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.
55 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
58 JH/02 The smtp transport option "multi_domain" is now expanded.
60 JH/03 The smtp transport now requests PRDR by default, if the server offers
63 JH/04 Certificate name checking on server certificates, when exim is a client,
64 is now done by default. The transport option tls_verify_cert_hostnames
65 can be used to disable this per-host. The build option
66 EXPERIMENTAL_CERTNAMES is withdrawn.
68 JH/05 The value of the tls_verify_certificates smtp transport and main options
69 default to the word "system" to access the system default CA bundle.
70 For GnuTLS, only version 3.0.20 or later.
72 JH/06 Verification of the server certificate for a TLS connection is now tried
73 (but not required) by default. The verification status is now logged by
74 default, for both outbound TLS and client-certificate supplying inbound
77 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
80 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
81 Status Notification (bounce) messages are now MIME format per RFC 3464.
82 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
83 under the control of the dsn_advertise_hosts option, and routers may
84 have a dsn_lasthop option.
86 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
87 default, modifiable by a malware= option. The list separator for
88 the options can now be changed in the usual way. Bug 68.
90 JH/10 The smtp_receive_timeout main option is now expanded before use.
92 JH/11 The incoming_interface log option now also enables logging of the
93 local interface on delivery outgoing connections.
95 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
96 if the interface and destination host and port all match.
98 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
101 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
102 Patch from Andrew Lewis.
104 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
105 now supports optional time-restrictions, weighting, and priority
106 modifiers per server. Patch originally by <rommer@active.by>.
108 JH/16 The spamd_address main option now supports a mixed list of local
109 and remote servers. Remote servers can be IPv6 addresses, and
110 specify a port-range.
112 JH/17 Bug 68: The spamd_address main option now supports an optional
113 timeout value per server.
115 JH/18 Bug 1581: Router and transport options headers_add/remove can
116 now have the list separator specified.
118 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
121 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
124 JH/21 Support for the A6 type of dns record is withdrawn.
126 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
127 rather than the verbs used.
129 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
130 from 255 to 1024 chars.
132 JH/24 Verification callouts now attempt to use TLS by default.
134 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
135 are generic router options now. The defaults didn't change.
137 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
138 Original patch from Alexander Shikoff, worked over by JH.
140 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
141 files if ZCAT_COMMAND is not executable.
143 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
145 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
147 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
148 Normally benign, it bites when the pair was led to by a CNAME;
149 modern usage is to not canoicalize the domain to a CNAME target
150 (and we were inconsistent anyway for A-only vs AAAA+A).
152 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
154 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
155 when evaluating $sender_host_dnssec.
157 JH/31 Check the HELO verification lookup for DNSSEC, adding new
158 $sender_helo_dnssec variable.
160 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
162 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
164 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
166 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
167 documented as working, but never had. Support all but $spam_report.
169 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
170 added for tls authenticator.
175 TL/01 When running the test suite, the README says that variables such as
176 no_msglog_check are global and can be placed anywhere in a specific
177 test's script, however it was observed that placement needed to be near
178 the beginning for it to behave that way. Changed the runtest perl
179 script to read through the entire script once to detect and set these
180 variables, reset to the beginning of the script, and then run through
181 the script parsing/test process like normal.
183 TL/02 The BSD's have an arc4random API. One of the functions to induce
184 adding randomness was arc4random_stir(), but it has been removed in
185 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
186 function when detected.
188 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
189 cause callback expansion.
191 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
192 syntax errors in an expansion can be treated as a string instead of
193 logging or causing an error, due to the internal use of bool_lax
194 instead of bool when processing it.
196 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
197 server certificates when making smtp deliveries.
199 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
201 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
203 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
205 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
206 Merged patch from Sebastian Wiedenroth.
208 JH/05 Fix results-pipe from transport process. Several recipients, combined
209 with certificate use, exposed issues where response data items split
210 over buffer boundaries were not parsed properly. This eventually
211 resulted in duplicates being sent. This issue only became common enough
212 to notice due to the introduction of conection certificate information,
213 the item size being so much larger. Found and fixed by Wolfgang Breyha.
215 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
216 size buffer was used, resulting in syntax errors when an expansion
219 JH/07 Add support for directories of certificates when compiled with a GnuTLS
220 version 3.3.6 or later.
222 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
223 is EXPERIMENTAL_EVENT, the main-configuration and transport options
224 both become "event_action", the variables become $event_name, $event_data
225 and $event_defer_errno. There is a new variable $verify_mode, usable in
226 routers, transports and related events. The tls:cert event is now also
227 raised for inbound connections, if the main configuration event_action
230 TL/06 In test suite, disable OCSP for old versions of openssl which contained
231 early OCSP support, but no stapling (appears to be less than 1.0.0).
233 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
234 server certificate names available under the smtp transport option
235 "tls_verify_cert_hostname" now do not permit multi-component wildcard
238 JH/10 Time-related extraction expansions from certificates now use the main
239 option "timezone" setting for output formatting, and are consistent
240 between OpenSSL and GnuTLS compilations. Bug 1541.
242 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
243 encoded parameter in the incoming message. Bug 1558.
245 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
246 include certificate info, eximon was claiming there were spoolfile
249 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
251 JH/14 Log delivery-related information more consistently, using the sequence
252 "H=<name> [<ip>]" wherever possible.
254 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
255 are problematic for Debian distribution, omit them from the release
258 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
260 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
263 JH/17 Fix a null-indirection in certextract expansions when a nondefault
264 output list separator was used.
269 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
270 checkers that were complaining about end of non-void function with no
273 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
274 This was a regression intruduced in 4.83 by another bugfix.
276 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
278 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
279 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
285 TF/01 Correctly close the server side of TLS when forking for delivery.
287 When a message was received over SMTP with TLS, Exim failed to clear up
288 the incoming connection properly after forking off the child process to
289 deliver the message. In some situations the subsequent outgoing
290 delivery connection happened to have the same fd number as the incoming
291 connection previously had. Exim would try to use TLS and fail, logging
292 a "Bad file descriptor" error.
294 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
295 utilities have not been installed.
297 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
298 temporary space as the ACL may create new global variables.
300 TL/01 LDAP support uses per connection or global context settings, depending
301 upon the detected version of the libraries at build time.
303 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
304 to extract and use the src ip:port in logging and expansions as if it
305 were a direct connection from the outside internet. PPv2 support was
306 updated based on HAProxy spec change in May 2014.
308 JH/02 Add ${listextract {number}{list}{success}{fail}}.
310 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
311 Properly escape header and check for NULL return.
313 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
316 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
318 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
319 characters in header names, implemented as a verify condition.
320 Contributed by Michael Fischer v. Mollard.
322 TL/05 Rename SPF condition results err_perm and err_temp to standardized
323 results permerror and temperror. Previous values are deprecated but
324 still accepted. In a future release, err_perm and err_temp will be
325 completely removed, which will be a backward incompatibility if the
326 ACL tests for either of these two old results. Patch contributed by
327 user bes-internal on the mailing list.
329 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
331 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
332 selectors, in both main and reject logs.
334 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
337 JH/07 Add malware type "sock" for talking to simple daemon.
339 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
341 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
342 routers/transports under cutthrough routing.
344 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
345 numbers. Touch up "bool" conditional to keep the same definition.
347 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
349 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
351 JH/12 Expand items in router/transport headers_add or headers_remove lists
352 individually rather than the list as a whole. Bug 1452.
354 Required for reasonable handling of multiple headers_ options when
355 they may be empty; requires that headers_remove items with embedded
356 colons must have them doubled (or the list-separator changed).
358 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
359 view the policy declared in the DMARC record. Currently, $dmarc_status
360 is a combined value of both the record presence and the result of the
363 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
365 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
366 dnslookup router and the smtp transport (applying to the forward
369 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
370 of ldap servers used for a specific lookup. Patch provided by Heiko
373 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
374 New variable $lookup_dnssec_authenticated for observability.
376 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
377 Patch submitted by Lars Timman.
379 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
381 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
382 Requires trusted mode and valid format message id, aborts otherwise.
383 Patch contributed by Heiko Schlichting.
385 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
386 certextract with support for various fields. Bug 1358.
388 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
389 is requested by default, modifiable by smtp transport option
392 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
393 operate on certificate variables to give certificate fingerprints
394 Also new ${sha256:cert_variable}.
396 JH/23 The PRDR feature is moved from being Experimental into the mainline.
398 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
399 Christian Aistleitner.
401 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
403 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
404 file. Patch from Wolfgang Breyha.
406 JH/25 Expand the coverage of the delivery $host and $host_address to
407 client authenticators run in verify callout. Bug 1476.
409 JH/26 Port service names are now accepted for tls_on_connect_ports, to
410 align with daemon_smtp_ports. Bug 72.
412 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
413 support and error reporting did not work properly.
415 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
416 and is readable. Patch from Andrew Colin Kissa.
418 TL/14 Enhance documentation of ${run expansion and how it parses the
419 commandline after expansion, particularly in the case when an
420 unquoted variable expansion results in an empty value.
422 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
424 PP/02 Fix internal collision of T_APL on systems which support RFC3123
425 by renaming away from it. Addresses GH issue 15, reported by
428 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
430 TL/15 SECURITY: prevent double expansion in math comparison functions
431 (can expand unsanitized data). Not remotely exploitable.
438 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
440 PP/02 Make -n do something, by making it not do something.
441 When combined with -bP, the name of an option is not output.
443 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
446 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
447 $sender_host_name and config options to manage this, and basic check
450 PP/05 DSCP support for outbound connections and control modifier for inbound.
452 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
453 (Only plugin which currently uses this is kerberos4, which nobody should
454 be using, but we should make it available and other future plugins might
455 conceivably use it, even though it would break NAT; stuff *should* be
456 using channel bindings instead).
458 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
459 name; added for Sendmail compatibility; requires admin caller.
460 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
461 just ignore it); requires trusted caller.
462 Also parse but ignore: -Ac -Am -X<logfile>
465 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
467 TL/02 Add +smtp_confirmation as a default logging option.
469 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
470 Patch by Magnus Holmgren from 2007-02-20.
472 TL/04 Bugzilla 1281 - Spec typo.
473 Bugzilla 1283 - Spec typo.
474 Bugzilla 1290 - Spec grammar fixes.
476 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
478 TL/06 Add Experimental DMARC support using libopendmarc libraries.
480 TL/07 Fix an out of order global option causing a segfault. Reported to dev
481 mailing list by by Dmitry Isaikin.
483 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
485 JH/02 Support "G" suffix to numbers in ${if comparisons.
487 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
489 NM/01 Bugzilla 1197 - Spec typo
490 Bugzilla 1196 - Spec examples corrections
492 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
494 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
495 gnutls_enable_pkcs11, but renamed to more accurately indicate its
498 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
499 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
501 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
502 "acl {{name}{arg}...}", and optional args on acl condition
505 JH/05 Permit multiple router/transport headers_add/remove lines.
507 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
509 JH/07 Avoid using a waiting database for a single-message-only transport.
510 Performance patch from Paul Fisher. Bugzilla 1262.
512 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
515 JH/09 Add $headers_added variable, with content from use of ACL modifier
516 add_header (but not yet added to the message). Bugzilla 199.
518 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
519 Pulled from Bugzilla 817 by Wolfgang Breyha.
521 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
523 (nb: this is the same fix as in Exim 4.80.1)
525 JH/11 Add A= logging on delivery lines, and a client_set_id option on
528 JH/12 Add optional authenticated_sender logging to A= and a log_selector
531 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
533 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
534 advertise SMTP AUTH mechanism to us, instead of a generic
535 protocol violation error. Also, make Exim more robust to bad
536 data from the Dovecot auth socket.
538 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
540 When a queue runner is handling a message, Exim first routes the
541 recipient addresses, during which it prunes them based on the retry
542 hints database. After that it attempts to deliver the message to
543 any remaining recipients. It then updates the hints database using
546 So if a recipient address works intermittently, it can get repeatedly
547 deferred at routing time. The retry hints record remains fresh so the
548 address never reaches the final cutoff time.
550 This is a fairly common occurrence when a user is bumping up against
551 their storage quota. Exim had some logic in its local delivery code
552 to deal with this. However it did not apply to per-recipient defers
553 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
555 This change adds a proper retry rule check during routing so that the
556 final cutoff time is checked against the message's age. We only do
557 this check if there is an address retry record and there is not a
558 domain retry record; this implies that previous attempts to handle
559 the address had the retry_use_local_parts option turned on. We use
560 this as an approximation for the destination being like a local
561 delivery, as in LMTP.
563 I suspect this new check makes the old local delivery cutoff check
564 redundant, but I have not verified this so I left the code in place.
566 TF/02 Correct gecos expansion when From: is a prefix of the username.
568 Test 0254 submits a message to Exim with the header
572 When I ran the test suite under the user fanf2, Exim expanded
573 the header to contain my full name, whereas it should have added
574 a Resent-Sender: header. It erroneously treats any prefix of the
575 username as equal to the username.
577 This change corrects that bug.
579 GF/01 DCC debug and logging tidyup
580 Error conditions log to paniclog rather than rejectlog.
581 Debug lines prefixed by "DCC: " to remove any ambiguity.
583 TF/03 Avoid unnecessary rebuilds of lookup-related code.
585 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
586 Bug spotted by Jeremy Harris; was flawed since initial commit.
587 Would have resulted in OCSP responses post-SNI triggering an Exim
588 NULL dereference and crash.
590 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
592 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
593 Bug detection, analysis and fix by Samuel Thibault.
594 Bugzilla 1331, Debian bug #698092.
596 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
598 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
599 Server implementation by Todd Lyons, client by JH.
600 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
601 config variable "prdr_enable" controls whether the server
602 advertises the facility. If the client requests PRDR a new
603 acl_data_smtp_prdr ACL is called once for each recipient, after
604 the body content is received and before the acl_smtp_data ACL.
605 The client is controlled by bolth of: a hosts_try_prdr option
606 on the smtp transport, and the server advertisement.
607 Default client logging of deliveries and rejections involving
608 PRDR are flagged with the string "PRDR".
610 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
611 fclose(). Diagnosis by Todd Lyons.
613 PP/17 Update configure.default to handle IPv6 localhost better.
614 Patch by Alain Williams (plus minor tweaks).
617 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
618 This is now consistent with GnuTLS, and is now documented: the
619 previous undocumented portable approach to treating the option as
620 unset was to force an expansion failure. That still works, and
621 an empty string is now equivalent.
623 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
624 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
625 not performing validation itself.
627 PP/20 Added force_command boolean option to pipe transport.
628 Patch from Nick Koston, of cPanel Inc.
630 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
633 TF/04 Added udpsend ACL modifer and hexquote expansion operator
635 PP/21 Fix eximon continuous updating with timestamped log-files.
636 Broken in a format-string cleanup in 4.80, missed when I repaired the
637 other false fix of the same issue.
638 Report and fix from Heiko Schlichting.
641 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
642 Report from Prashanth Katuri.
644 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
645 It's SecureTransport, so affects any MacOS clients which use the
646 system-integrated TLS libraries, including email clients.
648 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
649 using a MIME ACL for non-SMTP local injection.
650 Report and assistance in diagnosis by Warren Baker.
652 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
654 JH/16 Fix comparisons for 64b. Bugzilla 1385.
656 TL/09 Add expansion variable $authenticated_fail_id to keep track of
657 last id that failed so it may be referenced in subsequent ACL's.
659 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
662 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
663 ldap library initialization, allowing self-signed CA's to be
664 used. Also properly sets require_cert option later in code by
665 using NULL (global ldap config) instead of ldap handle (per
666 session). Bug diagnosis and testing by alxgomz.
668 TL/12 Enhanced documentation in the ratelimit.pl script provided in
669 the src/util/ subdirectory.
671 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
672 renamed to Transport Post Delivery Action by Jeremy Harris, as
675 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
676 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
677 redis_servers = needs to be configured which will be used by the redis
678 lookup. Patch from Warren Baker, of The Packet Hub.
680 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
682 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
683 hostname or reverse DNS when processing a host list. Used suggestions
684 from multiple comments on this bug.
686 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
688 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
689 Missed a few lines, added it to make the runtest require no keyboard
692 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
693 contains upper case chars. Make router use caseful_local_part.
695 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
696 support when GnuTLS has been built with p11-kit.
702 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
704 This, or similar/improved, will also be change PP/11 of 4.82.
710 PP/01 Handle short writes when writing local log-files.
711 In practice, only affects FreeBSD (8 onwards).
712 Bugzilla 1053, with thanks to Dmitry Isaikin.
714 NM/01 Bugzilla 949 - Documentation tweak
716 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
719 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
721 PP/02 Implemented gsasl authenticator.
723 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
725 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
726 `pkg-config foo` for cflags/libs.
728 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
729 with rest of GSASL and with heimdal_gssapi.
731 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
732 `pkg-config foo` for cflags/libs for the TLS implementation.
734 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
735 properties get this fed in as external SSF. A number of robustness
736 and debugging improvements to the cyrus_sasl authenticator.
738 PP/08 cyrus_sasl server now expands the server_realm option.
740 PP/09 Bugzilla 1214 - Log authentication information in reject log.
741 Patch by Jeremy Harris.
743 PP/10 Added dbmjz lookup type.
745 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
747 PP/12 MAIL args handles TAB as well as SP, for better interop with
748 non-compliant senders.
749 Analysis and variant patch by Todd Lyons.
751 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
752 Bug report from Lars Müller <lars@samba.org> (via SUSE),
753 Patch from Dirk Mueller <dmueller@suse.com>
755 PP/13 tls_peerdn now print-escaped for spool files.
756 Observed some $tls_peerdn in wild which contained \n, which resulted
757 in spool file corruption.
759 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
760 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
761 or write after TLS renegotiation, which otherwise led to messages
764 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
765 as a tracking header (ie: a signed header comes before the signature).
766 Patch from Wolfgang Breyha.
768 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
769 comma-sep list; embedded commas doubled.
771 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
773 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
775 Report and patch from Dmitry Banschikov.
777 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
778 Removed SSL_clear() after SSL_new() which led to protocol negotiation
779 failures. We appear to now support TLS1.1+ with Exim.
781 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
782 lets Exim select keys and certificates based upon TLS SNI from client.
783 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
784 before an outbound SMTP session. New log_selector, +tls_sni.
786 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
787 NULL dereference. Report and patch from Alun Jones.
789 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
790 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
791 Not seeing resolver debug output on NetBSD, but suspect this is a
792 resolver implementation change.
794 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
795 Left warnings. Added "eximon gdb" invocation mode.
797 PP/21 Defaulting "accept_8bitmime" to true, not false.
799 PP/22 Added -bw for inetd wait mode support.
801 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
802 locate the relevant includes and libraries. Made this the default.
804 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
805 Bugzilla 1246, report and most of solution from Tomasz Kusy.
807 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
808 This may cause build issues on older platforms.
810 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
811 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
812 gnutls_require_mac & gnutls_require_protocols (no longer supported).
813 Added SNI support via GnuTLS too.
814 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
816 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
818 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
819 Applied second patch from Janne, implementing suggestion to default
820 multiple-strings-in-record handling to match SPF spec.
822 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
824 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
825 read-only, out of scope).
826 Patch from Wolfgang Breyha, report from Stuart Northfield.
828 PP/29 Fix three issues highlighted by clang analyser static analysis.
829 Only crash-plausible issue would require the Cambridge-specific
830 iplookup router and a misconfiguration.
831 Report from Marcin Mirosław.
833 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
835 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
836 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
837 As part of this, removing so much warning spew let me fix some minor
838 real issues in debug logging.
840 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
841 assignment on my part. Fixed.
843 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
844 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
845 Janne Snabb (who went above and beyond: thank you).
847 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
848 string otherwise requires a connection and a bunch more work and it's
849 relatively easy to get wrong. Should also expose TLS library linkage
852 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
853 64-bit ${eval} (JH/03).
855 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
856 GNU libc to support some of the 64-bit stuff, should not lead to
857 conflicts. Defined before os.h is pulled in, so if a given platform
858 needs to override this, it can.
860 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
861 protection layer was required, which is not implemented.
862 Bugzilla 1254, patch from Wolfgang Breyha.
864 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
865 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
866 tls_dhparam take prime identifiers. Also unbreak combination of
867 OpenSSL+DH_params+TLSSNI.
869 PP/39 Disable SSLv2 by default in OpenSSL support.
875 PP/01 Solaris build fix for Oracle's LDAP libraries.
876 Bugzilla 1109, patch from Stephen Usher.
878 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
880 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
883 TF/02 Fix a couple more cases where we did not log the error message
884 when unlink() failed. See also change 4.74-TF/03.
886 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
887 lock up or crash if it happened to be inside a call to libc when it
888 got a SIGUSR1 from exiwhat.
890 The SIGUSR1 handler appends the current process status to the process
891 log which is later printed by exiwhat. It used to use the general
892 purpose logging code to do this, but several functions it calls are
893 not safe for signals.
895 The new output code in the SIGUSR1 handler is specific to the process
896 log, and simple enough that it's easy to inspect for signal safety.
897 Removing some special cases also simplifies the general logging code.
898 Removing the spurious timestamps from the process log simplifies
901 TF/04 Improved ratelimit ACL condition.
903 The /noupdate option has been deprecated in favour of /readonly which
904 has clearer semantics. The /leaky, /strict, and /readonly update modes
905 are mutually exclusive. The update mode is no longer included in the
906 database key; it just determines when the database is updated. (This
907 means that when you upgrde Exim will forget old rate measurements.)
909 Exim now checks that the per_* options are used with an update mode that
910 makes sense for the current ACL. For example, when Exim is processing a
911 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
912 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
913 must specify per_mail/readonly. If you omit the update mode it defaults to
914 /leaky where that makes sense (as before) or /readonly where required.
916 The /noupdate option is now undocumented but still supported for
917 backwards compatibility. It is equivalent to /readonly except that in
918 ACLs where /readonly is required you may specify /leaky/noupdate or
919 /strict/noupdate which are treated the same as /readonly.
921 A useful new feature is the /count= option. This is a generalization
922 of the per_byte option, so that you can measure the throughput of other
923 aggregate values. For example, the per_byte option is now equivalent
924 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
926 The per_rcpt option has been generalized using the /count= mechanism
927 (though it's more complicated than the per_byte equivalence). When it is
928 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
929 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
930 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
931 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
932 Note that using per_rcpt with a non-readonly update mode in more than
933 one ACL will cause the recipients to be double-counted. (The per_mail
934 and per_byte options don't have this problem.)
936 The handling of very low rates has changed slightly. If the computed rate
937 is less than the event's count (usually one) then this event is the first
938 after a long gap. In this case the rate is set to the same as this event's
939 count, so that the first message of a spam run is counted properly.
941 The major new feature is a mechanism for counting the rate of unique
942 events. The new per_addr option counts the number of different
943 recipients that someone has sent messages to in the last time period. It
944 behaves like per_rcpt if all the recipient addresses are different, but
945 duplicate recipient addresses do not increase the measured rate. Like
946 the /count= option this is a general mechanism, so the per_addr option
947 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
948 example, measure the rate that a client uses different sender addresses
949 with the options per_mail/unique=$sender_address. There are further
950 details in the main documentation.
952 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
954 TF/06 Removed a few PCRE remnants.
956 TF/07 Automatically extract Exim's version number from tags in the git
957 repository when doing development or release builds.
959 PP/02 Raise smtp_cmd_buffer_size to 16kB.
960 Bugzilla 879. Patch from Paul Fisher.
962 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
963 Heavily based on revision 40f9a89a from Simon Arlott's tree.
966 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
968 PP/05 Variable $av_failed, true if the AV scanner deferred.
969 Bugzilla 1078. Patch from John Horne.
971 PP/06 Stop make process more reliably on build failure.
972 Bugzilla 1087. Patch from Heiko Schlittermann.
974 PP/07 Make maildir_use_size_file an _expandable_ boolean.
975 Bugzilla 1089. Patch from Heiko Schlittermann.
977 PP/08 Handle ${run} returning more data than OS pipe buffer size.
978 Bugzilla 1131. Patch from Holger Weiß.
980 PP/09 Handle IPv6 addresses with SPF.
981 Bugzilla 860. Patch from Wolfgang Breyha.
983 PP/10 GnuTLS: support TLS 1.2 & 1.1.
985 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
988 PP/11 match_* no longer expand right-hand-side by default.
989 New compile-time build option, EXPAND_LISTMATCH_RHS.
990 New expansion conditions, "inlist", "inlisti".
992 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
994 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
996 PP/14 fix log_write() format string regression from TF/03.
997 Bugzilla 1152. Patch from Dmitry Isaikin.
1003 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1005 PP/02 Harmonised TLS library version reporting; only show if debugging.
1006 Layout now matches that introduced for other libraries in 4.74 PP/03.
1008 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1010 PP/04 New "dns_use_edns0" global option.
1012 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1015 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1016 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1018 TK/01 Updated PolarSSL code to 0.14.2.
1019 Bugzilla 1097. Patch from Andreas Metzler.
1021 PP/07 Catch divide-by-zero in ${eval:...}.
1022 Fixes bugzilla 1102.
1024 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1027 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1028 format-string attack -- SECURITY: remote arbitrary code execution.
1030 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1031 time unintentionally subject to list matching rules, letting the header
1032 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1033 arbitrary string expansion). This allowed for information disclosure.
1035 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1036 INT_MIN/-1 -- value coerced to INT_MAX.
1042 NM/01 Workround for PCRE version dependency in version reporting
1045 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1046 This fixes portability to compilers other than gcc, notably
1047 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1049 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1050 makefiles for portability to HP-UX and POSIX correctness.
1052 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1053 Also via indented variable definition in the Makefile.
1054 (Debugging by Oliver Heesakkers).
1056 PP/02 Restore caching of spamd results with expanded spamd_address.
1057 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1059 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1060 Improves build reliability. Fix from: Frank Elsner
1062 NM/02 Fix wide character breakage in the rfc2047 coding
1063 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1065 NM/03 Allow underscore in dnslist lookups
1066 Fixes bug 1026. Patch from Graeme Fowler
1068 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1069 Code patches from Adam Ciarcinski of NetBSD.
1071 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1074 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1075 is logged, to avoid truncation. Patch from John Horne.
1077 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1078 Patch from Jakob Hirsch.
1080 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1081 SQL string expansion failure details.
1082 Patch from Andrey Oktyabrski.
1084 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1085 Patch from Simon Arlott.
1087 PP/09 New lookups functionality failed to compile on old gcc which rejects
1088 extern declarations in function scope.
1089 Patch from Oliver Fleischmann
1091 PP/10 Use sig_atomic_t for flags set from signal handlers.
1092 Check getgroups() return and improve debugging.
1093 Fixed developed for diagnosis in bug 927 (which turned out to be
1096 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1097 Patch from Mark Zealey.
1099 PP/12 Bugzilla 1056: Improved spamd server selection.
1100 Patch from Mark Zealey.
1102 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1103 Based on patch from Heiko Schlittermann.
1105 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1106 Patch from Uwe Doering, sign-off by Michael Haardt.
1108 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1109 variable declaration deep within a block. Bug and patch from
1112 PP/15 lookups-Makefile IRIX compatibilty coercion.
1114 PP/16 Make DISABLE_DKIM build knob functional.
1116 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1117 Patch by Simon Arlott
1119 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1120 variable argument macros. Our copy now differs from upstream.
1126 TF/01 Failure to get a lock on a hints database can have serious
1127 consequences so log it to the panic log.
1129 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1130 controlled using the smtp_confirmation log selector.
1132 TF/03 Include the error message when we fail to unlink a spool file.
1134 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1135 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1136 for maintaining out-of-tree patches for some time.
1138 PP/01 Bugzilla 139: Documentation and portability issues.
1139 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1140 Handle per-OS dynamic-module compilation flags.
1142 PP/02 Let /dev/null have normal permissions.
1143 The 4.73 fixes were a little too stringent and complained about the
1144 permissions on /dev/null. Exempt it from some checks.
1145 Reported by Andreas M. Kirchwitz.
1147 PP/03 Report version information for many libraries, including
1148 Exim version information for dynamically loaded libraries. Created
1149 version.h, now support a version extension string for distributors
1150 who patch heavily. Dynamic module ABI change.
1152 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1153 privilege escalation vulnerability whereby the Exim run-time user
1154 can cause root to append content of the attacker's choosing to
1157 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1160 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1161 If dropping privileges for untrusted macros, we disabled normal logging
1162 on the basis that it would fail; for the Exim run-time user, this is not
1163 the case, and it resulted in successful deliveries going unlogged.
1164 Fixed. Reported by Andreas Metzler.
1170 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1171 only prepend for the Resent-* case. Fixes regression introduced in
1172 Exim 4.70 by NM/22 for Bugzilla 607.
1174 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1175 increasing numbers of administrators be bitten by this.
1177 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1179 PP/03 Bugzilla 994: added openssl_options main configuration option.
1181 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1183 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1185 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1187 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1188 without a peer certificate, leading to a segfault because of an
1189 assumption that peers always have certificates. Be a little more
1190 paranoid. Problem reported by Martin Tscholak.
1192 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1193 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1194 NB: ClamAV planning to remove STREAM in "middle of 2010".
1195 CL also introduces -bmalware, various -d+acl logging additions and
1196 more caution in buffer sizes.
1198 PP/09 Implemented reverse_ip expansion operator.
1200 PP/10 Bugzilla 937: provide a "debug" ACL control.
1202 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1204 PP/12 Bugzilla 973: Implement --version.
1206 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1208 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1210 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1212 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1213 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1214 ignore trailing whitespace.
1216 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1218 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1221 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1222 Notification from Dr Andrew Aitchison.
1224 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1225 ExtendedDetectionInfo response format.
1226 Notification from John Horne.
1228 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1231 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1232 XSL and documented dependency on system catalogs, with examples of how
1235 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1238 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1239 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1240 configuration file which is writeable by the Exim user or group.
1242 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1243 of configuration files to cover files specified with the -C option if
1244 they are going to be used with root privileges, not just the default
1247 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1248 option (effectively making it always true).
1250 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1251 files to be used while preserving root privileges.
1253 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1254 that rogue child processes cannot use them.
1256 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1257 run-time user, instead of root.
1259 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1260 Exim run-time user without dropping privileges.
1262 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1263 result string, instead of calling string_vformat() twice with the same
1266 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1267 for other users. Others should always drop root privileges if they use
1268 -C on the command line, even for a whitelisted configure file.
1270 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1272 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1278 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1279 $data_path, and $header_path variables; fixed documentation bugs and
1282 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1283 exipick to access non-standard spools, including the "frozen" queue
1286 NM/01 Bugzilla 965: Support mysql stored procedures.
1287 Patch from Alain Williams
1289 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1291 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1292 Patch from Andreas Metzler
1294 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1295 Patch from Kirill Miazine
1297 NM/05 Bugzilla 671: Added umask to procmail example.
1299 JJ/03 installed exipick 20100323.0, fixing doc bug
1301 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1302 directory. Notification and patch from Dan Rosenberg.
1304 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1306 TK/02 Improve log output when DKIM signing operation fails.
1308 MH/01 Treat the transport option dkim_domain as a colon separated
1309 list, not as a single string, and sign the message with each element,
1310 omitting multiple occurences of the same signer.
1312 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1313 Bugzilla 985, 986. Patch by Simon Arlott
1315 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1316 Patch by Simon Arlott
1318 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1319 MBX locking. Notification from Dan Rosenberg.
1325 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1327 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1329 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1331 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1333 NM/04 Bugzilla 915: Flag broken perl installation during build.
1339 TK/01 Added patch by Johannes Berg that expands the main option
1340 "spamd_address" if it starts with a dollar sign.
1342 TK/02 Write list of recipients to X-Envelope-Sender header when building
1343 the mbox-format spool file for content scanning (suggested by Jakob
1346 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1347 (http://www.dcc-servers.net/) support via dccifd. Activated by
1348 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1350 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1351 by Mark Daniel Reidel <mr@df.eu>.
1353 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1354 When building exim an external PCRE library is now needed -
1355 PCRE is a system library on the majority of modern systems.
1356 See entry on PCRE_LIBS in EDITME file.
1358 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1359 conversation. Added nologin parameter to request.
1360 Patch contributed by Kirill Miazine.
1362 TF/01 Do not log submission mode rewrites if they do not change the address.
1364 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1366 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1367 log files in place. Contributed by Roberto Lima.
1369 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1371 TF/03 Bugzilla 615: When checking the local_parts router precondition
1372 after a local_part_suffix or local_part_prefix option, Exim now
1373 does not use the address's named list lookup cache, since this
1374 contains cached lookups for the whole local part.
1376 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1377 Robert Millan. Documentation is in experimental-spec.txt.
1379 TF/04 Bugzilla 668: Fix parallel build (make -j).
1381 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1383 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1384 Patch provided by Jan Srzednicki.
1386 TF/05 Leading white space used to be stripped from $spam_report which
1387 wrecked the formatting. Now it is preserved.
1389 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1390 that they are available at delivery time.
1392 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1394 TF/08 TLS error reporting now respects the incoming_interface and
1395 incoming_port log selectors.
1397 TF/09 Produce a more useful error message if an SMTP transport's hosts
1398 setting expands to an empty string.
1400 NM/06 Bugzilla 744: EXPN did not work under TLS.
1401 Patch provided by Phil Pennock.
1403 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1404 Patch provided by Richard Godbee.
1406 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1407 acl_smtp_notquit, added index entry.
1409 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1410 Patch provided by Eugene Bujak.
1412 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1413 accept(). Patch provided by Maxim Dounin.
1415 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1416 Patch provided by Phil Pennock.
1418 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1420 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1421 Patch provided by Brad "anomie" Jorsch.
1423 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1424 Patch provided by Dean Brooks.
1426 TK/05 Add native DKIM support (does not depend on external libraries).
1428 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1429 Patch provided by Graeme Fowler.
1431 NM/16 Bugzilla 851: Documentation example syntax fix.
1433 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1435 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1438 NM/19 Bugzilla 745: TLS version reporting.
1439 Patch provided by Phil Pennock.
1441 NM/20 Bugzilla 167: bool: condition support.
1442 Patch provided by Phil Pennock.
1444 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1445 clients. Patch provided by Phil Pennock.
1447 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1448 Patch provided by Brad "anomie" Jorsch.
1450 NM/23 Bugzilla 687: Fix misparses in eximstats.
1451 Patch provided by Heiko Schlittermann.
1453 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1454 Patch provided by Heiko Schlittermann.
1456 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1457 plus update to original patch.
1459 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1461 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1462 Patch provided by David Brownlee.
1464 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1466 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1468 NM/30 Bugzilla 888: TLS documentation bugfixes.
1470 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1472 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1473 Unlike the original bugzilla I have changed all shell scripts in src tree.
1475 NM/33 Bugzilla 898: Transport filter timeout fix.
1476 Patch by Todd Rinaldo.
1478 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1479 Patch by Serge Demonchaux.
1481 NM/35 Bugzilla 39: Base64 decode bug fixes.
1482 Patch by Jakob Hirsch.
1484 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1486 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1488 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1490 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1496 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1497 ALT-N's libdkim that I have put here:
1498 http://duncanthrax.net/exim-experimental/
1500 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1501 were called 'true' and it seems that C99 defines that as a reserved
1502 keyword to be used with 'bool' variable types. That means you could
1503 not include C99-style headers which use bools without triggering
1504 build errors in sieve.c.
1506 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1507 as mailq or other aliases. Changed the --help handling significantly
1508 to do whats expected. exim_usage() emits usage/help information.
1510 SC/01 Added the -bylocaldomain option to eximstats.
1512 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1514 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1516 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1522 PH/01 Another patch from the Sieve maintainer.
1524 PH/02 When an IPv6 address is converted to a string for single-key lookup
1525 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1526 dots are used instead of colons so that keys in lsearch files need not
1527 contain colons. This was done some time before quoting was made available
1528 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1529 (notated using the quote facility) so as to distinguish them from IPv4
1530 keys. This meant that lookups for IP addresses in host lists did not work
1531 for iplsearch lookups.
1533 This has been fixed by arranging for IPv6 addresses to be expressed with
1534 colons if the lookup type is iplsearch. This is not incompatible, because
1535 previously such lookups could never work.
1537 The situation is now rather anomolous, since one *can* have colons in
1538 ordinary lsearch keys. However, making the change in all cases is
1539 incompatible and would probably break a number of configurations.
1541 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1544 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1545 conversion specification without a maximum field width, thereby enabling
1546 a rogue spamd server to cause a buffer overflow. While nobody in their
1547 right mind would setup Exim to query an untrusted spamd server, an
1548 attacker that gains access to a server running spamd could potentially
1549 exploit this vulnerability to run arbitrary code as the Exim user.
1551 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1552 $primary_hostname instead of what libspf2 thinks the hosts name is.
1554 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1555 a directory entry by the name of the lookup key. Previously, if a
1556 symlink pointed to a non-existing file or a file in a directory that
1557 Exim lacked permissions to read, a lookup for a key matching that
1558 symlink would fail. Now it is enough that a matching directory entry
1559 exists, symlink or not. (Bugzilla 503.)
1561 PH/03 The body_linecount and body_zerocount variables are now exported in the
1564 PH/04 Added the $dnslist_matched variable.
1566 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1567 This means they are set thereafter only if the connection becomes
1570 PH/06 Added the client_condition to authenticators so that some can be skipped
1571 by clients under certain conditions.
1573 PH/07 The error message for a badly-placed control=no_multiline_responses left
1574 "_responses" off the end of the name.
1576 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1578 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1579 (without spaces) instead of just copying the configuration text.
1581 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1583 PH/11 Added $max_received_linelength.
1585 PH/12 Added +ignore_defer and +include_defer to host lists.
1587 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1588 way in which PCRE > 7.0 is built.
1590 PH/14 Implemented queue_only_load_latch.
1592 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1593 MAIL command. The effect was to mangle the value on 64-bit systems.
1595 PH/16 Another patch from the Sieve maintainer.
1597 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1599 PH/18 If a system quota error occurred while trying to create the file for
1600 a maildir delivery, the message "Mailbox is full" was not appended to the
1601 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1602 only to a quota excession during the actual writing of the file.
1604 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1605 characters?) which causes problems in log lines. The DN values are now
1606 passed through string_printing() before being added to log lines.
1608 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1609 and InterBase are left for another time.)
1611 PH/21 Added message_body_newlines option.
1613 PH/22 Guard against possible overflow in moan_check_errorcopy().
1615 PH/23 POSIX allows open() to be a macro; guard against that.
1617 PH/24 If the recipient of an error message contained an @ in the local part
1618 (suitably quoted, of course), incorrect values were put in $domain and
1619 $local_part during the evaluation of errors_copy.
1625 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1626 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1629 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1630 issue a MAIL command.
1632 PH/02 In an ACL statement such as
1634 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1636 if a client was not listed at all, or was listed with a value other than
1637 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1638 the condition was not true (as it should be), so access was not denied.
1639 The bug was that the ! inversion was incorrectly passed on to the second
1640 item. This has been fixed.
1642 PH/03 Added additional dnslists conditions == and =& which are different from
1643 = and & when the dns lookup returns more than one IP address.
1645 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1646 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1648 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1649 FSYNC, which compiles an option called disable_fsync that allows for
1650 bypassing fsync(). The documentation is heavily laced with warnings.
1652 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1654 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1655 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1656 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1657 including adding "make clean"; (3) Added -fPIC when compiling the test
1658 dynamically loaded module, to get rid of a warning.
1660 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1661 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1662 = 0s. The bug is otherwise harmless.
1664 PH/07 There was a bug in the dovecot authenticator such that the value of
1665 $auth1 could be overwritten, and so not correctly preserved, after a
1666 successful authentication. This usually meant that the value preserved by
1667 the server_setid option was incorrect.
1669 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1671 PH/09 Installed PCRE release 7.0.
1673 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1674 run for batched SMTP input. It is now run at the start of every message
1675 in the batch. While fixing this I discovered that the process information
1676 (output by running exiwhat) was not always getting set for -bs and -bS
1677 input. This is fixed, and it now also says "batched" for BSMTP.
1679 PH/11 Added control=no_pipelining.
1681 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1682 patch, slightly modified), and move the expansion of helo_data till after
1683 the connection is made in the smtp transport (so it can use these
1686 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1688 PH/14 Added log_selector = +pid.
1690 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1692 PH/16 Add ${if forany and ${if forall.
1694 PH/17 Added dsn_from option to vary the From: line in DSNs.
1696 PH/18 Flush SMTP output before performing a callout, unless control =
1697 no_callout_flush is set.
1699 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1700 was true (the default) a successful delivery failed to delete the retry
1701 item, thus causing premature timeout of the address. The bug is now
1704 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1706 PH/21 Long custom messages for fakedefer and fakereject are now split up
1707 into multiline reponses in the same way that messages for "deny" and
1708 other ACL rejections are.
1710 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1711 with slight modification.
1713 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1714 draft, changing the syntax and factoring some duplicate code".
1716 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1717 for deliveries of the second and subsequent messages over the same SMTP
1720 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1721 ${reduce, with only minor "tidies".
1723 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1725 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1726 expansion side effects.
1728 PH/27 When a message times out after an over-quota error from an Exim-imposed
1729 quota, the bounce message says "mailbox is full". This message was not
1730 being given when it was a system quota that was exceeded. It now should
1733 MH/03 Made $recipients available in local_scan(). local_scan() already has
1734 better access to the recipient list through recipients_list[], but
1735 $recipients can be useful in postmaster-provided expansion strings.
1737 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1738 in the case of a MAIL command with additional options following the
1739 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1740 were accidentally chopped off.
1742 PH/29 SMTP synchronization checks are implemented when a command is read -
1743 there is a check that no more input is waiting when there shouldn't be
1744 any. However, for some commands, a delay in an ACL can mean that it is
1745 some time before the response is written. In this time, more input might
1746 arrive, invalidly. So now there are extra checks after an ACL has run for
1747 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1748 pipelining has not been advertised.
1750 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1752 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1753 returned in the SMTP response when smtp_return_error_details was false.
1754 This has been fixed.
1756 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1757 instead of the C I/O that was originally supplied, because problems were
1758 reported on Solaris.
1760 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1761 Exim which did not show up earlier: it was assuming that a call to
1762 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1763 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1764 was a macro that became an assignment, so it seemed to work. This has
1765 changed to a proper function call with a void return, hence the compile
1766 error. Exim's code has been fixed.
1768 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1771 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1773 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1775 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1776 because it turns out to be too big or there is a timeout) there is no
1777 "Completed" line in the log. When some messages of this type were
1778 selected by exigrep, they were listed as "not completed". Others were
1779 picked up by some special patterns. I have improved the selection
1780 criteria to be more general.
1782 PH/38 The host_find_failed option in the manualroute router can now be set
1783 to "ignore", to completely ignore a host whose IP address cannot be
1784 found. If all hosts are ignored, the behaviour is controlled by the new
1785 host_all_ignored option.
1787 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1788 homing or because of multiple MX records with /mx) generated more than
1789 one IP address, and the following item turned out to be the local host,
1790 all the secondary addresses of the first item were incorrectly removed
1791 from the list, along with the local host and any following hosts (which
1792 is what is supposed to happen).
1794 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1795 whoever called Exim into the -H file. In the case of the daemon it was
1796 behaving confusingly. When first started, it used values for whoever
1797 started the daemon, but after a SIGHUP it used the Exim user (because it
1798 calls itself on a restart). I have changed the code so that it now always
1801 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1802 message are rejected with the same error (e.g. no authentication or bad
1803 sender address), and a DATA command is nevertheless sent (as can happen
1804 with PIPELINING or a stupid MUA), the error message that was given to the
1805 RCPT commands is included in the rejection of the DATA command. This is
1806 intended to be helpful for MUAs that show only the final error to their
1809 PH/42 Another patch from the Sieve maintainer.
1811 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1812 Eximstats - Fixed some broken HTML links and added missing column headers
1814 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1815 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1817 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1819 PH/43 Yet another patch from the Sieve maintainer.
1821 PH/44 I found a way to check for a TCP/IP connection going away before sending
1822 the response to the final '.' that terminates a message, but only in the
1823 case where the client has not sent further data following the '.'
1824 (unfortunately, this is allowed). However, in many cases there won't be
1825 any further data because there won't be any more messages to send. A call
1826 to select() can be used: if it shows that the input is "ready", there is
1827 either input waiting, or the socket has been closed. An attempt to read
1828 the next input character can distinguish the two cases. Previously, Exim
1829 would have sent an OK response which the client would never have see.
1830 This could lead to message repetition. This fix should cure that, at
1831 least in a lot of common cases.
1833 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1834 advertised in response to EHLO.
1840 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1841 fixed by 4.65/MH/01 (is this a record?) are fixed:
1843 (i) An empty string was always treated as zero by the numeric comparison
1844 operators. This behaviour has been restored.
1846 (ii) It is documented that the numeric comparison operators always treat
1847 their arguments as decimal numbers. This was broken in that numbers
1848 starting with 0 were being interpreted as octal.
1850 While fixing these problems I realized that there was another issue that
1851 hadn't been noticed. Values of message_size_limit (both the global option
1852 and the transport option) were treated as octal if they started with 0.
1853 The documentation was vague. These values are now always treated as
1854 decimal, and I will make that clear in the documentation.
1860 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1861 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1864 MH/01 Don't check that the operands of numeric comparison operators are
1865 integers when their expansion is in "skipping" mode (fixes bug
1866 introduced by 4.64-PH/07).
1868 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1869 child addresses, Exim now panics and dies. Previously, because the count
1870 is held in a short int, deliveries were likely to be lost. As such a
1871 large number of recipients for a single message is ridiculous
1872 (performance will be very, very poor), I have chosen to impose a limit
1873 rather than extend the field.
1879 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1880 leftover -K file (the existence of which was triggered by #402).
1881 While we were at it, introduced process PID as part of the -K
1882 filename. This should rule out race conditions when creating
1885 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1886 processing considerably. Previous code took too long for large mails,
1887 triggering a timeout which in turn triggers #401.
1889 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1890 in the DK code in transports.c. sendfile() is not really portable,
1891 hence the _LINUX specificness.
1893 TF/01 In the add_headers option to the mail command in an Exim filter,
1894 there was a bug that Exim would claim a syntax error in any
1895 header after the first one which had an odd number of characters
1898 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1899 callout verification, Exim cached a "reject" for the entire domain. This
1900 is correct for most verifications, but it is not correct for a recipient
1901 verification with use_sender or use_postmaster set, because in that case
1902 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1903 case of MAIL FROM:<> rejection from other early rejections (e.g.
1904 rejection of HELO). When verifying a recipient using a non-null MAIL
1905 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1906 Whatever the result of the callout, the value of the domain cache is
1907 left unchanged (for any other kind of callout, getting as far as trying
1908 RCPT means that the domain itself is ok).
1910 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1913 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1914 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1915 session, the daemon ignores ECONNECT errors and logs others; it now
1916 ignores EPIPE as well.
1918 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1919 (quoted-printable decoding).
1921 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1922 later the small subsequent patch to fix an introduced bug.
1924 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1926 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1928 PH/08 An error is now given if message_size_limit is specified negative.
1930 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1931 to be given (somewhat) arbitrary names.
1933 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1936 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1937 miscellaneous code fixes
1939 PH/10 Added the log_reject_target ACL modifier to specify where to log
1942 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1943 hostname. This is wrong, because it relates to the incoming message (and
1944 probably the interface on which it is arriving) and not to the outgoing
1945 callout (which could be using a different interface). This has been
1946 changed to use the value of the helo_data option from the smtp transport
1947 instead - this is what is used when a message is actually being sent. If
1948 there is no remote transport (possible with a router that sets up host
1949 addresses), $smtp_active_hostname is used.
1951 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1952 tweaks were necessary in order to get it to work (see also 21 below):
1953 (a) The code assumed that strncpy() returns a negative number on buffer
1954 overflow, which isn't the case. Replaced with Exim's string_format()
1956 (b) There were several signed/unsigned issues. I just did the minimum
1957 hacking in of casts. There is scope for a larger refactoring.
1958 (c) The code used strcasecmp() which is not a standard C function.
1959 Replaced with Exim's strcmpic() function.
1960 (d) The code set only $1; it now sets $auth1 as well.
1961 (e) A simple test gave the error "authentication client didn't specify
1962 service in request". It would seem that Dovecot has changed its
1963 interface. Fortunately there's a specification; I followed it and
1964 changed what the client sends and it appears to be working now.
1966 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1969 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1970 address A is aliased to B and C, where B exists and C does not. Without
1971 -v the output is "A verified" because verification stops after a
1972 successful redirection if more than one address is generated. However,
1973 with -v the child addresses are also verified. Exim was outputting "A
1974 failed to verify" and then showing the successful verification for C,
1975 with its parentage. It now outputs "B failed to verify", showing B's
1976 parentage before showing the successful verification of C.
1978 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1979 look up a TXT record in a specific list after matching in a combined
1982 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1983 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1984 they consult the DNS. I had assumed they would set it the way they
1985 wanted; and indeed my experiments on Linux seem to show that in some
1986 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1987 To be on the safe side, however, I have now made the interface to
1988 host_find_byname() similar to host_find_bydns(), with an argument
1989 containing the DNS resolver options. The host_find_byname() function now
1990 sets these options at its start, just as host_find_bydns() does. The smtp
1991 transport options dns_qualify_single and dns_search_parents are passed to
1992 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1993 of host_find_byname() use the default settings of RES_DEFNAMES
1994 (qualify_single) but not RES_DNSRCH (search_parents).
1996 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1997 spool_read_header() do less string testing, by means of a preliminary
1998 switch on the second character of optional "-foo" lines. (This is
1999 overdue, caused by the large number of possibilities that now exist.
2000 Originally there were few.) While I was there, I also converted the
2001 str(n)cmp tests so they don't re-test the leading "-" and the first
2002 character, in the hope this might squeeze out yet more improvement.
2004 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2005 flag allowing group syntax was set by the header_syntax check but not
2006 turned off, possible causing trouble later; (2) The flag was not being
2007 set at all for the header_verify test, causing "group"-style headers to
2008 be rejected. I have now set it in this case, and also caused header_
2009 verify to ignore an empty address taken from a group. While doing this, I
2010 came across some other cases where the code for allowing group syntax
2011 while scanning a header line wasn't quite right (mostly, not resetting
2012 the flag correctly in the right place). These bugs could have caused
2013 trouble for malformed header lines. I hope it is now all correct.
2015 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2016 with the "reply" argument non-NULL. The code, however (which originally
2017 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2018 but it didn't always do it. This confused somebody who was copying the
2019 code for some other use. I have removed all the tests.
2021 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2022 feature that was used to support insecure browsers during the U.S. crypto
2023 embargo. It requires special client support, and Exim is probably the
2024 only MTA that supported it -- and would never use it because real RSA is
2025 always available. This code has been removed, because it had the bad
2026 effect of slowing Exim down by computing (never used) parameters for the
2027 RSA_EXPORT functionality.
2029 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2030 authenticator to fail if there's a tab character in the incoming data
2031 (there should never be unless someone is messing about, as it's supposed
2032 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2033 if the connection is using TLS or if the remote IP is the same as the
2034 local IP, and the "valid-client-cert option" if a client certificate has
2037 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2038 authenticators. This can be used for authorization after authentication
2039 succeeds. (In the case of plaintext, it servers for both authentication
2042 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2043 if any retry times were supplied.
2045 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2046 connection, where there is no client IP address to check. In this
2047 situation, the verify now always succeeds.
2049 PH/25 Applied John Jetmore's -Mset patch.
2051 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2053 PH/27 In a string expansion for a processed (not raw) header when multiple
2054 headers of the same name were present, leading whitespace was being
2055 removed from all of them, but trailing whitespace was being removed only
2056 from the last one. Now trailing whitespace is removed from each header
2057 before concatenation. Completely empty headers in a concatenation (as
2058 before) are ignored.
2060 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2061 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2063 PH/29 [Removed. This was a change that I later backed out, and forgot to
2064 correct the ChangeLog entry (that I had efficiently created) before
2065 committing the later change.]
2067 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2068 address errors (4xx response to RCPT) over the same connection as other
2069 messages routed to the same hosts. Such deliveries are always "forced",
2070 so retry times are not inspected. This resulted in far too many retries
2071 for the affected addresses. The effect occurred only when there were more
2072 hosts than the hosts_max_try setting in the smtp transport when it had
2073 the 4xx errors. Those hosts that it had tried were not added to the list
2074 of hosts for which the message was waiting, so if all were tried, there
2075 was no problem. Two fixes have been applied:
2077 (i) If there are any address or message errors in an SMTP delivery, none
2078 of the hosts (tried or untried) are now added to the list of hosts
2079 for which the message is waiting, so the message should not be a
2080 candidate for sending over the same connection that was used for a
2081 successful delivery of some other message. This seems entirely
2082 reasonable: after all the message is NOT "waiting for some host".
2083 This is so "obvious" that I'm not sure why it wasn't done
2084 previously. Hope I haven't missed anything, but it can't do any
2085 harm, as the worst effect is to miss an optimization.
2087 (ii) If, despite (i), such a delivery is accidentally attempted, the
2088 routing retry time is respected, so at least it doesn't keep
2089 hammering the server.
2091 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2092 in ${readsocket because some servers need this prod.
2094 PH/32 Added some extra debug output when updating a wait-xxx database.
2096 PH/33 The hint "could be header name not terminated by colon", which has been
2097 given for certain expansion errors for a long time, was not being given
2098 for the ${if def:h_colon_omitted{... case.
2100 PH/34 The spec says: "With one important exception, whenever a domain list is
2101 being scanned, $domain contains the subject domain." There was at least
2102 one case where this was not true.
2104 PH/35 The error "getsockname() failed: connection reset by peer" was being
2105 written to the panic log as well as the main log, but it isn't really
2106 panic-worthy as it just means the connection died rather early on. I have
2107 removed the panic log writing for the ECONNRESET error when getsockname()
2110 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2111 runs only) independently of the message's sender address. This meant
2112 that, if the 4xx error was in fact related to the sender, a different
2113 message to the same recipient with a different sender could confuse
2114 things. In particualar, this can happen when sending to a greylisting
2115 server, but other circumstances could also provoke similar problems.
2116 I have changed the default so that the retry time for these errors is now
2117 based a combination of the sender and recipient addresses. This change
2118 can be overridden by setting address_retry_include_sender=false in the
2121 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2122 remote server are returned as part of bounce messages. This was not
2123 happening for LMTP over a pipe (the lmtp transport), but now it is the
2124 same for both kinds of LMTP.
2126 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2127 in header lines that were in fact CNAMEs. This is no longer the case.
2129 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2130 and queue runs started by the daemon processed all messages. This has
2131 been fixed so that -R and -S can now usefully be given with -q<time>.
2133 PH/40 Import PCRE release 6.7 (fixes some bugs).
2135 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2137 PH/42 Give an error if -q is specified more than once.
2139 PH/43 Renamed the variables $interface_address and $interface_port as
2140 $received_ip_address and $received_port, to make it clear that these
2141 values apply to message reception, and not to the outgoing interface when
2142 a message is delivered. (The old names remain recognized, of course.)
2144 PH/44 There was no timeout on the connect() call when using a Unix domain
2145 socket in the ${readsocket expansion. There now is.
2147 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2148 be meaningful with "accept".
2150 SC/01 Eximstats V1.43
2151 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2153 SC/02 Eximstats V1.44
2154 Use a glob alias rather than an array ref in the generated
2155 parser. This improves both readability and performance.
2157 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2158 Collect SpamAssassin and rejection statistics.
2159 Don't display local sender or destination tables unless
2160 there is data to show.
2161 Added average volumes into the top table text output.
2163 SC/04 Eximstats V1.46
2164 Collect data on the number of addresses (recipients)
2165 as well as the number of messages.
2167 SC/05 Eximstats V1.47
2168 Added 'Message too big' to the list of mail rejection
2169 reasons (thanks to Marco Gaiarin).
2171 SC/06 Eximstats V1.48
2172 Mainlog lines which have GMT offsets and are too short to
2173 have a flag are now skipped.
2175 SC/07 Eximstats V1.49 (Alain Williams)
2176 Added the -emptyok flag.
2178 SC/08 Eximstats V1.50
2179 Fixes for obtaining the IP address from reject messages.
2181 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2182 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2183 whitesspace changes from 4.64-PH/27
2185 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2188 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2189 are found, allow negative numbers in numeric criteria)
2191 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2193 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2194 to match changes made in 4.64-PH/43
2196 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2198 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2199 "domain" is actually an IP address, and force a failure. This locks out
2200 those revolvers/nameservers that support "A-for-A" lookups, in
2201 contravention of the specifications.
2203 PH/48 When a host name was looked up from an IP address, and the subsequent
2204 forward lookup of the name timed out, the host name was left in
2205 $sender_host_name, contrary to the specification.
2207 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2208 restricted to single-key lookups, Exim was not diagnosing an error if
2209 * or *@ was used with a query-style lookup.
2211 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2213 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2214 long ago, but noone interested enough thought of it. Let's just say that
2215 the "1.1" means that there are some new functions that weren't there at
2216 some point in the past.
2218 PH/51 Error processing for expansion failure of helo_data from an smtp
2219 transport during callout processing was broken.
2221 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2222 tested/used via the -bh/-bhc/-bs options.
2224 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2225 bug, fixed in subsequent PCRE releases).
2227 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2228 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2230 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2236 SC/01 Use a glob alias rather than an array ref in eximstats generated
2237 parser. This improves both readability and performance.
2239 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2240 Don't display local sender or destination tables in eximstats unless
2241 there is data to show.
2242 Added average volumes into the eximstats top table text output.
2244 SC/03 Collect data on the number of addresses (recipients) as well
2245 as the number of messages in eximstats.
2247 TF/01 Correct an error in the documentation for the redirect router. Exim
2248 does (usually) call initgroups() when daemonizing.
2250 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2251 with consistent privilege compared to when running as a daemon.
2253 TF/03 Note in the spec that $authenticated_id is not set for local
2254 submissions from trusted users.
2256 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2257 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2259 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2260 by adding some example configuration directives to the default
2261 configuration file. A little bit of work is required to uncomment the
2262 directives and define how usernames and passwords are checked, but
2263 there is now a framework to start from.
2265 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2266 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2267 without this. I don't know how relevant this is to other LDAP libraries.
2269 PH/02 Add the verb name to the "unknown ACL verb" error.
2271 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2273 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2275 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2276 directory not expanded when it should be if an expanded home directory
2277 was set for the address (which is overridden by the transport).
2279 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2282 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2283 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2284 because it is too late at that time, and has no effect.
2286 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2287 security issue with \' (bugzilla #107). I could not use the
2288 PQescapeStringConn() function, because it needs a PGconn value as one of
2291 PH/08 When testing addresses using -bt, indicate those final addresses that
2292 are duplicates that would not cause an additional delivery. At least one
2293 person was confused, thinking that -bt output corresponded to deliveries.
2294 (Suppressing duplicates isn't a good idea as you lose the information
2295 about possibly different redirections that led to the duplicates.)
2297 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2298 systems where poll() doesn't work, in particular OS X.
2300 PH/10 Added more information to debugging output for retry time not reached.
2302 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2303 operations in malware.c.
2305 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2308 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2309 syslog_duplication set false, log lines that would normally be written
2310 both the the main log and to the reject log were not written to syslog at
2313 PH/14 In the default configuration, change the use of "message" in ACL warn
2314 statements to "add_header".
2316 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2317 not followed by a command (e.g. "seen endif").
2319 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2320 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2323 PH/17 Added extra conditions to the default value of delay_warning_condition
2327 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2328 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2329 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2332 The Auto-Submitted: and various List- headers are standardised, whereas I
2333 don't think Precedence: ever was.
2335 PH/18 Refactored debugging code in route_finduser() to show more information,
2336 in particular, the error code if getpwnam() issues one.
2338 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2339 This is apparently needed in addition to the PH/07 change above to avoid
2340 any possible encoding problems.
2342 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2343 but not after initializing Perl.
2345 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2346 output them only if debugging. By default they are written stderr,
2347 apparently, which is not desirable.
2349 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2352 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2355 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2357 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2358 authenticated or an ident call has been made. Suppress the default
2359 values for $authenticated_id and $authenticated_sender (but permit -oMai
2360 and -oMas) when testing with -bh.
2362 PH/24 Re-jigged the order of the tests in the default configuration so that the
2363 tests for valid domains and recipients precede the DNS black list and CSA
2364 tests, on the grounds that those ones are more expensive.
2366 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2367 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2368 This bug exists in every version of Exim that I still have, right back to
2371 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2372 However, an attempt to turn on case-sensitivity in a regex key by
2373 including (?-i) didn't work because the subject string was already
2374 lowercased, and the effects were non-intuitive. It turns out that a
2375 one-line patch can be used to allow (?-i) to work as expected.
2381 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2382 other effects) broke the use of negated acl sub-conditions.
2384 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2387 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2388 "Deny" causes Exim to reject the incoming connection with a 554 error.
2389 Unfortunately, if there is a major crisis, such as a disk failure,
2390 tcp-wrappers gives "deny", whereas what one would like would be some
2391 kind of temporary error. A kludge has been added to help with this.
2392 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2393 554 error is used if errno is still zero or contains ENOENT (which occurs
2394 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2397 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2399 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2400 errors. Otherwise a message that provokes a temporary error (when other
2401 messages do not) can cause a whole host to time out.
2403 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2404 addresses were routed directly to files or pipes from a redirect router.
2405 File deliveries just didn't batch; pipe deliveries might have suffered
2408 PH/06 A failure to get a lock for a hints database would erroneously always say
2409 "Failed to get write lock", even when it was really a read lock.
2411 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2412 of 0600. This has been changed to use the value of the lockfile_mode
2413 option (which defaults to 0600).
2415 PH/08 Applied small patch from the Sieve maintainer.
2417 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2418 folder from quota calculations, a direct delivery into this folder messed
2419 up the contents of the maildirsize file. This was because the regex was
2420 used only to exclude .Trash (or whatever) when the size of the mailbox
2421 was calculated. There was no check that a delivery was happening into an
2422 excluded directory. This bug has been fixed by ignoring all quota
2423 processing for deliveries into excluded directories.
2425 PH/10 Added the maildirfolder_create_regex option to appendfile.
2431 PH/01 The code for finding all the local interface addresses on a FreeBSD
2432 system running IPv6 was broken. This may well have applied to all BSD
2433 systems, as well as to others that have similar system calls. The broken
2434 code found IPv4 interfaces correctly, but gave incorrect values for the
2435 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2436 that it would not match correctly against @[] and not recognize the IPv6
2439 PH/02 The ipliteral router was not recognizing addresses of the form user@
2440 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2442 PH/03 Added disable_ipv6.
2444 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2445 decoded form, because it is most often used to construct To: headers
2446 lines in autoreplies, and the decoded form may well be syntactically
2447 invalid. However, $reply_address has leading white space removed, and all
2448 newlines turned into spaces so that the autoreply transport does not
2451 PH/05 If group was specified without a user on a router, and no group or user
2452 was specified on a transport, the group from the router was ignored.
2454 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2455 for visible compile-time settings that can be used to change these
2456 numbers, for those that want even more. Backwards compatibility with old
2457 spool files has been maintained. However, going back to a previous Exim
2458 release will lost any variables that are in spool files.
2460 PH/07 Two small changes when running in the test harness: increase delay when
2461 passing a TCP/IP connection to a new process, in case the original
2462 process has to generate a bounce, and remove special handling of
2463 127.0.0.2 (sic), which is no longer necessary.
2465 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2466 be the same on different OS.
2468 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2471 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2472 whether --show-vars was specified or not
2474 JJ/02 exipick: Added support for new ACL variable spool format introduced
2477 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2478 syntactically invalid From: or Reply-to: line, and a filter used this to
2479 generate an autoreply, and therefore failed to obtain an address for the
2480 autoreply, Exim could try to deliver to a non-existent relative file
2481 name, causing unrelated and misleading errors. What now happens is that
2482 it logs this as a hard delivery error, but does not attempt to create a
2485 PH/11 The exinext utility has a -C option for testing purposes, but although
2486 the given file was scanned by exinext itself; it wasn't being passed on
2487 when Exim was called.
2489 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2490 an end-of-file indication when reading a command response.
2492 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2493 compiled. In many other places in Exim, IPv6 addresses are always
2494 recognized, so I have changed this. It also means that IPv4 domain
2495 literals of the form [IPV4:n.n.n.n] are now always recognized.
2497 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2498 used if the router is not running as root, for example, when verifying at
2499 ACL time, or when using -bh. The debugging output from this situation was
2500 non-existent - all you got was a failure to exec. I have made two
2503 (a) Failures to set uid/gid, the current directory, or a process leader
2504 in a subprocess such as that created by queryprogram now generate
2505 suitable debugging ouput when -d is set.
2507 (b) The queryprogram router detects when it is not running as root,
2508 outputs suitable debugging information if -d is set, and then runs
2509 the subprocess without attempting to change uid/gid.
2511 PH/15 Minor change to Makefile for building test_host (undocumented testing
2514 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2515 additional section of a DNS packet that returns MX or SRV records.
2516 Instead, it always explicitly searches for A/AAAA records. This avoids
2517 major problems that occur when a DNS server includes only records of one
2518 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2519 fixed another bug: if SRV records were looked up and the corresponding
2520 address records were *not* found in the additional section, the port
2521 values from the SRV records were lost.
2523 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2524 using the correct key (the original address) when searching the retry
2525 rules in order to find which one to use for generating the retry hint.
2527 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2528 adding the default one. Similarly, if it contains a Reply-To: header, the
2529 errors_reply_to option, if set, is not used.
2531 PH/19 When calculating a retry time, Exim used to measure the "time since
2532 failure" by looking at the "first failed" field in the retry record. Now
2533 it does not use this if it is later than than the arrival time of the
2534 message. Instead it uses the arrival time. This makes for better
2535 behaviour in cases where some deliveries succeed, thus re-setting the
2536 "first failed" field. An example is a quota failure for a huge message
2537 when small messages continue to be delivered. Without this change, the
2538 "time since failure" will always be short, possible causing more frequent
2539 delivery attempts for the huge message than are intended.
2540 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2542 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2543 $1, $2, $3) because the numerical variables can be reset during some
2544 expansion items (e.g. "match"), thereby losing the authentication data.
2546 PH/21 Make -bV show the size of off_t variables so that the test suite can
2547 decide whether to run tests for quotas > 2G.
2549 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2550 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2551 filecount value is greater than 2G or if a quota value is greater than 2G
2552 on a system where the size of off_t is not greater than 4, a panic error
2555 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2556 never match. The debug and -bh output now contains an explicit error
2557 message indicating a malformed IPv4 address or mask.
2559 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2560 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2561 PH/23 above applies.
2563 PH/25 Do not write to syslog when running in the test harness. The only
2564 occasion when this arises is a failure to open the main or panic logs
2565 (for which there is an explicit test).
2567 PH/26 Added the /no_tell option to "control=freeze".
2569 PH/27 If a host name lookup failed very early in a connection, for example, if
2570 the IP address matched host_lookup and the reverse lookup yielded a name
2571 that did not have a forward lookup, an error message of the form "no IP
2572 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2573 could be logged. Now it outputs the IP address instead of "NULL".
2575 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2576 allows the sender and the authenticated sender to be set when
2577 submitting a message from within Exim. Since child_open_exim() is
2578 documented for local_scan(), the new function should be too.
2580 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2581 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2582 results in an empty string is now treated as unset.
2584 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2586 PH/31 Added sender_verify_fail logging option.
2588 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2589 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2590 tidied the source and removed it altogether.
2592 PH/33 When a queue run was abandoned because the load average was too high, a
2593 log line was always written; now it is written only if the queue_run log
2594 selector is set. In addition, the log line for abandonment now contains
2595 information about the queue run such as the pid. This is always present
2596 in "start" and "stop" lines but was omitted from the "abandon" line.
2598 PH/34 Omit spaces between a header name and the colon in the error message that
2599 is given when verify = headers_syntax fails (if there are lots of them,
2600 the message gets confusing).
2602 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2603 names, as there are now some PTR records that contain slashes. This check
2604 is only to protect against broken name servers that fall over on strange
2605 characters, so the fact that it applies to all lookups doesn't matter.
2607 PH/36 Now that the new test suite is complete, we can remove some of the
2608 special code in Exim that was needed for the old test suite. For example,
2609 sorting DNS records because real resolvers return them in an arbitrary
2610 order. The new test suite's fake resolver always returns records in the
2613 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2614 bounces) except when queue_only is set, to avoid logging races between
2615 the different processes.
2617 PH/38 Panic-die if .include specifies a non-absolute path.
2619 PH/39 A tweak to the "H" retry rule from its user.
2621 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2622 a label. They prevented compilation on older perls.
2624 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2625 a warning to be raised on newish perls.
2627 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2628 on queue. Changes to match documented behaviour of showing count of
2629 messages matching specified criteria.
2631 PH/40 Changed the default ident timeout from 30s to 5s.
2633 PH/41 Added support for the use of login_cap features, on those BSD systems
2634 that have them, for controlling the resources used by pipe deliveries.
2636 PH/42 The content-scanning code uses fopen() to create files in which to put
2637 message data. Previously it was not paying any attention to the mode of
2638 the files. Exim runs with umask(0) because the rest of the code creates
2639 files with open(), and sets the required mode explicitly. Thus, these
2640 files were ending up world-writeable. This was not a big issue, because,
2641 being within the spool directory, they were not world-accessible. I have
2642 created a function called modefopen, which takes an additional mode
2643 argument. It sets umask(777), creates the file, chmods it to the required
2644 mode, then resets the umask. All the relevant calls to fopen() in the
2645 content scanning code have been changed to use this function.
2647 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2648 to 24 hours. This avoids potential overflow problems when processing G
2649 and H retry rules. I suspect nobody ever tinkers with this value.
2651 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2653 PH/45 When the plaintext authenticator is running as a client, the server's
2654 challenges are checked to ensure they are valid base64 strings. By
2655 default, the authentication attempt is cancelled if an invalid string is
2656 received. Setting client_ignore_invalid_base64 true ignores these errors.
2657 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2658 they are received. Thus, the responses can be made to depend on the
2659 challenges. If an invalid string is ignored, an empty string is placed in
2662 PH/46 Messages that are created by the autoreply transport now contains a
2663 References: header, in accordance with RFCs 2822 and 3834.
2665 PH/47 Added authenticated_sender_force to the smtp transport.
2667 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2669 PH/49 Installed latest patch from the Sieve maintainer.
2671 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2672 was also set, the appendfile transport was unnecessarily scanning a
2673 directory of message files (e.g. for maildir delivery) to find the count
2674 of files (along with the size), even though it did not need this
2675 information. It now does the scan only if it needs to find either the
2676 size of the count of files.
2678 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2680 PH/52 Two bugs concerned with error handling when the smtp transport is
2683 (i) Exim was not creating retry information for temporary errors given
2684 for individual recipients after the DATA command when the smtp transport
2685 was used in LMTP mode. This meant that they could be retried too
2686 frequently, and not timed out correctly.
2688 (ii) Exim was setting the flag that allows error details to be returned
2689 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2690 recipients that were returned after the DATA command.
2692 PH/53 This is related to PH/52, but is more general: for any failing address,
2693 when detailed error information was permitted to be returned to the
2694 sender, but the error was temporary, then after the final timeout, only
2695 "retry timeout exceeded" was returned. Now it returns the full error as
2696 well as "retry timeout exceeded".
2698 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2699 do this, and (what is worse) MTAs that accept it.
2701 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2702 will now be deprecated.
2704 PH/56 New os.c-cygwin from the Cygwin maintainer.
2706 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2707 formats (previously only available in exim formats via -bpr, -bpru,
2708 and -bpra. Now also available in native and exiqgrep formats)
2710 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2711 with very large, slow to parse queues
2713 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2715 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2717 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2718 responses to authentication challenges, though it was showing the
2719 challenges; (ii) I've removed the CR characters from the debug output for
2722 PH/58 Allow for the insertion of a newline as well as a space when a string
2723 is turned into more than one encoded-word during RFC 2047 encoding. The
2724 Sieve code now uses this.
2726 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2727 data_4xx, lost_connection, tls_required.
2729 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2730 message was being sent as an SMTP response.
2732 PH/61 Add -l and -k options to exicyclog.
2734 PH/62 When verifying, if an address was redirected to one new address, so that
2735 verification continued, and the new address failed or deferred after
2736 having set something in $address_data, the value of $address_data was not
2737 passed back to the ACL. This was different to the case when no
2738 redirection occurred. The value is now passed back in both cases.
2740 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2741 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2742 use login_cap.h, so on its own it isn't the distinguishing feature. The
2743 new name refers directly to the setclassresources() function.
2745 PH/65 Added configuration files for NetBSD3.
2747 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2749 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2750 is preferred over IPv4.
2752 PH/68 The bounce_return_message and bounce_return_body options were not being
2753 honoured for bounces generated during the reception of non-SMTP messages.
2754 In particular, this applied to messages rejected by the ACL. This bug has
2755 been fixed. However, if bounce_return_message is true and bounce_return_
2756 body is false, the headers that are returned for a non-SMTP message
2757 include only those that have been read before the error was detected.
2758 (In the case of an ACL rejection, they have all been read.)
2760 PH/69 The HTML version of the specification is now built in a directory called
2761 spec_html instead of spec.html, because the latter looks like a path with
2762 a MIME-type, and this confuses some software.
2764 PH/70 Catch two compiler warnings in sieve.c.
2766 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2767 function verify_get_ident() calls ip_connect() to connect a socket, but
2768 if the "connect()" function timed out, ip_connect() used to close the
2769 socket. However, verify_get_ident() also closes the socket later, and in
2770 between Exim writes to the log, which may get opened at this point. When
2771 the socket was closed in ip_connect(), the log could get the same file
2772 descriptor number as the socket. This naturally causes chaos. The fix is
2773 not to close the socket in ip_connect(); the socket should be closed by
2774 the function that creates it. There was only one place in the code where
2775 this was missing, in the iplookup router, which I don't think anybody now
2776 uses, but I've fixed it anyway.
2778 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2779 well as to direct DNS lookups. Otherwise the handling of names in host
2780 lists is inconsistent and therefore confusing.
2786 PH/01 Two changes to the default runtime configuration:
2788 (1) Move the checks for relay_from_hosts and authenticated clients from
2789 after to before the (commented out) DNS black list checks.
2791 (2) Add control=submission to the relay_from_hosts and authenticated
2792 clients checks, on the grounds that messages accepted by these
2793 statements are most likely to be submissions.
2795 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2797 (1) Generate an error if the third argument for the ${prvs expansion is
2800 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2803 (3) Reset the variables that are obtained from the first argument of
2804 ${prvscheck and used in the second argument before leaving the code,
2805 because their memory is reclaimed, so using them afterwards may do
2808 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2809 one (it's much easier than Tom thought :-).
2811 (5) Because of (4), we can now allow for the use of $prvscheck_result
2812 inside the third argument.
2814 PH/03 For some reason, the default setting of PATH when running a command from
2815 a pipe transport was just "/usr/bin". I have changed it to
2818 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2819 anything to be listed in the output from -bV.
2821 PH/05 When a filter generated an autoreply, the entire To: header line was
2822 quoted in the delivery log line, like this:
2824 => >A.N.Other <ano@some.domain> <original@ddress> ...
2826 This has been changed so that it extracts the operative address. There
2827 may be more than one such address. If so, they are comma-separated, like
2830 => >ano@some.domain,ona@other.domain <original@ddress> ...
2832 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2833 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2834 not being looked up in the rDNS to get a host name, Exim was showing the
2835 IP address twice in Received: lines, even though the IP addresses were
2836 identical. For example:
2838 Received: from [1.2.3.4] (helo=[1.2.3.4])
2840 However, if the real host name was known, it was omitting the HELO data
2841 if it matched the actual IP address. This has been tidied up so that it
2842 doesn't show the same IP address twice.
2844 PH/07 When both +timestamp and +memory debugging was on, the value given by
2845 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2846 called by the debug printing, thereby overwriting the timestamp buffer.
2847 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2849 PH/08 When the original message was included in an autoreply transport, it
2850 always said "this is a copy of the message, including all the headers",
2851 even if body_only or headers_only was set. It now gives an appropriate
2854 PH/09 Applied a patch from the Sieve maintainer which:
2856 o fixes some comments
2857 o adds the (disabled) notify extension core
2858 o adds some debug output for the result of if/elsif tests
2859 o points to the current vacation draft in the documentation
2860 and documents the missing references header update
2864 o fixes a bug in processing the envelope test (when testing
2865 multiple envelope elements, the last element determinted the
2868 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2869 Electronic Mail") by including:
2871 Auto-submitted: auto-generated
2873 in the messages that it generates (bounce messages and others, such as
2874 warnings). In the case of bounce messages for non-SMTP mesages, there was
2875 also a typo: it was using "Auto_submitted" (underscore instead of
2876 hyphen). Since every message generated by Exim is necessarily in response
2877 to another message, thes have all been changed to:
2879 Auto-Submitted: auto-replied
2881 in accordance with these statements in the RFC:
2883 The auto-replied keyword:
2885 - SHOULD be used on messages sent in direct response to another
2886 message by an automatic process,
2888 - MUST NOT be used on manually-generated messages,
2890 - MAY be used on Delivery Status Notifications (DSNs) and Message
2891 Disposition Notifications (MDNs),
2893 - MUST NOT be used on messages generated by automatic or periodic
2894 processes, except for messages which are automatic responses to
2897 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2898 to the default Received: header definition.
2900 PH/12 Added log selector acl_warn_skipped (default on).
2902 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2903 variables because (a) they are in the wrong storage pool and (b) even if
2904 they were copied, it wouldn't work properly because of the caching.
2906 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2907 checking when decoding. Apparently there are clients that generate
2908 overlong encoded strings. Why am I not surprised?
2910 PH/15 If the first argument of "${if match_address" was not empty, but did not
2911 contain an "@" character, Exim crashed. Now it writes a panic log message
2912 and treats the condition as false.
2914 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2916 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2917 extension "envelope-auth". The code is finished and in agreement with
2918 other implementations, but there is no documentation so far and in fact,
2919 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2920 not changing the active code.
2922 Print executed "if" and "elsif" statements when debugging is used. This
2923 helps a great deal to understand what a filter does.
2925 Document more things not specified clearly in RFC3028. I had all this
2926 sorted out, when out of a sudden new issues came to my mind. Oops."
2928 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2931 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2932 canonical form (as documented). However, after a host name lookup from
2933 the IP address, check_host() was doing a simple string comparison with
2934 addresses acquired from the DNS when checking that the found name did
2935 have the original IP as one of its addresses. Since any found IPv6
2936 addresses are likely to be in abbreviated form, the comparison could
2937 fail. Luckily, there already exists a function for doing the comparison
2938 by converting both addresses to binary, so now that is used instead of
2939 the text comparison.
2941 PH/20 There was another similar case to PH/19, when a complete host name was
2942 given in a host list; looking up its IP address could give an abbreviated
2943 form, whereas the current host's name might or might not be abbreviated.
2944 The same fix has been applied.
2950 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2951 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2954 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2955 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2957 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2959 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2960 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2961 Also, for "Auto-Submitted", treat anything other than "no" as
2962 non-personal, in accordance with RFC 3834. (Previously it treated
2963 anything starting "auto-" as non-personal.)
2965 TF/01 The control=submission/name=... option had a problem with syntax
2966 errors if the name included a slash character. The /name= option
2967 now slurps the rest of the string, so it can include any characters
2968 but it must come last in the list of options (after /sender_retain
2971 PH/05 Some modifications to the interface to the fake nameserver for the new
2979 TK/01 Added the "success_on_redirect" address verification option. See
2980 NewStuff for rationale and an example.
2982 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2984 PH/02 Patch to exigrep to allow it to work on syslog lines.
2986 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2987 fread() to skip over the body file's header line, because in Cygwin the
2988 header line is locked and is inaccessible.
2990 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2991 co-exist for some time) to make it clear that it is the Exim ID that is
2992 referenced, not the Message-ID: header line.
2994 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2995 string_format() function, because snprintf() does not exist on all
2998 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2999 ${stat: expansion item.
3001 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3002 protocol synchronization error", to keep the pedants happy.
3004 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3005 well as for IRIX systems, when gcc is being used. See the host.c source
3008 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3010 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3013 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3014 transport and to the smtp transport in LMTP mode.
3016 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3018 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3019 run a filter in a subprocess. This could lead to confusion in subsequent
3020 lookups in the parent process. There should also be a search_tidyup() at
3021 the end of the subprocess.
3023 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3024 only if the host matched helo_try_verify_hosts, which caused the
3025 verification to occur when the EHLO/HELO command was issued. The ACL just
3026 tested the remembered result. Now, if a previous verification attempt has
3027 not happened, "verify = helo" does it there and then.
3029 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3031 TK/03 Fix log output including CR from clamd.
3033 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3034 exist provoked a memory error which could cause a segfault.
3036 PH/15 Installed PCRE 6.2
3038 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3040 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3041 of the problem. Specifically, suggested +O2 rather than +O1 for the
3044 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3046 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3047 record, the port was not correctly logged when the outgoing_port log
3048 selector was set (it logged the transort's default port).
3050 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3051 fallback_hosts, and "hosts" in the smtp transport.
3053 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3054 host errors such as "Connection refused".
3056 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3057 authentication with radiusclient 0.4.9:
3059 - Error returned from rc_read_config was caught wrongly
3060 - Username/password not passed on to radius server due to wrong length.
3062 The presumption is that some radiusclient API changes for 4.51/PH/17
3063 were not taken care of correctly. The code is still untested by me (my
3064 Linux distribution still has 0.3.2 of radiusclient), but it was
3065 contributed by a Radius user.
3067 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3068 expanding the "port" option of the smtp transport.
3070 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3071 while reading a MIME header. Thanks to Tom Hughes for a patch.
3073 PH/24 Include config.h inside local_scan.h so that configuration settings are
3076 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3077 that in an ACL for RCPT (for example), you can examine exactly what was
3080 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3081 commands, but it was not correctly comparing the address with the actual
3082 client host address. Thus, it would show the EHLO address in Received:
3083 header lines when this was not necessary.
3085 PH/27 Added the % operator to ${eval:}.
3087 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3088 it should be ignoring failures (because with -C, for example, it has lost
3089 privilege). It wasn't ignoring creation failures other than "already
3092 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3095 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3096 because an input line was too long, either on its own, or by virtue of
3097 too many continuations, the temporary file was not being removed, and the
3098 return code was incorrect.
3100 PH/31 Missing "BOOL" in function definition in filtertest.c.
3102 PH/32 Applied Sieve patches from the maintainer.
3104 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3106 PH/33 Added "verify = not_blind".
3108 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3109 Local/Makefile (with some defaults set). These are used in built scripts
3110 such as exicyclog, but they have never been used in the exim_install
3111 script (though there are many overriding facilities there). I have
3112 arranged that the exim_install script now takes note of these two
3115 PH/35 Installed configuration files for Dragonfly.
3117 PH/36 When a locally submitted message by a trusted user did not contain a
3118 From: header, and the sender address was obtained from -f or from an SMTP
3119 MAIL command, and the trusted user did not use -F to supply a sender
3120 name, $originator_name was incorrectly used when constructing a From:
3121 header. Furthermore, $originator_name was used for submission mode
3122 messages from external hosts without From: headers in a similar way,
3123 which is clearly wrong.
3125 PH/37 Added control=suppress_local_fixups.
3127 PH/38 When log_selector = +received_sender was set, and the addition of the
3128 sender made the log line's construction buffer exactly full, or one byte
3129 less than full, an overflow happened when the terminating "\n" was
3132 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3133 when the result of a list match is failure because a DNS lookup failed.
3135 PH/40 RM_COMMAND is now used in the building process.
3137 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3138 the "build-* directories that it finds.
3140 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3141 address in a domain literal was a prefix of an interface address.
3143 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3144 when verifying a sender address, unless rewrite_headers is false.
3146 PH/44 Wrote a long comment about why errors_to addresses are verified as
3147 recipients, not senders.
3149 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3150 the ratelimit ACL was added.
3152 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3154 PH/46 Added extra information about PostgreSQL errors to the error string.
3156 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3157 avoiding the need to install special zones in a real server. This is
3158 backwards compatible; if it can't find the fake resolver, it drops back.
3159 Thus, both old and new test suites can be run.
3161 TF/02 Added util/ratelimit.pl
3163 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3164 clock is set back in time.
3166 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3167 Candler <B.Candler@pobox.com>.
3169 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3170 OK for addresses that are the result of redirections.
3172 PH/48 A number of further additions for the benefit of the new test suite,
3173 including a fake gethostbyname() that interfaces to the fake DNS resolver
3176 TF/06 The fix for widen_domains has also been applied to qualify_single and
3177 search_parents which are the other dnslookup options that can cause
3180 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3183 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3185 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3186 service" if it encountered an MX record with an empty target hostname.
3187 The message is now "an MX or SRV record indicated no SMTP service".
3189 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3190 if the DNS of the sending site is misconfigured. This is quite a
3191 common situation. This change restores the behaviour of treating a
3192 helo verification defer as a failure.
3194 PH/51 If self=fail was set on a router, the bounce message did not include the
3195 actual error message.
3201 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3203 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3204 command itself ended in error, the underlying message about the transport
3205 filter timeout was being overwritten with the pipe command error. Now the
3206 underlying error message should be appended to the second error message.
3208 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3210 PH/02 Reduce the amount of output that "make" produces by default. Full output
3211 can still be requested.
3213 PH/03 The warning log line about a condition test deferring for a "warn" verb
3214 was being output only once per connection, rather than after each
3215 occurrence (because it was using the same function as for successful
3216 "warn" verbs). This seems wrong, so I have changed it.
3218 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3219 it should not have, which might have caused a crash in the right
3220 circumstances, but probably never did.
3222 PH/04 Installed a modified version of Tony Finch's patch to make submission
3223 mode fix the return path as well as the Sender: header line, and to
3224 add a /name= option so that you can make the user's friendly name appear
3227 TF/03 Added the control = fakedefer ACL modifier.
3229 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3230 Mark Lowes for thorough testing.
3232 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3234 TK/03 Merged latest SRS patch from Miles Wilton.
3236 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3237 with the definition in sysexits.h (which is #included earlier).
3238 Fortunately, Exim does not actually use EX_OK. The code used to try to
3239 preserve the sysexits.h value, by assumimg that macro definitions were
3240 scanned for macro replacements. I have been disabused of this notion,
3241 so now the code just undefines EX_OK before #including unistd.h.
3243 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3244 in the smtp transport. When a block could not be written in a single
3245 write() function, the timeout was being re-applied to each part-write.
3246 This seems wrong - if the receiver was accepting one byte at a time it
3247 would take for ever. The timeout is now adjusted when this happens. It
3248 doesn't have to be particularly precise.
3250 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3251 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3253 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3254 without a domain if the check to <postmaster@domain> fails.
3256 SC/01 Eximstats: added -xls and the ability to specify output files
3257 (patch written by Frank Heydlauf).
3259 SC/02 Eximstats: use FileHandles for outputing results.
3261 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3263 SC/04 Eximstats: fixed display of large numbers with -nvr option
3265 SC/05 Eximstats: fixed merging of reports with empty tables.
3267 SC/06 Eximstats: added the -include_original_destination flag
3269 SC/07 Eximstats: removed tabs and trailing whitespace.
3271 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3273 TK/06 MBOX spool code: Add real "From " MBOX separator line
3274 so the .eml file is really in mbox format (even though
3275 most programs do not really care). Patch from Alex Miller.
3277 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3278 The latter is generated from $received_to and is only set if the
3279 message has one envelope recipient. SA can use these headers,
3280 obviously out-of-the-box. Patch from Alex Miller.
3282 PH/08 The ${def test on a variable was returning false if the variable's
3283 value was "0", contrary to what the specification has always said!
3284 The result should be true unless the variable is empty.
3286 PH/09 The syntax error of a character other than { following "${if
3287 def:variable_name" (after optional whitespace) was not being diagnosed.
3288 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3289 accidental colon was present, for example, could give incorrect results.
3291 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3292 result is used (not including appendfile, where other changes are about
3295 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3296 This involved changing a lot of size variables from int to off_t. It
3297 should work with maildirs and everything.
3299 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3300 spamd dying while we are connected to it.
3302 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3305 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3306 function for BDB 4.3.
3308 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3310 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3311 the sources for PCRE have been re-arranged and more files are now
3314 PH/15 The code I had for printing potentially long long variables in PH/11
3315 above was not the best (it lost precision). The length of off_t variables
3316 is now inspected at build time, and an appropriate printing format (%ld
3317 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3318 to be "long long int" or "long int". This is needed for the internal
3319 formatting function string_vformat().
3321 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3322 the configuration file to be ":syslog", then the script "guesses" where
3323 the logs files are, rather than using the compiled in default. In our
3324 case the guess is not the same as the compiled default, so the script
3325 suddenly stopped working when I started to use syslog. The patch checks
3326 to see if log_file_path is "". If so, it attempts to read it from exim
3327 with no configuration file to get the compiled in version, before it
3328 falls back to the previous guessing code."
3330 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3331 implementing BATV in an Exim configuration. See NewStuff for the gory
3334 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3335 Makefile that are specific to HP-UX.
3337 PH/18 If the "use_postmaster" option was set for a recipient callout together
3338 with the "random" option, the postmaster address was used as the MAIL
3339 FROM address for the random test, but not for the subsequent recipient
3340 test. It is now used for both.
3342 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3343 patch removes a few documentation additions to RFC 3028, because the
3344 latest draft now contains them. It adds the new en;ascii-case comparator
3345 and a new error check for 8bit text in MIME parts. Comparator and
3346 require names are now matched exactly. I enabled the subaddress
3347 extension, but it is not well tested yet (read: it works for me)."
3349 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3350 rework some of the code of TK/09 above to avoid the hardwired use of
3351 "%lld" and "long long". Replaced the call to snprintf() with a call to
3354 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3355 records point to non-existent hosts", "retry timeout exceeded", and
3356 "retry time not reached for any host after a long failure period".
3358 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3359 experimental DomainKeys support:
3361 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3362 (2) On an error such as an illegally used "control", the wrong name for
3363 the control was given.
3365 These problems did NOT occur unless DomainKeys support was compiled.
3367 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3369 PH/24 Added ${if match_ip condition.
3371 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3372 databases so that it will be absolutely obvious if a crash occurs in the
3373 DB library. This is a regular occurrence (often caused by mis-matched
3376 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3377 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3378 picked up on a user's system that detects such things. There doesn't seem
3379 to be a gcc warning option for this - only an attribute that has to be
3380 put on the function's prototype. It seems that in Fedora Core 4 they have
3381 set this on a number of new functions. No doubt there will be more in due
3384 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3385 specify a transport. However, if an address that was verified by such a
3386 router was the subject of a callout, Exim crashed because it tried to
3387 read the rcpt_include_affixes from the non-existent transport. Now it
3388 just assumes that the setting of that option is false. This bug was
3389 introduced by 4.51/PH/31.
3391 PH/28 Changed -d+all to exclude +memory, because that information is very
3392 rarely of interest, but it makes the output a lot bigger. People tend to
3393 do -d+all out of habit.
3395 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3396 code in os-type was giving problems when libc.so lives in lib64, like on
3399 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3400 aren't the modern standard, and it seems that some systems' include files
3401 don't always have them. Exim was already checking for some of the newer
3402 ones like T_AAAA, and defining it itself. I've added checks for all the
3403 record types that Exim uses.
3405 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3406 not automatically generating a new one, as it is supposed to. This
3407 prevented TLS from working. If the file did exist, but contained invalid
3408 data, a new version was generated, as expected. It was only the case of a
3409 non-existent file that was broken.
3411 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3412 with a change in libdomainkeys > 0.64.
3414 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3415 from DNS. If the selector record carries the flag, it now has
3416 precedence over the domain-wide flag.
3418 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3420 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3421 the use of a port name that isn't defined in /etc/services) occurred, the
3422 message was deferred as in a normal delivery, and thus remained on the
3423 spool, instead of being failed because of the mua_wrapper setting. This
3424 is now fixed, and I tidied up some of the mua_wrapper messages at the
3427 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3428 the messages in a hash of arrays rather than using individual hashes.
3429 This is a bit cleaner and results in dramatic memory savings, albeit
3430 at a slight CPU cost.
3432 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3433 as requested by Marc Sherman.
3435 SC/10 Eximstats: added histograms for user specified patterns as requested
3438 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3440 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3441 fopen() in the content-scanning modules that did not already have it.
3447 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3448 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3450 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3452 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3454 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3455 to test Sieve filters that use "vacation".
3457 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3458 that changes the way the GnuTLS parameters are stored in the cache file.
3459 The new format can be generated externally. For backward compatibility,
3460 if the data in the cache doesn't make sense, Exim assumes it has read an
3461 old-format file, and it generates new data and writes a new file. This
3462 means that you can't go back to an older release without removing the
3465 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3466 work if there are any delivery delays because "one_time" forces the
3467 parent to be marked "delivered", so its unseen clone is never tried
3468 again. For this reason, Exim now forbids the simultaneous setting of
3471 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3472 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3473 entry if you want to know the details. The fix, however, neglected to
3474 consider the case when local delivery batching is involved. The test for
3475 "previously delivered" was not happening when checking to see if an
3476 address could be batched with a previous (undelivered) one; under
3477 certain circumstances this could lead to multiple deliveries to the same
3480 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3481 in its include files, and this causes problems building Exim.
3483 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3484 header_syntax) but Exim was just ignoring anything given after a slash.
3485 In particular, this caused confusion with an attempt to use "verify =
3486 reverse_host_lookup/defer_ok". An error is now given when options are
3487 supplied for verify items that do not have them. (Maybe reverse_host_
3488 lookup should have a defer_ok option, but that's a different point.)
3490 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3491 defined by RFC 821) to 2048, because there were problems with some AUTH
3492 commands, and RFC 1869 says the size should be increased for extended
3493 SMTP commands that take arguments.
3495 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3498 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3499 "unknown" error; now it says that the functionality isn't in the binary.
3501 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3502 an address' error message when a string expansion fails (syntax or
3503 whatever). Otherwise the password may appear in the log. Following change
3504 PH/42 below, there is no longer a chance of it appearing in a bounce
3507 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3509 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3510 of its addresses was ever tried. (Bugzilla bug #2.)
3512 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3513 the result incorrectly in the debug output. (It correctly added a newline
3514 to what was transported.)
3516 TF/01 Added $received_time.
3518 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3519 commented out examples of how to interface to a virus scanner and to
3520 SpamAssassin. Also added commented examples of av_scanner and
3521 spamd_address settings.
3523 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3524 and controls are allowed in which ACLs. There were a couple of minor
3525 errors. Some of the entries in the conditions table (which is a table of
3526 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3527 negation of where the condition IS allowed.
3529 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3531 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3532 header file does not have a version number, so I've had to invent a new
3533 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3534 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3535 radiusclient), but it was contributed by a Radius user.
3537 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3538 files or directories, for OpenSSL.
3540 PH/19 When an Exim process that is running as root has to create an Exim log
3541 file, it does so in a subprocess that runs as exim:exim so as to get the
3542 ownership right at creation (otherwise, other Exim processes might see
3543 the file with the wrong ownership). There was no test for failure of this
3544 fork() call, which would lead to the process getting stuck as it waited
3545 for a non-existent subprocess. Forks do occasionally fail when resources
3546 run out. I reviewed all the other calls to fork(); they all seem to check
3549 PH/20 When checking for unexpected SMTP input at connect time (before writing
3550 the banner), Exim was not dealing correctly with a non-positive return
3551 from the read() function. If the client had disconnected by this time,
3552 the result was a log entry for a synchronization error with an empty
3553 string after "input=" when read() returned zero. If read() returned -1
3554 (an event I could not check), uninitialized data bytes were printed.
3555 There were reports of junk text (parts of files, etc) appearing after
3558 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3560 PH/22 Added support for macro redefinition, and (re)definition in between
3561 driver and ACL definitions.
3563 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3564 forgetting to use the resulting value; it was using the unexpanded value.
3566 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3567 hadn't been configured. The fix is from Juergen Kreileder, who
3568 understands it better than I do:
3570 "Here's what I see happening with three configured cyrus_sasl
3571 authenticators configured (plain, login, cram-md5):
3573 On startup auth_cyrus_sasl_init() gets called for each of these.
3574 This means three calls to sasl_listmech() without a specified mech_list.
3575 => SASL tests which mechs of all available mechs actually work
3576 => three warnings about OTP not working
3577 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3579 With the patch, sasl_listmech() also gets called three times. But now
3580 SASL's mech_list option is set to the server_mech specified in the the
3581 authenticator. Or in other words, the answer from sasl_listmech()
3582 gets limited to just the mech you're testing for (which is different
3584 => the return list contains just 'plain' or 'login', 'cram-md5' or
3585 nothing depending on the value of ob->server_mech.
3587 I've just tested the patch: Authentication still works fine,
3588 unavailable mechs specified in the exim configuration are still
3589 caught, and the auth.log warnings about OTP are gone."
3591 PH/25 When debugging is enabled, the contents of the command line are added
3592 to the debugging output, even when log_selector=+arguments is not
3595 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3596 answer is "GNU", and only if the return is "GNU/something" is the answer
3599 PH/27 $acl_verify_message is now set immediately after the failure of a
3600 verification in an ACL, and so is available in subsequent modifiers. In
3601 particular, the message can be preserved by coding like this:
3603 warn !verify = sender
3604 set acl_m0 = $acl_verify_message
3606 Previously, $acl_verify_message was set only while expanding "message"
3607 and "log_message" when a very denied access.
3609 PH/28 Modified OS/os.c-Linux with
3611 -#ifndef OS_LOAD_AVERAGE
3612 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3614 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3615 nomenclature these days.)
3617 PH/29 Installed patch from the Sieve maintainer that adds the options
3618 sieve_useraddress and sieve_subaddress to the redirect router.
3620 PH/30 In these circumstances:
3621 . Two addresses routed to the same list of hosts;
3622 . First host does not offer TLS;
3623 . First host accepts first address;
3624 . First host gives temporary error to second address;
3625 . Second host offers TLS and a TLS session is established;
3626 . Second host accepts second address.
3627 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3628 and peerdn, if requested) that were in fact used only for the second
3631 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3632 attention to any local part prefix or suffix that was matched by the
3633 router that accepted the address. It now behaves in the same way as it
3634 does for delivery: the affixes are removed from the local part unless
3635 rcpt_include_affixes is set on the transport.
3637 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3638 timeout during the DATA phase of an incoming message.
3640 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3641 applied a patch sanctioned by the Sieve maintainer.
3643 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3644 the uid or gid is negative. A case of a negative gid caused this to be
3645 noticed. The fix allows for either to be negative.
3647 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3648 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3651 PH/36 The change PH/12 above was broken. Fixed it.
3653 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3654 the grounds that routing the same address twice would always produce the
3655 same answer. This might have been true once, but it is certainly no
3656 longer true now. Routing a child address may depend on the previous
3657 routing that produced that child. Some complicated redirection strategies
3658 went wrong when messages had multiple recipients, and made Exim's
3659 behaviour dependent on the order in which the addresses were given.
3661 I have moved the duplicate checking until after the routing is complete.
3662 Exim scans the addresses that are assigned to local and remote
3663 transports, and removes any duplicates. This means that more work will be
3664 done, as duplicates will always all be routed, but duplicates are
3665 presumably rare, so I don't expect this is of any significance.
3667 For deliveries to pipes, files, and autoreplies, the duplicate checking
3668 still happens during the routing process, since they are not going to be
3671 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3672 It corrects a timeout issue with spamd. This is Ian's comment: "The
3673 background is that sometimes spamd either never reads data from a
3674 connection it has accepted, or it never writes response data. The exiscan
3675 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3676 blindly assumes that writes won't block so it may never time out."
3678 PH/39 Allow G after quota size as well as K and M.
3680 PH/40 The value set for $authenticated_id in an authenticator may not contain
3681 binary zeroes or newlines because the value is written to log lines and
3682 to spool files. There was no check on this. Now the value is run through
3683 the string_printing() function so that such characters are converted to
3684 printable escape sequences.
3686 PH/41 $message_linecount is a new variable that contains the total number of
3687 lines in the message. Compare $body_linecount, which is the count for the
3690 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3691 bounce and delay warning messages, except in certain special cases, which
3694 (a) An SMTP error message from a remote host;
3695 (b) A message specified in a :fail: redirection;
3696 (c) A message specified in a "fail" command in a system filter;
3697 (d) A message specified in a FAIL return from the queryprogram router;
3698 (e) A message specified by the cannot_route_message router option.
3700 In these cases only, Exim does include the error details in bounce and
3701 warning messages. There are also a few cases where bland messages such
3702 as "unrouteable address" or "local delivery error" are given.
3704 PH/43 $value is now also set for the "else" part of a ${run expansion.
3706 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3707 being worked on, but at least Exim now implements the latest version to
3710 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3711 process to complete was treated as a delivery failure, a timeout while
3712 writing the message to the pipe was logged, but erroneously treated as a
3713 successful delivery. Such timeouts include transport filter timeouts. For
3714 consistency with the overall process timeout, these timeouts are now
3715 treated as errors, giving rise to delivery failures by default. However,
3716 there is now a new Boolean option for the pipe transport called
3717 timeout_defer, which, if set TRUE, converts the failures into defers for
3718 both kinds of timeout. A transport filter timeout is now identified in
3721 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3722 systems where "make" and "gmake" are different, calling "gmake" at top
3723 level broke things. I've arranged for the value of $(MAKE) to be passed
3724 from the Makefile to this script so that it can call the same version of
3728 A note about Exim versions 4.44 and 4.50
3729 ----------------------------------------
3731 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3732 changes of various kinds. As a consequence, a big documentation update was
3733 needed. This delayed the release for rather longer than seemed good, especially
3734 in the light of a couple of (minor) security issues. Therefore, the changes
3735 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3736 release. So 4.44 and 4.50 are in effect two different branches that both start
3739 I have left the 4.50 change log unchanged; it contains all the changes since
3740 4.43. The change log for 4.44 is below; many of its items are identical to
3741 those for 4.50. This seems to be the most sensible way to preserve the
3742 historical information.
3748 1. Minor wording change to the doc/README.SIEVE file.
3750 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3751 computation of the current number of files was incorrect.
3753 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3754 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3757 4. Give more explanation in the error message when the command for a transport
3758 filter fails to execute.
3760 5. There are several places where Exim runs a non-Exim command in a
3761 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3762 was being done only for the command run by the queryprogram router. It is
3763 now done for all such subprocesses. The other cases are: ${run, transport
3764 filters, and the commands run by the lmtp and pipe transports.
3766 6. Added CONFIGURE_GROUP build-time option.
3768 7. Some older OS have a limit of 256 on the maximum number of file
3769 descriptors. Exim was using setrlimit() to set 1000 as a large value
3770 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3771 systems. I've change it so that if it can't get 1000, it tries for 256.
3773 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3774 was an oversight, and furthermore, ever since the addition of extra
3775 controls (e.g. 4.43/32), the checks on when to allow different forms of
3776 "control" were broken. There should now be diagnostics for all cases when a
3777 control that does not make sense is encountered.
3779 9. Added the /retain_sender option to "control=submission".
3781 10. $recipients is now available in the predata ACL (oversight).
3783 11. Tidy the search cache before the fork to do a delivery from a message
3784 received from the command line. Otherwise the child will trigger a lookup
3785 failure and thereby defer the delivery if it tries to use (for example) a
3786 cached ldap connection that the parent has called unbind on.
3788 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3789 of $address_data from the recipient verification was clobbered by the
3790 sender verification.
3792 13. The value of address_data from a sender verification is now available in
3793 $sender_address_data in subsequent conditions in the ACL statement.
3795 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3797 15. Added a new option "connect=<time>" to callout options, to set a different
3800 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3801 was its contents. (It was OK if the option was not defined at all.)
3803 17. A "Completed" log line is now written for messages that are removed from
3804 the spool by the -Mrm option.
3806 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3807 information about exactly what failed.
3809 19. Added -dd to debug only the daemon process.
3811 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3812 handles timeouts, both on the server side and network timeouts. Renamed the
3813 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3815 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3816 It is now set to "smtps".
3818 22. $host_address is now set to the target address during the checking of
3819 ignore_target_hosts.
3821 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3822 being passed; this would have caused $sender_host_name to have been used if
3823 matching the list had actually called for a host name (not very likely,
3824 since this list is usually IP addresses). A host name is now passed as
3827 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3828 code that specifies a non-restarting handler (typically sigaction() in
3829 modern systems) in an attempt to fix a rare and obscure crash bug.
3831 25. Narrowed the window for a race in the daemon that could cause it to ignore
3832 SIGCHLD signals. This is not a major problem, because they are used only to
3833 wake it up if nothing else does.
3835 26. A malformed maildirsize file could cause Exim to calculate negative values
3836 for the mailbox size or file count. Odd effects could occur as a result.
3837 The maildirsize information is now recalculated if the size or filecount
3840 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3841 support for a long time. Removed HAVE_SYS_VFS_H.
3843 28. Installed the latest version of exipick from John Jetmore.
3845 29. In an address list, if the pattern was not a regular expression, an empty
3846 subject address (from a bounce message) matched only if the pattern was an
3847 empty string. Non-empty patterns were not even tested. This was the wrong
3848 because it is perfectly reasonable to use an empty address as part of a
3849 database query. An empty address is now tested by patterns that are
3850 lookups. However, all the other forms of pattern expect the subject to
3851 contain a local part and a domain, and therefore, for them, an empty
3852 address still always fails if the pattern is not itself empty.
3854 30. Exim went into a mad DNS loop when attempting to do a callout where the
3855 host was specified on an smtp transport, and looking it up yielded more
3856 than one IP address.
3858 31. Re-factored the code for checking spool and log partition space into a
3859 function that finds that data and another that does the check. The former
3860 is then used to implement four new variables: $spool_space, $log_space,
3861 $spool_inodes, and $log_inodes.
3863 32. The RFC2047 encoding function was originally intended for short strings
3864 such as real names; it was not keeping to the 75-character limit for
3865 encoded words that the RFC imposes. It now respects the limit, and
3866 generates multiple encoded words if necessary. To be on the safe side, I
3867 have increased the buffer size for the ${rfc2047: expansion operator from
3870 33. It is now permitted to omit both strings after an "if" condition; if the
3871 condition is true, the result is "true". As before, when the second string
3872 is omitted, a false condition yields an empty string. This makes it less
3873 cumbersome to write custom ACL and router conditions.
3875 34. Failure to deliver a bounce message always caused it to be frozen, even if
3876 there was an errors_to setting on the router. The errors_to setting is now
3879 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3880 canonical form (fully expanded) before being placed in
3881 $sender_host_address.
3883 36. The table in the code that translates DNS record types into text (T_A to
3884 "A" for instance) was missing entries for NS and CNAME. It is just possible
3885 that this could have caused confusion if both these types were looked up
3886 for the same domain, because the text type is used as part of Exim's
3887 per-process caching. But the chance of anyone hitting this buglet seems
3890 37. The dnsdb lookup has been extended in a number of ways.
3892 (1) There is a new type, "zns", which walks up the domain tree until it
3893 finds some nameserver records. It should be used with care.
3895 (2) There is a new type, "mxh", which is like "mx" except that it returns
3896 just the host names, not the priorities.
3898 (3) It is now possible to give a list of domains (or IP addresses) to be
3899 looked up. The behaviour when one of the lookups defers can be
3900 controlled by a keyword.
3902 (4) It is now possible to specify the separator character for use when
3903 multiple records are returned.
3905 38. The dnslists ACL condition has been extended: it is now possible to supply
3906 a list of IP addresses and/or domains to be looked up in a particular DNS
3909 39. Added log_selector=+queue_time_overall.
3911 40. When running the queue in the test harness, wait just a tad after forking a
3912 delivery process, to get repeatability of debugging output.
3914 41. Include certificate and key file names in error message when GnuTLS fails
3915 to set them up, because the GnuTLS error message doesn't include the name
3916 of the failing file when there is a problem reading it.
3918 42. Allow both -bf and -bF in the same test run.
3920 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3922 44. The "Exiscan patch" is now merged into the mainline Exim source.
3924 45. Sometimes the final signoff response after QUIT could fail to get
3925 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3926 < 0 before doing a fflush(). This bug looks as though it goes back to the
3927 introduction of TLS in release 3.20, but "sometimes" must have been rare
3928 because the tests only now provoked it.
3930 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3931 (this can affect the format of dates).
3933 47. exim_tidydb, when checking for the continued existence of a message for
3934 which it has found a message-specific retry record, was not finding
3935 messages that were in split spool directories. Consequently, it was
3936 deleting retry records that should have stayed in existence.
3938 48. Steve fixed some bugs in eximstats.
3940 49. The SPA authentication driver was not abandoning authentication and moving
3941 on to the next authenticator when an expansion was forced to fail,
3942 contradicting the general specification for all authenticators. Instead it
3943 was generating a temporary error. It now behaves as specified.
3945 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3946 (the order specifies the preference for clients). The order is now AES256,
3947 AES128, 3DES, ARCFOUR128.
3949 51. Small patch to Sieve code - explicitly set From: when generating an
3952 52. Exim crashed if a remote delivery caused a very long error message to be
3953 recorded - for instance if somebody sent an entire SpamAssassin report back
3954 as a large number of 550 error lines. This bug was coincidentally fixed by
3955 increasing the size of one of Exim's internal buffers (big_buffer) that
3956 happened as part of the Exiscan merge. However, to be on the safe side, I
3957 have made the code more robust (and fixed the comments that describe what
3960 53. Now that there can be additional text after "Completed" in log lines (if
3961 the queue_time_overall log selector is set), a one-byte patch to exigrep
3962 was needed to allow it to recognize "Completed" as not the last thing in
3965 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3966 patch that reportedly fixes this has been added. I am not expert enough to
3967 create a test for it. This is what the patch creator wrote:
3969 "I found a little strange behaviour of ldap code when working with
3970 Windows 2003 AD Domain, where users was placed in more than one
3971 Organization Units. When I tried to give exim partial DN, the exit code
3972 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3973 But simultaneously result of request was absolutely normal ldap result,
3974 so I produce this patch..."
3976 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3977 REFERENCE, so I have modified the code to exclude the patch when that macro
3980 55. Some experimental protocols are using DNS PTR records for new purposes. The
3981 keys for these records are domain names, not reversed IP addresses. The
3982 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3983 leaves it alone. Component reversal etc. now happens only for IP addresses.
3986 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3988 57. Double the size of the debug message buffer (to 2048) so that more of very
3989 long debug lines gets shown.
3991 58. The exicyclog utility now does better if the number of log files to keep
3992 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3994 59. Two changes related to the smtp_active_hostname option:
3996 (1) $smtp_active_hostname is now available as a variable.
3997 (2) The default for smtp_banner uses $smtp_active_hostname instead
3998 of $primary_hostname.
4000 60. The host_aton() function is supposed to be passed a string that is known
4001 to be a valid IP address. However, in the case of IPv6 addresses, it was
4002 not checking this. This is a hostage to fortune. Exim now panics and dies
4003 if the condition is not met. A case was found where this could be provoked
4004 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4005 components; fortuitously, this particular loophole had already been fixed
4006 by change 4.50/55 above.
4008 If there are any other similar loopholes, the new check in host_aton()
4009 itself should stop them being exploited. The report I received stated that
4010 data on the command line could provoke the exploit when Exim was running as
4011 exim, but did not say which command line option was involved. All I could
4012 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4013 running as the user.
4016 61. There was a buffer overflow vulnerability in the SPA authentication code
4017 (which came originally from the Samba project). I have added a test to the
4018 spa_base64_to_bits() function which I hope fixes it.
4021 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4022 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4024 63. The daemon start-up calls getloadavg() while still root for those OS that
4025 need the first call to be done as root, but it missed one case: when
4026 deliver_queue_load_max is set with deliver_drop_privilege. This is
4027 necessary for the benefit of the queue runner, because there is no re-exec
4028 when deliver_drop_privilege is set.
4030 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4031 This has been fixed.
4033 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4034 was in use, was not putting the data itself into the right store pool;
4035 consequently, it could be overwritten for a subsequent message in the same
4036 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4039 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4041 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4042 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4043 the function were treating the return as a boolean value, which happened to
4044 work because 0=false and not-0=true, but is not correct code.
4046 68. The host_aton() function was not handling scoped IPv6 addresses (those
4047 with, for example, "%eth0" on the end) correctly.
4049 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4050 negated items (that is, ~something) in unsigned ints. Some compilers
4051 apparently mutter when there is no cast.
4053 70. If an address verification called from an ACL failed, and did not produce a
4054 user-specific message (i.e. there was only a "system" message), nothing was
4055 put in $acl_verify_message. In this situation, it now puts the system
4058 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4059 session; change 4.31/43 added the unwanted input to the log line - except
4060 that it did not do this in the start of session case. It now does.
4062 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4063 This is wrong and can cause the other end to generate a synchronization
4064 error if it is another Exim or anything else that does the synchronization
4065 check. A QUIT command is no longer sent after a timeout.
4067 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4068 during host lookups.
4070 74. The defer_ok option of callout verification was not working if it was used
4071 when verifying addresses in header lines, that is, for this case:
4073 verify = header_sender/callout=defer_ok
4075 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4076 those file descriptors could be used for SMTP connections. If anything
4077 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4078 could be sent to the SMTP client, causing chaos. The daemon now opens
4079 stdin, stdout, and stderr to /dev/null when it puts itself into the
4082 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4083 log by default. The user can override this with suitable Perl magic.
4085 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4086 the log message when discard triggers, was not working for the DATA ACL or
4087 for the non-SMTP ACL.
4089 78. Error message wording change in sieve.c.
4091 79. If smtp_accept_max_per_host was set, the number of connections could be
4092 restricted to fewer than expected, because the daemon was trying to set up
4093 a new connection before checking whether the processes handling previous
4094 connections had finished. The check for completed processes is now done
4095 earlier. On busy systems, this bug wouldn't be noticed because something
4096 else would have woken the daemon, and it would have reaped the completed
4099 80. If a message was submitted locally by a user whose login name contained one
4100 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4101 It caused a spool format error. I have fixed the spool reading code. A
4102 related problem was that the "from" clause in the Received: line became
4103 illegal because of the space(s). It is now covered by ${quote_local_part.
4105 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4108 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4110 83. Patch from Sieve maintainer to fix unterminated string problem in
4111 "vacation" handling.
4113 84. Some minor changes to the Linux configuration files to help with other
4114 OS variants using glibc.
4116 85. One more patch for Sieve to update vacation handling to latest spec.
4119 ----------------------------------------------------
4120 See the note above about the 4.44 and 4.50 releases.
4121 ----------------------------------------------------
4127 1. Change 4.43/35 introduced a bug that caused file counts to be
4128 incorrectly computed when quota_filecount was set in an appendfile
4131 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4132 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4135 3. Give more explanation in the error message when the command for a transport
4136 filter fails to execute.
4138 4. There are several places where Exim runs a non-Exim command in a
4139 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4140 was being done only for the command run by the queryprogram router. It is
4141 now done for all such subprocesses. The other cases are: ${run, transport
4142 filters, and the commands run by the lmtp and pipe transports.
4144 5. Some older OS have a limit of 256 on the maximum number of file
4145 descriptors. Exim was using setrlimit() to set 1000 as a large value
4146 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4147 systems. I've change it so that if it can't get 1000, it tries for 256.
4149 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4150 was an oversight, and furthermore, ever since the addition of extra
4151 controls (e.g. 4.43/32), the checks on when to allow different forms of
4152 "control" were broken. There should now be diagnostics for all cases when a
4153 control that does not make sense is encountered.
4155 7. $recipients is now available in the predata ACL (oversight).
4157 8. Tidy the search cache before the fork to do a delivery from a message
4158 received from the command line. Otherwise the child will trigger a lookup
4159 failure and thereby defer the delivery if it tries to use (for example) a
4160 cached ldap connection that the parent has called unbind on.
4162 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4163 of $address_data from the recipient verification was clobbered by the
4164 sender verification.
4166 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4167 was its contents. (It was OK if the option was not defined at all.)
4169 11. A "Completed" log line is now written for messages that are removed from
4170 the spool by the -Mrm option.
4172 12. $host_address is now set to the target address during the checking of
4173 ignore_target_hosts.
4175 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4176 being passed; this would have caused $sender_host_name to have been used if
4177 matching the list had actually called for a host name (not very likely,
4178 since this list is usually IP addresses). A host name is now passed as
4181 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4182 code that specifies a non-restarting handler (typically sigaction() in
4183 modern systems) in an attempt to fix a rare and obscure crash bug.
4185 15. Narrowed the window for a race in the daemon that could cause it to ignore
4186 SIGCHLD signals. This is not a major problem, because they are used only to
4187 wake it up if nothing else does.
4189 16. A malformed maildirsize file could cause Exim to calculate negative values
4190 for the mailbox size or file count. Odd effects could occur as a result.
4191 The maildirsize information is now recalculated if the size or filecount
4194 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4195 support for a long time. Removed HAVE_SYS_VFS_H.
4197 18. Updated exipick to current release from John Jetmore.
4199 19. Allow an empty sender to be matched against a lookup in an address list.
4200 Previously the only cases considered were a regular expression, or an
4203 20. Exim went into a mad DNS lookup loop when doing a callout where the
4204 host was specified on the transport, if the DNS lookup yielded more than
4207 21. The RFC2047 encoding function was originally intended for short strings
4208 such as real names; it was not keeping to the 75-character limit for
4209 encoded words that the RFC imposes. It now respects the limit, and
4210 generates multiple encoded words if necessary. To be on the safe side, I
4211 have increased the buffer size for the ${rfc2047: expansion operator from
4214 22. Failure to deliver a bounce message always caused it to be frozen, even if
4215 there was an errors_to setting on the router. The errors_to setting is now
4218 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4219 canonical form (fully expanded) before being placed in
4220 $sender_host_address.
4222 24. Updated eximstats to version 1.33
4224 25. Include certificate and key file names in error message when GnuTLS fails
4225 to set them up, because the GnuTLS error message doesn't include the name
4226 of the failing file when there is a problem reading it.
4228 26. Expand error message when OpenSSL has problems setting up cert/key files.
4231 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4232 (this can affect the format of dates).
4234 28. exim_tidydb, when checking for the continued existence of a message for
4235 which it has found a message-specific retry record, was not finding
4236 messages that were in split spool directories. Consequently, it was
4237 deleting retry records that should have stayed in existence.
4239 29. eximstats updated to version 1.35
4240 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4241 1.35 - bugfix such that pie charts by volume are generated correctly
4243 30. The SPA authentication driver was not abandoning authentication and moving
4244 on to the next authenticator when an expansion was forced to fail,
4245 contradicting the general specification for all authenticators. Instead it
4246 was generating a temporary error. It now behaves as specified.
4248 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4249 (the order specifies the preference for clients). The order is now AES256,
4250 AES128, 3DES, ARCFOUR128.
4252 31. Small patch to Sieve code - explicitly set From: when generating an
4255 32. Exim crashed if a remote delivery caused a very long error message to be
4256 recorded - for instance if somebody sent an entire SpamAssassin report back
4257 as a large number of 550 error lines. This bug was coincidentally fixed by
4258 increasing the size of one of Exim's internal buffers (big_buffer) that
4259 happened as part of the Exiscan merge. However, to be on the safe side, I
4260 have made the code more robust (and fixed the comments that describe what
4263 33. Some experimental protocols are using DNS PTR records for new purposes. The
4264 keys for these records are domain names, not reversed IP addresses. The
4265 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4266 leaves it alone. Component reversal etc. now happens only for IP addresses.
4269 34. The host_aton() function is supposed to be passed a string that is known
4270 to be a valid IP address. However, in the case of IPv6 addresses, it was
4271 not checking this. This is a hostage to fortune. Exim now panics and dies
4272 if the condition is not met. A case was found where this could be provoked
4273 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4274 components; fortuitously, this particular loophole had already been fixed
4275 by change 4.50/55 or 4.44/33 above.
4277 If there are any other similar loopholes, the new check in host_aton()
4278 itself should stop them being exploited. The report I received stated that
4279 data on the command line could provoke the exploit when Exim was running as
4280 exim, but did not say which command line option was involved. All I could
4281 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4282 running as the user.
4285 35. There was a buffer overflow vulnerability in the SPA authentication code
4286 (which came originally from the Samba project). I have added a test to the
4287 spa_base64_to_bits() function which I hope fixes it.
4290 36. The daemon start-up calls getloadavg() while still root for those OS that
4291 need the first call to be done as root, but it missed one case: when
4292 deliver_queue_load_max is set with deliver_drop_privilege. This is
4293 necessary for the benefit of the queue runner, because there is no re-exec
4294 when deliver_drop_privilege is set.
4296 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4297 was in use, was not putting the data itself into the right store pool;
4298 consequently, it could be overwritten for a subsequent message in the same
4299 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4302 38. Sometimes the final signoff response after QUIT could fail to get
4303 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4304 < 0 before doing a fflush(). This bug looks as though it goes back to the
4305 introduction of TLS in release 3.20, but "sometimes" must have been rare
4306 because the tests only now provoked it.
4312 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4313 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4314 Now it returns an integer. A number of calls were still expecting a T/F
4315 return. Fortuitously, in all cases, the tests worked in OK situations,
4316 which is the norm. However, things would have gone wrong on any write
4317 failures on the smtp file descriptor. This function is used when sending
4318 messages over SMTP and also when doing verify callouts.
4320 2. When Exim is called to do synchronous delivery of a locally submitted
4321 message (the -odf or -odi options), it no longer closes stderr before doing
4324 3. Implemented the mua_wrapper option.
4326 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4328 5. Implemented the functions header_remove(), header_testname(),
4329 header_add_at_position(), and receive_remove_recipient(), and exported them
4332 6. If an ACL "warn" statement specified the addition of headers, Exim already
4333 inserted X-ACL-Warn: at the start if there was no header name. However, it
4334 was not making this test for the second and subsequent header lines if
4335 there were newlines in the string. This meant that an invalid header could
4336 be inserted if Exim was badly configured.
4338 7. Allow an ACL "warn" statement to add header lines at the start or after all
4339 the Received: headers, as well as at the end.
4341 8. Added the rcpt_4xx retry error code.
4343 9. Added postmaster_mailfrom=xxx to callout verification option.
4345 10. Added mailfrom=xxxx to the callout verification option, for verify=
4348 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4349 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4351 12. Inserted some casts to stop certain compilers warning when using pointer
4352 differences as field lengths or precisions in printf-type calls (mostly
4353 affecting debugging statements).
4355 13. Added optional readline() support for -be (dynamically loaded).
4357 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4358 same clock tick as a message's arrival, so that its received time was the
4359 same as the "first fail" time on the retry record, and that message
4360 remained on the queue past the ultimate address timeout, every queue runner
4361 would try a delivery (because it was past the ultimate address timeout) but
4362 after another failure, the ultimate address timeout, which should have then
4363 bounced the address, did not kick in. This was a "< instead of <=" error;
4364 in most cases the first failure would have been in the next clock tick
4365 after the received time, and all would be well.
4367 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4368 being recognized when the domain list was tested by the match_domain
4369 condition in an expansion string.
4371 16. Added the ${str2b64: operator.
4373 17. Exim was always calling setrlimit() to set a large limit for the number of
4374 processes, without checking whether the existing limit was already
4375 adequate. (It did check for the limit on file descriptors.) Furthermore,
4376 errors from getrlimit() and setrlimit() were being ignored. Now they are
4377 logged to the main and panic logs, but Exim does carry on, to try to do its
4378 job under whatever limits there are.
4380 18. Imported PCRE 5.0.
4382 19. Trivial typo in log message " temporarily refused connection" (the leading
4385 20. If the log selector return_path_on_delivery was set and an address was
4386 redirected to /dev/null, the delivery process crashed because it assumed
4387 that a return path would always be set for a "successful" delivery. In this
4388 case, the whole delivery is bypassed as an optimization, and therefore no
4391 21. Internal re-arrangement: the function for sending a challenge and reading
4392 a response while authentication was assuming a zero-terminated challenge
4393 string. It's now changed to take a pointer and a length, to allow for
4394 binary data in such strings.
4396 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4398 23. Exim was not respecting finduser_retries when seeking the login of the
4399 uid under which it was called; it was always trying 10 times. (The default
4400 setting of finduser_retries is zero.) Also, it was sleeping after the final
4401 failure, which is pointless.
4403 24. Implemented tls_on_connect_ports.
4405 25. Implemented acl_smtp_predata.
4407 26. If the domain in control=submission is set empty, Exim assumes that the
4408 authenticated id is a complete email address when it generates From: or
4409 Sender: header lines.
4411 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4412 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4413 chown and chgrp in /bin and hostname in /usr/bin.
4415 28. Exim was keeping the "process log" file open after each use, just as it
4416 does for the main log. This opens the possibility of it remaining open for
4417 long periods when the USR1 signal hits a daemon. Occasional processlog
4418 errors were reported, that could have been caused by this. Anyway, it seems
4419 much more sensible not to leave this file open at all, so that is what now
4422 29. The long-running daemon process does not normally write to the log once it
4423 has entered its main loop, and it closes the log before doing so. This is
4424 so that log files can straightforwardly be renamed and moved. However,
4425 there are a couple of unusual error situations where the daemon does write
4426 log entries, and I had neglected to close the log afterwards.
4428 30. The text of an SMTP error response that was received during a remote
4429 delivery was being truncated at 512 bytes. This is too short for some of
4430 the long messages that one sometimes sees. I've increased the limit to
4433 31. It is now possible to make retry rules that apply only when a message has a
4434 specific sender, in particular, an empty sender.
4436 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4437 it possible to be selective about when SMTP synchronization is enforced.
4439 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4441 32. Implemented hosts_connection_nolog.
4443 33. Added an ACL for QUIT.
4445 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4448 35. Added mailbox_size and mailbox_filecount to appendfile.
4450 36. Added control = no_multiline_responses to ACLs.
4452 37. There was a bug in the logic of the code that waits for the clock to tick
4453 in the case where the clock went backwards by a substantial amount such
4454 that the microsecond fraction of "now" was more than the microsecond
4455 fraction of "then" (but the whole seconds number was less).
4457 38. Added support for the libradius Radius client library this is found on
4458 FreeBSD (previously only the radiusclient library was supported).
4464 1. When certain lookups returned multiple values in the form name=value, the
4465 quoting of the values was not always being done properly. Specifically:
4466 (a) If the value started with a double quote, but contained no whitespace,
4468 (b) If the value contained whitespace other than a space character (i.e.
4469 tabs or newlines or carriage returns) it was not quoted.
4470 This fix has been applied to the mysql and pgsql lookups by writing a
4471 separate quoting function and calling it from the lookup code. The fix
4472 should probably also be applied to nisplus, ibase and oracle lookups, but
4473 since I cannot test any of those, I have not disturbed their existing code.
4475 2. A hit in the callout cache for a specific address caused a log line with no
4476 reason for rejecting RCPT. Now it says "Previous (cached) callout
4477 verification failure".
4479 3. There was an off-by-one bug in the queryprogram router. An over-long
4480 return line was truncated at 256 instead of 255 characters, thereby
4481 overflowing its buffer with the terminating zero. As well as fixing this, I
4482 have increased the buffer size to 1024 (and made a note to document this).
4484 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4485 when Exim is waiting for an SMTP response from a remote server, Exim
4486 restarts its select() call on the socket, thereby resetting its timeout.
4487 This is not a problem when such interrupts are rare. Somebody set up a cron
4488 job to run exiwhat every 2 minutes, which is less than the normal select()
4489 timeout (5 or 10 minutes). This meant that the select() timeout never
4490 kicked in because it was always reset. I have fixed this by comparing the
4491 time when an interrupt arrives with the time at the start of the first call
4492 to select(). If more time than the timeout has elapsed, the interrupt is
4493 treated as a timeout.
4495 5. Some internal re-factoring in preparation for the addition of Sieve
4496 extensions (by MH). In particular, the "personal" test is moved to a
4497 separate function, and given an option for scanning Cc: and Bcc: (which is
4498 not set for Exim filters).
4500 6. When Exim created an email address using the login of the caller as the
4501 local part (e.g. when creating a From: or Sender: header line), it was not
4502 quoting the local part when it contained special characters such as @.
4504 7. Installed new OpenBSD configuration files.
4506 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4507 try to make them clearer.
4509 9. Callout options, other than the timeout value, were being ignored when
4510 verifying sender addresses in header lines. For example, when using
4512 verify = header_sender/callout=no_cache
4514 the cache was (incorrectly) being used.
4516 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4517 only the Cygwin environment.
4519 11. When return_path_on_delivery was set as a log selector, if different remote
4520 addresses in the same message used different return paths and parallel
4521 remote delivery occurred, the wrong values would sometimes be logged.
4522 (Whenever a remote delivery process finished, the return path value from
4523 the most recently started remote delivery process was logged.)
4525 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4526 lines when AUTH and/or TLS are in use. This is the "received protocol"
4527 field. Exim used to use "asmtp" for authenticated SMTP, without any
4528 indication (in the protocol name) for TLS use. Now it follows the RFC and
4529 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4530 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4531 names appear in log lines as well as in Received: header lines.
4533 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4534 extensions, and comparison tests, and to fix some bugs.
4536 14. Changes to the "personal" filter test:
4538 (1) The test was buggy in that it was just doing the equivalent of
4539 "contains" tests on header lines. For example, if a user's address was
4540 anne@some.where, the "personal" test would incorrectly be true for
4542 To: susanne@some.where
4544 This test is now done by extracting each address from the header in turn,
4545 and checking the entire address. Other tests that are part of "personal"
4546 are now done using regular expressions (for example, to check local parts
4547 of addresses in From: header lines).
4549 (2) The list of non-personal local parts in From: addresses has been
4550 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4551 taken from the Sieve specification recommendations.
4553 (3) If the message contains any header line starting with "List-" it is
4554 treated as non-personal.
4556 (4) The test for "circular" in the Subject: header line has been removed
4557 because it now seems ill-conceived.
4559 15. Minor typos in src/EDITME comments corrected.
4561 16. Installed latest exipick from John Jetmore.
4563 17. If headers_add on a router specified a text string that was too long for
4564 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4565 of string_sprintf() is now avoided.
4567 18. $message_body_size was not set (it was always zero) when running the DATA
4568 ACL and the local_scan() function.
4570 19. For the "mail" command in an Exim filter, no default was being set for
4571 the once_repeat time, causing a random time value to be used if "once" was
4572 specified. (If the value happened to be <= 0, no repeat happened.) The
4573 default is now 0s, meaning "never repeat". The "vacation" command was OK
4574 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4575 (I found it when inspecting the code).
4577 20. There is now an overall timeout for performing a callout verification. It
4578 defaults to 4 times the callout timeout, which applies to individual SMTP
4579 commands during the callout. The overall timeout applies when there is more
4580 than one host that can be tried. The timeout is checked before trying the
4581 next host. This prevents very long delays if there are a large number of
4582 hosts and all are timing out (e.g. when the network connections are timing
4583 out). The value of the overall timeout can be changed by specifying an
4584 additional sub-option for "callout", called "maxwait". For example:
4586 verify = sender/callout=5s,maxwait=20s
4588 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4589 to the end before writing, but this should make it even safer).
4591 22. Exim was forgetting that it had advertised PIPELINING for the second and
4592 subsequent messages on an SMTP connection. It was also not resetting its
4593 memory on STARTTLS and an internal HELO.
4595 23. When Exim logs an SMTP synchronization error within a session, it now
4596 records whether PIPELINING has been advertised or not.
4598 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4599 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4600 rather than long int.
4602 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4604 26. Added the never_mail option to autoreply.
4610 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4611 crash if the getsockname() call failed; this can happen if a connection is
4612 closed very soon after it is established. The problem was simply in the
4613 order in which certain operations were done, causing Exim to try to write
4614 to the SMTP stream before it had set up the file descriptor. The bug has
4615 been fixed by making things happen in the correct order.
4621 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4622 before the connection was closed, thus losing the rejection response.
4624 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4625 some early Solaris releases, but causes trouble in current releases where
4626 socklen_t is defined.
4628 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4631 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4634 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4635 that releasing the top of it at the end releases what was used for sub-
4636 expansions (unless the block got too big). However, discard this block if
4637 the first thing is a variable or header, so that we can use its block when
4638 it is dynamic (useful for very large $message_headers, for example).
4640 6. Lookups now cache *every* query, not just the most recent. A new, separate
4641 store pool is used for this. It can be recovered when all lookup caches are
4642 flushed. Lookups now release memory at the end of their result strings.
4643 This has involved some general refactoring of the lookup sources.
4645 7. Some code has been added to the store_xxx() functions to reduce the amount
4646 of flapping under certain conditions.
4648 8. log_incoming_interface used to affect only the <= reception log lines. Now
4649 it causes the local interface and port to be added to several more SMTP log
4650 lines, for example "SMTP connection from", and rejection lines.
4652 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4654 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4656 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4657 could be overwritten at the end of the current message (or the start of a
4658 new message if it was set in a HELO ACL). The value is now preserved for
4659 the duration of the SMTP connection.
4661 12. If a transport had a headers_rewrite setting, and a matching header line
4662 contained an unqualified address, that address was qualified, even if it
4663 did not match any rewriting rules. The underlying bug was that the values
4664 of the flags that permit the existence of unqualified sender and recipient
4665 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4666 non-local messages, and by -bnq for local messages) were not being
4667 preserved with the message after it was received.
4669 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4670 "next input=" as part of the text comprising the host identity instead of
4671 the correct text. The code was using the same buffer for two different
4672 strings. However, depending on which order the printing function evaluated
4673 its arguments, the bug did not always show up. Under Linux, for example, my
4674 test suite worked just fine.
4676 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4677 doesn't work with some older versions of Perl. It has been changed to "my",
4678 which in any case is probably the better facility to use.
4680 15. A really picky compiler found some instances of statements for creating
4681 error messages that either had too many or two few arguments for the format
4684 16. The size of the buffer for calls to the DNS resolver has been increased
4685 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4686 for addresses that have a lot of PTR records. This alleviates a problem; it
4687 does not fully solve it.
4689 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4690 buffer now truncates the list and logs the incident, which is the same
4691 action as happens when Exim is looking up a host name and its aliases.
4692 Previously in this situation something unpredictable would happen;
4693 sometimes it was "internal error: store_reset failed".
4695 18. If a server dropped the connection unexpectedly when an Exim client was
4696 using GnuTLS and trying to read a response, the client delivery process
4697 crashed while trying to generate an error log message.
4699 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4700 string, for example:
4702 warn message = H1: something\nH2: something
4704 the text was added as a single header line from Exim's point of view
4705 though it ended up OK in the delivered message. However, searching for the
4706 second and subsequent header lines using $h_h2: did not work. This has been
4707 fixed. Similarly, if a system filter added multiple headers in this way,
4708 the routers could not see them.
4710 20. Expanded the error message when iplsearch is called with an invalid key to
4711 suggest using net-iplsearch in a host list.
4713 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4714 ACLs are no longer actually imposed (and a message to that effect is
4717 22. If a "gecos" field in a passwd entry contained escaped characters, in
4718 particular, if it contained a \" sequence, Exim got it wrong when building
4719 a From: or a Sender: header from that name. A second bug also caused
4720 incorrect handling when an unquoted " was present following a character
4721 that needed quoting.
4723 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4724 was not being matched caselessly.
4726 24. Arranged for all hyphens in the exim.8 source to be escaped with
4729 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4730 properly. Recipient callout cache records were still being keyed to include
4731 the sender, even when use_sender was set false. This led to far more
4732 callouts that were necessary. The sender is no longer included in the key
4733 when use_sender is false.
4735 26. Added "control = submission" modifier to ACLs.
4737 27. Added the ${base62d: operator to decode base 62 numbers.
4739 28. dnsdb lookups can now access SRV records.
4741 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4742 the configuration file.
4744 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4745 (-v) mode. This makes the output for a verbose queue run more intelligible.
4747 31. Added a use_postmaster feature to recipient callouts.
4749 32. Added the $body_zerocount variable, containing the number of binary zero
4750 bytes in the message body.
4752 33. The time of last modification of the "new" subdirectory is now used as the
4753 "mailbox time last read" when there is a quota error for a maildir
4756 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4758 35. Added +ignore_unknown as a special item in host lists.
4760 36. Code for decoding IPv6 addresses in host lists is now included, even if
4761 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4762 address was recognized as an IP address, but was then not correctly decoded
4763 into binary, causing unexpected and incorrect effects when compared with
4770 1. Very minor rewording of debugging text in manualroute to say "list of
4771 hosts" instead of "hostlist".
4773 2. If verify=header_syntax was set, and a header line with an unqualified
4774 address (no domain) and a large number of spaces between the end of the
4775 name and the colon was received, the reception process suffered a buffer
4776 overflow, and (when I tested it) crashed. This was caused by some obsolete
4777 code that should have been removed. The fix is to remove it!
4779 3. When running in the test harness, delay a bit after writing a bounce
4780 message to get a bit more predictability in the log output.
4782 4. Added a call to search_tidyup() just before forking a reception process. In
4783 theory, someone could use a lookup in the expansion of smtp_accept_max_
4784 per_host which, without the tidyup, could leave open a database connection.
4786 5. Added the variables $recipient_data and $sender_data which get set from a
4787 lookup success in an ACL "recipients" or "senders" condition, or a router
4788 "senders" option, similar to $domain_data and $local_part_data.
4790 6. Moved the writing of debug_print from before to after the "senders" test
4793 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4794 problems for message scanning, either using a data ACL, or using
4795 local_scan() because the Received: header was not generated till after they
4796 were called (in order to set the time as the time of reception completion).
4797 I have revised the way this works. The header is now generated after the
4798 body is received, but before the ACL or local_scan() are called. After they
4799 are run, the timestamp in the header is updated.
4805 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4806 before starting a queue runner without re-exec. This happened only when
4807 deliver_drop_privilege was set or when the Exim user was set to root. The
4808 effect of the bug was that timeouts during subsequent deliveries caused
4809 crashes instead of being properly handled. The handler is now left at its
4810 default (and expected) setting.
4812 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4813 message, again when deliver_drop_privilege is set or Exim is run as root.
4814 The bug described in (1) was not present in this case, but the tidying up
4815 of the other signals was missing. I have made the two cases consistent.
4817 3. The ignore_target_hosts setting on a manualroute router was being ignored
4818 for hosts that were looked up using the /MX notation.
4820 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4823 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4824 operated on the sender address. After changing the $sender_address to <>
4825 for the sender address verify, Exim was re-instated it as the original
4826 (before rewriting) address, but remembering that it had rewritten it, so it
4827 wasn't rewriting it again. This bug also had the effect of breaking the
4828 sender address verification caching when the sender address was rewritten.
4830 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4831 This has been changed so that if the ip literal address matches
4832 ignore_target_hosts, the router declines.
4834 7. Added expansion conditions match_domain, match_address, and match_local_
4835 part (NOT match_host).
4837 8. The placeholder for the Received: header didn't have a length field set.
4839 9. Added code to Exim itself and to exim_lock to test for a specific race
4840 condition that could lead to file corruption when using MBX delivery. The
4841 issue is with the lockfile that is created in /tmp. If this file is removed
4842 after a process has opened it but before that process has acquired a lock,
4843 there is the potential for a second process to recreate the file and also
4844 acquire a lock. This could lead to two Exim processes writing to the file
4845 at the same time. The added code performs the same test as UW imapd; it
4846 checks after acquiring the lock that its file descriptor still refers to
4847 the same named file.
4849 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4850 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4853 11. Added the smtp_active_hostname option. If used, this will typically be made
4854 to depend on the incoming interface address. Because $interface_address is
4855 not set up until the daemon has forked a reception process, error responses
4856 that can happen earlier (such as "too many connections") no longer contain
4859 12. If an expansion in a condition on a "warn" statement fails because a lookup
4860 defers, the "warn" statement is abandoned, and the next ACL statement is
4861 processed. Previously this caused the whole ACL to be aborted.
4863 13. Added the iplsearch lookup type.
4865 14. Added ident_timeout as a log selector.
4867 15. Added tls_certificate_verified as a log selector.
4869 16. Added a global option tls_require_ciphers (compare the smtp transport
4870 option of the same name). This controls incoming TLS connections.
4872 17. I finally figured out how to make tls_require_ciphers do a similar thing
4873 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4874 before starting the TLS session.
4876 18. Tabs are now shown as \t in -bP output.
4878 19. If the log selector return_path_on_delivery was set, Exim crashed when
4879 bouncing a message because it had too many Received: header lines.
4881 20. If two routers both had headers_remove settings, and the first one included
4882 a superfluous trailing colon, the final name in the first list and the
4883 first name in the second list were incorrectly joined into one item (with a
4884 colon in the middle).
4890 1. Added -C and -D options to the exinext utility, mainly to make it easier
4891 to include in the automated testing, but these could be helpful when
4892 multiple configurations are in use.
4894 2. The exinext utility was not formatting the output nicely when there was
4895 an alternate port involved in the retry record key, nor when there was a
4896 message id as well (for retries that were specific to a specific message
4897 and a specific host). It was also confused by IPv6 addresses, because of
4898 the additional colons they contain. I have fixed the IPv4 problem, and
4899 patched it up to do a reasonable job for IPv6.
4901 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4902 delivery, the log line now contains "pipelined" if PIPELINING was used.
4904 4. An SMTP transport process used to panic and die if the bind() call to set
4905 an explicit outgoing interface failed. This has been changed; it is now
4906 treated in the same way as a connect() failure.
4908 5. A reference to $sender_host_name in the part of a conditional expansion
4909 that was being skipped was still causing a DNS lookup. This no longer
4912 6. The def: expansion condition was not recognizing references to header lines
4913 that used bh_ and bheader_.
4915 7. Added the _cache feature to named lists.
4917 8. The code for checking quota_filecount in the appendfile transport was
4918 allowing one more file than it should have been.
4920 9. For compatibility with Sendmail, the command line option
4928 and sets the incoming protocol and host name (for trusted callers). The
4929 host name and its colon can be omitted when only the protocol is to be set.
4930 Note the Exim already has two private options, -pd and -ps, that refer to
4931 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4932 "s", but I don't think that's a major issue.
4934 10. A number of refactoring changes to the code, none of which should affect
4937 (a) The number of logging options was getting close to filling up the
4938 32-bit word that was used as a bit map. I have split them into two classes:
4939 those that are passed in the argument to log_write(), and those that are
4940 only ever tested independently outside of that function. These are now in
4941 separate 32-bit words, so there is plenty of room for expansion again.
4942 There is no change in the user interface or the logging behaviour.
4944 (b) When building, for example, log lines, the code previously used a
4945 macro that called string_cat() twice, in order to add two strings. This is
4946 not really sufficiently general. Furthermore, there was one instance where
4947 it was actually wrong because one of the argument was used twice, and in
4948 one call a function was used. (As it happened, calling the function twice
4949 did not affect the overall behaviour.) The macro has been replaced by a
4950 function that can join an arbitrary number of extra strings onto a growing
4953 (c) The code for expansion conditions now uses a table and a binary chop
4954 instead of a serial search (which was left over from when there were very
4955 few conditions). Also, it now recognizes conditions like "pam" even when
4956 the relevant support is not compiled in: a suitably worded error message is
4957 given if an attempt is made to use such a condition.
4959 11. Added ${time_interval:xxxxx}.
4961 12. A bug was causing one of the ddress fields not to be passed back correctly
4962 from remote delivery subprocesses. The field in question was not being
4963 subsequently used, so this caused to problems in practice.
4965 13. Added new log selectors queue_time and deliver_time.
4967 14. Might have fixed a bug in maildirsizefile handling that threw up
4968 "unexpected character" debug warnings, and recalculated the data
4969 unnecessarily. In any case, I expanded the warning message to give more
4972 15. Added the message "Restricted characters in address" to the statements in
4973 the default ACL that block characters like @ and % in local parts.
4975 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4976 Three changes have been made:
4978 (a) There was a serious bug; a negative response to MAIL caused the whole
4979 recipient domain to be cached as invalid, thereby blocking all messages
4980 to all local parts at the same domain, from all senders. This bug has
4981 been fixed. The domain is no longer cached after a negative response to
4982 MAIL if the sender used is not empty.
4984 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4987 (c) A new callout option, "use_sender" has been added for people who want
4988 the modified behaviour.
4994 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4997 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4998 indeed breaks things for older releases.
5000 3. Added additional logging to the case where there is a problem reading data
5001 from a filter that is running in a subprocess using a pipe, in order to
5002 try to track down a specific problem.
5004 4. Testing facility fudge: when running in the test harness and attempting
5005 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5006 getting "No route to host". Convert this to a timeout.
5008 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5011 6. Some OS don't have socklen_t but use size_t instead. This affects the
5012 fifth argument of getsockopt() amongst other things. This is now
5013 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5014 can be set for individual OS. I have set it for SunOS5, OSF1, and
5015 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5016 some earlier ones do not.
5018 7. Change 4.30/15 was not doing the test caselessly.
5020 8. The standard form for an IPv6 address literal was being rejected by address
5021 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5022 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5023 this, as well as the form without the "IPv6" on the front (but only when
5024 address literals are enabled, of course).
5026 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5028 10. Exim crashed if a message with an empty sender address specified by -f
5029 encountered a router with an errors_to setting. This could be provoked only
5030 by a command such as
5034 where an empty string was supplied; "<>" did not hit this bug.
5036 11. Installed PCRE release 4.5.
5038 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5039 remained set. It is now erased.
5041 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5042 times from message ids (which are base 36 rather than the normal 62).
5044 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5045 were being counted as actual protocol errors, and logged if the log
5046 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5047 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5048 and DATA following a set of rejected RCPTs do not count as protocol errors.
5049 In other words, Exim assumes they were pipelined, though this may not
5050 actually be the case. Of course, in all cases the client gets an
5051 appropriate error code.
5053 15. If a lookup fails in an ACL condition, a message about the failure may
5054 be available; it is used if testing the ACL cannot continue, because most
5055 such messages specify what the cause of the deferral is. However, some
5056 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5057 that caused an old message to be retained and used if a later statement
5058 caused a defer, replacing the real cause of the deferral.
5060 16. If an IP address had so many PTR records that the DNS lookup buffer
5061 was not large enough to hold them, Exim could crash while trying to process
5062 the truncated data. It now detects and logs this case.
5064 17. Further to 4.21/58, another change has been made: if (and only if) the
5065 first line of a message (the first header line) ends with CRLF, a bare LF
5066 in a subsequent header line has a space inserted after it, so as not to
5067 terminate the header.
5069 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5070 unnecessarily, used atoi() instead of strtol(), and didn't check the
5071 termination when getting file sizes from file names by regex.
5073 19. Completely re-implemented the support for maildirsize files, in the light
5074 of a number of problems with the previous contributed implementation
5075 (4.30/29). In particular:
5077 . If the quota is zero, the maildirsize file is maintained, but no quota is
5080 . If the maildir directory does not exist, it is created before any attempt
5081 to write a maildirsize file.
5083 . The quota value in the file is just a cache; if the quota is changed in
5084 the transport, the new value overrides.
5086 . A regular expression is available for excluding directories from the
5089 20. The autoreply transport checks the characters in options that define the
5090 message's headers; it allows continued headers, but it was checking with
5091 isspace() after an embedded newline instead of explicitly looking for a
5094 21. If all the "regular" hosts to which an address was routed had passed their
5095 expiry times, and had not reached their retry times, the address was
5096 bounced, even if fallback hosts were defined. Now Exim should go on to try
5099 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5100 equivalent code in the SMTP transport. Some hosts send humungous responses
5101 to HELO/EHLO, more than 1024 it seems.
5103 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5104 gives compiler warnings in some environments. I've now done it "properly",
5107 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5108 (because of problems with the built-in one) was declared to return uschar *
5109 instead of char *, causing compiler failure.
5111 25. Fixed a file descriptor leak when processing alias/forward files.
5113 26. Fixed a minor format string issue in dbfn.c.
5115 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5117 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5118 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5121 29. When a sender address is verified, it is cached, to save repeating the test
5122 when there is more than one recipient in a message. However, when the
5123 verification involves a callout, it is possible for different callout
5124 options to be set for different recipients. It is too complicated to keep
5125 track of this in the cache, so now Exim always runs a verification when a
5126 callout is required, relying on the callout cache for the optimization.
5127 The overhead is duplication of the address routing, but this should not be
5130 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5131 to be verified with callout=postmaster, and the main callout worked but the
5132 postmaster check failed, the verification correctly failed. However, if a
5133 subsequent RCPT command asked for sender verification *without* the
5134 postmaster check, incorrect caching caused this verification also to fail,
5137 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5138 it was not caching the DNS options (qualify_single, search_parents) that
5139 were used when the lookup failed. A subsequent lookup with different
5140 options therefore always gave the same answer, though there were cases
5141 where it should not have. (Example: a "domains = !$mx_any" option on a
5142 dnslookup router: the "domains" option is always processed without any
5143 widening, but the router might have qualify_single set.) Now Exim uses the
5144 cached value only when the same options are set.
5146 32. Added John Jetmore's "exipick" utility to the distribution.
5148 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5149 than a timeout (e.g. a certificate is required, and is not provided), an
5150 Exim server now closes the connection immediately. Previously it waited for
5151 the client to close - but if the client is SSL, it seems that they each
5152 wait for each other, leading to a delay before one of them times out.
5154 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5155 maintained 0.8.x compatibility because I don't think many are using it, and
5156 it is clearly obsolete.
5158 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5161 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5162 client certificate was expired. A simple patch fixes this, though I don't
5163 understand the full logic of why the verify callback is called multiple
5166 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5167 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5168 which causes problems with some clients (such as the Certicom SSL Plus
5169 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5170 disables the coutermeasure allowing Eudora to connect."
5172 38. Exim was not checking that a write() to a log file succeeded. This could
5173 lead to Bad Things if a log got too big, in particular if it hit a file
5174 size limit. Exim now panics and dies if it cannot write to a log file, just
5175 as it does if it cannot open a log file.
5177 39. Modified OS/Makefile-Linux so that it now contains
5179 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5181 The two -D definitions ensure that Exim is compiled with large file
5182 support, which makes it possible to handle log files that are bigger than
5185 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5186 instance) a domain was checked against a named list that involved a lookup,
5187 causing $domain_data to be set, then another domain was checked against the
5188 same list, then the first domain was re-checked, the value of $domain_data
5189 after the final check could be wrong. In particular, if the second check
5190 failed, it could be set empty. This bug probably also applied to
5193 41. The strip_trailing_dot option was not being applied to the address given
5194 with the -f command-line option.
5196 42. The code for reading a message's header from the spool was incrementing
5197 $received_count, but never initializing it. This meant that the value was
5198 incorrect (doubled) while delivering a message in the same process in which
5199 it was received. In the most common configuration of Exim, this never
5200 happens - a fresh exec is done - but it can happen when
5201 deliver_drop_privilege is set.
5203 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5204 it now includes up to 150 characters of the unexpected data in the log
5207 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5208 and building data strings. The size of both of these buffers was 10 000
5209 bytes - far larger than anybody would *ever* want, thought I. Needless to
5210 say, somebody hit the limit. I have increased the maximum line length to
5211 20 000 and the maximum data length of concatenated lines to 100 000. I have
5212 also fixed two bugs, because there was no checking on these buffers. Tsk,
5213 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5214 buffer is too small.
5216 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5217 lsearch lookups. Now it does.
5219 46. When parsing a route_list item in a manualroute router, a fixed-length
5220 buffer was used for the list of hosts. I made this 1024 bytes long,
5221 thinking that nobody would ever have a list of hosts that long. Wrong.
5222 Somebody had a whole pile of complicated expansion conditions, and the
5223 string was silently truncated, leading to an expansion error. It turns out
5224 that it is easier to change to an unlimited length (owing to other changes
5225 that have happened since this code was originally written) than to build
5226 structure for giving a limitation error. The length of the item that
5227 expands into the list of hosts is now unlimited.
5229 47. The lsearch lookup could not handle data where the length of text line was
5230 more than 4095 characters. Such lines were truncated, leading to shortened
5231 data being returned. It should now handle lines of any length.
5233 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5234 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5237 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5238 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5239 no longer happens in comment lines. A list of replacements is now placed
5240 at the head of all of the source files, except those whose only change is
5241 to replace PERL_COMMAND in the very first #! line.
5243 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5244 messages on the queue, with a bottom-up merge sort, using code contributed
5245 by Michael Haardt. This should make operations like -bp somewhat faster on
5246 large queues. It won't affect queue runners, except when queue_run_in_order
5249 51. Installed eximstats 1.31 in the distribution.
5251 52. Added support for SRV lookups to the dnslookup router.
5253 53. If an ACL referred to $message_body or $message_body_end, the value was not
5254 reset for any messages that followed in the same SMTP session.
5256 54. The store-handling optimization for building very long strings was not
5257 differentiating between the different store pools. I don't think this
5258 actually made any difference in practice, but I've tidied it.
5260 55. While running the routers to verify a sender address, $sender_address
5261 was still set to the sender address. This is wrong, because when routing to
5262 send a bounce to the sender, it would be empty. Therefore, I have changed
5263 it so that, while verifying a sender address, $sender_address is set to <>.
5264 (There is no change to what happens when verifying a recipient address.)
5266 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5267 target A or AAAA records (if not already returned) without resetting the
5268 qualify_single or search_parents options of the DNS resolver. These are
5269 inappropriate in this case because the targets of MX and SRV records must
5270 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5271 target that, when qualified, matched something in the local domain. These
5272 two options are now turned off when doing these lookups.
5274 57. It seems that at least some releases of Reiserfs (which does not have the
5275 concept of a fixed number of inodes) returns zero and not -1 for the
5276 number of available inodes. This interacted badly with check_spool_inodes,
5277 which assumed that -1 was the "no such thing" setting. What I have done is
5278 to check that the total number of inodes is greater than zero before doing
5279 the test of how many are available.
5281 58. When a "warn" ACL statement has a log_message modifier, the message is
5282 remembered, and not repeated. This is to avoid a lot of repetition when a
5283 message has many recipients that cause the same warning to be written.
5284 Howewer, Exim was preserving the list of already written lines for an
5285 entire SMTP session, which doesn't seem right. The memory is now reset if a
5286 new message is started.
5288 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5289 debugging output unless log_rewrite was also set.
5291 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5292 of (int)(handle) when we know that handle contains (void *)(-1).
5294 61. The Exim daemon panic-logs an error return when it closes the incoming
5295 connection. However "connection reset by peer" seems to be common, and
5296 isn't really an error worthy of noting specially, so that particular error
5299 62. When Exim is trying to find all the local interfaces, it used to panic and
5300 die if the ioctl to get the interface flags failed. However, it seems that
5301 on at least one OS (Solaris 9) it is possible to have an interface that is
5302 included in the list of interfaces, but for which you get a failure error
5303 for this call. This happens when the interface is not "plumbed" into a
5304 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5305 failure of the "get flags" call assumes that the interface is down.
5307 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5308 makes life easier for people who are doing arithmetic on fields extracted
5309 from dates, where you often get leading zeros that should not be
5310 interpreted as octal.
5312 64. Added qualify_domain to the redirect router, to override the global
5315 65. If a pathologically long header line contained very many addresses (the
5316 report of this problem mentioned 10 000) and each of them was rewritten,
5317 Exim could use up a very large amount of memory. (It kept on making new
5318 copies of the header line as it rewrote, and never released the old ones.)
5319 At the expense of a bit more processing, the header rewriting function has
5320 been changed so that it no longer eats memory in this way.
5322 66. The generation of the Received: header has been moved from the time that a
5323 message starts to be received, to the time that it finishes. The timestamp
5324 in the Received: header should now be very close to that of the <= log
5325 line. There are two side-effects of this change:
5327 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5328 logged header lines no longer include the local Received: line, because
5329 it has not yet been created. The same applies to a copy of the message
5330 that is returned to a non-SMTP sender when a message is rejected.
5332 (b) When a filter file is tested using -bf, no additional Received: header
5333 is added to the test message. After some thought, I decided that this
5336 This change does not affect the value of $received_for. It is still set
5337 after address rewriting, but before local_scan() is called.
5339 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5341 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5342 gave an unhelpful panic error message, and a defer error. I have managed to
5343 change this behaviour so that it now rejects any supplied certificate,
5344 which seems right, as the list of acceptable certificates is empty.
5346 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5347 gave an unhelpful defer error. I have not managed to make this reject any
5348 supplied certificates, but the error message it gives is "no certificate
5349 supplied", which is not helpful.
5351 70. exigrep's output now also includes lines that are not associated with any
5352 message, but which match the given pattern. Implemented by a patch from
5353 Martin Sluka, which also tidied up the Perl a bit.
5355 71. Recipient callout verification, like sender verification, was using <> in
5356 the MAIL FROM command. This isn't really the right thing, since the actual
5357 sender may affect whether the remote host accepts the recipient or not. I
5358 have changed it to use the actual sender in the callout; this means that
5359 the cache record is now keyed on a recipient/sender pair, not just the
5360 recipient address. There doesn't seem to be a real danger of callout loops,
5361 since a callout by the remote host to check the sender would use <>.
5362 [SEE ABOVE: changed after hitting problems.]
5364 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5365 temporary errors. However, in the case of such a code being given after
5366 the end of a data transmission (i.e. after ".") Exim was failing to write
5367 a retry record for the message. (Yes, there was some broken host that was
5368 actually sending 8xx at this point.)
5370 73. An unknown lookup type in a host list could cause Exim to panic-die when
5371 the list was checked. (An example that provoked this was putting <; in the
5372 middle of a list instead of at the start.) If this happened during a DATA
5373 ACL check, a -D file could be left lying around. This kind of configuration
5374 error no longer causes Exim to die; instead it causes a defer errror. The
5375 incident is still logged to the main and panic logs.
5377 74. Buglet left over from Exim 3 conversion. The message "too many messages
5378 in one connection" was written to the rejectlog but not the mainlog, except
5379 when address rewriting (yes!) was being logged.
5381 75. Added write_rejectlog option.
5383 76. When a system filter was run not as root (that is, when system_filter_user
5384 was set), the values of the $n variables were not being returned to the
5385 main process; thus, they were not subsequently available in the $sn
5388 77. Added +return_path_on_delivery log selector.
5390 78. A connection timeout was being treated differently from recipients deferred
5391 when testing hosts_max_try with a message that was older than the host's
5392 retry timeout. (The host should not be counted, thus allowing all hosts to
5393 be tried at least once before bouncing.) This may have been the cause of an
5394 occasionally reported bug whereby a message would remain on the queue
5395 longer than the retry timeout, but would be bounced if a delivery was
5396 forced. I say "may" because I never totally pinned down the problem;
5397 setting up timeout/retry tests is difficult. See also the next item.
5399 79. The ultimate address timeout was not being applied to errors that involved
5400 a combination of host plus message (for example, a timeout on a MAIL
5401 command). When an address resolved to a number of possible hosts, and they
5402 were not all tried for each delivery (e.g. because of hosts_max_try), a
5403 message could remain on the queue longer than the retry timeout.
5405 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5408 81. Fixed an obscure SMTP outgoing bug which required at least the following
5409 conditions: (a) there was another message waiting for the same server;
5410 (b) the server returned 5xx to all RCPT commands in the first message so
5411 that the message was not completed; (c) the server dropped the connection
5412 or gave a negative response to the RSET that Exim sends to abort the
5413 transaction. The observed case was a dropped connection after DATA that had
5414 been sent in pipelining mode. That is, the server had advertised PIPELINING
5415 but was not implementing it correctly. The effect of the bug was incorrect
5416 behaviour, such as trying another host, and this could lead to a crash.
5422 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5423 and daemon.c were passed as pointers to ints; they should have been
5424 pointers to socklen_t variables (which are typically unsigned ints).
5426 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5429 3. Fixed a really odd bug that affected only the testing scheme; patching a
5430 certain fixed string in the binary changed the value of another string that
5431 happened to be identical to the end of the original first string.
5433 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5434 name", it returns that address as the IP address. On some operating
5435 systems (e.g. Solaris), it also passes back the IP address string as the
5436 "host name". However, on others (e.g. Linux), it passes back an empty
5437 string. Exim wasn't checking for this, and was changing the host name to an
5438 empty string, assuming it had been canonicized.
5440 5. Although rare, it is permitted to have more than one PTR record for a given
5441 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5442 all the names associated with an address, because they do in Solaris.
5443 However, it seems that they do not in Linux for data that comes from the
5444 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5445 I found this out when I moved to a new Linux workstation and tried to run
5446 the Exim test suite.
5448 To get round this problem I have changed the code so that it now does its
5449 own call to the DNS to look up PTR records when searching for a host name.
5450 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5451 addresses that are only in /etc/hosts are still found.
5453 This behaviour is, however, controlled by an option called host_lookup_
5454 order, which defaults to "bydns:byaddr". If people want to use the other
5455 order, or indeed, just use one or the other means of lookup, they can
5456 specify it in this variable.
5458 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5459 some operating systems, this comes back from gethostbyaddr() as an empty
5460 string, and this is what Exim used to test for. However, it seems that in
5461 other systems, "." is yielded. Exim now tests for this case too.
5463 7. The values of check_spool_space and check_log_space are now held internally
5464 as a number of kilobytes instead of an absolute number of bytes. If a
5465 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5466 kilobyte. This means that much larger values can be stored.
5468 8. Exim monitor: an attempt to get the action menu when not actually pointing
5469 at a message produces an empty menu entitled "No message selected". This
5470 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5471 no entries in it ("Shell widget menu has zero width and/or height"). So I
5472 have added a single, blank menu entry in this case.
5474 9. Added ${quote_local_part.
5476 10. MIME decoding is now applied to the contents of Subject: header lines when
5479 11. Now that a reference to $sender_host_address automatically causes a reverse
5480 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5481 host lookup before query-style lookups in lists that might use this
5482 variable. This has therefore been abolished, and the "net-" prefix is no
5483 longer necessary for query-style lookups.
5485 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5486 have been a typo for LFLAGS, so it has been changed.
5488 13. The install script calls Exim with "-C /dev/null" in order to find the
5489 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5490 to be output. Howeve, since Exim outputs its version number before the
5491 error, it didn't break the script. It just looked ugly. I fixed this by
5492 always allowing "-C /dev/null" if the caller is root.
5494 14. Ignore overlarge ACL variable number when reading spool file - insurance
5495 against a later release with more variables having written the file.
5497 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5498 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5499 this, as well as the form without the "IPv6" on the front.
5501 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5502 OS/Makefile-Darwin file.
5504 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5505 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5507 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5508 variable. However, some OS use ints for inodes. I've added cast to long int
5509 to get rid of the compiler warning.
5511 19. I had forgotten to lock out "/../" in configuration file names when
5512 ALT_CONFIG_PREFIX was set.
5514 20. Routers used for verification do not need to specify transports. However,
5515 if such a router generated a host list, and callout was configured, Exim
5516 crashed, because it could not find a port number from the (non-existent)
5517 transport. It now assumes port 25 in this circumstance.
5519 21. Added the -t option to exigrep.
5521 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5522 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5523 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5524 them is set, LOOKUP_LSEARCH is forced.
5526 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5528 24. Added sender and host information to the "rejected by local_scan()" log
5529 line; previously there was no indication of these.
5531 25. Added .include_if_exists.
5533 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5534 Linux. It turns out that not all file systems support this. Apparently some
5535 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5536 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5537 not supported on the file descriptor, is now ignored when Exim is trying to
5538 sync a directory. This applies only to Linux.
5540 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5542 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5543 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5546 29. Added support for maildirsize files from supplied patch (modified a bit).
5548 30. The use of :fail: followed by an empty string could lead Exim to respond to
5549 sender verification failures with (e.g.):
5551 550 Verification failed for <xxx>
5552 550 Sender verify failed
5554 where the first response line was missing the '-' that indicates it is not
5555 the final line of the response.
5557 31. The loop for finding the name of the user that called Exim had a hardwired
5558 limit of 10; it now uses the value of finduser_retries, which is used for
5559 all other user lookups.
5561 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5564 33. Exim was neglecting to zero errno before one call of strtol() when
5565 expanding a string and expecting an integer value. On some systems this
5566 resulted in spurious "integer overflow" errors. Also, it was casting the
5567 result into an int without checking.
5569 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5570 did not work. The code looks as if it has *never* worked, though it appears
5571 to have been documented since at least releast 1.62. I have made it work.
5573 35. The "timeout_DNS" error in retry rules, also documented since at least
5574 1.62, also never worked. As it isn't clear exactly what this means, and
5575 clearly it isn't a major issue, I have abolished the feature by treating it
5576 as "timeout", and writing a warning to the main and panic logs.
5578 36. The display of retry rules for -brt wasn't always showing the error code
5581 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5582 timeout_connect_A, timeout_connect_MX.
5584 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5585 to the empty sender.
5587 39. The daemon was not analysing the content of -oX till after it had closed
5588 stderr and disconnected from the controlling terminal. This meant that any
5589 syntax errors were only noted on the panic log, and the return code from
5590 the command was 0. By re-arranging the code a little, I've made the
5591 decoding happen first, so such errors now appear on stderr, and the return
5592 code is 1. However, the actual setting up of the sockets still happens in
5593 the disconnected process, so errors there are still only recorded on the
5596 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5597 connections (as happens on some IP stacks) was logged at start up time as
5598 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5599 it from "IPv6 and IPv4", which means that two separate sockets are being
5602 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5603 says whether AF_INET or AF_INET6 was passed as an argument.
5605 42. Exiwhat output was messed up when time zones were included in log
5608 43. Exiwhat now gives more information about the daemon's listening ports,
5609 and whether -tls-on-connect was used.
5611 44. The "port" option of the smtp transport is now expanded.
5613 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5614 silently ignored. Now an error message is written to the main and panic
5617 46. There's a new ACL modifier called "logwrite" which writes to a log file
5618 as soon as it is encountered.
5620 47. Added $local_user_uid and $local_user_gid at routing time.
5622 48. Exim crashed when trying to verify a sender address that was being
5625 49. Exim was recognizing only a space character after ".include". It now also
5626 recognizes a tab character.
5628 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5629 extracting the relevant information from the specification. The man page no
5630 longer contains scrambled data for the -d option, and I've added a section
5631 at the front about calling Exim under different names.
5633 51. Added "extra_headers" argument to the "mail" command in filter files.
5635 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5638 53. Installed eximstats 1.29.
5640 54. Added transport_filter_timeout as a generic transport option.
5642 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5643 Cc: header lines. This was required by RFC 822, but it not required by RFC
5646 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5647 incoming messages that did not have them. Now it does so only if the
5648 message originates locally, that is, if there is no associated remote host
5649 address. When Resent- header lines are present, this applies to the Resent-
5650 lines rather than the non-Resent- lines.
5652 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5653 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5655 58. Messages for configuration errors now include the name of the main
5656 configuration file - useful now that there may be more than one file in a
5657 list (.included file names were always shown).
5659 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5660 for those rare installations that do not start the daemon as root or run it
5661 setuid root. I've cut out the call to initgroups() if the daemon is not
5664 60. The Exim user and group can now be bound into the binary as text strings
5665 that are looked up at the start of Exim's processing.
5667 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5669 62. Added $mailstore_basename variable.
5671 63. Installed patch to sieve.c from Michael Haardt.
5673 64. When Exim failed to open the panic log after failing to open the main log,
5674 the original message it was trying to log was written to stderr and debug
5675 output, but if they were not available (the usual case in production), it
5676 was lost. Now it is written to syslog before the two lines that record the
5677 failures to open the logs.
5679 65. Users' Exim filters run in subprocesses under the user's uid. It is
5680 possible for a "deliver" command or an alias in a "personal" command to
5681 provoke an address rewrite. If logging of address rewriting is configured,
5682 this fails because the process is not running as root or exim. There may be
5683 a better way of dealing with this, but for the moment (because 4.30 needs
5684 to be released), I have disabled address rewrite logging when running a
5685 filter in a non-root, non-exim process.
5691 1. The buildconfig auxiliary program wasn't quoting the value set for
5692 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5693 not defined. This bug was masked in 4.22 by the effect that was fixed in
5696 2. Some messages that were rejected after a message id was allocated were
5697 shown as "incomplete" by exigrep. It no longer does this for messages that
5698 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5700 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5701 have allow_domain_literals set, the ID did not get logged in the <= line.
5702 Domain literals are now always recognized in Message-ID: header lines.
5704 4. The first argument for a ${extract expansion item is the key name or field
5705 number. Leading and trailing spaces in this item were not being ignored,
5706 causing some misleading effects.
5708 5. When deliver_drop_privilege was set, single queue runner processes started
5709 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5710 same command in the process it spins off) were not dropping privilege.
5712 6. When the daemon running as "exim" started a queue runner, it always
5713 re-executed Exim in the spun-off process. This is a waste of effort when
5714 deliver_drop_privilege is set. The new process now just calls the
5715 queue-runner function directly.
5721 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5724 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5725 ignored. Though the use of -oP was forcing the writing of a pid file, it
5726 was always written to the default place.
5728 3. If the message "no IP address found for host xxxx" is generated during
5729 incoming verification, it is now followed by identification of the incoming
5730 connection (so you can more easily find what provoked it).
5732 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5734 5. Added some features to "harden" Exim a bit more against certain attacks:
5736 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5737 be put in Local/Makefile. This is like the never_users runtime option,
5738 but it cannot be overridden. The default setting is "root".
5740 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5741 prefix string with which any file named in a -C command line option
5744 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5745 is retained for -C and -D only if the caller of Exim is root. Without
5746 it, the exim user may also use -C and -D and retain privilege.
5748 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5749 command line option is disabled.
5751 6. Macro names set by the -D option must start with an upper case letter, just
5752 like macro names defined in the configuration file.
5754 7. Added "dereference=" facility to LDAP.
5756 8. Two instances of the typo "uknown" in the source files are fixed.
5758 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5759 the Configure-Makefile script screwed up while processing it.
5761 10. Incorporated PCRE 4.4.
5763 11. The SMTP synchronization check was not operating right at the start of an
5764 SMTP session. For example, it could not catch a HELO sent before the client
5765 waited for the greeting. There is now a check for outstanding input at the
5766 point when the greeting is written. Because of the duplex, asynchronous
5767 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5768 way, but not yet received, when the check is performed.
5770 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5771 on TCP/IP sockets, because this apparently causes some broken clients to
5774 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5775 unchanged) from the Cygwin maintainer.
5777 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5778 was supported instead of testing for mbx. Effectively a typo.
5780 15. The spa authenticator server code was not checking that the input it
5781 received was valid base64.
5783 16. The debug output line for the "set" modifier in ACLs was not showing the
5784 name of the variable that was being set.
5786 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5788 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5789 lookup on its own. Something else was needed to trigger the lookup. For
5790 example, a match in host_lookup or the need for a host name in a host list.
5791 Now, if $sender_host_name is referenced and the host name has not yet been
5792 looked up, a lookup is performed. If the lookup fails, the variable remains
5793 empty, and $host_lookup_failed is set to "1".
5795 19. Added "eqi" as a case-independent comparison operator.
5797 20. The saslauthd authentication condition could segfault if neither service
5798 nor realm was specified.
5800 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5801 error message that was logged was misleading, and incoming SMTP
5802 connections were dropped. The message is now more accurate, and temporary
5803 errors are given to SMTP connections.
5805 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5806 (except RSET) with a 421 error, until QUIT is received. However, it was
5807 failing to send a response to QUIT.
5809 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5810 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5811 ACL is now run after the helo_try_verify_hosts code.
5813 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5814 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5815 case-independent, so other case variants are also recognized). Apparently
5816 some systems use these upper case variants.
5818 25. If more than two messages were waiting for the same host, and a transport
5819 filter was specified for the transport, Exim sent two messages over the
5820 same TCP/IP connection, and then failed with "socket operation on non-
5821 socket" when it tried to send the third.
5823 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5825 27. The extern definition of crypt16() in expand.c was not being excluded when
5826 the OS had its own crypt16() function.
5828 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5829 as a preferred synonym for return_size_limit, both as an option and as an
5832 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5834 30. Changed the default configuration ACL to relax the local part checking rule
5835 for addresses that are not in any local domains. For these addresses,
5836 slashes and pipe symbols are allowed within local parts, but the sequence
5837 /../ is explicitly forbidden.
5839 31. SPA server authentication was not clearing the challenge buffer before
5842 32. log_message in a "warn" ACL statement was writing to the reject log as
5843 well as to the main log, which contradicts the documentation and doesn't
5844 seem right (because no rejection is happening). So I have stopped it.
5846 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5847 However, I am unable to do any testing of this.
5849 34. Fixed an infelicity in the appendfile transport. When checking directories
5850 for a mailbox, to see if any needed to be created, it was accidentally
5851 using path names with one or more superfluous leading slashes; tracing
5852 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5854 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5855 contents are added to the log line that is written for every discarded
5856 recipient. (Previously a log_message setting was ignored.)
5858 36. The ${quote: operator now quotes the string if it is empty.
5860 37. The install script runs exim in order to find its version number. If for
5861 some reason other than non-existence or emptiness, which it checks, it
5862 could not run './exim', it was installing it with an empty version number,
5863 i.e. as "exim-". This error state is now caught, and the installation is
5866 38. An argument was missing from the function that creates an error message
5867 when Exim fails to connect to the socket for saslauthd authentication.
5868 This could cause Exim to crash, or give a corrupted message.
5870 39. Added isip, isip4, and isip6 to ${if conditions.
5872 40. The ACL variables $acl_xx are now saved with the message, and can be
5873 accessed later in routers, transports, and filters.
5875 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5876 strings in the file are not string-expanded.
5878 42. If a MAIL command specified a SIZE value that was too large to fit into an
5879 int variable, the check against message_size_limit failed. Such values are
5880 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5881 day this will have to be increased, but I don't think I want to be around
5882 when emails are that large.
5889 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5890 iconv() is not standard in FreeBSD.
5892 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5893 IPv6 enabled. The observed symptom was a segmentation fault on return from
5894 the function os_common_find_running_interfaces() in src/os.c.
5896 3. In the check_special_case() function in daemon.c I had used "errno" as an
5897 argument name, which causes warnings on some systems. This was basically a
5898 typo, since it was named "eno" in the comments!
5900 4. The code that waits for the clock to tick (at a resolution of some fraction
5901 of a second) so as to ensure message-id uniqueness was always waiting for
5902 at least one whole tick, when it could have waited for less. [This is
5903 almost certainly not relevant at current processor speeds, where it is
5904 unlikely to ever wait at all. But we try to future-proof.]
5906 5. The function that sleeps for a time interval that includes fractions of a
5907 second contained a race. It did not block SIGALRM between setting the
5908 timer, and suspending (a couple of lines later). If the interval was short
5909 and the sigsuspend() was delayed until after it had expired, the suspension
5910 never ended. On busy systems this could lead to processes getting stuck for
5913 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5914 process, before it forks any delivery processes. The open lookup caching
5915 mechanism meant that the open file or database connection was passed into
5916 the delivery process. The problem was that delivery processes always tidy
5917 up cached lookup data. This could cause a problem for the next delivery
5918 process started by the queue runner, because the external queue runner
5919 process does not know about the closure. So the next delivery process
5920 still has data in the lookup cache. In the case of a file lookup, there was
5921 no problem because closing a file descriptor in a subprocess doesn't affect
5922 the parent. However, if the lookup was caching a connection to a database,
5923 the connection was closed, and the second delivery process was likely to
5924 see errors such as "PGSQL: query failed: server closed the connection
5925 unexpectedly". The problem has been fixed by closing all cached lookups
5926 in a queue runner before running a delivery process.
5928 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5929 seem to have the "const" qualifier which it has on other OS. I've
5932 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5933 *of the same type* (client or server) with the same public name that an
5934 error should be diagnosed.
5936 9. When Exim looked up a host name for an IP address, but failed to find the
5937 original IP address when looking up the host name (a safety check), it
5938 output the message "<ip address> does not match any IP for NULL", which was
5939 confusing, to say the least. The bug was that the host name should have
5940 appeared instead of "NULL".
5942 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5943 user that is built into the binary, and the -C or -D options is used, root
5944 privilege is dropped before the configuration file is read. In addition,
5945 logging is switched to stderr instead of the normal log files. If the
5946 configuration then re-defines the Exim user, the unprivileged environment
5947 is probably not what is expected, so Exim logs a panic warning message (but
5950 However, if deliver_drop_privilege is set, the unprivileged state may well
5951 be exactly what is intended, so the warning has been cut out in that case,
5952 and Exim is allowed to try to write to its normal log files.
5958 1. smtp_return_error_details was not giving details for temporary sender
5959 or receiver verification errors.
5961 2. Diagnose a configuration error if two authenticators have the same public
5964 3. Exim used not to create the message log file for a message until the first
5965 delivery attempt. This could be confusing when incoming messages were held
5966 for policy or load reasons. The message log file is now created at the time
5967 the message is received, and an initial "Received" line is written to it.
5969 4. The automatically generated man page for command line options had a minor
5970 bug that caused no ill effects; however, a more serious problem was that
5971 the procedure for building the man page automatically didn't always
5972 operate. Consequently, release 4.20 contains an out-of-date version. This
5973 shouldn't happen again.
5975 5. When building Exim with embedded Perl support, the script that builds the
5976 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5977 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5979 6. The freeze_tell option was not being used for messages that were frozen on
5980 arrival, either by an ACL or by local_scan().
5982 7. Added the smtp_incomplete_transaction log selector.
5984 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5985 was accepting AUTH without a new EHLO.
5987 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5988 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5991 10. Logging of TCP/IP connections (when configured) now happens in the main
5992 daemon process instead of the child process, so that the TCP/IP connection
5993 count is more accurate (but it can never be perfect).
5995 11. The use of "drop" in a nested ACL was not being handled correctly in the
5996 outer ACL. Now, if condition failure induced by the nested "drop" causes
5997 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5998 or "require"), the connection is dropped.
6000 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6001 that yield "discard" can now be used with an "accept" or a "discard" verb,
6002 but an error is generated for any others (because I can't see a useful way
6003 to define what should happen).
6005 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6006 are now processed in the same way as lines in the Exim configuration file.
6007 In particular, continuation lines are supported.
6009 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6011 15. Added -ti meaning -t -i.
6013 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6014 domains, and warn by logging if others are found.
6016 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6017 returned from GIFCONF when Exim is trying to find the list of interfaces on
6018 a host. The code in os.c has been modified to copy each ifreq to an aligned
6019 structure in all cases.
6021 Also, in some cases, the returned ifreq's were being copied to a 'struct
6022 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6023 means the last couple of bytes of an IPv6 address could be chopped if the
6024 ifreq contained only a normal sockaddr (14 bytes storage).
6026 18. Named domain lists were not supported in the hosts_treat_as_local option.
6027 An entry such as +xxxx was not recognized, and was treated as a literal
6030 19. Ensure that header lines added by a DATA ACL are included in the reject log
6031 if the ACL subsequently rejects the message.
6033 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6034 MD5 (which is deprecated).
6036 21. When testing a filter file using -bf, Exim was writing a message when it
6037 took the sender from a "From " line in the message, but it was not doing so
6038 when it took $return_path from a Return-Path: header line. It now does.
6040 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6041 with a valid header line field name (a series of printing characters
6042 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6044 23. Changed "disc" in the source to "disk" to conform to the documentation and
6045 the book and for uniformity.
6047 24. Ignore Sendmail's -Ooption=value command line item.
6049 25. When execve() failed while trying to run a command in a pipe transport,
6050 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6051 could be confused with a return value of 69 from the command itself. This
6052 has been changed to 127, the value the shell returns if it is asked to run
6053 a non-existent command. The wording for the related log line suggests a
6054 non-existent command as the problem.
6056 26. If received_header_text expands to an empty string, do not add a Received:
6057 header line to the message. (Well, it adds a token one on the spool, but
6058 marks it "old" so that it doesn't get used or transmitted.)
6060 27. Installed eximstats 1.28 (addition of -nt option).
6062 28. There was no check for failure on the call to getsockname() in the daemon
6063 code. This can fail if there is a shortage of resources on the system, with
6064 ENOMEM, for example. A temporary error is now given on failure.
6066 29. Contrary to the C standard, it seems that in some environments, the
6067 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6068 program. Exim now does this explicitly; it affects the formatting of
6069 timestamps using strftime().
6071 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6072 complaints. I've now initialized all the variables, to avoid this.
6074 32. Header lines added by a system filter were not being "seen" during
6075 transport-time rewrites.
6077 33. The info_callback() function passed to OpenSSL is set up with type void
6078 (*)(SSL *, int, int), as described somewhere. However, when calling the
6079 function (actually a macro) that sets it up, the type void(*)() is
6080 expected. I've put in a cast to prevent warnings from picky compilers.
6082 34. If a DNS black list lookup found a CNAME record, but there were no A
6083 records associated with the domain it pointed at, Exim crashed.
6085 35. If a DNS black list lookup returned more than one A record, Exim ignored
6086 all but the first. It now scans all returned addresses if a particular IP
6087 value is being sought. In this situation, the contents of the
6088 $dnslist_value variable are a list of all the addresses, separated by a
6091 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6092 to accept a host name and all its aliases if the forward lookup for any of
6093 them yielded the IP address of the incoming connection. Now it accepts only
6094 those names whose forward lookup yields the correct IP address. Any other
6095 names are discarded. This closes a loophole whereby a rogue DNS
6096 administrator could create reverse DNS records to break through a
6097 wildcarded host restriction in an ACL.
6099 37. If a user filter or a system filter that ran in a subprocess used any of
6100 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6101 the wrong values were passed to the pipe command ($thisaddress had the
6102 value of $0, $0 had the value of $1, etc). This bug was introduced by
6103 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6105 38. Improved the line breaking for long SMTP error messages from ACLs.
6106 Previously, if there was no break point between 40 and 75 characters, Exim
6107 left the rest of the message alone. Two changes have been made: (a) I've
6108 reduced the minimum length to 35 characters; (b) if it can't find a break
6109 point between 35 and 75 characters, it looks ahead and uses the first one
6110 that it finds. This may give the occasional overlong line, but at least the
6111 remaining text gets split now.
6113 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6114 file descriptors might be low, and that setting 1000 would always raise it.
6115 It turns out that in some environments, the limit is already over 1000 and
6116 that lowering it causes trouble. So now Exim takes care not to decrease it.
6118 40. When delivering a message, the value of $return_path is set to $sender_
6119 address at the start of routing (routers may change the value). By an
6120 oversight, this default was not being set up when an address was tested by
6121 -bt or -bv, which affected the outcome if any router or filter referred to
6124 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6125 log only when "message" or "log_message" are set. However, if one of the
6126 conditions was an address verification, or a call to a nested ACL, the
6127 messages generated by the underlying test were being passed through. This
6128 no longer happens. The underlying message is available in $acl_verify_
6129 message for both "message" and "log_message" expansions, so it can be
6130 passed through if needed.
6132 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6133 new expansion $bh_ to give the encoded byte string without charset
6134 translation. Translation happens only if iconv() is available; HAVE_ICONV
6135 indicates this at build time. HEADERS_CHARSET gives the charset to
6136 translate to; headers_charset can change it in the configuration, and
6137 "headers charset" can change it in an individual filter file.
6139 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6140 that creates RFC 2047 encoded "words" labels them as that charset instead
6141 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6142 expansion operator; (ii) when Exim creates a From: line for a local
6143 message; (iii) when a header line is rewritten to include a "phrase" part.
6145 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6146 buggy, causing it to skip the first lines of messages whose message ID
6147 ended in 'D'. This would not have bitten before Exim release 4.14, because
6148 message IDs were unlikely to end in 'D' before then. The effect was to have
6149 incorrect size information for certain domains.
6151 45. #include "config.h" was missing at the start of the crypt16.c module. This
6152 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6155 46. If there was a timeout during a "random" callout check, Exim treated it as
6156 a failure of the random address, and carried on sending RSET and the real
6157 address. If the delay was just some slowness somewhere, the response to the
6158 original RCPT would be taken as a response to RSET and so on, causing
6159 mayhem of various kinds.
6161 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6162 when I implemented it. It didn't allow for the fact that some option values
6163 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6164 the right test for positive values.
6166 48. Domain names in DNS records are case-independent. Exim always looks them up
6167 in lower case. Some resolvers return domain names in exactly the case they
6168 appear in the zone file, that is, they may contain uppercase letters. Not
6169 all resolvers do this - some return always lower case. Exim was treating a
6170 change of case by a resolver as a change of domain, similar to a widening
6171 of a domain abbreviation. This triggered its re-routing code and so it was
6172 trying to route what was effectively the same domain again. This normally
6173 caused routing to fail (because the router wouldn't handle the domain
6174 twice). Now Exim checks for this case specially, and just changes the
6175 casing of the domain that it ultimately uses when it transmits the message
6178 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6181 50. If a filter generated a file delivery with a non-absolute name (possible if
6182 no home directory exists for the router), the forbid_file option was not
6185 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6186 the existing equality matching.
6188 52. Exim was using ints instead of ino_t variables in some places where it was
6189 dealing with inode numbers.
6191 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6192 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6193 called TMPDIR, and if it finds it is different, it changes its value.
6195 54. The smtp_printf() function is now made available to local_scan() so
6196 additional output lines can be written before returning. There is also an
6197 smtp_fflush() function to enable the detection of a dropped connection.
6198 The variables smtp_input and smtp_batched_input are exported to
6201 55. Changed the default runtime configuration: the message "Unknown user"
6202 has been removed from the ACL, and instead placed on the localuser router,
6203 using the cannot_route_message feature. This means that any verification
6204 failures that generate their own messages won't get overridden. Similarly,
6205 the "Unrouteable address" message that was in the ACL for unverifiable
6206 relay addresses has also been removed.
6208 56. Added hosts_avoid_esmtp to the smtp transport.
6210 57. The exicyclog script was not checking for the esoteric option
6211 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6212 will work only if exicyclog is run under the appropriate euid.
6214 58. Following a discussion on the list, the rules by which Exim recognises line
6215 endings on incoming messages have been changed. The -dropcr and drop_cr
6216 options are now no-ops, retained only for backwards compatibility. The
6217 following line terminators are recognized: LF CRLF CR. However, special
6218 processing applies to CR:
6220 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6221 nor a local message in the state where . is a terminator.
6223 (ii) If a bare CR is encountered in a header line, an extra space is added
6224 after the line terminator so as not to end the header. The reasoning
6225 behind this is that bare CRs in header lines are most likely either
6226 to be mistakes, or people trying to play silly games.
6228 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6229 was being incorrectly given as 18 bytes larger than it should have been.
6230 This is a VOB (very old bug).
6232 60. This may never have affected anything current, but just in case it has:
6233 When the local host is found other than at the start of a list of hosts,
6234 the local host, those with the same MX, and any that follow, are discarded.
6235 When the list in question was part of a longer list of hosts, the following
6236 hosts (not currently being processed) were also being discarded. This no
6237 longer happens. I'm not sure if this situation could ever has previously
6240 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6243 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6244 header. This is something that is recommended in a new Internet Draft, and
6245 is something that is documented as being done by Sendmail. There are two
6246 possible values. For messages generated by the autoreply transport, Exim
6249 Auto-Submitted: auto-replied
6251 whereas for all other generated messages (e.g. bounces) it adds
6253 Auto-Submitted: auto-generated
6255 63. The "personal" condition in filters now includes a test for the
6256 Auto-Submitted: header. If it contains the string "auto-" the message it
6257 not considered personal.
6259 64. Added rcpt_include_affixes as a generic transport option.
6261 65. Added queue_only_override (default true).
6263 66. Added the syslog_duplication option.
6265 67. If what should have been the first header line of a message consisted of
6266 a space followed by a colon, Exim was mis-interpreting it as a header line.
6267 It isn't of course - it is syntactically invalid and should therefore be
6268 treated as the start of the message body. The misbehaviour could have
6269 caused a number of strange effects, including loss of data in subsequent
6270 header lines, and spool format errors.
6272 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6273 client host had authenticated. This control can now be exercised by an ACL
6274 for more flexibility.
6276 69. By default, callouts do not happen when testing with -bh. There is now a
6277 variant, -bhc, which does actually run the callout code, including
6278 consulting and updating the callout cache.
6280 70. Added support for saslauthd authentication, courtesy of Alexander
6283 71. If statvfs() failed on the spool or log directories while checking their
6284 size for availability, Exim confusingly gave the error "space shortage".
6285 Furthermore, in debugging mode it crashed with a floating point exception.
6286 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6287 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6288 serious problem, Exim now writes to the main and panic logs when this
6289 happens, with details of the failure. It then refuses to accept the
6290 incoming message, giving the message "spool directory problem" or "log
6291 directory problem" with a 421 code for SMTP messages.
6293 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6294 0, 1, and 2 exist, because some OS complain for execs without them (see
6295 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6296 descriptors. However, the code omitted to check that the open succeeded,
6297 causing mysterious errors if for some reason the permissions on /dev/null
6298 got screwed. Now Exim writes a message to the main and panic logs, and
6299 bombs out if it can't open /dev/null.
6301 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6302 interact so that it is all more flexible. It is supposed to remain
6303 backwards compatible. Also added extra_local_interfaces.
6305 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6306 to bomb out with an assertion failure - to the client this appears as a
6307 connection drop. This problem occurs in the part of the code that was taken
6308 from the Samba project. Fortunately, the assertion is in a very simple
6309 function, so I have fixed this by reproducing the function inline in the
6310 one place where it is called, and arranging for authentication to fail
6311 instead of killing the process with assert().
6313 75. The SPA client code was not working when the server requested OEM rather
6314 than Unicode encoding.
6316 76. Added code to make require_files with a specific uid setting more usable in
6317 the case where statting the file as root fails - usually a non-root-mounted
6318 NFS file system. When this happens and the failure is EACCES, Exim now
6319 forks a subprocess and does the per-uid checking as the relevant uid.
6321 77. Added process_log_path.
6323 78. If log_file_path was not explicitly set, a setting of check_log_space or
6324 check_log_inodes was ignored.
6326 79. If a space check for the spool or log partitions fails, the incident is now
6327 logged. Of course, in the latter case the data may get lost...
6329 80. Added the %p formatting code to string_format() so that it can be used to
6330 print addresses in debug_print(). Adjusted all the address printing in the
6331 debugging in store.c to use %p rather than %d.
6333 81. There was a concern that a line of code in smtp_in.c could overflow a
6334 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6335 initially expressed, the concern was not well-founded, because trailing
6336 spaces are removed early. However, if the trailing spaces were followed by
6337 a NULL, they did not get removed, so the overflow was possible. Two fixes
6340 (a) I re-wrote the offending code in a cleaner fashion.
6341 (b) If an incoming SMTP command contains a NULL character, it is rejected
6344 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6345 runs initgroups(), so that if the Exim user is in any additional groups,
6346 they will be used during message reception.
6352 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob