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5194 to be verified with callout=postmaster, and the main callout worked but the
5195 postmaster check failed, the verification correctly failed. However, if a
5196 subsequent RCPT command asked for sender verification *without* the
5197 postmaster check, incorrect caching caused this verification also to fail,
5200 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5201 it was not caching the DNS options (qualify_single, search_parents) that
5202 were used when the lookup failed. A subsequent lookup with different
5203 options therefore always gave the same answer, though there were cases
5204 where it should not have. (Example: a "domains = !$mx_any" option on a
5205 dnslookup router: the "domains" option is always processed without any
5206 widening, but the router might have qualify_single set.) Now Exim uses the
5207 cached value only when the same options are set.
5209 32. Added John Jetmore's "exipick" utility to the distribution.
5211 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5212 than a timeout (e.g. a certificate is required, and is not provided), an
5213 Exim server now closes the connection immediately. Previously it waited for
5214 the client to close - but if the client is SSL, it seems that they each
5215 wait for each other, leading to a delay before one of them times out.
5217 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5218 maintained 0.8.x compatibility because I don't think many are using it, and
5219 it is clearly obsolete.
5221 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5224 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5225 client certificate was expired. A simple patch fixes this, though I don't
5226 understand the full logic of why the verify callback is called multiple
5229 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5230 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5231 which causes problems with some clients (such as the Certicom SSL Plus
5232 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5233 disables the coutermeasure allowing Eudora to connect."
5235 38. Exim was not checking that a write() to a log file succeeded. This could
5236 lead to Bad Things if a log got too big, in particular if it hit a file
5237 size limit. Exim now panics and dies if it cannot write to a log file, just
5238 as it does if it cannot open a log file.
5240 39. Modified OS/Makefile-Linux so that it now contains
5242 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5244 The two -D definitions ensure that Exim is compiled with large file
5245 support, which makes it possible to handle log files that are bigger than
5248 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5249 instance) a domain was checked against a named list that involved a lookup,
5250 causing $domain_data to be set, then another domain was checked against the
5251 same list, then the first domain was re-checked, the value of $domain_data
5252 after the final check could be wrong. In particular, if the second check
5253 failed, it could be set empty. This bug probably also applied to
5256 41. The strip_trailing_dot option was not being applied to the address given
5257 with the -f command-line option.
5259 42. The code for reading a message's header from the spool was incrementing
5260 $received_count, but never initializing it. This meant that the value was
5261 incorrect (doubled) while delivering a message in the same process in which
5262 it was received. In the most common configuration of Exim, this never
5263 happens - a fresh exec is done - but it can happen when
5264 deliver_drop_privilege is set.
5266 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5267 it now includes up to 150 characters of the unexpected data in the log
5270 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5271 and building data strings. The size of both of these buffers was 10 000
5272 bytes - far larger than anybody would *ever* want, thought I. Needless to
5273 say, somebody hit the limit. I have increased the maximum line length to
5274 20 000 and the maximum data length of concatenated lines to 100 000. I have
5275 also fixed two bugs, because there was no checking on these buffers. Tsk,
5276 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5277 buffer is too small.
5279 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5280 lsearch lookups. Now it does.
5282 46. When parsing a route_list item in a manualroute router, a fixed-length
5283 buffer was used for the list of hosts. I made this 1024 bytes long,
5284 thinking that nobody would ever have a list of hosts that long. Wrong.
5285 Somebody had a whole pile of complicated expansion conditions, and the
5286 string was silently truncated, leading to an expansion error. It turns out
5287 that it is easier to change to an unlimited length (owing to other changes
5288 that have happened since this code was originally written) than to build
5289 structure for giving a limitation error. The length of the item that
5290 expands into the list of hosts is now unlimited.
5292 47. The lsearch lookup could not handle data where the length of text line was
5293 more than 4095 characters. Such lines were truncated, leading to shortened
5294 data being returned. It should now handle lines of any length.
5296 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5297 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5300 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5301 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5302 no longer happens in comment lines. A list of replacements is now placed
5303 at the head of all of the source files, except those whose only change is
5304 to replace PERL_COMMAND in the very first #! line.
5306 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5307 messages on the queue, with a bottom-up merge sort, using code contributed
5308 by Michael Haardt. This should make operations like -bp somewhat faster on
5309 large queues. It won't affect queue runners, except when queue_run_in_order
5312 51. Installed eximstats 1.31 in the distribution.
5314 52. Added support for SRV lookups to the dnslookup router.
5316 53. If an ACL referred to $message_body or $message_body_end, the value was not
5317 reset for any messages that followed in the same SMTP session.
5319 54. The store-handling optimization for building very long strings was not
5320 differentiating between the different store pools. I don't think this
5321 actually made any difference in practice, but I've tidied it.
5323 55. While running the routers to verify a sender address, $sender_address
5324 was still set to the sender address. This is wrong, because when routing to
5325 send a bounce to the sender, it would be empty. Therefore, I have changed
5326 it so that, while verifying a sender address, $sender_address is set to <>.
5327 (There is no change to what happens when verifying a recipient address.)
5329 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5330 target A or AAAA records (if not already returned) without resetting the
5331 qualify_single or search_parents options of the DNS resolver. These are
5332 inappropriate in this case because the targets of MX and SRV records must
5333 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5334 target that, when qualified, matched something in the local domain. These
5335 two options are now turned off when doing these lookups.
5337 57. It seems that at least some releases of Reiserfs (which does not have the
5338 concept of a fixed number of inodes) returns zero and not -1 for the
5339 number of available inodes. This interacted badly with check_spool_inodes,
5340 which assumed that -1 was the "no such thing" setting. What I have done is
5341 to check that the total number of inodes is greater than zero before doing
5342 the test of how many are available.
5344 58. When a "warn" ACL statement has a log_message modifier, the message is
5345 remembered, and not repeated. This is to avoid a lot of repetition when a
5346 message has many recipients that cause the same warning to be written.
5347 Howewer, Exim was preserving the list of already written lines for an
5348 entire SMTP session, which doesn't seem right. The memory is now reset if a
5349 new message is started.
5351 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5352 debugging output unless log_rewrite was also set.
5354 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5355 of (int)(handle) when we know that handle contains (void *)(-1).
5357 61. The Exim daemon panic-logs an error return when it closes the incoming
5358 connection. However "connection reset by peer" seems to be common, and
5359 isn't really an error worthy of noting specially, so that particular error
5362 62. When Exim is trying to find all the local interfaces, it used to panic and
5363 die if the ioctl to get the interface flags failed. However, it seems that
5364 on at least one OS (Solaris 9) it is possible to have an interface that is
5365 included in the list of interfaces, but for which you get a failure error
5366 for this call. This happens when the interface is not "plumbed" into a
5367 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5368 failure of the "get flags" call assumes that the interface is down.
5370 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5371 makes life easier for people who are doing arithmetic on fields extracted
5372 from dates, where you often get leading zeros that should not be
5373 interpreted as octal.
5375 64. Added qualify_domain to the redirect router, to override the global
5378 65. If a pathologically long header line contained very many addresses (the
5379 report of this problem mentioned 10 000) and each of them was rewritten,
5380 Exim could use up a very large amount of memory. (It kept on making new
5381 copies of the header line as it rewrote, and never released the old ones.)
5382 At the expense of a bit more processing, the header rewriting function has
5383 been changed so that it no longer eats memory in this way.
5385 66. The generation of the Received: header has been moved from the time that a
5386 message starts to be received, to the time that it finishes. The timestamp
5387 in the Received: header should now be very close to that of the <= log
5388 line. There are two side-effects of this change:
5390 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5391 logged header lines no longer include the local Received: line, because
5392 it has not yet been created. The same applies to a copy of the message
5393 that is returned to a non-SMTP sender when a message is rejected.
5395 (b) When a filter file is tested using -bf, no additional Received: header
5396 is added to the test message. After some thought, I decided that this
5399 This change does not affect the value of $received_for. It is still set
5400 after address rewriting, but before local_scan() is called.
5402 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5404 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5405 gave an unhelpful panic error message, and a defer error. I have managed to
5406 change this behaviour so that it now rejects any supplied certificate,
5407 which seems right, as the list of acceptable certificates is empty.
5409 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5410 gave an unhelpful defer error. I have not managed to make this reject any
5411 supplied certificates, but the error message it gives is "no certificate
5412 supplied", which is not helpful.
5414 70. exigrep's output now also includes lines that are not associated with any
5415 message, but which match the given pattern. Implemented by a patch from
5416 Martin Sluka, which also tidied up the Perl a bit.
5418 71. Recipient callout verification, like sender verification, was using <> in
5419 the MAIL FROM command. This isn't really the right thing, since the actual
5420 sender may affect whether the remote host accepts the recipient or not. I
5421 have changed it to use the actual sender in the callout; this means that
5422 the cache record is now keyed on a recipient/sender pair, not just the
5423 recipient address. There doesn't seem to be a real danger of callout loops,
5424 since a callout by the remote host to check the sender would use <>.
5425 [SEE ABOVE: changed after hitting problems.]
5427 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5428 temporary errors. However, in the case of such a code being given after
5429 the end of a data transmission (i.e. after ".") Exim was failing to write
5430 a retry record for the message. (Yes, there was some broken host that was
5431 actually sending 8xx at this point.)
5433 73. An unknown lookup type in a host list could cause Exim to panic-die when
5434 the list was checked. (An example that provoked this was putting <; in the
5435 middle of a list instead of at the start.) If this happened during a DATA
5436 ACL check, a -D file could be left lying around. This kind of configuration
5437 error no longer causes Exim to die; instead it causes a defer errror. The
5438 incident is still logged to the main and panic logs.
5440 74. Buglet left over from Exim 3 conversion. The message "too many messages
5441 in one connection" was written to the rejectlog but not the mainlog, except
5442 when address rewriting (yes!) was being logged.
5444 75. Added write_rejectlog option.
5446 76. When a system filter was run not as root (that is, when system_filter_user
5447 was set), the values of the $n variables were not being returned to the
5448 main process; thus, they were not subsequently available in the $sn
5451 77. Added +return_path_on_delivery log selector.
5453 78. A connection timeout was being treated differently from recipients deferred
5454 when testing hosts_max_try with a message that was older than the host's
5455 retry timeout. (The host should not be counted, thus allowing all hosts to
5456 be tried at least once before bouncing.) This may have been the cause of an
5457 occasionally reported bug whereby a message would remain on the queue
5458 longer than the retry timeout, but would be bounced if a delivery was
5459 forced. I say "may" because I never totally pinned down the problem;
5460 setting up timeout/retry tests is difficult. See also the next item.
5462 79. The ultimate address timeout was not being applied to errors that involved
5463 a combination of host plus message (for example, a timeout on a MAIL
5464 command). When an address resolved to a number of possible hosts, and they
5465 were not all tried for each delivery (e.g. because of hosts_max_try), a
5466 message could remain on the queue longer than the retry timeout.
5468 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5471 81. Fixed an obscure SMTP outgoing bug which required at least the following
5472 conditions: (a) there was another message waiting for the same server;
5473 (b) the server returned 5xx to all RCPT commands in the first message so
5474 that the message was not completed; (c) the server dropped the connection
5475 or gave a negative response to the RSET that Exim sends to abort the
5476 transaction. The observed case was a dropped connection after DATA that had
5477 been sent in pipelining mode. That is, the server had advertised PIPELINING
5478 but was not implementing it correctly. The effect of the bug was incorrect
5479 behaviour, such as trying another host, and this could lead to a crash.
5485 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5486 and daemon.c were passed as pointers to ints; they should have been
5487 pointers to socklen_t variables (which are typically unsigned ints).
5489 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5492 3. Fixed a really odd bug that affected only the testing scheme; patching a
5493 certain fixed string in the binary changed the value of another string that
5494 happened to be identical to the end of the original first string.
5496 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5497 name", it returns that address as the IP address. On some operating
5498 systems (e.g. Solaris), it also passes back the IP address string as the
5499 "host name". However, on others (e.g. Linux), it passes back an empty
5500 string. Exim wasn't checking for this, and was changing the host name to an
5501 empty string, assuming it had been canonicized.
5503 5. Although rare, it is permitted to have more than one PTR record for a given
5504 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5505 all the names associated with an address, because they do in Solaris.
5506 However, it seems that they do not in Linux for data that comes from the
5507 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5508 I found this out when I moved to a new Linux workstation and tried to run
5509 the Exim test suite.
5511 To get round this problem I have changed the code so that it now does its
5512 own call to the DNS to look up PTR records when searching for a host name.
5513 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5514 addresses that are only in /etc/hosts are still found.
5516 This behaviour is, however, controlled by an option called host_lookup_
5517 order, which defaults to "bydns:byaddr". If people want to use the other
5518 order, or indeed, just use one or the other means of lookup, they can
5519 specify it in this variable.
5521 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5522 some operating systems, this comes back from gethostbyaddr() as an empty
5523 string, and this is what Exim used to test for. However, it seems that in
5524 other systems, "." is yielded. Exim now tests for this case too.
5526 7. The values of check_spool_space and check_log_space are now held internally
5527 as a number of kilobytes instead of an absolute number of bytes. If a
5528 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5529 kilobyte. This means that much larger values can be stored.
5531 8. Exim monitor: an attempt to get the action menu when not actually pointing
5532 at a message produces an empty menu entitled "No message selected". This
5533 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5534 no entries in it ("Shell widget menu has zero width and/or height"). So I
5535 have added a single, blank menu entry in this case.
5537 9. Added ${quote_local_part.
5539 10. MIME decoding is now applied to the contents of Subject: header lines when
5542 11. Now that a reference to $sender_host_address automatically causes a reverse
5543 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5544 host lookup before query-style lookups in lists that might use this
5545 variable. This has therefore been abolished, and the "net-" prefix is no
5546 longer necessary for query-style lookups.
5548 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5549 have been a typo for LFLAGS, so it has been changed.
5551 13. The install script calls Exim with "-C /dev/null" in order to find the
5552 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5553 to be output. Howeve, since Exim outputs its version number before the
5554 error, it didn't break the script. It just looked ugly. I fixed this by
5555 always allowing "-C /dev/null" if the caller is root.
5557 14. Ignore overlarge ACL variable number when reading spool file - insurance
5558 against a later release with more variables having written the file.
5560 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5561 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5562 this, as well as the form without the "IPv6" on the front.
5564 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5565 OS/Makefile-Darwin file.
5567 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5568 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5570 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5571 variable. However, some OS use ints for inodes. I've added cast to long int
5572 to get rid of the compiler warning.
5574 19. I had forgotten to lock out "/../" in configuration file names when
5575 ALT_CONFIG_PREFIX was set.
5577 20. Routers used for verification do not need to specify transports. However,
5578 if such a router generated a host list, and callout was configured, Exim
5579 crashed, because it could not find a port number from the (non-existent)
5580 transport. It now assumes port 25 in this circumstance.
5582 21. Added the -t option to exigrep.
5584 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5585 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5586 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5587 them is set, LOOKUP_LSEARCH is forced.
5589 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5591 24. Added sender and host information to the "rejected by local_scan()" log
5592 line; previously there was no indication of these.
5594 25. Added .include_if_exists.
5596 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5597 Linux. It turns out that not all file systems support this. Apparently some
5598 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5599 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5600 not supported on the file descriptor, is now ignored when Exim is trying to
5601 sync a directory. This applies only to Linux.
5603 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5605 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5606 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5609 29. Added support for maildirsize files from supplied patch (modified a bit).
5611 30. The use of :fail: followed by an empty string could lead Exim to respond to
5612 sender verification failures with (e.g.):
5614 550 Verification failed for <xxx>
5615 550 Sender verify failed
5617 where the first response line was missing the '-' that indicates it is not
5618 the final line of the response.
5620 31. The loop for finding the name of the user that called Exim had a hardwired
5621 limit of 10; it now uses the value of finduser_retries, which is used for
5622 all other user lookups.
5624 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5627 33. Exim was neglecting to zero errno before one call of strtol() when
5628 expanding a string and expecting an integer value. On some systems this
5629 resulted in spurious "integer overflow" errors. Also, it was casting the
5630 result into an int without checking.
5632 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5633 did not work. The code looks as if it has *never* worked, though it appears
5634 to have been documented since at least releast 1.62. I have made it work.
5636 35. The "timeout_DNS" error in retry rules, also documented since at least
5637 1.62, also never worked. As it isn't clear exactly what this means, and
5638 clearly it isn't a major issue, I have abolished the feature by treating it
5639 as "timeout", and writing a warning to the main and panic logs.
5641 36. The display of retry rules for -brt wasn't always showing the error code
5644 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5645 timeout_connect_A, timeout_connect_MX.
5647 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5648 to the empty sender.
5650 39. The daemon was not analysing the content of -oX till after it had closed
5651 stderr and disconnected from the controlling terminal. This meant that any
5652 syntax errors were only noted on the panic log, and the return code from
5653 the command was 0. By re-arranging the code a little, I've made the
5654 decoding happen first, so such errors now appear on stderr, and the return
5655 code is 1. However, the actual setting up of the sockets still happens in
5656 the disconnected process, so errors there are still only recorded on the
5659 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5660 connections (as happens on some IP stacks) was logged at start up time as
5661 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5662 it from "IPv6 and IPv4", which means that two separate sockets are being
5665 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5666 says whether AF_INET or AF_INET6 was passed as an argument.
5668 42. Exiwhat output was messed up when time zones were included in log
5671 43. Exiwhat now gives more information about the daemon's listening ports,
5672 and whether -tls-on-connect was used.
5674 44. The "port" option of the smtp transport is now expanded.
5676 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5677 silently ignored. Now an error message is written to the main and panic
5680 46. There's a new ACL modifier called "logwrite" which writes to a log file
5681 as soon as it is encountered.
5683 47. Added $local_user_uid and $local_user_gid at routing time.
5685 48. Exim crashed when trying to verify a sender address that was being
5688 49. Exim was recognizing only a space character after ".include". It now also
5689 recognizes a tab character.
5691 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5692 extracting the relevant information from the specification. The man page no
5693 longer contains scrambled data for the -d option, and I've added a section
5694 at the front about calling Exim under different names.
5696 51. Added "extra_headers" argument to the "mail" command in filter files.
5698 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5701 53. Installed eximstats 1.29.
5703 54. Added transport_filter_timeout as a generic transport option.
5705 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5706 Cc: header lines. This was required by RFC 822, but it not required by RFC
5709 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5710 incoming messages that did not have them. Now it does so only if the
5711 message originates locally, that is, if there is no associated remote host
5712 address. When Resent- header lines are present, this applies to the Resent-
5713 lines rather than the non-Resent- lines.
5715 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5716 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5718 58. Messages for configuration errors now include the name of the main
5719 configuration file - useful now that there may be more than one file in a
5720 list (.included file names were always shown).
5722 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5723 for those rare installations that do not start the daemon as root or run it
5724 setuid root. I've cut out the call to initgroups() if the daemon is not
5727 60. The Exim user and group can now be bound into the binary as text strings
5728 that are looked up at the start of Exim's processing.
5730 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5732 62. Added $mailstore_basename variable.
5734 63. Installed patch to sieve.c from Michael Haardt.
5736 64. When Exim failed to open the panic log after failing to open the main log,
5737 the original message it was trying to log was written to stderr and debug
5738 output, but if they were not available (the usual case in production), it
5739 was lost. Now it is written to syslog before the two lines that record the
5740 failures to open the logs.
5742 65. Users' Exim filters run in subprocesses under the user's uid. It is
5743 possible for a "deliver" command or an alias in a "personal" command to
5744 provoke an address rewrite. If logging of address rewriting is configured,
5745 this fails because the process is not running as root or exim. There may be
5746 a better way of dealing with this, but for the moment (because 4.30 needs
5747 to be released), I have disabled address rewrite logging when running a
5748 filter in a non-root, non-exim process.
5754 1. The buildconfig auxiliary program wasn't quoting the value set for
5755 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5756 not defined. This bug was masked in 4.22 by the effect that was fixed in
5759 2. Some messages that were rejected after a message id was allocated were
5760 shown as "incomplete" by exigrep. It no longer does this for messages that
5761 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5763 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5764 have allow_domain_literals set, the ID did not get logged in the <= line.
5765 Domain literals are now always recognized in Message-ID: header lines.
5767 4. The first argument for a ${extract expansion item is the key name or field
5768 number. Leading and trailing spaces in this item were not being ignored,
5769 causing some misleading effects.
5771 5. When deliver_drop_privilege was set, single queue runner processes started
5772 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5773 same command in the process it spins off) were not dropping privilege.
5775 6. When the daemon running as "exim" started a queue runner, it always
5776 re-executed Exim in the spun-off process. This is a waste of effort when
5777 deliver_drop_privilege is set. The new process now just calls the
5778 queue-runner function directly.
5784 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5787 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5788 ignored. Though the use of -oP was forcing the writing of a pid file, it
5789 was always written to the default place.
5791 3. If the message "no IP address found for host xxxx" is generated during
5792 incoming verification, it is now followed by identification of the incoming
5793 connection (so you can more easily find what provoked it).
5795 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5797 5. Added some features to "harden" Exim a bit more against certain attacks:
5799 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5800 be put in Local/Makefile. This is like the never_users runtime option,
5801 but it cannot be overridden. The default setting is "root".
5803 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5804 prefix string with which any file named in a -C command line option
5807 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5808 is retained for -C and -D only if the caller of Exim is root. Without
5809 it, the exim user may also use -C and -D and retain privilege.
5811 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5812 command line option is disabled.
5814 6. Macro names set by the -D option must start with an upper case letter, just
5815 like macro names defined in the configuration file.
5817 7. Added "dereference=" facility to LDAP.
5819 8. Two instances of the typo "uknown" in the source files are fixed.
5821 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5822 the Configure-Makefile script screwed up while processing it.
5824 10. Incorporated PCRE 4.4.
5826 11. The SMTP synchronization check was not operating right at the start of an
5827 SMTP session. For example, it could not catch a HELO sent before the client
5828 waited for the greeting. There is now a check for outstanding input at the
5829 point when the greeting is written. Because of the duplex, asynchronous
5830 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5831 way, but not yet received, when the check is performed.
5833 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5834 on TCP/IP sockets, because this apparently causes some broken clients to
5837 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5838 unchanged) from the Cygwin maintainer.
5840 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5841 was supported instead of testing for mbx. Effectively a typo.
5843 15. The spa authenticator server code was not checking that the input it
5844 received was valid base64.
5846 16. The debug output line for the "set" modifier in ACLs was not showing the
5847 name of the variable that was being set.
5849 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5851 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5852 lookup on its own. Something else was needed to trigger the lookup. For
5853 example, a match in host_lookup or the need for a host name in a host list.
5854 Now, if $sender_host_name is referenced and the host name has not yet been
5855 looked up, a lookup is performed. If the lookup fails, the variable remains
5856 empty, and $host_lookup_failed is set to "1".
5858 19. Added "eqi" as a case-independent comparison operator.
5860 20. The saslauthd authentication condition could segfault if neither service
5861 nor realm was specified.
5863 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5864 error message that was logged was misleading, and incoming SMTP
5865 connections were dropped. The message is now more accurate, and temporary
5866 errors are given to SMTP connections.
5868 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5869 (except RSET) with a 421 error, until QUIT is received. However, it was
5870 failing to send a response to QUIT.
5872 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5873 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5874 ACL is now run after the helo_try_verify_hosts code.
5876 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5877 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5878 case-independent, so other case variants are also recognized). Apparently
5879 some systems use these upper case variants.
5881 25. If more than two messages were waiting for the same host, and a transport
5882 filter was specified for the transport, Exim sent two messages over the
5883 same TCP/IP connection, and then failed with "socket operation on non-
5884 socket" when it tried to send the third.
5886 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5888 27. The extern definition of crypt16() in expand.c was not being excluded when
5889 the OS had its own crypt16() function.
5891 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5892 as a preferred synonym for return_size_limit, both as an option and as an
5895 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5897 30. Changed the default configuration ACL to relax the local part checking rule
5898 for addresses that are not in any local domains. For these addresses,
5899 slashes and pipe symbols are allowed within local parts, but the sequence
5900 /../ is explicitly forbidden.
5902 31. SPA server authentication was not clearing the challenge buffer before
5905 32. log_message in a "warn" ACL statement was writing to the reject log as
5906 well as to the main log, which contradicts the documentation and doesn't
5907 seem right (because no rejection is happening). So I have stopped it.
5909 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5910 However, I am unable to do any testing of this.
5912 34. Fixed an infelicity in the appendfile transport. When checking directories
5913 for a mailbox, to see if any needed to be created, it was accidentally
5914 using path names with one or more superfluous leading slashes; tracing
5915 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5917 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5918 contents are added to the log line that is written for every discarded
5919 recipient. (Previously a log_message setting was ignored.)
5921 36. The ${quote: operator now quotes the string if it is empty.
5923 37. The install script runs exim in order to find its version number. If for
5924 some reason other than non-existence or emptiness, which it checks, it
5925 could not run './exim', it was installing it with an empty version number,
5926 i.e. as "exim-". This error state is now caught, and the installation is
5929 38. An argument was missing from the function that creates an error message
5930 when Exim fails to connect to the socket for saslauthd authentication.
5931 This could cause Exim to crash, or give a corrupted message.
5933 39. Added isip, isip4, and isip6 to ${if conditions.
5935 40. The ACL variables $acl_xx are now saved with the message, and can be
5936 accessed later in routers, transports, and filters.
5938 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5939 strings in the file are not string-expanded.
5941 42. If a MAIL command specified a SIZE value that was too large to fit into an
5942 int variable, the check against message_size_limit failed. Such values are
5943 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5944 day this will have to be increased, but I don't think I want to be around
5945 when emails are that large.
5952 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5953 iconv() is not standard in FreeBSD.
5955 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5956 IPv6 enabled. The observed symptom was a segmentation fault on return from
5957 the function os_common_find_running_interfaces() in src/os.c.
5959 3. In the check_special_case() function in daemon.c I had used "errno" as an
5960 argument name, which causes warnings on some systems. This was basically a
5961 typo, since it was named "eno" in the comments!
5963 4. The code that waits for the clock to tick (at a resolution of some fraction
5964 of a second) so as to ensure message-id uniqueness was always waiting for
5965 at least one whole tick, when it could have waited for less. [This is
5966 almost certainly not relevant at current processor speeds, where it is
5967 unlikely to ever wait at all. But we try to future-proof.]
5969 5. The function that sleeps for a time interval that includes fractions of a
5970 second contained a race. It did not block SIGALRM between setting the
5971 timer, and suspending (a couple of lines later). If the interval was short
5972 and the sigsuspend() was delayed until after it had expired, the suspension
5973 never ended. On busy systems this could lead to processes getting stuck for
5976 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5977 process, before it forks any delivery processes. The open lookup caching
5978 mechanism meant that the open file or database connection was passed into
5979 the delivery process. The problem was that delivery processes always tidy
5980 up cached lookup data. This could cause a problem for the next delivery
5981 process started by the queue runner, because the external queue runner
5982 process does not know about the closure. So the next delivery process
5983 still has data in the lookup cache. In the case of a file lookup, there was
5984 no problem because closing a file descriptor in a subprocess doesn't affect
5985 the parent. However, if the lookup was caching a connection to a database,
5986 the connection was closed, and the second delivery process was likely to
5987 see errors such as "PGSQL: query failed: server closed the connection
5988 unexpectedly". The problem has been fixed by closing all cached lookups
5989 in a queue runner before running a delivery process.
5991 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5992 seem to have the "const" qualifier which it has on other OS. I've
5995 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5996 *of the same type* (client or server) with the same public name that an
5997 error should be diagnosed.
5999 9. When Exim looked up a host name for an IP address, but failed to find the
6000 original IP address when looking up the host name (a safety check), it
6001 output the message "<ip address> does not match any IP for NULL", which was
6002 confusing, to say the least. The bug was that the host name should have
6003 appeared instead of "NULL".
6005 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6006 user that is built into the binary, and the -C or -D options is used, root
6007 privilege is dropped before the configuration file is read. In addition,
6008 logging is switched to stderr instead of the normal log files. If the
6009 configuration then re-defines the Exim user, the unprivileged environment
6010 is probably not what is expected, so Exim logs a panic warning message (but
6013 However, if deliver_drop_privilege is set, the unprivileged state may well
6014 be exactly what is intended, so the warning has been cut out in that case,
6015 and Exim is allowed to try to write to its normal log files.
6021 1. smtp_return_error_details was not giving details for temporary sender
6022 or receiver verification errors.
6024 2. Diagnose a configuration error if two authenticators have the same public
6027 3. Exim used not to create the message log file for a message until the first
6028 delivery attempt. This could be confusing when incoming messages were held
6029 for policy or load reasons. The message log file is now created at the time
6030 the message is received, and an initial "Received" line is written to it.
6032 4. The automatically generated man page for command line options had a minor
6033 bug that caused no ill effects; however, a more serious problem was that
6034 the procedure for building the man page automatically didn't always
6035 operate. Consequently, release 4.20 contains an out-of-date version. This
6036 shouldn't happen again.
6038 5. When building Exim with embedded Perl support, the script that builds the
6039 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6040 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6042 6. The freeze_tell option was not being used for messages that were frozen on
6043 arrival, either by an ACL or by local_scan().
6045 7. Added the smtp_incomplete_transaction log selector.
6047 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6048 was accepting AUTH without a new EHLO.
6050 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6051 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6054 10. Logging of TCP/IP connections (when configured) now happens in the main
6055 daemon process instead of the child process, so that the TCP/IP connection
6056 count is more accurate (but it can never be perfect).
6058 11. The use of "drop" in a nested ACL was not being handled correctly in the
6059 outer ACL. Now, if condition failure induced by the nested "drop" causes
6060 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6061 or "require"), the connection is dropped.
6063 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6064 that yield "discard" can now be used with an "accept" or a "discard" verb,
6065 but an error is generated for any others (because I can't see a useful way
6066 to define what should happen).
6068 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6069 are now processed in the same way as lines in the Exim configuration file.
6070 In particular, continuation lines are supported.
6072 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6074 15. Added -ti meaning -t -i.
6076 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6077 domains, and warn by logging if others are found.
6079 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6080 returned from GIFCONF when Exim is trying to find the list of interfaces on
6081 a host. The code in os.c has been modified to copy each ifreq to an aligned
6082 structure in all cases.
6084 Also, in some cases, the returned ifreq's were being copied to a 'struct
6085 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6086 means the last couple of bytes of an IPv6 address could be chopped if the
6087 ifreq contained only a normal sockaddr (14 bytes storage).
6089 18. Named domain lists were not supported in the hosts_treat_as_local option.
6090 An entry such as +xxxx was not recognized, and was treated as a literal
6093 19. Ensure that header lines added by a DATA ACL are included in the reject log
6094 if the ACL subsequently rejects the message.
6096 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6097 MD5 (which is deprecated).
6099 21. When testing a filter file using -bf, Exim was writing a message when it
6100 took the sender from a "From " line in the message, but it was not doing so
6101 when it took $return_path from a Return-Path: header line. It now does.
6103 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6104 with a valid header line field name (a series of printing characters
6105 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6107 23. Changed "disc" in the source to "disk" to conform to the documentation and
6108 the book and for uniformity.
6110 24. Ignore Sendmail's -Ooption=value command line item.
6112 25. When execve() failed while trying to run a command in a pipe transport,
6113 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6114 could be confused with a return value of 69 from the command itself. This
6115 has been changed to 127, the value the shell returns if it is asked to run
6116 a non-existent command. The wording for the related log line suggests a
6117 non-existent command as the problem.
6119 26. If received_header_text expands to an empty string, do not add a Received:
6120 header line to the message. (Well, it adds a token one on the spool, but
6121 marks it "old" so that it doesn't get used or transmitted.)
6123 27. Installed eximstats 1.28 (addition of -nt option).
6125 28. There was no check for failure on the call to getsockname() in the daemon
6126 code. This can fail if there is a shortage of resources on the system, with
6127 ENOMEM, for example. A temporary error is now given on failure.
6129 29. Contrary to the C standard, it seems that in some environments, the
6130 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6131 program. Exim now does this explicitly; it affects the formatting of
6132 timestamps using strftime().
6134 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6135 complaints. I've now initialized all the variables, to avoid this.
6137 32. Header lines added by a system filter were not being "seen" during
6138 transport-time rewrites.
6140 33. The info_callback() function passed to OpenSSL is set up with type void
6141 (*)(SSL *, int, int), as described somewhere. However, when calling the
6142 function (actually a macro) that sets it up, the type void(*)() is
6143 expected. I've put in a cast to prevent warnings from picky compilers.
6145 34. If a DNS black list lookup found a CNAME record, but there were no A
6146 records associated with the domain it pointed at, Exim crashed.
6148 35. If a DNS black list lookup returned more than one A record, Exim ignored
6149 all but the first. It now scans all returned addresses if a particular IP
6150 value is being sought. In this situation, the contents of the
6151 $dnslist_value variable are a list of all the addresses, separated by a
6154 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6155 to accept a host name and all its aliases if the forward lookup for any of
6156 them yielded the IP address of the incoming connection. Now it accepts only
6157 those names whose forward lookup yields the correct IP address. Any other
6158 names are discarded. This closes a loophole whereby a rogue DNS
6159 administrator could create reverse DNS records to break through a
6160 wildcarded host restriction in an ACL.
6162 37. If a user filter or a system filter that ran in a subprocess used any of
6163 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6164 the wrong values were passed to the pipe command ($thisaddress had the
6165 value of $0, $0 had the value of $1, etc). This bug was introduced by
6166 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6168 38. Improved the line breaking for long SMTP error messages from ACLs.
6169 Previously, if there was no break point between 40 and 75 characters, Exim
6170 left the rest of the message alone. Two changes have been made: (a) I've
6171 reduced the minimum length to 35 characters; (b) if it can't find a break
6172 point between 35 and 75 characters, it looks ahead and uses the first one
6173 that it finds. This may give the occasional overlong line, but at least the
6174 remaining text gets split now.
6176 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6177 file descriptors might be low, and that setting 1000 would always raise it.
6178 It turns out that in some environments, the limit is already over 1000 and
6179 that lowering it causes trouble. So now Exim takes care not to decrease it.
6181 40. When delivering a message, the value of $return_path is set to $sender_
6182 address at the start of routing (routers may change the value). By an
6183 oversight, this default was not being set up when an address was tested by
6184 -bt or -bv, which affected the outcome if any router or filter referred to
6187 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6188 log only when "message" or "log_message" are set. However, if one of the
6189 conditions was an address verification, or a call to a nested ACL, the
6190 messages generated by the underlying test were being passed through. This
6191 no longer happens. The underlying message is available in $acl_verify_
6192 message for both "message" and "log_message" expansions, so it can be
6193 passed through if needed.
6195 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6196 new expansion $bh_ to give the encoded byte string without charset
6197 translation. Translation happens only if iconv() is available; HAVE_ICONV
6198 indicates this at build time. HEADERS_CHARSET gives the charset to
6199 translate to; headers_charset can change it in the configuration, and
6200 "headers charset" can change it in an individual filter file.
6202 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6203 that creates RFC 2047 encoded "words" labels them as that charset instead
6204 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6205 expansion operator; (ii) when Exim creates a From: line for a local
6206 message; (iii) when a header line is rewritten to include a "phrase" part.
6208 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6209 buggy, causing it to skip the first lines of messages whose message ID
6210 ended in 'D'. This would not have bitten before Exim release 4.14, because
6211 message IDs were unlikely to end in 'D' before then. The effect was to have
6212 incorrect size information for certain domains.
6214 45. #include "config.h" was missing at the start of the crypt16.c module. This
6215 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6218 46. If there was a timeout during a "random" callout check, Exim treated it as
6219 a failure of the random address, and carried on sending RSET and the real
6220 address. If the delay was just some slowness somewhere, the response to the
6221 original RCPT would be taken as a response to RSET and so on, causing
6222 mayhem of various kinds.
6224 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6225 when I implemented it. It didn't allow for the fact that some option values
6226 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6227 the right test for positive values.
6229 48. Domain names in DNS records are case-independent. Exim always looks them up
6230 in lower case. Some resolvers return domain names in exactly the case they
6231 appear in the zone file, that is, they may contain uppercase letters. Not
6232 all resolvers do this - some return always lower case. Exim was treating a
6233 change of case by a resolver as a change of domain, similar to a widening
6234 of a domain abbreviation. This triggered its re-routing code and so it was
6235 trying to route what was effectively the same domain again. This normally
6236 caused routing to fail (because the router wouldn't handle the domain
6237 twice). Now Exim checks for this case specially, and just changes the
6238 casing of the domain that it ultimately uses when it transmits the message
6241 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6244 50. If a filter generated a file delivery with a non-absolute name (possible if
6245 no home directory exists for the router), the forbid_file option was not
6248 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6249 the existing equality matching.
6251 52. Exim was using ints instead of ino_t variables in some places where it was
6252 dealing with inode numbers.
6254 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6255 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6256 called TMPDIR, and if it finds it is different, it changes its value.
6258 54. The smtp_printf() function is now made available to local_scan() so
6259 additional output lines can be written before returning. There is also an
6260 smtp_fflush() function to enable the detection of a dropped connection.
6261 The variables smtp_input and smtp_batched_input are exported to
6264 55. Changed the default runtime configuration: the message "Unknown user"
6265 has been removed from the ACL, and instead placed on the localuser router,
6266 using the cannot_route_message feature. This means that any verification
6267 failures that generate their own messages won't get overridden. Similarly,
6268 the "Unrouteable address" message that was in the ACL for unverifiable
6269 relay addresses has also been removed.
6271 56. Added hosts_avoid_esmtp to the smtp transport.
6273 57. The exicyclog script was not checking for the esoteric option
6274 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6275 will work only if exicyclog is run under the appropriate euid.
6277 58. Following a discussion on the list, the rules by which Exim recognises line
6278 endings on incoming messages have been changed. The -dropcr and drop_cr
6279 options are now no-ops, retained only for backwards compatibility. The
6280 following line terminators are recognized: LF CRLF CR. However, special
6281 processing applies to CR:
6283 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6284 nor a local message in the state where . is a terminator.
6286 (ii) If a bare CR is encountered in a header line, an extra space is added
6287 after the line terminator so as not to end the header. The reasoning
6288 behind this is that bare CRs in header lines are most likely either
6289 to be mistakes, or people trying to play silly games.
6291 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6292 was being incorrectly given as 18 bytes larger than it should have been.
6293 This is a VOB (very old bug).
6295 60. This may never have affected anything current, but just in case it has:
6296 When the local host is found other than at the start of a list of hosts,
6297 the local host, those with the same MX, and any that follow, are discarded.
6298 When the list in question was part of a longer list of hosts, the following
6299 hosts (not currently being processed) were also being discarded. This no
6300 longer happens. I'm not sure if this situation could ever has previously
6303 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6306 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6307 header. This is something that is recommended in a new Internet Draft, and
6308 is something that is documented as being done by Sendmail. There are two
6309 possible values. For messages generated by the autoreply transport, Exim
6312 Auto-Submitted: auto-replied
6314 whereas for all other generated messages (e.g. bounces) it adds
6316 Auto-Submitted: auto-generated
6318 63. The "personal" condition in filters now includes a test for the
6319 Auto-Submitted: header. If it contains the string "auto-" the message it
6320 not considered personal.
6322 64. Added rcpt_include_affixes as a generic transport option.
6324 65. Added queue_only_override (default true).
6326 66. Added the syslog_duplication option.
6328 67. If what should have been the first header line of a message consisted of
6329 a space followed by a colon, Exim was mis-interpreting it as a header line.
6330 It isn't of course - it is syntactically invalid and should therefore be
6331 treated as the start of the message body. The misbehaviour could have
6332 caused a number of strange effects, including loss of data in subsequent
6333 header lines, and spool format errors.
6335 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6336 client host had authenticated. This control can now be exercised by an ACL
6337 for more flexibility.
6339 69. By default, callouts do not happen when testing with -bh. There is now a
6340 variant, -bhc, which does actually run the callout code, including
6341 consulting and updating the callout cache.
6343 70. Added support for saslauthd authentication, courtesy of Alexander
6346 71. If statvfs() failed on the spool or log directories while checking their
6347 size for availability, Exim confusingly gave the error "space shortage".
6348 Furthermore, in debugging mode it crashed with a floating point exception.
6349 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6350 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6351 serious problem, Exim now writes to the main and panic logs when this
6352 happens, with details of the failure. It then refuses to accept the
6353 incoming message, giving the message "spool directory problem" or "log
6354 directory problem" with a 421 code for SMTP messages.
6356 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6357 0, 1, and 2 exist, because some OS complain for execs without them (see
6358 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6359 descriptors. However, the code omitted to check that the open succeeded,
6360 causing mysterious errors if for some reason the permissions on /dev/null
6361 got screwed. Now Exim writes a message to the main and panic logs, and
6362 bombs out if it can't open /dev/null.
6364 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6365 interact so that it is all more flexible. It is supposed to remain
6366 backwards compatible. Also added extra_local_interfaces.
6368 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6369 to bomb out with an assertion failure - to the client this appears as a
6370 connection drop. This problem occurs in the part of the code that was taken
6371 from the Samba project. Fortunately, the assertion is in a very simple
6372 function, so I have fixed this by reproducing the function inline in the
6373 one place where it is called, and arranging for authentication to fail
6374 instead of killing the process with assert().
6376 75. The SPA client code was not working when the server requested OEM rather
6377 than Unicode encoding.
6379 76. Added code to make require_files with a specific uid setting more usable in
6380 the case where statting the file as root fails - usually a non-root-mounted
6381 NFS file system. When this happens and the failure is EACCES, Exim now
6382 forks a subprocess and does the per-uid checking as the relevant uid.
6384 77. Added process_log_path.
6386 78. If log_file_path was not explicitly set, a setting of check_log_space or
6387 check_log_inodes was ignored.
6389 79. If a space check for the spool or log partitions fails, the incident is now
6390 logged. Of course, in the latter case the data may get lost...
6392 80. Added the %p formatting code to string_format() so that it can be used to
6393 print addresses in debug_print(). Adjusted all the address printing in the
6394 debugging in store.c to use %p rather than %d.
6396 81. There was a concern that a line of code in smtp_in.c could overflow a
6397 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6398 initially expressed, the concern was not well-founded, because trailing
6399 spaces are removed early. However, if the trailing spaces were followed by
6400 a NULL, they did not get removed, so the overflow was possible. Two fixes
6403 (a) I re-wrote the offending code in a cleaner fashion.
6404 (b) If an incoming SMTP command contains a NULL character, it is rejected
6407 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6408 runs initgroups(), so that if the Exim user is in any additional groups,
6409 they will be used during message reception.
6415 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob