1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
8 and 3.4.4 - once the server is enabled to respond to an OCSP request
9 it does even when not requested, resulting in a stapling non-aware
10 client dropping the TLS connection.
12 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
13 support variable-length bit vectors. No functional change.
15 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
16 The I= interface field on outgoing lines is now after the H= remote
17 host field, same as incoming lines. There is a separate
18 outgoing_interface log selector which allows you to disable the
21 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
22 If not running log_selector +smtp_connection the mainlog would be held
23 open indefinitely after a "too many connections" event, including to a
24 deleted file after a log rotate. Leave the per net connection logging
25 leaving it open for efficiency as that will be quickly detected by the
26 check on the next write.
28 HS/01 Bug 1671: Fix post transport crash.
29 Processing the wait-<transport> messages could crash the delivery
30 process if the message IDs didn't exist for some reason. When
31 using 'split_spool_directory=yes' the construction of the spool
32 file name failed already, exposing the same netto behaviour.
34 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
35 mime_regex ACL conditions.
37 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
38 to DSN fail messages (bounces): remote IP, remote greeting, remote response
39 to HELO, local diagnostic string.
41 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
42 log line to debug. Even when configured with a tls authenticator many
43 client connections are expected to not authenticate in this way, so
44 an authenticate fail is not an error.
46 HS/02 Add the Exim version string to the process info. This way exiwhat
47 gives some more detail about the running daemon.
49 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
50 matter for fast-change records such as DNSBLs.
52 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
53 retry record, even if constant. There may be multiple transports with
54 different interface settings and the retry behaviour needs to be kept
57 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
59 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
61 JH/10 Bug 840: fix log_defer_output option of pipe transport
63 JH/11 Bug 830: use same host for all RCPTS of a message, even under
64 hosts_randomize. This matters a lot when combined with mua_wrapper.
66 JH/12 Bug 1706: percent and underbar characters are no longer excaped by the
67 ${quote_pgsql:<string>} operator.
72 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
75 JH/02 The smtp transport option "multi_domain" is now expanded.
77 JH/03 The smtp transport now requests PRDR by default, if the server offers
80 JH/04 Certificate name checking on server certificates, when exim is a client,
81 is now done by default. The transport option tls_verify_cert_hostnames
82 can be used to disable this per-host. The build option
83 EXPERIMENTAL_CERTNAMES is withdrawn.
85 JH/05 The value of the tls_verify_certificates smtp transport and main options
86 default to the word "system" to access the system default CA bundle.
87 For GnuTLS, only version 3.0.20 or later.
89 JH/06 Verification of the server certificate for a TLS connection is now tried
90 (but not required) by default. The verification status is now logged by
91 default, for both outbound TLS and client-certificate supplying inbound
94 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
97 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
98 Status Notification (bounce) messages are now MIME format per RFC 3464.
99 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
100 under the control of the dsn_advertise_hosts option, and routers may
101 have a dsn_lasthop option.
103 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
104 default, modifiable by a malware= option. The list separator for
105 the options can now be changed in the usual way. Bug 68.
107 JH/10 The smtp_receive_timeout main option is now expanded before use.
109 JH/11 The incoming_interface log option now also enables logging of the
110 local interface on delivery outgoing connections.
112 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
113 if the interface and destination host and port all match.
115 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
118 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
119 Patch from Andrew Lewis.
121 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
122 now supports optional time-restrictions, weighting, and priority
123 modifiers per server. Patch originally by <rommer@active.by>.
125 JH/16 The spamd_address main option now supports a mixed list of local
126 and remote servers. Remote servers can be IPv6 addresses, and
127 specify a port-range.
129 JH/17 Bug 68: The spamd_address main option now supports an optional
130 timeout value per server.
132 JH/18 Bug 1581: Router and transport options headers_add/remove can
133 now have the list separator specified.
135 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
138 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
141 JH/21 Support for the A6 type of dns record is withdrawn.
143 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
144 rather than the verbs used.
146 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
147 from 255 to 1024 chars.
149 JH/24 Verification callouts now attempt to use TLS by default.
151 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
152 are generic router options now. The defaults didn't change.
154 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
155 Original patch from Alexander Shikoff, worked over by JH.
157 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
158 files if ZCAT_COMMAND is not executable.
160 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
162 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
164 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
165 Normally benign, it bites when the pair was led to by a CNAME;
166 modern usage is to not canoicalize the domain to a CNAME target
167 (and we were inconsistent anyway for A-only vs AAAA+A).
169 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
171 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
172 when evaluating $sender_host_dnssec.
174 JH/31 Check the HELO verification lookup for DNSSEC, adding new
175 $sender_helo_dnssec variable.
177 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
179 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
181 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
183 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
184 documented as working, but never had. Support all but $spam_report.
186 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
187 added for tls authenticator.
192 TL/01 When running the test suite, the README says that variables such as
193 no_msglog_check are global and can be placed anywhere in a specific
194 test's script, however it was observed that placement needed to be near
195 the beginning for it to behave that way. Changed the runtest perl
196 script to read through the entire script once to detect and set these
197 variables, reset to the beginning of the script, and then run through
198 the script parsing/test process like normal.
200 TL/02 The BSD's have an arc4random API. One of the functions to induce
201 adding randomness was arc4random_stir(), but it has been removed in
202 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
203 function when detected.
205 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
206 cause callback expansion.
208 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
209 syntax errors in an expansion can be treated as a string instead of
210 logging or causing an error, due to the internal use of bool_lax
211 instead of bool when processing it.
213 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
214 server certificates when making smtp deliveries.
216 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
218 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
220 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
222 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
223 Merged patch from Sebastian Wiedenroth.
225 JH/05 Fix results-pipe from transport process. Several recipients, combined
226 with certificate use, exposed issues where response data items split
227 over buffer boundaries were not parsed properly. This eventually
228 resulted in duplicates being sent. This issue only became common enough
229 to notice due to the introduction of conection certificate information,
230 the item size being so much larger. Found and fixed by Wolfgang Breyha.
232 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
233 size buffer was used, resulting in syntax errors when an expansion
236 JH/07 Add support for directories of certificates when compiled with a GnuTLS
237 version 3.3.6 or later.
239 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
240 is EXPERIMENTAL_EVENT, the main-configuration and transport options
241 both become "event_action", the variables become $event_name, $event_data
242 and $event_defer_errno. There is a new variable $verify_mode, usable in
243 routers, transports and related events. The tls:cert event is now also
244 raised for inbound connections, if the main configuration event_action
247 TL/06 In test suite, disable OCSP for old versions of openssl which contained
248 early OCSP support, but no stapling (appears to be less than 1.0.0).
250 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
251 server certificate names available under the smtp transport option
252 "tls_verify_cert_hostname" now do not permit multi-component wildcard
255 JH/10 Time-related extraction expansions from certificates now use the main
256 option "timezone" setting for output formatting, and are consistent
257 between OpenSSL and GnuTLS compilations. Bug 1541.
259 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
260 encoded parameter in the incoming message. Bug 1558.
262 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
263 include certificate info, eximon was claiming there were spoolfile
266 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
268 JH/14 Log delivery-related information more consistently, using the sequence
269 "H=<name> [<ip>]" wherever possible.
271 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
272 are problematic for Debian distribution, omit them from the release
275 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
277 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
280 JH/17 Fix a null-indirection in certextract expansions when a nondefault
281 output list separator was used.
286 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
287 checkers that were complaining about end of non-void function with no
290 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
291 This was a regression intruduced in 4.83 by another bugfix.
293 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
295 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
296 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
302 TF/01 Correctly close the server side of TLS when forking for delivery.
304 When a message was received over SMTP with TLS, Exim failed to clear up
305 the incoming connection properly after forking off the child process to
306 deliver the message. In some situations the subsequent outgoing
307 delivery connection happened to have the same fd number as the incoming
308 connection previously had. Exim would try to use TLS and fail, logging
309 a "Bad file descriptor" error.
311 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
312 utilities have not been installed.
314 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
315 temporary space as the ACL may create new global variables.
317 TL/01 LDAP support uses per connection or global context settings, depending
318 upon the detected version of the libraries at build time.
320 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
321 to extract and use the src ip:port in logging and expansions as if it
322 were a direct connection from the outside internet. PPv2 support was
323 updated based on HAProxy spec change in May 2014.
325 JH/02 Add ${listextract {number}{list}{success}{fail}}.
327 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
328 Properly escape header and check for NULL return.
330 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
333 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
335 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
336 characters in header names, implemented as a verify condition.
337 Contributed by Michael Fischer v. Mollard.
339 TL/05 Rename SPF condition results err_perm and err_temp to standardized
340 results permerror and temperror. Previous values are deprecated but
341 still accepted. In a future release, err_perm and err_temp will be
342 completely removed, which will be a backward incompatibility if the
343 ACL tests for either of these two old results. Patch contributed by
344 user bes-internal on the mailing list.
346 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
348 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
349 selectors, in both main and reject logs.
351 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
354 JH/07 Add malware type "sock" for talking to simple daemon.
356 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
358 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
359 routers/transports under cutthrough routing.
361 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
362 numbers. Touch up "bool" conditional to keep the same definition.
364 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
366 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
368 JH/12 Expand items in router/transport headers_add or headers_remove lists
369 individually rather than the list as a whole. Bug 1452.
371 Required for reasonable handling of multiple headers_ options when
372 they may be empty; requires that headers_remove items with embedded
373 colons must have them doubled (or the list-separator changed).
375 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
376 view the policy declared in the DMARC record. Currently, $dmarc_status
377 is a combined value of both the record presence and the result of the
380 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
382 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
383 dnslookup router and the smtp transport (applying to the forward
386 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
387 of ldap servers used for a specific lookup. Patch provided by Heiko
390 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
391 New variable $lookup_dnssec_authenticated for observability.
393 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
394 Patch submitted by Lars Timman.
396 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
398 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
399 Requires trusted mode and valid format message id, aborts otherwise.
400 Patch contributed by Heiko Schlichting.
402 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
403 certextract with support for various fields. Bug 1358.
405 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
406 is requested by default, modifiable by smtp transport option
409 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
410 operate on certificate variables to give certificate fingerprints
411 Also new ${sha256:cert_variable}.
413 JH/23 The PRDR feature is moved from being Experimental into the mainline.
415 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
416 Christian Aistleitner.
418 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
420 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
421 file. Patch from Wolfgang Breyha.
423 JH/25 Expand the coverage of the delivery $host and $host_address to
424 client authenticators run in verify callout. Bug 1476.
426 JH/26 Port service names are now accepted for tls_on_connect_ports, to
427 align with daemon_smtp_ports. Bug 72.
429 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
430 support and error reporting did not work properly.
432 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
433 and is readable. Patch from Andrew Colin Kissa.
435 TL/14 Enhance documentation of ${run expansion and how it parses the
436 commandline after expansion, particularly in the case when an
437 unquoted variable expansion results in an empty value.
439 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
441 PP/02 Fix internal collision of T_APL on systems which support RFC3123
442 by renaming away from it. Addresses GH issue 15, reported by
445 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
447 TL/15 SECURITY: prevent double expansion in math comparison functions
448 (can expand unsanitized data). Not remotely exploitable.
455 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
457 PP/02 Make -n do something, by making it not do something.
458 When combined with -bP, the name of an option is not output.
460 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
463 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
464 $sender_host_name and config options to manage this, and basic check
467 PP/05 DSCP support for outbound connections and control modifier for inbound.
469 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
470 (Only plugin which currently uses this is kerberos4, which nobody should
471 be using, but we should make it available and other future plugins might
472 conceivably use it, even though it would break NAT; stuff *should* be
473 using channel bindings instead).
475 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
476 name; added for Sendmail compatibility; requires admin caller.
477 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
478 just ignore it); requires trusted caller.
479 Also parse but ignore: -Ac -Am -X<logfile>
482 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
484 TL/02 Add +smtp_confirmation as a default logging option.
486 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
487 Patch by Magnus Holmgren from 2007-02-20.
489 TL/04 Bugzilla 1281 - Spec typo.
490 Bugzilla 1283 - Spec typo.
491 Bugzilla 1290 - Spec grammar fixes.
493 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
495 TL/06 Add Experimental DMARC support using libopendmarc libraries.
497 TL/07 Fix an out of order global option causing a segfault. Reported to dev
498 mailing list by by Dmitry Isaikin.
500 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
502 JH/02 Support "G" suffix to numbers in ${if comparisons.
504 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
506 NM/01 Bugzilla 1197 - Spec typo
507 Bugzilla 1196 - Spec examples corrections
509 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
511 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
512 gnutls_enable_pkcs11, but renamed to more accurately indicate its
515 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
516 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
518 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
519 "acl {{name}{arg}...}", and optional args on acl condition
522 JH/05 Permit multiple router/transport headers_add/remove lines.
524 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
526 JH/07 Avoid using a waiting database for a single-message-only transport.
527 Performance patch from Paul Fisher. Bugzilla 1262.
529 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
532 JH/09 Add $headers_added variable, with content from use of ACL modifier
533 add_header (but not yet added to the message). Bugzilla 199.
535 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
536 Pulled from Bugzilla 817 by Wolfgang Breyha.
538 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
540 (nb: this is the same fix as in Exim 4.80.1)
542 JH/11 Add A= logging on delivery lines, and a client_set_id option on
545 JH/12 Add optional authenticated_sender logging to A= and a log_selector
548 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
550 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
551 advertise SMTP AUTH mechanism to us, instead of a generic
552 protocol violation error. Also, make Exim more robust to bad
553 data from the Dovecot auth socket.
555 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
557 When a queue runner is handling a message, Exim first routes the
558 recipient addresses, during which it prunes them based on the retry
559 hints database. After that it attempts to deliver the message to
560 any remaining recipients. It then updates the hints database using
563 So if a recipient address works intermittently, it can get repeatedly
564 deferred at routing time. The retry hints record remains fresh so the
565 address never reaches the final cutoff time.
567 This is a fairly common occurrence when a user is bumping up against
568 their storage quota. Exim had some logic in its local delivery code
569 to deal with this. However it did not apply to per-recipient defers
570 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
572 This change adds a proper retry rule check during routing so that the
573 final cutoff time is checked against the message's age. We only do
574 this check if there is an address retry record and there is not a
575 domain retry record; this implies that previous attempts to handle
576 the address had the retry_use_local_parts option turned on. We use
577 this as an approximation for the destination being like a local
578 delivery, as in LMTP.
580 I suspect this new check makes the old local delivery cutoff check
581 redundant, but I have not verified this so I left the code in place.
583 TF/02 Correct gecos expansion when From: is a prefix of the username.
585 Test 0254 submits a message to Exim with the header
589 When I ran the test suite under the user fanf2, Exim expanded
590 the header to contain my full name, whereas it should have added
591 a Resent-Sender: header. It erroneously treats any prefix of the
592 username as equal to the username.
594 This change corrects that bug.
596 GF/01 DCC debug and logging tidyup
597 Error conditions log to paniclog rather than rejectlog.
598 Debug lines prefixed by "DCC: " to remove any ambiguity.
600 TF/03 Avoid unnecessary rebuilds of lookup-related code.
602 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
603 Bug spotted by Jeremy Harris; was flawed since initial commit.
604 Would have resulted in OCSP responses post-SNI triggering an Exim
605 NULL dereference and crash.
607 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
609 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
610 Bug detection, analysis and fix by Samuel Thibault.
611 Bugzilla 1331, Debian bug #698092.
613 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
615 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
616 Server implementation by Todd Lyons, client by JH.
617 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
618 config variable "prdr_enable" controls whether the server
619 advertises the facility. If the client requests PRDR a new
620 acl_data_smtp_prdr ACL is called once for each recipient, after
621 the body content is received and before the acl_smtp_data ACL.
622 The client is controlled by bolth of: a hosts_try_prdr option
623 on the smtp transport, and the server advertisement.
624 Default client logging of deliveries and rejections involving
625 PRDR are flagged with the string "PRDR".
627 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
628 fclose(). Diagnosis by Todd Lyons.
630 PP/17 Update configure.default to handle IPv6 localhost better.
631 Patch by Alain Williams (plus minor tweaks).
634 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
635 This is now consistent with GnuTLS, and is now documented: the
636 previous undocumented portable approach to treating the option as
637 unset was to force an expansion failure. That still works, and
638 an empty string is now equivalent.
640 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
641 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
642 not performing validation itself.
644 PP/20 Added force_command boolean option to pipe transport.
645 Patch from Nick Koston, of cPanel Inc.
647 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
650 TF/04 Added udpsend ACL modifer and hexquote expansion operator
652 PP/21 Fix eximon continuous updating with timestamped log-files.
653 Broken in a format-string cleanup in 4.80, missed when I repaired the
654 other false fix of the same issue.
655 Report and fix from Heiko Schlichting.
658 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
659 Report from Prashanth Katuri.
661 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
662 It's SecureTransport, so affects any MacOS clients which use the
663 system-integrated TLS libraries, including email clients.
665 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
666 using a MIME ACL for non-SMTP local injection.
667 Report and assistance in diagnosis by Warren Baker.
669 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
671 JH/16 Fix comparisons for 64b. Bugzilla 1385.
673 TL/09 Add expansion variable $authenticated_fail_id to keep track of
674 last id that failed so it may be referenced in subsequent ACL's.
676 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
679 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
680 ldap library initialization, allowing self-signed CA's to be
681 used. Also properly sets require_cert option later in code by
682 using NULL (global ldap config) instead of ldap handle (per
683 session). Bug diagnosis and testing by alxgomz.
685 TL/12 Enhanced documentation in the ratelimit.pl script provided in
686 the src/util/ subdirectory.
688 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
689 renamed to Transport Post Delivery Action by Jeremy Harris, as
692 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
693 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
694 redis_servers = needs to be configured which will be used by the redis
695 lookup. Patch from Warren Baker, of The Packet Hub.
697 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
699 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
700 hostname or reverse DNS when processing a host list. Used suggestions
701 from multiple comments on this bug.
703 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
705 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
706 Missed a few lines, added it to make the runtest require no keyboard
709 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
710 contains upper case chars. Make router use caseful_local_part.
712 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
713 support when GnuTLS has been built with p11-kit.
719 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
721 This, or similar/improved, will also be change PP/11 of 4.82.
727 PP/01 Handle short writes when writing local log-files.
728 In practice, only affects FreeBSD (8 onwards).
729 Bugzilla 1053, with thanks to Dmitry Isaikin.
731 NM/01 Bugzilla 949 - Documentation tweak
733 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
736 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
738 PP/02 Implemented gsasl authenticator.
740 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
742 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
743 `pkg-config foo` for cflags/libs.
745 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
746 with rest of GSASL and with heimdal_gssapi.
748 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
749 `pkg-config foo` for cflags/libs for the TLS implementation.
751 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
752 properties get this fed in as external SSF. A number of robustness
753 and debugging improvements to the cyrus_sasl authenticator.
755 PP/08 cyrus_sasl server now expands the server_realm option.
757 PP/09 Bugzilla 1214 - Log authentication information in reject log.
758 Patch by Jeremy Harris.
760 PP/10 Added dbmjz lookup type.
762 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
764 PP/12 MAIL args handles TAB as well as SP, for better interop with
765 non-compliant senders.
766 Analysis and variant patch by Todd Lyons.
768 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
769 Bug report from Lars Müller <lars@samba.org> (via SUSE),
770 Patch from Dirk Mueller <dmueller@suse.com>
772 PP/13 tls_peerdn now print-escaped for spool files.
773 Observed some $tls_peerdn in wild which contained \n, which resulted
774 in spool file corruption.
776 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
777 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
778 or write after TLS renegotiation, which otherwise led to messages
781 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
782 as a tracking header (ie: a signed header comes before the signature).
783 Patch from Wolfgang Breyha.
785 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
786 comma-sep list; embedded commas doubled.
788 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
790 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
792 Report and patch from Dmitry Banschikov.
794 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
795 Removed SSL_clear() after SSL_new() which led to protocol negotiation
796 failures. We appear to now support TLS1.1+ with Exim.
798 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
799 lets Exim select keys and certificates based upon TLS SNI from client.
800 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
801 before an outbound SMTP session. New log_selector, +tls_sni.
803 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
804 NULL dereference. Report and patch from Alun Jones.
806 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
807 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
808 Not seeing resolver debug output on NetBSD, but suspect this is a
809 resolver implementation change.
811 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
812 Left warnings. Added "eximon gdb" invocation mode.
814 PP/21 Defaulting "accept_8bitmime" to true, not false.
816 PP/22 Added -bw for inetd wait mode support.
818 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
819 locate the relevant includes and libraries. Made this the default.
821 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
822 Bugzilla 1246, report and most of solution from Tomasz Kusy.
824 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
825 This may cause build issues on older platforms.
827 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
828 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
829 gnutls_require_mac & gnutls_require_protocols (no longer supported).
830 Added SNI support via GnuTLS too.
831 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
833 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
835 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
836 Applied second patch from Janne, implementing suggestion to default
837 multiple-strings-in-record handling to match SPF spec.
839 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
841 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
842 read-only, out of scope).
843 Patch from Wolfgang Breyha, report from Stuart Northfield.
845 PP/29 Fix three issues highlighted by clang analyser static analysis.
846 Only crash-plausible issue would require the Cambridge-specific
847 iplookup router and a misconfiguration.
848 Report from Marcin Mirosław.
850 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
852 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
853 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
854 As part of this, removing so much warning spew let me fix some minor
855 real issues in debug logging.
857 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
858 assignment on my part. Fixed.
860 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
861 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
862 Janne Snabb (who went above and beyond: thank you).
864 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
865 string otherwise requires a connection and a bunch more work and it's
866 relatively easy to get wrong. Should also expose TLS library linkage
869 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
870 64-bit ${eval} (JH/03).
872 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
873 GNU libc to support some of the 64-bit stuff, should not lead to
874 conflicts. Defined before os.h is pulled in, so if a given platform
875 needs to override this, it can.
877 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
878 protection layer was required, which is not implemented.
879 Bugzilla 1254, patch from Wolfgang Breyha.
881 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
882 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
883 tls_dhparam take prime identifiers. Also unbreak combination of
884 OpenSSL+DH_params+TLSSNI.
886 PP/39 Disable SSLv2 by default in OpenSSL support.
892 PP/01 Solaris build fix for Oracle's LDAP libraries.
893 Bugzilla 1109, patch from Stephen Usher.
895 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
897 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
900 TF/02 Fix a couple more cases where we did not log the error message
901 when unlink() failed. See also change 4.74-TF/03.
903 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
904 lock up or crash if it happened to be inside a call to libc when it
905 got a SIGUSR1 from exiwhat.
907 The SIGUSR1 handler appends the current process status to the process
908 log which is later printed by exiwhat. It used to use the general
909 purpose logging code to do this, but several functions it calls are
910 not safe for signals.
912 The new output code in the SIGUSR1 handler is specific to the process
913 log, and simple enough that it's easy to inspect for signal safety.
914 Removing some special cases also simplifies the general logging code.
915 Removing the spurious timestamps from the process log simplifies
918 TF/04 Improved ratelimit ACL condition.
920 The /noupdate option has been deprecated in favour of /readonly which
921 has clearer semantics. The /leaky, /strict, and /readonly update modes
922 are mutually exclusive. The update mode is no longer included in the
923 database key; it just determines when the database is updated. (This
924 means that when you upgrde Exim will forget old rate measurements.)
926 Exim now checks that the per_* options are used with an update mode that
927 makes sense for the current ACL. For example, when Exim is processing a
928 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
929 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
930 must specify per_mail/readonly. If you omit the update mode it defaults to
931 /leaky where that makes sense (as before) or /readonly where required.
933 The /noupdate option is now undocumented but still supported for
934 backwards compatibility. It is equivalent to /readonly except that in
935 ACLs where /readonly is required you may specify /leaky/noupdate or
936 /strict/noupdate which are treated the same as /readonly.
938 A useful new feature is the /count= option. This is a generalization
939 of the per_byte option, so that you can measure the throughput of other
940 aggregate values. For example, the per_byte option is now equivalent
941 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
943 The per_rcpt option has been generalized using the /count= mechanism
944 (though it's more complicated than the per_byte equivalence). When it is
945 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
946 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
947 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
948 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
949 Note that using per_rcpt with a non-readonly update mode in more than
950 one ACL will cause the recipients to be double-counted. (The per_mail
951 and per_byte options don't have this problem.)
953 The handling of very low rates has changed slightly. If the computed rate
954 is less than the event's count (usually one) then this event is the first
955 after a long gap. In this case the rate is set to the same as this event's
956 count, so that the first message of a spam run is counted properly.
958 The major new feature is a mechanism for counting the rate of unique
959 events. The new per_addr option counts the number of different
960 recipients that someone has sent messages to in the last time period. It
961 behaves like per_rcpt if all the recipient addresses are different, but
962 duplicate recipient addresses do not increase the measured rate. Like
963 the /count= option this is a general mechanism, so the per_addr option
964 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
965 example, measure the rate that a client uses different sender addresses
966 with the options per_mail/unique=$sender_address. There are further
967 details in the main documentation.
969 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
971 TF/06 Removed a few PCRE remnants.
973 TF/07 Automatically extract Exim's version number from tags in the git
974 repository when doing development or release builds.
976 PP/02 Raise smtp_cmd_buffer_size to 16kB.
977 Bugzilla 879. Patch from Paul Fisher.
979 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
980 Heavily based on revision 40f9a89a from Simon Arlott's tree.
983 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
985 PP/05 Variable $av_failed, true if the AV scanner deferred.
986 Bugzilla 1078. Patch from John Horne.
988 PP/06 Stop make process more reliably on build failure.
989 Bugzilla 1087. Patch from Heiko Schlittermann.
991 PP/07 Make maildir_use_size_file an _expandable_ boolean.
992 Bugzilla 1089. Patch from Heiko Schlittermann.
994 PP/08 Handle ${run} returning more data than OS pipe buffer size.
995 Bugzilla 1131. Patch from Holger Weiß.
997 PP/09 Handle IPv6 addresses with SPF.
998 Bugzilla 860. Patch from Wolfgang Breyha.
1000 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1002 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1005 PP/11 match_* no longer expand right-hand-side by default.
1006 New compile-time build option, EXPAND_LISTMATCH_RHS.
1007 New expansion conditions, "inlist", "inlisti".
1009 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1011 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1013 PP/14 fix log_write() format string regression from TF/03.
1014 Bugzilla 1152. Patch from Dmitry Isaikin.
1020 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1022 PP/02 Harmonised TLS library version reporting; only show if debugging.
1023 Layout now matches that introduced for other libraries in 4.74 PP/03.
1025 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1027 PP/04 New "dns_use_edns0" global option.
1029 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1032 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1033 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1035 TK/01 Updated PolarSSL code to 0.14.2.
1036 Bugzilla 1097. Patch from Andreas Metzler.
1038 PP/07 Catch divide-by-zero in ${eval:...}.
1039 Fixes bugzilla 1102.
1041 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1044 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1045 format-string attack -- SECURITY: remote arbitrary code execution.
1047 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1048 time unintentionally subject to list matching rules, letting the header
1049 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1050 arbitrary string expansion). This allowed for information disclosure.
1052 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1053 INT_MIN/-1 -- value coerced to INT_MAX.
1059 NM/01 Workround for PCRE version dependency in version reporting
1062 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1063 This fixes portability to compilers other than gcc, notably
1064 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1066 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1067 makefiles for portability to HP-UX and POSIX correctness.
1069 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1070 Also via indented variable definition in the Makefile.
1071 (Debugging by Oliver Heesakkers).
1073 PP/02 Restore caching of spamd results with expanded spamd_address.
1074 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1076 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1077 Improves build reliability. Fix from: Frank Elsner
1079 NM/02 Fix wide character breakage in the rfc2047 coding
1080 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1082 NM/03 Allow underscore in dnslist lookups
1083 Fixes bug 1026. Patch from Graeme Fowler
1085 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1086 Code patches from Adam Ciarcinski of NetBSD.
1088 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1091 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1092 is logged, to avoid truncation. Patch from John Horne.
1094 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1095 Patch from Jakob Hirsch.
1097 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1098 SQL string expansion failure details.
1099 Patch from Andrey Oktyabrski.
1101 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1102 Patch from Simon Arlott.
1104 PP/09 New lookups functionality failed to compile on old gcc which rejects
1105 extern declarations in function scope.
1106 Patch from Oliver Fleischmann
1108 PP/10 Use sig_atomic_t for flags set from signal handlers.
1109 Check getgroups() return and improve debugging.
1110 Fixed developed for diagnosis in bug 927 (which turned out to be
1113 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1114 Patch from Mark Zealey.
1116 PP/12 Bugzilla 1056: Improved spamd server selection.
1117 Patch from Mark Zealey.
1119 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1120 Based on patch from Heiko Schlittermann.
1122 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1123 Patch from Uwe Doering, sign-off by Michael Haardt.
1125 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1126 variable declaration deep within a block. Bug and patch from
1129 PP/15 lookups-Makefile IRIX compatibilty coercion.
1131 PP/16 Make DISABLE_DKIM build knob functional.
1133 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1134 Patch by Simon Arlott
1136 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1137 variable argument macros. Our copy now differs from upstream.
1143 TF/01 Failure to get a lock on a hints database can have serious
1144 consequences so log it to the panic log.
1146 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1147 controlled using the smtp_confirmation log selector.
1149 TF/03 Include the error message when we fail to unlink a spool file.
1151 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1152 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1153 for maintaining out-of-tree patches for some time.
1155 PP/01 Bugzilla 139: Documentation and portability issues.
1156 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1157 Handle per-OS dynamic-module compilation flags.
1159 PP/02 Let /dev/null have normal permissions.
1160 The 4.73 fixes were a little too stringent and complained about the
1161 permissions on /dev/null. Exempt it from some checks.
1162 Reported by Andreas M. Kirchwitz.
1164 PP/03 Report version information for many libraries, including
1165 Exim version information for dynamically loaded libraries. Created
1166 version.h, now support a version extension string for distributors
1167 who patch heavily. Dynamic module ABI change.
1169 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1170 privilege escalation vulnerability whereby the Exim run-time user
1171 can cause root to append content of the attacker's choosing to
1174 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1177 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1178 If dropping privileges for untrusted macros, we disabled normal logging
1179 on the basis that it would fail; for the Exim run-time user, this is not
1180 the case, and it resulted in successful deliveries going unlogged.
1181 Fixed. Reported by Andreas Metzler.
1187 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1188 only prepend for the Resent-* case. Fixes regression introduced in
1189 Exim 4.70 by NM/22 for Bugzilla 607.
1191 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1192 increasing numbers of administrators be bitten by this.
1194 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1196 PP/03 Bugzilla 994: added openssl_options main configuration option.
1198 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1200 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1202 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1204 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1205 without a peer certificate, leading to a segfault because of an
1206 assumption that peers always have certificates. Be a little more
1207 paranoid. Problem reported by Martin Tscholak.
1209 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1210 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1211 NB: ClamAV planning to remove STREAM in "middle of 2010".
1212 CL also introduces -bmalware, various -d+acl logging additions and
1213 more caution in buffer sizes.
1215 PP/09 Implemented reverse_ip expansion operator.
1217 PP/10 Bugzilla 937: provide a "debug" ACL control.
1219 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1221 PP/12 Bugzilla 973: Implement --version.
1223 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1225 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1227 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1229 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1230 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1231 ignore trailing whitespace.
1233 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1235 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1238 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1239 Notification from Dr Andrew Aitchison.
1241 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1242 ExtendedDetectionInfo response format.
1243 Notification from John Horne.
1245 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1248 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1249 XSL and documented dependency on system catalogs, with examples of how
1252 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1255 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1256 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1257 configuration file which is writeable by the Exim user or group.
1259 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1260 of configuration files to cover files specified with the -C option if
1261 they are going to be used with root privileges, not just the default
1264 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1265 option (effectively making it always true).
1267 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1268 files to be used while preserving root privileges.
1270 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1271 that rogue child processes cannot use them.
1273 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1274 run-time user, instead of root.
1276 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1277 Exim run-time user without dropping privileges.
1279 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1280 result string, instead of calling string_vformat() twice with the same
1283 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1284 for other users. Others should always drop root privileges if they use
1285 -C on the command line, even for a whitelisted configure file.
1287 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1289 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1295 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1296 $data_path, and $header_path variables; fixed documentation bugs and
1299 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1300 exipick to access non-standard spools, including the "frozen" queue
1303 NM/01 Bugzilla 965: Support mysql stored procedures.
1304 Patch from Alain Williams
1306 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1308 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1309 Patch from Andreas Metzler
1311 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1312 Patch from Kirill Miazine
1314 NM/05 Bugzilla 671: Added umask to procmail example.
1316 JJ/03 installed exipick 20100323.0, fixing doc bug
1318 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1319 directory. Notification and patch from Dan Rosenberg.
1321 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1323 TK/02 Improve log output when DKIM signing operation fails.
1325 MH/01 Treat the transport option dkim_domain as a colon separated
1326 list, not as a single string, and sign the message with each element,
1327 omitting multiple occurences of the same signer.
1329 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1330 Bugzilla 985, 986. Patch by Simon Arlott
1332 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1333 Patch by Simon Arlott
1335 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1336 MBX locking. Notification from Dan Rosenberg.
1342 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1344 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1346 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1348 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1350 NM/04 Bugzilla 915: Flag broken perl installation during build.
1356 TK/01 Added patch by Johannes Berg that expands the main option
1357 "spamd_address" if it starts with a dollar sign.
1359 TK/02 Write list of recipients to X-Envelope-Sender header when building
1360 the mbox-format spool file for content scanning (suggested by Jakob
1363 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1364 (http://www.dcc-servers.net/) support via dccifd. Activated by
1365 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1367 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1368 by Mark Daniel Reidel <mr@df.eu>.
1370 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1371 When building exim an external PCRE library is now needed -
1372 PCRE is a system library on the majority of modern systems.
1373 See entry on PCRE_LIBS in EDITME file.
1375 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1376 conversation. Added nologin parameter to request.
1377 Patch contributed by Kirill Miazine.
1379 TF/01 Do not log submission mode rewrites if they do not change the address.
1381 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1383 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1384 log files in place. Contributed by Roberto Lima.
1386 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1388 TF/03 Bugzilla 615: When checking the local_parts router precondition
1389 after a local_part_suffix or local_part_prefix option, Exim now
1390 does not use the address's named list lookup cache, since this
1391 contains cached lookups for the whole local part.
1393 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1394 Robert Millan. Documentation is in experimental-spec.txt.
1396 TF/04 Bugzilla 668: Fix parallel build (make -j).
1398 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1400 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1401 Patch provided by Jan Srzednicki.
1403 TF/05 Leading white space used to be stripped from $spam_report which
1404 wrecked the formatting. Now it is preserved.
1406 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1407 that they are available at delivery time.
1409 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1411 TF/08 TLS error reporting now respects the incoming_interface and
1412 incoming_port log selectors.
1414 TF/09 Produce a more useful error message if an SMTP transport's hosts
1415 setting expands to an empty string.
1417 NM/06 Bugzilla 744: EXPN did not work under TLS.
1418 Patch provided by Phil Pennock.
1420 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1421 Patch provided by Richard Godbee.
1423 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1424 acl_smtp_notquit, added index entry.
1426 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1427 Patch provided by Eugene Bujak.
1429 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1430 accept(). Patch provided by Maxim Dounin.
1432 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1433 Patch provided by Phil Pennock.
1435 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1437 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1438 Patch provided by Brad "anomie" Jorsch.
1440 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1441 Patch provided by Dean Brooks.
1443 TK/05 Add native DKIM support (does not depend on external libraries).
1445 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1446 Patch provided by Graeme Fowler.
1448 NM/16 Bugzilla 851: Documentation example syntax fix.
1450 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1452 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1455 NM/19 Bugzilla 745: TLS version reporting.
1456 Patch provided by Phil Pennock.
1458 NM/20 Bugzilla 167: bool: condition support.
1459 Patch provided by Phil Pennock.
1461 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1462 clients. Patch provided by Phil Pennock.
1464 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1465 Patch provided by Brad "anomie" Jorsch.
1467 NM/23 Bugzilla 687: Fix misparses in eximstats.
1468 Patch provided by Heiko Schlittermann.
1470 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1471 Patch provided by Heiko Schlittermann.
1473 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1474 plus update to original patch.
1476 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1478 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1479 Patch provided by David Brownlee.
1481 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1483 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1485 NM/30 Bugzilla 888: TLS documentation bugfixes.
1487 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1489 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1490 Unlike the original bugzilla I have changed all shell scripts in src tree.
1492 NM/33 Bugzilla 898: Transport filter timeout fix.
1493 Patch by Todd Rinaldo.
1495 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1496 Patch by Serge Demonchaux.
1498 NM/35 Bugzilla 39: Base64 decode bug fixes.
1499 Patch by Jakob Hirsch.
1501 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1503 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1505 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1507 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1513 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1514 ALT-N's libdkim that I have put here:
1515 http://duncanthrax.net/exim-experimental/
1517 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1518 were called 'true' and it seems that C99 defines that as a reserved
1519 keyword to be used with 'bool' variable types. That means you could
1520 not include C99-style headers which use bools without triggering
1521 build errors in sieve.c.
1523 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1524 as mailq or other aliases. Changed the --help handling significantly
1525 to do whats expected. exim_usage() emits usage/help information.
1527 SC/01 Added the -bylocaldomain option to eximstats.
1529 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1531 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1533 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1539 PH/01 Another patch from the Sieve maintainer.
1541 PH/02 When an IPv6 address is converted to a string for single-key lookup
1542 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1543 dots are used instead of colons so that keys in lsearch files need not
1544 contain colons. This was done some time before quoting was made available
1545 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1546 (notated using the quote facility) so as to distinguish them from IPv4
1547 keys. This meant that lookups for IP addresses in host lists did not work
1548 for iplsearch lookups.
1550 This has been fixed by arranging for IPv6 addresses to be expressed with
1551 colons if the lookup type is iplsearch. This is not incompatible, because
1552 previously such lookups could never work.
1554 The situation is now rather anomolous, since one *can* have colons in
1555 ordinary lsearch keys. However, making the change in all cases is
1556 incompatible and would probably break a number of configurations.
1558 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1561 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1562 conversion specification without a maximum field width, thereby enabling
1563 a rogue spamd server to cause a buffer overflow. While nobody in their
1564 right mind would setup Exim to query an untrusted spamd server, an
1565 attacker that gains access to a server running spamd could potentially
1566 exploit this vulnerability to run arbitrary code as the Exim user.
1568 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1569 $primary_hostname instead of what libspf2 thinks the hosts name is.
1571 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1572 a directory entry by the name of the lookup key. Previously, if a
1573 symlink pointed to a non-existing file or a file in a directory that
1574 Exim lacked permissions to read, a lookup for a key matching that
1575 symlink would fail. Now it is enough that a matching directory entry
1576 exists, symlink or not. (Bugzilla 503.)
1578 PH/03 The body_linecount and body_zerocount variables are now exported in the
1581 PH/04 Added the $dnslist_matched variable.
1583 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1584 This means they are set thereafter only if the connection becomes
1587 PH/06 Added the client_condition to authenticators so that some can be skipped
1588 by clients under certain conditions.
1590 PH/07 The error message for a badly-placed control=no_multiline_responses left
1591 "_responses" off the end of the name.
1593 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1595 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1596 (without spaces) instead of just copying the configuration text.
1598 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1600 PH/11 Added $max_received_linelength.
1602 PH/12 Added +ignore_defer and +include_defer to host lists.
1604 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1605 way in which PCRE > 7.0 is built.
1607 PH/14 Implemented queue_only_load_latch.
1609 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1610 MAIL command. The effect was to mangle the value on 64-bit systems.
1612 PH/16 Another patch from the Sieve maintainer.
1614 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1616 PH/18 If a system quota error occurred while trying to create the file for
1617 a maildir delivery, the message "Mailbox is full" was not appended to the
1618 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1619 only to a quota excession during the actual writing of the file.
1621 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1622 characters?) which causes problems in log lines. The DN values are now
1623 passed through string_printing() before being added to log lines.
1625 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1626 and InterBase are left for another time.)
1628 PH/21 Added message_body_newlines option.
1630 PH/22 Guard against possible overflow in moan_check_errorcopy().
1632 PH/23 POSIX allows open() to be a macro; guard against that.
1634 PH/24 If the recipient of an error message contained an @ in the local part
1635 (suitably quoted, of course), incorrect values were put in $domain and
1636 $local_part during the evaluation of errors_copy.
1642 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1643 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1646 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1647 issue a MAIL command.
1649 PH/02 In an ACL statement such as
1651 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1653 if a client was not listed at all, or was listed with a value other than
1654 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1655 the condition was not true (as it should be), so access was not denied.
1656 The bug was that the ! inversion was incorrectly passed on to the second
1657 item. This has been fixed.
1659 PH/03 Added additional dnslists conditions == and =& which are different from
1660 = and & when the dns lookup returns more than one IP address.
1662 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1663 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1665 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1666 FSYNC, which compiles an option called disable_fsync that allows for
1667 bypassing fsync(). The documentation is heavily laced with warnings.
1669 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1671 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1672 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1673 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1674 including adding "make clean"; (3) Added -fPIC when compiling the test
1675 dynamically loaded module, to get rid of a warning.
1677 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1678 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1679 = 0s. The bug is otherwise harmless.
1681 PH/07 There was a bug in the dovecot authenticator such that the value of
1682 $auth1 could be overwritten, and so not correctly preserved, after a
1683 successful authentication. This usually meant that the value preserved by
1684 the server_setid option was incorrect.
1686 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1688 PH/09 Installed PCRE release 7.0.
1690 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1691 run for batched SMTP input. It is now run at the start of every message
1692 in the batch. While fixing this I discovered that the process information
1693 (output by running exiwhat) was not always getting set for -bs and -bS
1694 input. This is fixed, and it now also says "batched" for BSMTP.
1696 PH/11 Added control=no_pipelining.
1698 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1699 patch, slightly modified), and move the expansion of helo_data till after
1700 the connection is made in the smtp transport (so it can use these
1703 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1705 PH/14 Added log_selector = +pid.
1707 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1709 PH/16 Add ${if forany and ${if forall.
1711 PH/17 Added dsn_from option to vary the From: line in DSNs.
1713 PH/18 Flush SMTP output before performing a callout, unless control =
1714 no_callout_flush is set.
1716 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1717 was true (the default) a successful delivery failed to delete the retry
1718 item, thus causing premature timeout of the address. The bug is now
1721 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1723 PH/21 Long custom messages for fakedefer and fakereject are now split up
1724 into multiline reponses in the same way that messages for "deny" and
1725 other ACL rejections are.
1727 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1728 with slight modification.
1730 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1731 draft, changing the syntax and factoring some duplicate code".
1733 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1734 for deliveries of the second and subsequent messages over the same SMTP
1737 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1738 ${reduce, with only minor "tidies".
1740 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1742 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1743 expansion side effects.
1745 PH/27 When a message times out after an over-quota error from an Exim-imposed
1746 quota, the bounce message says "mailbox is full". This message was not
1747 being given when it was a system quota that was exceeded. It now should
1750 MH/03 Made $recipients available in local_scan(). local_scan() already has
1751 better access to the recipient list through recipients_list[], but
1752 $recipients can be useful in postmaster-provided expansion strings.
1754 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1755 in the case of a MAIL command with additional options following the
1756 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1757 were accidentally chopped off.
1759 PH/29 SMTP synchronization checks are implemented when a command is read -
1760 there is a check that no more input is waiting when there shouldn't be
1761 any. However, for some commands, a delay in an ACL can mean that it is
1762 some time before the response is written. In this time, more input might
1763 arrive, invalidly. So now there are extra checks after an ACL has run for
1764 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1765 pipelining has not been advertised.
1767 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1769 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1770 returned in the SMTP response when smtp_return_error_details was false.
1771 This has been fixed.
1773 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1774 instead of the C I/O that was originally supplied, because problems were
1775 reported on Solaris.
1777 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1778 Exim which did not show up earlier: it was assuming that a call to
1779 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1780 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1781 was a macro that became an assignment, so it seemed to work. This has
1782 changed to a proper function call with a void return, hence the compile
1783 error. Exim's code has been fixed.
1785 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1788 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1790 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1792 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1793 because it turns out to be too big or there is a timeout) there is no
1794 "Completed" line in the log. When some messages of this type were
1795 selected by exigrep, they were listed as "not completed". Others were
1796 picked up by some special patterns. I have improved the selection
1797 criteria to be more general.
1799 PH/38 The host_find_failed option in the manualroute router can now be set
1800 to "ignore", to completely ignore a host whose IP address cannot be
1801 found. If all hosts are ignored, the behaviour is controlled by the new
1802 host_all_ignored option.
1804 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1805 homing or because of multiple MX records with /mx) generated more than
1806 one IP address, and the following item turned out to be the local host,
1807 all the secondary addresses of the first item were incorrectly removed
1808 from the list, along with the local host and any following hosts (which
1809 is what is supposed to happen).
1811 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1812 whoever called Exim into the -H file. In the case of the daemon it was
1813 behaving confusingly. When first started, it used values for whoever
1814 started the daemon, but after a SIGHUP it used the Exim user (because it
1815 calls itself on a restart). I have changed the code so that it now always
1818 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1819 message are rejected with the same error (e.g. no authentication or bad
1820 sender address), and a DATA command is nevertheless sent (as can happen
1821 with PIPELINING or a stupid MUA), the error message that was given to the
1822 RCPT commands is included in the rejection of the DATA command. This is
1823 intended to be helpful for MUAs that show only the final error to their
1826 PH/42 Another patch from the Sieve maintainer.
1828 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1829 Eximstats - Fixed some broken HTML links and added missing column headers
1831 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1832 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1834 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1836 PH/43 Yet another patch from the Sieve maintainer.
1838 PH/44 I found a way to check for a TCP/IP connection going away before sending
1839 the response to the final '.' that terminates a message, but only in the
1840 case where the client has not sent further data following the '.'
1841 (unfortunately, this is allowed). However, in many cases there won't be
1842 any further data because there won't be any more messages to send. A call
1843 to select() can be used: if it shows that the input is "ready", there is
1844 either input waiting, or the socket has been closed. An attempt to read
1845 the next input character can distinguish the two cases. Previously, Exim
1846 would have sent an OK response which the client would never have see.
1847 This could lead to message repetition. This fix should cure that, at
1848 least in a lot of common cases.
1850 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1851 advertised in response to EHLO.
1857 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1858 fixed by 4.65/MH/01 (is this a record?) are fixed:
1860 (i) An empty string was always treated as zero by the numeric comparison
1861 operators. This behaviour has been restored.
1863 (ii) It is documented that the numeric comparison operators always treat
1864 their arguments as decimal numbers. This was broken in that numbers
1865 starting with 0 were being interpreted as octal.
1867 While fixing these problems I realized that there was another issue that
1868 hadn't been noticed. Values of message_size_limit (both the global option
1869 and the transport option) were treated as octal if they started with 0.
1870 The documentation was vague. These values are now always treated as
1871 decimal, and I will make that clear in the documentation.
1877 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1878 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1881 MH/01 Don't check that the operands of numeric comparison operators are
1882 integers when their expansion is in "skipping" mode (fixes bug
1883 introduced by 4.64-PH/07).
1885 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1886 child addresses, Exim now panics and dies. Previously, because the count
1887 is held in a short int, deliveries were likely to be lost. As such a
1888 large number of recipients for a single message is ridiculous
1889 (performance will be very, very poor), I have chosen to impose a limit
1890 rather than extend the field.
1896 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1897 leftover -K file (the existence of which was triggered by #402).
1898 While we were at it, introduced process PID as part of the -K
1899 filename. This should rule out race conditions when creating
1902 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1903 processing considerably. Previous code took too long for large mails,
1904 triggering a timeout which in turn triggers #401.
1906 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1907 in the DK code in transports.c. sendfile() is not really portable,
1908 hence the _LINUX specificness.
1910 TF/01 In the add_headers option to the mail command in an Exim filter,
1911 there was a bug that Exim would claim a syntax error in any
1912 header after the first one which had an odd number of characters
1915 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1916 callout verification, Exim cached a "reject" for the entire domain. This
1917 is correct for most verifications, but it is not correct for a recipient
1918 verification with use_sender or use_postmaster set, because in that case
1919 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1920 case of MAIL FROM:<> rejection from other early rejections (e.g.
1921 rejection of HELO). When verifying a recipient using a non-null MAIL
1922 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1923 Whatever the result of the callout, the value of the domain cache is
1924 left unchanged (for any other kind of callout, getting as far as trying
1925 RCPT means that the domain itself is ok).
1927 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1930 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1931 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1932 session, the daemon ignores ECONNECT errors and logs others; it now
1933 ignores EPIPE as well.
1935 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1936 (quoted-printable decoding).
1938 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1939 later the small subsequent patch to fix an introduced bug.
1941 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1943 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1945 PH/08 An error is now given if message_size_limit is specified negative.
1947 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1948 to be given (somewhat) arbitrary names.
1950 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1953 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1954 miscellaneous code fixes
1956 PH/10 Added the log_reject_target ACL modifier to specify where to log
1959 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1960 hostname. This is wrong, because it relates to the incoming message (and
1961 probably the interface on which it is arriving) and not to the outgoing
1962 callout (which could be using a different interface). This has been
1963 changed to use the value of the helo_data option from the smtp transport
1964 instead - this is what is used when a message is actually being sent. If
1965 there is no remote transport (possible with a router that sets up host
1966 addresses), $smtp_active_hostname is used.
1968 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1969 tweaks were necessary in order to get it to work (see also 21 below):
1970 (a) The code assumed that strncpy() returns a negative number on buffer
1971 overflow, which isn't the case. Replaced with Exim's string_format()
1973 (b) There were several signed/unsigned issues. I just did the minimum
1974 hacking in of casts. There is scope for a larger refactoring.
1975 (c) The code used strcasecmp() which is not a standard C function.
1976 Replaced with Exim's strcmpic() function.
1977 (d) The code set only $1; it now sets $auth1 as well.
1978 (e) A simple test gave the error "authentication client didn't specify
1979 service in request". It would seem that Dovecot has changed its
1980 interface. Fortunately there's a specification; I followed it and
1981 changed what the client sends and it appears to be working now.
1983 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1986 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1987 address A is aliased to B and C, where B exists and C does not. Without
1988 -v the output is "A verified" because verification stops after a
1989 successful redirection if more than one address is generated. However,
1990 with -v the child addresses are also verified. Exim was outputting "A
1991 failed to verify" and then showing the successful verification for C,
1992 with its parentage. It now outputs "B failed to verify", showing B's
1993 parentage before showing the successful verification of C.
1995 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1996 look up a TXT record in a specific list after matching in a combined
1999 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2000 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2001 they consult the DNS. I had assumed they would set it the way they
2002 wanted; and indeed my experiments on Linux seem to show that in some
2003 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2004 To be on the safe side, however, I have now made the interface to
2005 host_find_byname() similar to host_find_bydns(), with an argument
2006 containing the DNS resolver options. The host_find_byname() function now
2007 sets these options at its start, just as host_find_bydns() does. The smtp
2008 transport options dns_qualify_single and dns_search_parents are passed to
2009 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2010 of host_find_byname() use the default settings of RES_DEFNAMES
2011 (qualify_single) but not RES_DNSRCH (search_parents).
2013 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2014 spool_read_header() do less string testing, by means of a preliminary
2015 switch on the second character of optional "-foo" lines. (This is
2016 overdue, caused by the large number of possibilities that now exist.
2017 Originally there were few.) While I was there, I also converted the
2018 str(n)cmp tests so they don't re-test the leading "-" and the first
2019 character, in the hope this might squeeze out yet more improvement.
2021 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2022 flag allowing group syntax was set by the header_syntax check but not
2023 turned off, possible causing trouble later; (2) The flag was not being
2024 set at all for the header_verify test, causing "group"-style headers to
2025 be rejected. I have now set it in this case, and also caused header_
2026 verify to ignore an empty address taken from a group. While doing this, I
2027 came across some other cases where the code for allowing group syntax
2028 while scanning a header line wasn't quite right (mostly, not resetting
2029 the flag correctly in the right place). These bugs could have caused
2030 trouble for malformed header lines. I hope it is now all correct.
2032 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2033 with the "reply" argument non-NULL. The code, however (which originally
2034 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2035 but it didn't always do it. This confused somebody who was copying the
2036 code for some other use. I have removed all the tests.
2038 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2039 feature that was used to support insecure browsers during the U.S. crypto
2040 embargo. It requires special client support, and Exim is probably the
2041 only MTA that supported it -- and would never use it because real RSA is
2042 always available. This code has been removed, because it had the bad
2043 effect of slowing Exim down by computing (never used) parameters for the
2044 RSA_EXPORT functionality.
2046 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2047 authenticator to fail if there's a tab character in the incoming data
2048 (there should never be unless someone is messing about, as it's supposed
2049 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2050 if the connection is using TLS or if the remote IP is the same as the
2051 local IP, and the "valid-client-cert option" if a client certificate has
2054 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2055 authenticators. This can be used for authorization after authentication
2056 succeeds. (In the case of plaintext, it servers for both authentication
2059 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2060 if any retry times were supplied.
2062 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2063 connection, where there is no client IP address to check. In this
2064 situation, the verify now always succeeds.
2066 PH/25 Applied John Jetmore's -Mset patch.
2068 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2070 PH/27 In a string expansion for a processed (not raw) header when multiple
2071 headers of the same name were present, leading whitespace was being
2072 removed from all of them, but trailing whitespace was being removed only
2073 from the last one. Now trailing whitespace is removed from each header
2074 before concatenation. Completely empty headers in a concatenation (as
2075 before) are ignored.
2077 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2078 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2080 PH/29 [Removed. This was a change that I later backed out, and forgot to
2081 correct the ChangeLog entry (that I had efficiently created) before
2082 committing the later change.]
2084 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2085 address errors (4xx response to RCPT) over the same connection as other
2086 messages routed to the same hosts. Such deliveries are always "forced",
2087 so retry times are not inspected. This resulted in far too many retries
2088 for the affected addresses. The effect occurred only when there were more
2089 hosts than the hosts_max_try setting in the smtp transport when it had
2090 the 4xx errors. Those hosts that it had tried were not added to the list
2091 of hosts for which the message was waiting, so if all were tried, there
2092 was no problem. Two fixes have been applied:
2094 (i) If there are any address or message errors in an SMTP delivery, none
2095 of the hosts (tried or untried) are now added to the list of hosts
2096 for which the message is waiting, so the message should not be a
2097 candidate for sending over the same connection that was used for a
2098 successful delivery of some other message. This seems entirely
2099 reasonable: after all the message is NOT "waiting for some host".
2100 This is so "obvious" that I'm not sure why it wasn't done
2101 previously. Hope I haven't missed anything, but it can't do any
2102 harm, as the worst effect is to miss an optimization.
2104 (ii) If, despite (i), such a delivery is accidentally attempted, the
2105 routing retry time is respected, so at least it doesn't keep
2106 hammering the server.
2108 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2109 in ${readsocket because some servers need this prod.
2111 PH/32 Added some extra debug output when updating a wait-xxx database.
2113 PH/33 The hint "could be header name not terminated by colon", which has been
2114 given for certain expansion errors for a long time, was not being given
2115 for the ${if def:h_colon_omitted{... case.
2117 PH/34 The spec says: "With one important exception, whenever a domain list is
2118 being scanned, $domain contains the subject domain." There was at least
2119 one case where this was not true.
2121 PH/35 The error "getsockname() failed: connection reset by peer" was being
2122 written to the panic log as well as the main log, but it isn't really
2123 panic-worthy as it just means the connection died rather early on. I have
2124 removed the panic log writing for the ECONNRESET error when getsockname()
2127 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2128 runs only) independently of the message's sender address. This meant
2129 that, if the 4xx error was in fact related to the sender, a different
2130 message to the same recipient with a different sender could confuse
2131 things. In particualar, this can happen when sending to a greylisting
2132 server, but other circumstances could also provoke similar problems.
2133 I have changed the default so that the retry time for these errors is now
2134 based a combination of the sender and recipient addresses. This change
2135 can be overridden by setting address_retry_include_sender=false in the
2138 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2139 remote server are returned as part of bounce messages. This was not
2140 happening for LMTP over a pipe (the lmtp transport), but now it is the
2141 same for both kinds of LMTP.
2143 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2144 in header lines that were in fact CNAMEs. This is no longer the case.
2146 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2147 and queue runs started by the daemon processed all messages. This has
2148 been fixed so that -R and -S can now usefully be given with -q<time>.
2150 PH/40 Import PCRE release 6.7 (fixes some bugs).
2152 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2154 PH/42 Give an error if -q is specified more than once.
2156 PH/43 Renamed the variables $interface_address and $interface_port as
2157 $received_ip_address and $received_port, to make it clear that these
2158 values apply to message reception, and not to the outgoing interface when
2159 a message is delivered. (The old names remain recognized, of course.)
2161 PH/44 There was no timeout on the connect() call when using a Unix domain
2162 socket in the ${readsocket expansion. There now is.
2164 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2165 be meaningful with "accept".
2167 SC/01 Eximstats V1.43
2168 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2170 SC/02 Eximstats V1.44
2171 Use a glob alias rather than an array ref in the generated
2172 parser. This improves both readability and performance.
2174 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2175 Collect SpamAssassin and rejection statistics.
2176 Don't display local sender or destination tables unless
2177 there is data to show.
2178 Added average volumes into the top table text output.
2180 SC/04 Eximstats V1.46
2181 Collect data on the number of addresses (recipients)
2182 as well as the number of messages.
2184 SC/05 Eximstats V1.47
2185 Added 'Message too big' to the list of mail rejection
2186 reasons (thanks to Marco Gaiarin).
2188 SC/06 Eximstats V1.48
2189 Mainlog lines which have GMT offsets and are too short to
2190 have a flag are now skipped.
2192 SC/07 Eximstats V1.49 (Alain Williams)
2193 Added the -emptyok flag.
2195 SC/08 Eximstats V1.50
2196 Fixes for obtaining the IP address from reject messages.
2198 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2199 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2200 whitesspace changes from 4.64-PH/27
2202 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2205 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2206 are found, allow negative numbers in numeric criteria)
2208 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2210 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2211 to match changes made in 4.64-PH/43
2213 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2215 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2216 "domain" is actually an IP address, and force a failure. This locks out
2217 those revolvers/nameservers that support "A-for-A" lookups, in
2218 contravention of the specifications.
2220 PH/48 When a host name was looked up from an IP address, and the subsequent
2221 forward lookup of the name timed out, the host name was left in
2222 $sender_host_name, contrary to the specification.
2224 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2225 restricted to single-key lookups, Exim was not diagnosing an error if
2226 * or *@ was used with a query-style lookup.
2228 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2230 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2231 long ago, but noone interested enough thought of it. Let's just say that
2232 the "1.1" means that there are some new functions that weren't there at
2233 some point in the past.
2235 PH/51 Error processing for expansion failure of helo_data from an smtp
2236 transport during callout processing was broken.
2238 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2239 tested/used via the -bh/-bhc/-bs options.
2241 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2242 bug, fixed in subsequent PCRE releases).
2244 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2245 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2247 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2253 SC/01 Use a glob alias rather than an array ref in eximstats generated
2254 parser. This improves both readability and performance.
2256 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2257 Don't display local sender or destination tables in eximstats unless
2258 there is data to show.
2259 Added average volumes into the eximstats top table text output.
2261 SC/03 Collect data on the number of addresses (recipients) as well
2262 as the number of messages in eximstats.
2264 TF/01 Correct an error in the documentation for the redirect router. Exim
2265 does (usually) call initgroups() when daemonizing.
2267 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2268 with consistent privilege compared to when running as a daemon.
2270 TF/03 Note in the spec that $authenticated_id is not set for local
2271 submissions from trusted users.
2273 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2274 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2276 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2277 by adding some example configuration directives to the default
2278 configuration file. A little bit of work is required to uncomment the
2279 directives and define how usernames and passwords are checked, but
2280 there is now a framework to start from.
2282 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2283 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2284 without this. I don't know how relevant this is to other LDAP libraries.
2286 PH/02 Add the verb name to the "unknown ACL verb" error.
2288 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2290 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2292 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2293 directory not expanded when it should be if an expanded home directory
2294 was set for the address (which is overridden by the transport).
2296 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2299 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2300 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2301 because it is too late at that time, and has no effect.
2303 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2304 security issue with \' (bugzilla #107). I could not use the
2305 PQescapeStringConn() function, because it needs a PGconn value as one of
2308 PH/08 When testing addresses using -bt, indicate those final addresses that
2309 are duplicates that would not cause an additional delivery. At least one
2310 person was confused, thinking that -bt output corresponded to deliveries.
2311 (Suppressing duplicates isn't a good idea as you lose the information
2312 about possibly different redirections that led to the duplicates.)
2314 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2315 systems where poll() doesn't work, in particular OS X.
2317 PH/10 Added more information to debugging output for retry time not reached.
2319 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2320 operations in malware.c.
2322 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2325 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2326 syslog_duplication set false, log lines that would normally be written
2327 both the the main log and to the reject log were not written to syslog at
2330 PH/14 In the default configuration, change the use of "message" in ACL warn
2331 statements to "add_header".
2333 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2334 not followed by a command (e.g. "seen endif").
2336 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2337 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2340 PH/17 Added extra conditions to the default value of delay_warning_condition
2344 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2345 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2346 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2349 The Auto-Submitted: and various List- headers are standardised, whereas I
2350 don't think Precedence: ever was.
2352 PH/18 Refactored debugging code in route_finduser() to show more information,
2353 in particular, the error code if getpwnam() issues one.
2355 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2356 This is apparently needed in addition to the PH/07 change above to avoid
2357 any possible encoding problems.
2359 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2360 but not after initializing Perl.
2362 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2363 output them only if debugging. By default they are written stderr,
2364 apparently, which is not desirable.
2366 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2369 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2372 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2374 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2375 authenticated or an ident call has been made. Suppress the default
2376 values for $authenticated_id and $authenticated_sender (but permit -oMai
2377 and -oMas) when testing with -bh.
2379 PH/24 Re-jigged the order of the tests in the default configuration so that the
2380 tests for valid domains and recipients precede the DNS black list and CSA
2381 tests, on the grounds that those ones are more expensive.
2383 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2384 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2385 This bug exists in every version of Exim that I still have, right back to
2388 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2389 However, an attempt to turn on case-sensitivity in a regex key by
2390 including (?-i) didn't work because the subject string was already
2391 lowercased, and the effects were non-intuitive. It turns out that a
2392 one-line patch can be used to allow (?-i) to work as expected.
2398 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2399 other effects) broke the use of negated acl sub-conditions.
2401 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2404 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2405 "Deny" causes Exim to reject the incoming connection with a 554 error.
2406 Unfortunately, if there is a major crisis, such as a disk failure,
2407 tcp-wrappers gives "deny", whereas what one would like would be some
2408 kind of temporary error. A kludge has been added to help with this.
2409 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2410 554 error is used if errno is still zero or contains ENOENT (which occurs
2411 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2414 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2416 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2417 errors. Otherwise a message that provokes a temporary error (when other
2418 messages do not) can cause a whole host to time out.
2420 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2421 addresses were routed directly to files or pipes from a redirect router.
2422 File deliveries just didn't batch; pipe deliveries might have suffered
2425 PH/06 A failure to get a lock for a hints database would erroneously always say
2426 "Failed to get write lock", even when it was really a read lock.
2428 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2429 of 0600. This has been changed to use the value of the lockfile_mode
2430 option (which defaults to 0600).
2432 PH/08 Applied small patch from the Sieve maintainer.
2434 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2435 folder from quota calculations, a direct delivery into this folder messed
2436 up the contents of the maildirsize file. This was because the regex was
2437 used only to exclude .Trash (or whatever) when the size of the mailbox
2438 was calculated. There was no check that a delivery was happening into an
2439 excluded directory. This bug has been fixed by ignoring all quota
2440 processing for deliveries into excluded directories.
2442 PH/10 Added the maildirfolder_create_regex option to appendfile.
2448 PH/01 The code for finding all the local interface addresses on a FreeBSD
2449 system running IPv6 was broken. This may well have applied to all BSD
2450 systems, as well as to others that have similar system calls. The broken
2451 code found IPv4 interfaces correctly, but gave incorrect values for the
2452 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2453 that it would not match correctly against @[] and not recognize the IPv6
2456 PH/02 The ipliteral router was not recognizing addresses of the form user@
2457 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2459 PH/03 Added disable_ipv6.
2461 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2462 decoded form, because it is most often used to construct To: headers
2463 lines in autoreplies, and the decoded form may well be syntactically
2464 invalid. However, $reply_address has leading white space removed, and all
2465 newlines turned into spaces so that the autoreply transport does not
2468 PH/05 If group was specified without a user on a router, and no group or user
2469 was specified on a transport, the group from the router was ignored.
2471 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2472 for visible compile-time settings that can be used to change these
2473 numbers, for those that want even more. Backwards compatibility with old
2474 spool files has been maintained. However, going back to a previous Exim
2475 release will lost any variables that are in spool files.
2477 PH/07 Two small changes when running in the test harness: increase delay when
2478 passing a TCP/IP connection to a new process, in case the original
2479 process has to generate a bounce, and remove special handling of
2480 127.0.0.2 (sic), which is no longer necessary.
2482 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2483 be the same on different OS.
2485 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2488 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2489 whether --show-vars was specified or not
2491 JJ/02 exipick: Added support for new ACL variable spool format introduced
2494 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2495 syntactically invalid From: or Reply-to: line, and a filter used this to
2496 generate an autoreply, and therefore failed to obtain an address for the
2497 autoreply, Exim could try to deliver to a non-existent relative file
2498 name, causing unrelated and misleading errors. What now happens is that
2499 it logs this as a hard delivery error, but does not attempt to create a
2502 PH/11 The exinext utility has a -C option for testing purposes, but although
2503 the given file was scanned by exinext itself; it wasn't being passed on
2504 when Exim was called.
2506 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2507 an end-of-file indication when reading a command response.
2509 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2510 compiled. In many other places in Exim, IPv6 addresses are always
2511 recognized, so I have changed this. It also means that IPv4 domain
2512 literals of the form [IPV4:n.n.n.n] are now always recognized.
2514 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2515 used if the router is not running as root, for example, when verifying at
2516 ACL time, or when using -bh. The debugging output from this situation was
2517 non-existent - all you got was a failure to exec. I have made two
2520 (a) Failures to set uid/gid, the current directory, or a process leader
2521 in a subprocess such as that created by queryprogram now generate
2522 suitable debugging ouput when -d is set.
2524 (b) The queryprogram router detects when it is not running as root,
2525 outputs suitable debugging information if -d is set, and then runs
2526 the subprocess without attempting to change uid/gid.
2528 PH/15 Minor change to Makefile for building test_host (undocumented testing
2531 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2532 additional section of a DNS packet that returns MX or SRV records.
2533 Instead, it always explicitly searches for A/AAAA records. This avoids
2534 major problems that occur when a DNS server includes only records of one
2535 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2536 fixed another bug: if SRV records were looked up and the corresponding
2537 address records were *not* found in the additional section, the port
2538 values from the SRV records were lost.
2540 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2541 using the correct key (the original address) when searching the retry
2542 rules in order to find which one to use for generating the retry hint.
2544 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2545 adding the default one. Similarly, if it contains a Reply-To: header, the
2546 errors_reply_to option, if set, is not used.
2548 PH/19 When calculating a retry time, Exim used to measure the "time since
2549 failure" by looking at the "first failed" field in the retry record. Now
2550 it does not use this if it is later than than the arrival time of the
2551 message. Instead it uses the arrival time. This makes for better
2552 behaviour in cases where some deliveries succeed, thus re-setting the
2553 "first failed" field. An example is a quota failure for a huge message
2554 when small messages continue to be delivered. Without this change, the
2555 "time since failure" will always be short, possible causing more frequent
2556 delivery attempts for the huge message than are intended.
2557 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2559 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2560 $1, $2, $3) because the numerical variables can be reset during some
2561 expansion items (e.g. "match"), thereby losing the authentication data.
2563 PH/21 Make -bV show the size of off_t variables so that the test suite can
2564 decide whether to run tests for quotas > 2G.
2566 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2567 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2568 filecount value is greater than 2G or if a quota value is greater than 2G
2569 on a system where the size of off_t is not greater than 4, a panic error
2572 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2573 never match. The debug and -bh output now contains an explicit error
2574 message indicating a malformed IPv4 address or mask.
2576 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2577 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2578 PH/23 above applies.
2580 PH/25 Do not write to syslog when running in the test harness. The only
2581 occasion when this arises is a failure to open the main or panic logs
2582 (for which there is an explicit test).
2584 PH/26 Added the /no_tell option to "control=freeze".
2586 PH/27 If a host name lookup failed very early in a connection, for example, if
2587 the IP address matched host_lookup and the reverse lookup yielded a name
2588 that did not have a forward lookup, an error message of the form "no IP
2589 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2590 could be logged. Now it outputs the IP address instead of "NULL".
2592 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2593 allows the sender and the authenticated sender to be set when
2594 submitting a message from within Exim. Since child_open_exim() is
2595 documented for local_scan(), the new function should be too.
2597 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2598 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2599 results in an empty string is now treated as unset.
2601 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2603 PH/31 Added sender_verify_fail logging option.
2605 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2606 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2607 tidied the source and removed it altogether.
2609 PH/33 When a queue run was abandoned because the load average was too high, a
2610 log line was always written; now it is written only if the queue_run log
2611 selector is set. In addition, the log line for abandonment now contains
2612 information about the queue run such as the pid. This is always present
2613 in "start" and "stop" lines but was omitted from the "abandon" line.
2615 PH/34 Omit spaces between a header name and the colon in the error message that
2616 is given when verify = headers_syntax fails (if there are lots of them,
2617 the message gets confusing).
2619 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2620 names, as there are now some PTR records that contain slashes. This check
2621 is only to protect against broken name servers that fall over on strange
2622 characters, so the fact that it applies to all lookups doesn't matter.
2624 PH/36 Now that the new test suite is complete, we can remove some of the
2625 special code in Exim that was needed for the old test suite. For example,
2626 sorting DNS records because real resolvers return them in an arbitrary
2627 order. The new test suite's fake resolver always returns records in the
2630 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2631 bounces) except when queue_only is set, to avoid logging races between
2632 the different processes.
2634 PH/38 Panic-die if .include specifies a non-absolute path.
2636 PH/39 A tweak to the "H" retry rule from its user.
2638 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2639 a label. They prevented compilation on older perls.
2641 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2642 a warning to be raised on newish perls.
2644 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2645 on queue. Changes to match documented behaviour of showing count of
2646 messages matching specified criteria.
2648 PH/40 Changed the default ident timeout from 30s to 5s.
2650 PH/41 Added support for the use of login_cap features, on those BSD systems
2651 that have them, for controlling the resources used by pipe deliveries.
2653 PH/42 The content-scanning code uses fopen() to create files in which to put
2654 message data. Previously it was not paying any attention to the mode of
2655 the files. Exim runs with umask(0) because the rest of the code creates
2656 files with open(), and sets the required mode explicitly. Thus, these
2657 files were ending up world-writeable. This was not a big issue, because,
2658 being within the spool directory, they were not world-accessible. I have
2659 created a function called modefopen, which takes an additional mode
2660 argument. It sets umask(777), creates the file, chmods it to the required
2661 mode, then resets the umask. All the relevant calls to fopen() in the
2662 content scanning code have been changed to use this function.
2664 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2665 to 24 hours. This avoids potential overflow problems when processing G
2666 and H retry rules. I suspect nobody ever tinkers with this value.
2668 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2670 PH/45 When the plaintext authenticator is running as a client, the server's
2671 challenges are checked to ensure they are valid base64 strings. By
2672 default, the authentication attempt is cancelled if an invalid string is
2673 received. Setting client_ignore_invalid_base64 true ignores these errors.
2674 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2675 they are received. Thus, the responses can be made to depend on the
2676 challenges. If an invalid string is ignored, an empty string is placed in
2679 PH/46 Messages that are created by the autoreply transport now contains a
2680 References: header, in accordance with RFCs 2822 and 3834.
2682 PH/47 Added authenticated_sender_force to the smtp transport.
2684 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2686 PH/49 Installed latest patch from the Sieve maintainer.
2688 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2689 was also set, the appendfile transport was unnecessarily scanning a
2690 directory of message files (e.g. for maildir delivery) to find the count
2691 of files (along with the size), even though it did not need this
2692 information. It now does the scan only if it needs to find either the
2693 size of the count of files.
2695 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2697 PH/52 Two bugs concerned with error handling when the smtp transport is
2700 (i) Exim was not creating retry information for temporary errors given
2701 for individual recipients after the DATA command when the smtp transport
2702 was used in LMTP mode. This meant that they could be retried too
2703 frequently, and not timed out correctly.
2705 (ii) Exim was setting the flag that allows error details to be returned
2706 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2707 recipients that were returned after the DATA command.
2709 PH/53 This is related to PH/52, but is more general: for any failing address,
2710 when detailed error information was permitted to be returned to the
2711 sender, but the error was temporary, then after the final timeout, only
2712 "retry timeout exceeded" was returned. Now it returns the full error as
2713 well as "retry timeout exceeded".
2715 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2716 do this, and (what is worse) MTAs that accept it.
2718 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2719 will now be deprecated.
2721 PH/56 New os.c-cygwin from the Cygwin maintainer.
2723 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2724 formats (previously only available in exim formats via -bpr, -bpru,
2725 and -bpra. Now also available in native and exiqgrep formats)
2727 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2728 with very large, slow to parse queues
2730 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2732 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2734 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2735 responses to authentication challenges, though it was showing the
2736 challenges; (ii) I've removed the CR characters from the debug output for
2739 PH/58 Allow for the insertion of a newline as well as a space when a string
2740 is turned into more than one encoded-word during RFC 2047 encoding. The
2741 Sieve code now uses this.
2743 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2744 data_4xx, lost_connection, tls_required.
2746 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2747 message was being sent as an SMTP response.
2749 PH/61 Add -l and -k options to exicyclog.
2751 PH/62 When verifying, if an address was redirected to one new address, so that
2752 verification continued, and the new address failed or deferred after
2753 having set something in $address_data, the value of $address_data was not
2754 passed back to the ACL. This was different to the case when no
2755 redirection occurred. The value is now passed back in both cases.
2757 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2758 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2759 use login_cap.h, so on its own it isn't the distinguishing feature. The
2760 new name refers directly to the setclassresources() function.
2762 PH/65 Added configuration files for NetBSD3.
2764 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2766 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2767 is preferred over IPv4.
2769 PH/68 The bounce_return_message and bounce_return_body options were not being
2770 honoured for bounces generated during the reception of non-SMTP messages.
2771 In particular, this applied to messages rejected by the ACL. This bug has
2772 been fixed. However, if bounce_return_message is true and bounce_return_
2773 body is false, the headers that are returned for a non-SMTP message
2774 include only those that have been read before the error was detected.
2775 (In the case of an ACL rejection, they have all been read.)
2777 PH/69 The HTML version of the specification is now built in a directory called
2778 spec_html instead of spec.html, because the latter looks like a path with
2779 a MIME-type, and this confuses some software.
2781 PH/70 Catch two compiler warnings in sieve.c.
2783 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2784 function verify_get_ident() calls ip_connect() to connect a socket, but
2785 if the "connect()" function timed out, ip_connect() used to close the
2786 socket. However, verify_get_ident() also closes the socket later, and in
2787 between Exim writes to the log, which may get opened at this point. When
2788 the socket was closed in ip_connect(), the log could get the same file
2789 descriptor number as the socket. This naturally causes chaos. The fix is
2790 not to close the socket in ip_connect(); the socket should be closed by
2791 the function that creates it. There was only one place in the code where
2792 this was missing, in the iplookup router, which I don't think anybody now
2793 uses, but I've fixed it anyway.
2795 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2796 well as to direct DNS lookups. Otherwise the handling of names in host
2797 lists is inconsistent and therefore confusing.
2803 PH/01 Two changes to the default runtime configuration:
2805 (1) Move the checks for relay_from_hosts and authenticated clients from
2806 after to before the (commented out) DNS black list checks.
2808 (2) Add control=submission to the relay_from_hosts and authenticated
2809 clients checks, on the grounds that messages accepted by these
2810 statements are most likely to be submissions.
2812 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2814 (1) Generate an error if the third argument for the ${prvs expansion is
2817 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2820 (3) Reset the variables that are obtained from the first argument of
2821 ${prvscheck and used in the second argument before leaving the code,
2822 because their memory is reclaimed, so using them afterwards may do
2825 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2826 one (it's much easier than Tom thought :-).
2828 (5) Because of (4), we can now allow for the use of $prvscheck_result
2829 inside the third argument.
2831 PH/03 For some reason, the default setting of PATH when running a command from
2832 a pipe transport was just "/usr/bin". I have changed it to
2835 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2836 anything to be listed in the output from -bV.
2838 PH/05 When a filter generated an autoreply, the entire To: header line was
2839 quoted in the delivery log line, like this:
2841 => >A.N.Other <ano@some.domain> <original@ddress> ...
2843 This has been changed so that it extracts the operative address. There
2844 may be more than one such address. If so, they are comma-separated, like
2847 => >ano@some.domain,ona@other.domain <original@ddress> ...
2849 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2850 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2851 not being looked up in the rDNS to get a host name, Exim was showing the
2852 IP address twice in Received: lines, even though the IP addresses were
2853 identical. For example:
2855 Received: from [1.2.3.4] (helo=[1.2.3.4])
2857 However, if the real host name was known, it was omitting the HELO data
2858 if it matched the actual IP address. This has been tidied up so that it
2859 doesn't show the same IP address twice.
2861 PH/07 When both +timestamp and +memory debugging was on, the value given by
2862 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2863 called by the debug printing, thereby overwriting the timestamp buffer.
2864 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2866 PH/08 When the original message was included in an autoreply transport, it
2867 always said "this is a copy of the message, including all the headers",
2868 even if body_only or headers_only was set. It now gives an appropriate
2871 PH/09 Applied a patch from the Sieve maintainer which:
2873 o fixes some comments
2874 o adds the (disabled) notify extension core
2875 o adds some debug output for the result of if/elsif tests
2876 o points to the current vacation draft in the documentation
2877 and documents the missing references header update
2881 o fixes a bug in processing the envelope test (when testing
2882 multiple envelope elements, the last element determinted the
2885 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2886 Electronic Mail") by including:
2888 Auto-submitted: auto-generated
2890 in the messages that it generates (bounce messages and others, such as
2891 warnings). In the case of bounce messages for non-SMTP mesages, there was
2892 also a typo: it was using "Auto_submitted" (underscore instead of
2893 hyphen). Since every message generated by Exim is necessarily in response
2894 to another message, thes have all been changed to:
2896 Auto-Submitted: auto-replied
2898 in accordance with these statements in the RFC:
2900 The auto-replied keyword:
2902 - SHOULD be used on messages sent in direct response to another
2903 message by an automatic process,
2905 - MUST NOT be used on manually-generated messages,
2907 - MAY be used on Delivery Status Notifications (DSNs) and Message
2908 Disposition Notifications (MDNs),
2910 - MUST NOT be used on messages generated by automatic or periodic
2911 processes, except for messages which are automatic responses to
2914 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2915 to the default Received: header definition.
2917 PH/12 Added log selector acl_warn_skipped (default on).
2919 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2920 variables because (a) they are in the wrong storage pool and (b) even if
2921 they were copied, it wouldn't work properly because of the caching.
2923 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2924 checking when decoding. Apparently there are clients that generate
2925 overlong encoded strings. Why am I not surprised?
2927 PH/15 If the first argument of "${if match_address" was not empty, but did not
2928 contain an "@" character, Exim crashed. Now it writes a panic log message
2929 and treats the condition as false.
2931 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2933 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2934 extension "envelope-auth". The code is finished and in agreement with
2935 other implementations, but there is no documentation so far and in fact,
2936 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2937 not changing the active code.
2939 Print executed "if" and "elsif" statements when debugging is used. This
2940 helps a great deal to understand what a filter does.
2942 Document more things not specified clearly in RFC3028. I had all this
2943 sorted out, when out of a sudden new issues came to my mind. Oops."
2945 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2948 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2949 canonical form (as documented). However, after a host name lookup from
2950 the IP address, check_host() was doing a simple string comparison with
2951 addresses acquired from the DNS when checking that the found name did
2952 have the original IP as one of its addresses. Since any found IPv6
2953 addresses are likely to be in abbreviated form, the comparison could
2954 fail. Luckily, there already exists a function for doing the comparison
2955 by converting both addresses to binary, so now that is used instead of
2956 the text comparison.
2958 PH/20 There was another similar case to PH/19, when a complete host name was
2959 given in a host list; looking up its IP address could give an abbreviated
2960 form, whereas the current host's name might or might not be abbreviated.
2961 The same fix has been applied.
2967 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2968 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2971 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2972 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2974 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2976 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2977 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2978 Also, for "Auto-Submitted", treat anything other than "no" as
2979 non-personal, in accordance with RFC 3834. (Previously it treated
2980 anything starting "auto-" as non-personal.)
2982 TF/01 The control=submission/name=... option had a problem with syntax
2983 errors if the name included a slash character. The /name= option
2984 now slurps the rest of the string, so it can include any characters
2985 but it must come last in the list of options (after /sender_retain
2988 PH/05 Some modifications to the interface to the fake nameserver for the new
2996 TK/01 Added the "success_on_redirect" address verification option. See
2997 NewStuff for rationale and an example.
2999 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3001 PH/02 Patch to exigrep to allow it to work on syslog lines.
3003 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3004 fread() to skip over the body file's header line, because in Cygwin the
3005 header line is locked and is inaccessible.
3007 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3008 co-exist for some time) to make it clear that it is the Exim ID that is
3009 referenced, not the Message-ID: header line.
3011 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3012 string_format() function, because snprintf() does not exist on all
3015 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3016 ${stat: expansion item.
3018 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3019 protocol synchronization error", to keep the pedants happy.
3021 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3022 well as for IRIX systems, when gcc is being used. See the host.c source
3025 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3027 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3030 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3031 transport and to the smtp transport in LMTP mode.
3033 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3035 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3036 run a filter in a subprocess. This could lead to confusion in subsequent
3037 lookups in the parent process. There should also be a search_tidyup() at
3038 the end of the subprocess.
3040 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3041 only if the host matched helo_try_verify_hosts, which caused the
3042 verification to occur when the EHLO/HELO command was issued. The ACL just
3043 tested the remembered result. Now, if a previous verification attempt has
3044 not happened, "verify = helo" does it there and then.
3046 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3048 TK/03 Fix log output including CR from clamd.
3050 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3051 exist provoked a memory error which could cause a segfault.
3053 PH/15 Installed PCRE 6.2
3055 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3057 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3058 of the problem. Specifically, suggested +O2 rather than +O1 for the
3061 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3063 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3064 record, the port was not correctly logged when the outgoing_port log
3065 selector was set (it logged the transort's default port).
3067 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3068 fallback_hosts, and "hosts" in the smtp transport.
3070 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3071 host errors such as "Connection refused".
3073 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3074 authentication with radiusclient 0.4.9:
3076 - Error returned from rc_read_config was caught wrongly
3077 - Username/password not passed on to radius server due to wrong length.
3079 The presumption is that some radiusclient API changes for 4.51/PH/17
3080 were not taken care of correctly. The code is still untested by me (my
3081 Linux distribution still has 0.3.2 of radiusclient), but it was
3082 contributed by a Radius user.
3084 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3085 expanding the "port" option of the smtp transport.
3087 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3088 while reading a MIME header. Thanks to Tom Hughes for a patch.
3090 PH/24 Include config.h inside local_scan.h so that configuration settings are
3093 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3094 that in an ACL for RCPT (for example), you can examine exactly what was
3097 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3098 commands, but it was not correctly comparing the address with the actual
3099 client host address. Thus, it would show the EHLO address in Received:
3100 header lines when this was not necessary.
3102 PH/27 Added the % operator to ${eval:}.
3104 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3105 it should be ignoring failures (because with -C, for example, it has lost
3106 privilege). It wasn't ignoring creation failures other than "already
3109 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3112 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3113 because an input line was too long, either on its own, or by virtue of
3114 too many continuations, the temporary file was not being removed, and the
3115 return code was incorrect.
3117 PH/31 Missing "BOOL" in function definition in filtertest.c.
3119 PH/32 Applied Sieve patches from the maintainer.
3121 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3123 PH/33 Added "verify = not_blind".
3125 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3126 Local/Makefile (with some defaults set). These are used in built scripts
3127 such as exicyclog, but they have never been used in the exim_install
3128 script (though there are many overriding facilities there). I have
3129 arranged that the exim_install script now takes note of these two
3132 PH/35 Installed configuration files for Dragonfly.
3134 PH/36 When a locally submitted message by a trusted user did not contain a
3135 From: header, and the sender address was obtained from -f or from an SMTP
3136 MAIL command, and the trusted user did not use -F to supply a sender
3137 name, $originator_name was incorrectly used when constructing a From:
3138 header. Furthermore, $originator_name was used for submission mode
3139 messages from external hosts without From: headers in a similar way,
3140 which is clearly wrong.
3142 PH/37 Added control=suppress_local_fixups.
3144 PH/38 When log_selector = +received_sender was set, and the addition of the
3145 sender made the log line's construction buffer exactly full, or one byte
3146 less than full, an overflow happened when the terminating "\n" was
3149 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3150 when the result of a list match is failure because a DNS lookup failed.
3152 PH/40 RM_COMMAND is now used in the building process.
3154 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3155 the "build-* directories that it finds.
3157 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3158 address in a domain literal was a prefix of an interface address.
3160 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3161 when verifying a sender address, unless rewrite_headers is false.
3163 PH/44 Wrote a long comment about why errors_to addresses are verified as
3164 recipients, not senders.
3166 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3167 the ratelimit ACL was added.
3169 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3171 PH/46 Added extra information about PostgreSQL errors to the error string.
3173 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3174 avoiding the need to install special zones in a real server. This is
3175 backwards compatible; if it can't find the fake resolver, it drops back.
3176 Thus, both old and new test suites can be run.
3178 TF/02 Added util/ratelimit.pl
3180 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3181 clock is set back in time.
3183 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3184 Candler <B.Candler@pobox.com>.
3186 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3187 OK for addresses that are the result of redirections.
3189 PH/48 A number of further additions for the benefit of the new test suite,
3190 including a fake gethostbyname() that interfaces to the fake DNS resolver
3193 TF/06 The fix for widen_domains has also been applied to qualify_single and
3194 search_parents which are the other dnslookup options that can cause
3197 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3200 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3202 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3203 service" if it encountered an MX record with an empty target hostname.
3204 The message is now "an MX or SRV record indicated no SMTP service".
3206 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3207 if the DNS of the sending site is misconfigured. This is quite a
3208 common situation. This change restores the behaviour of treating a
3209 helo verification defer as a failure.
3211 PH/51 If self=fail was set on a router, the bounce message did not include the
3212 actual error message.
3218 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3220 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3221 command itself ended in error, the underlying message about the transport
3222 filter timeout was being overwritten with the pipe command error. Now the
3223 underlying error message should be appended to the second error message.
3225 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3227 PH/02 Reduce the amount of output that "make" produces by default. Full output
3228 can still be requested.
3230 PH/03 The warning log line about a condition test deferring for a "warn" verb
3231 was being output only once per connection, rather than after each
3232 occurrence (because it was using the same function as for successful
3233 "warn" verbs). This seems wrong, so I have changed it.
3235 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3236 it should not have, which might have caused a crash in the right
3237 circumstances, but probably never did.
3239 PH/04 Installed a modified version of Tony Finch's patch to make submission
3240 mode fix the return path as well as the Sender: header line, and to
3241 add a /name= option so that you can make the user's friendly name appear
3244 TF/03 Added the control = fakedefer ACL modifier.
3246 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3247 Mark Lowes for thorough testing.
3249 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3251 TK/03 Merged latest SRS patch from Miles Wilton.
3253 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3254 with the definition in sysexits.h (which is #included earlier).
3255 Fortunately, Exim does not actually use EX_OK. The code used to try to
3256 preserve the sysexits.h value, by assumimg that macro definitions were
3257 scanned for macro replacements. I have been disabused of this notion,
3258 so now the code just undefines EX_OK before #including unistd.h.
3260 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3261 in the smtp transport. When a block could not be written in a single
3262 write() function, the timeout was being re-applied to each part-write.
3263 This seems wrong - if the receiver was accepting one byte at a time it
3264 would take for ever. The timeout is now adjusted when this happens. It
3265 doesn't have to be particularly precise.
3267 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3268 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3270 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3271 without a domain if the check to <postmaster@domain> fails.
3273 SC/01 Eximstats: added -xls and the ability to specify output files
3274 (patch written by Frank Heydlauf).
3276 SC/02 Eximstats: use FileHandles for outputing results.
3278 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3280 SC/04 Eximstats: fixed display of large numbers with -nvr option
3282 SC/05 Eximstats: fixed merging of reports with empty tables.
3284 SC/06 Eximstats: added the -include_original_destination flag
3286 SC/07 Eximstats: removed tabs and trailing whitespace.
3288 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3290 TK/06 MBOX spool code: Add real "From " MBOX separator line
3291 so the .eml file is really in mbox format (even though
3292 most programs do not really care). Patch from Alex Miller.
3294 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3295 The latter is generated from $received_to and is only set if the
3296 message has one envelope recipient. SA can use these headers,
3297 obviously out-of-the-box. Patch from Alex Miller.
3299 PH/08 The ${def test on a variable was returning false if the variable's
3300 value was "0", contrary to what the specification has always said!
3301 The result should be true unless the variable is empty.
3303 PH/09 The syntax error of a character other than { following "${if
3304 def:variable_name" (after optional whitespace) was not being diagnosed.
3305 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3306 accidental colon was present, for example, could give incorrect results.
3308 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3309 result is used (not including appendfile, where other changes are about
3312 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3313 This involved changing a lot of size variables from int to off_t. It
3314 should work with maildirs and everything.
3316 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3317 spamd dying while we are connected to it.
3319 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3322 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3323 function for BDB 4.3.
3325 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3327 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3328 the sources for PCRE have been re-arranged and more files are now
3331 PH/15 The code I had for printing potentially long long variables in PH/11
3332 above was not the best (it lost precision). The length of off_t variables
3333 is now inspected at build time, and an appropriate printing format (%ld
3334 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3335 to be "long long int" or "long int". This is needed for the internal
3336 formatting function string_vformat().
3338 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3339 the configuration file to be ":syslog", then the script "guesses" where
3340 the logs files are, rather than using the compiled in default. In our
3341 case the guess is not the same as the compiled default, so the script
3342 suddenly stopped working when I started to use syslog. The patch checks
3343 to see if log_file_path is "". If so, it attempts to read it from exim
3344 with no configuration file to get the compiled in version, before it
3345 falls back to the previous guessing code."
3347 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3348 implementing BATV in an Exim configuration. See NewStuff for the gory
3351 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3352 Makefile that are specific to HP-UX.
3354 PH/18 If the "use_postmaster" option was set for a recipient callout together
3355 with the "random" option, the postmaster address was used as the MAIL
3356 FROM address for the random test, but not for the subsequent recipient
3357 test. It is now used for both.
3359 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3360 patch removes a few documentation additions to RFC 3028, because the
3361 latest draft now contains them. It adds the new en;ascii-case comparator
3362 and a new error check for 8bit text in MIME parts. Comparator and
3363 require names are now matched exactly. I enabled the subaddress
3364 extension, but it is not well tested yet (read: it works for me)."
3366 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3367 rework some of the code of TK/09 above to avoid the hardwired use of
3368 "%lld" and "long long". Replaced the call to snprintf() with a call to
3371 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3372 records point to non-existent hosts", "retry timeout exceeded", and
3373 "retry time not reached for any host after a long failure period".
3375 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3376 experimental DomainKeys support:
3378 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3379 (2) On an error such as an illegally used "control", the wrong name for
3380 the control was given.
3382 These problems did NOT occur unless DomainKeys support was compiled.
3384 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3386 PH/24 Added ${if match_ip condition.
3388 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3389 databases so that it will be absolutely obvious if a crash occurs in the
3390 DB library. This is a regular occurrence (often caused by mis-matched
3393 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3394 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3395 picked up on a user's system that detects such things. There doesn't seem
3396 to be a gcc warning option for this - only an attribute that has to be
3397 put on the function's prototype. It seems that in Fedora Core 4 they have
3398 set this on a number of new functions. No doubt there will be more in due
3401 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3402 specify a transport. However, if an address that was verified by such a
3403 router was the subject of a callout, Exim crashed because it tried to
3404 read the rcpt_include_affixes from the non-existent transport. Now it
3405 just assumes that the setting of that option is false. This bug was
3406 introduced by 4.51/PH/31.
3408 PH/28 Changed -d+all to exclude +memory, because that information is very
3409 rarely of interest, but it makes the output a lot bigger. People tend to
3410 do -d+all out of habit.
3412 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3413 code in os-type was giving problems when libc.so lives in lib64, like on
3416 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3417 aren't the modern standard, and it seems that some systems' include files
3418 don't always have them. Exim was already checking for some of the newer
3419 ones like T_AAAA, and defining it itself. I've added checks for all the
3420 record types that Exim uses.
3422 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3423 not automatically generating a new one, as it is supposed to. This
3424 prevented TLS from working. If the file did exist, but contained invalid
3425 data, a new version was generated, as expected. It was only the case of a
3426 non-existent file that was broken.
3428 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3429 with a change in libdomainkeys > 0.64.
3431 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3432 from DNS. If the selector record carries the flag, it now has
3433 precedence over the domain-wide flag.
3435 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3437 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3438 the use of a port name that isn't defined in /etc/services) occurred, the
3439 message was deferred as in a normal delivery, and thus remained on the
3440 spool, instead of being failed because of the mua_wrapper setting. This
3441 is now fixed, and I tidied up some of the mua_wrapper messages at the
3444 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3445 the messages in a hash of arrays rather than using individual hashes.
3446 This is a bit cleaner and results in dramatic memory savings, albeit
3447 at a slight CPU cost.
3449 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3450 as requested by Marc Sherman.
3452 SC/10 Eximstats: added histograms for user specified patterns as requested
3455 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3457 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3458 fopen() in the content-scanning modules that did not already have it.
3464 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3465 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3467 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3469 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3471 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3472 to test Sieve filters that use "vacation".
3474 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3475 that changes the way the GnuTLS parameters are stored in the cache file.
3476 The new format can be generated externally. For backward compatibility,
3477 if the data in the cache doesn't make sense, Exim assumes it has read an
3478 old-format file, and it generates new data and writes a new file. This
3479 means that you can't go back to an older release without removing the
3482 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3483 work if there are any delivery delays because "one_time" forces the
3484 parent to be marked "delivered", so its unseen clone is never tried
3485 again. For this reason, Exim now forbids the simultaneous setting of
3488 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3489 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3490 entry if you want to know the details. The fix, however, neglected to
3491 consider the case when local delivery batching is involved. The test for
3492 "previously delivered" was not happening when checking to see if an
3493 address could be batched with a previous (undelivered) one; under
3494 certain circumstances this could lead to multiple deliveries to the same
3497 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3498 in its include files, and this causes problems building Exim.
3500 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3501 header_syntax) but Exim was just ignoring anything given after a slash.
3502 In particular, this caused confusion with an attempt to use "verify =
3503 reverse_host_lookup/defer_ok". An error is now given when options are
3504 supplied for verify items that do not have them. (Maybe reverse_host_
3505 lookup should have a defer_ok option, but that's a different point.)
3507 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3508 defined by RFC 821) to 2048, because there were problems with some AUTH
3509 commands, and RFC 1869 says the size should be increased for extended
3510 SMTP commands that take arguments.
3512 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3515 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3516 "unknown" error; now it says that the functionality isn't in the binary.
3518 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3519 an address' error message when a string expansion fails (syntax or
3520 whatever). Otherwise the password may appear in the log. Following change
3521 PH/42 below, there is no longer a chance of it appearing in a bounce
3524 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3526 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3527 of its addresses was ever tried. (Bugzilla bug #2.)
3529 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3530 the result incorrectly in the debug output. (It correctly added a newline
3531 to what was transported.)
3533 TF/01 Added $received_time.
3535 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3536 commented out examples of how to interface to a virus scanner and to
3537 SpamAssassin. Also added commented examples of av_scanner and
3538 spamd_address settings.
3540 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3541 and controls are allowed in which ACLs. There were a couple of minor
3542 errors. Some of the entries in the conditions table (which is a table of
3543 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3544 negation of where the condition IS allowed.
3546 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3548 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3549 header file does not have a version number, so I've had to invent a new
3550 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3551 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3552 radiusclient), but it was contributed by a Radius user.
3554 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3555 files or directories, for OpenSSL.
3557 PH/19 When an Exim process that is running as root has to create an Exim log
3558 file, it does so in a subprocess that runs as exim:exim so as to get the
3559 ownership right at creation (otherwise, other Exim processes might see
3560 the file with the wrong ownership). There was no test for failure of this
3561 fork() call, which would lead to the process getting stuck as it waited
3562 for a non-existent subprocess. Forks do occasionally fail when resources
3563 run out. I reviewed all the other calls to fork(); they all seem to check
3566 PH/20 When checking for unexpected SMTP input at connect time (before writing
3567 the banner), Exim was not dealing correctly with a non-positive return
3568 from the read() function. If the client had disconnected by this time,
3569 the result was a log entry for a synchronization error with an empty
3570 string after "input=" when read() returned zero. If read() returned -1
3571 (an event I could not check), uninitialized data bytes were printed.
3572 There were reports of junk text (parts of files, etc) appearing after
3575 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3577 PH/22 Added support for macro redefinition, and (re)definition in between
3578 driver and ACL definitions.
3580 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3581 forgetting to use the resulting value; it was using the unexpanded value.
3583 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3584 hadn't been configured. The fix is from Juergen Kreileder, who
3585 understands it better than I do:
3587 "Here's what I see happening with three configured cyrus_sasl
3588 authenticators configured (plain, login, cram-md5):
3590 On startup auth_cyrus_sasl_init() gets called for each of these.
3591 This means three calls to sasl_listmech() without a specified mech_list.
3592 => SASL tests which mechs of all available mechs actually work
3593 => three warnings about OTP not working
3594 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3596 With the patch, sasl_listmech() also gets called three times. But now
3597 SASL's mech_list option is set to the server_mech specified in the the
3598 authenticator. Or in other words, the answer from sasl_listmech()
3599 gets limited to just the mech you're testing for (which is different
3601 => the return list contains just 'plain' or 'login', 'cram-md5' or
3602 nothing depending on the value of ob->server_mech.
3604 I've just tested the patch: Authentication still works fine,
3605 unavailable mechs specified in the exim configuration are still
3606 caught, and the auth.log warnings about OTP are gone."
3608 PH/25 When debugging is enabled, the contents of the command line are added
3609 to the debugging output, even when log_selector=+arguments is not
3612 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3613 answer is "GNU", and only if the return is "GNU/something" is the answer
3616 PH/27 $acl_verify_message is now set immediately after the failure of a
3617 verification in an ACL, and so is available in subsequent modifiers. In
3618 particular, the message can be preserved by coding like this:
3620 warn !verify = sender
3621 set acl_m0 = $acl_verify_message
3623 Previously, $acl_verify_message was set only while expanding "message"
3624 and "log_message" when a very denied access.
3626 PH/28 Modified OS/os.c-Linux with
3628 -#ifndef OS_LOAD_AVERAGE
3629 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3631 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3632 nomenclature these days.)
3634 PH/29 Installed patch from the Sieve maintainer that adds the options
3635 sieve_useraddress and sieve_subaddress to the redirect router.
3637 PH/30 In these circumstances:
3638 . Two addresses routed to the same list of hosts;
3639 . First host does not offer TLS;
3640 . First host accepts first address;
3641 . First host gives temporary error to second address;
3642 . Second host offers TLS and a TLS session is established;
3643 . Second host accepts second address.
3644 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3645 and peerdn, if requested) that were in fact used only for the second
3648 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3649 attention to any local part prefix or suffix that was matched by the
3650 router that accepted the address. It now behaves in the same way as it
3651 does for delivery: the affixes are removed from the local part unless
3652 rcpt_include_affixes is set on the transport.
3654 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3655 timeout during the DATA phase of an incoming message.
3657 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3658 applied a patch sanctioned by the Sieve maintainer.
3660 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3661 the uid or gid is negative. A case of a negative gid caused this to be
3662 noticed. The fix allows for either to be negative.
3664 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3665 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3668 PH/36 The change PH/12 above was broken. Fixed it.
3670 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3671 the grounds that routing the same address twice would always produce the
3672 same answer. This might have been true once, but it is certainly no
3673 longer true now. Routing a child address may depend on the previous
3674 routing that produced that child. Some complicated redirection strategies
3675 went wrong when messages had multiple recipients, and made Exim's
3676 behaviour dependent on the order in which the addresses were given.
3678 I have moved the duplicate checking until after the routing is complete.
3679 Exim scans the addresses that are assigned to local and remote
3680 transports, and removes any duplicates. This means that more work will be
3681 done, as duplicates will always all be routed, but duplicates are
3682 presumably rare, so I don't expect this is of any significance.
3684 For deliveries to pipes, files, and autoreplies, the duplicate checking
3685 still happens during the routing process, since they are not going to be
3688 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3689 It corrects a timeout issue with spamd. This is Ian's comment: "The
3690 background is that sometimes spamd either never reads data from a
3691 connection it has accepted, or it never writes response data. The exiscan
3692 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3693 blindly assumes that writes won't block so it may never time out."
3695 PH/39 Allow G after quota size as well as K and M.
3697 PH/40 The value set for $authenticated_id in an authenticator may not contain
3698 binary zeroes or newlines because the value is written to log lines and
3699 to spool files. There was no check on this. Now the value is run through
3700 the string_printing() function so that such characters are converted to
3701 printable escape sequences.
3703 PH/41 $message_linecount is a new variable that contains the total number of
3704 lines in the message. Compare $body_linecount, which is the count for the
3707 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3708 bounce and delay warning messages, except in certain special cases, which
3711 (a) An SMTP error message from a remote host;
3712 (b) A message specified in a :fail: redirection;
3713 (c) A message specified in a "fail" command in a system filter;
3714 (d) A message specified in a FAIL return from the queryprogram router;
3715 (e) A message specified by the cannot_route_message router option.
3717 In these cases only, Exim does include the error details in bounce and
3718 warning messages. There are also a few cases where bland messages such
3719 as "unrouteable address" or "local delivery error" are given.
3721 PH/43 $value is now also set for the "else" part of a ${run expansion.
3723 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3724 being worked on, but at least Exim now implements the latest version to
3727 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3728 process to complete was treated as a delivery failure, a timeout while
3729 writing the message to the pipe was logged, but erroneously treated as a
3730 successful delivery. Such timeouts include transport filter timeouts. For
3731 consistency with the overall process timeout, these timeouts are now
3732 treated as errors, giving rise to delivery failures by default. However,
3733 there is now a new Boolean option for the pipe transport called
3734 timeout_defer, which, if set TRUE, converts the failures into defers for
3735 both kinds of timeout. A transport filter timeout is now identified in
3738 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3739 systems where "make" and "gmake" are different, calling "gmake" at top
3740 level broke things. I've arranged for the value of $(MAKE) to be passed
3741 from the Makefile to this script so that it can call the same version of
3745 A note about Exim versions 4.44 and 4.50
3746 ----------------------------------------
3748 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3749 changes of various kinds. As a consequence, a big documentation update was
3750 needed. This delayed the release for rather longer than seemed good, especially
3751 in the light of a couple of (minor) security issues. Therefore, the changes
3752 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3753 release. So 4.44 and 4.50 are in effect two different branches that both start
3756 I have left the 4.50 change log unchanged; it contains all the changes since
3757 4.43. The change log for 4.44 is below; many of its items are identical to
3758 those for 4.50. This seems to be the most sensible way to preserve the
3759 historical information.
3765 1. Minor wording change to the doc/README.SIEVE file.
3767 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3768 computation of the current number of files was incorrect.
3770 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3771 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3774 4. Give more explanation in the error message when the command for a transport
3775 filter fails to execute.
3777 5. There are several places where Exim runs a non-Exim command in a
3778 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3779 was being done only for the command run by the queryprogram router. It is
3780 now done for all such subprocesses. The other cases are: ${run, transport
3781 filters, and the commands run by the lmtp and pipe transports.
3783 6. Added CONFIGURE_GROUP build-time option.
3785 7. Some older OS have a limit of 256 on the maximum number of file
3786 descriptors. Exim was using setrlimit() to set 1000 as a large value
3787 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3788 systems. I've change it so that if it can't get 1000, it tries for 256.
3790 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3791 was an oversight, and furthermore, ever since the addition of extra
3792 controls (e.g. 4.43/32), the checks on when to allow different forms of
3793 "control" were broken. There should now be diagnostics for all cases when a
3794 control that does not make sense is encountered.
3796 9. Added the /retain_sender option to "control=submission".
3798 10. $recipients is now available in the predata ACL (oversight).
3800 11. Tidy the search cache before the fork to do a delivery from a message
3801 received from the command line. Otherwise the child will trigger a lookup
3802 failure and thereby defer the delivery if it tries to use (for example) a
3803 cached ldap connection that the parent has called unbind on.
3805 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3806 of $address_data from the recipient verification was clobbered by the
3807 sender verification.
3809 13. The value of address_data from a sender verification is now available in
3810 $sender_address_data in subsequent conditions in the ACL statement.
3812 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3814 15. Added a new option "connect=<time>" to callout options, to set a different
3817 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3818 was its contents. (It was OK if the option was not defined at all.)
3820 17. A "Completed" log line is now written for messages that are removed from
3821 the spool by the -Mrm option.
3823 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3824 information about exactly what failed.
3826 19. Added -dd to debug only the daemon process.
3828 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3829 handles timeouts, both on the server side and network timeouts. Renamed the
3830 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3832 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3833 It is now set to "smtps".
3835 22. $host_address is now set to the target address during the checking of
3836 ignore_target_hosts.
3838 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3839 being passed; this would have caused $sender_host_name to have been used if
3840 matching the list had actually called for a host name (not very likely,
3841 since this list is usually IP addresses). A host name is now passed as
3844 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3845 code that specifies a non-restarting handler (typically sigaction() in
3846 modern systems) in an attempt to fix a rare and obscure crash bug.
3848 25. Narrowed the window for a race in the daemon that could cause it to ignore
3849 SIGCHLD signals. This is not a major problem, because they are used only to
3850 wake it up if nothing else does.
3852 26. A malformed maildirsize file could cause Exim to calculate negative values
3853 for the mailbox size or file count. Odd effects could occur as a result.
3854 The maildirsize information is now recalculated if the size or filecount
3857 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3858 support for a long time. Removed HAVE_SYS_VFS_H.
3860 28. Installed the latest version of exipick from John Jetmore.
3862 29. In an address list, if the pattern was not a regular expression, an empty
3863 subject address (from a bounce message) matched only if the pattern was an
3864 empty string. Non-empty patterns were not even tested. This was the wrong
3865 because it is perfectly reasonable to use an empty address as part of a
3866 database query. An empty address is now tested by patterns that are
3867 lookups. However, all the other forms of pattern expect the subject to
3868 contain a local part and a domain, and therefore, for them, an empty
3869 address still always fails if the pattern is not itself empty.
3871 30. Exim went into a mad DNS loop when attempting to do a callout where the
3872 host was specified on an smtp transport, and looking it up yielded more
3873 than one IP address.
3875 31. Re-factored the code for checking spool and log partition space into a
3876 function that finds that data and another that does the check. The former
3877 is then used to implement four new variables: $spool_space, $log_space,
3878 $spool_inodes, and $log_inodes.
3880 32. The RFC2047 encoding function was originally intended for short strings
3881 such as real names; it was not keeping to the 75-character limit for
3882 encoded words that the RFC imposes. It now respects the limit, and
3883 generates multiple encoded words if necessary. To be on the safe side, I
3884 have increased the buffer size for the ${rfc2047: expansion operator from
3887 33. It is now permitted to omit both strings after an "if" condition; if the
3888 condition is true, the result is "true". As before, when the second string
3889 is omitted, a false condition yields an empty string. This makes it less
3890 cumbersome to write custom ACL and router conditions.
3892 34. Failure to deliver a bounce message always caused it to be frozen, even if
3893 there was an errors_to setting on the router. The errors_to setting is now
3896 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3897 canonical form (fully expanded) before being placed in
3898 $sender_host_address.
3900 36. The table in the code that translates DNS record types into text (T_A to
3901 "A" for instance) was missing entries for NS and CNAME. It is just possible
3902 that this could have caused confusion if both these types were looked up
3903 for the same domain, because the text type is used as part of Exim's
3904 per-process caching. But the chance of anyone hitting this buglet seems
3907 37. The dnsdb lookup has been extended in a number of ways.
3909 (1) There is a new type, "zns", which walks up the domain tree until it
3910 finds some nameserver records. It should be used with care.
3912 (2) There is a new type, "mxh", which is like "mx" except that it returns
3913 just the host names, not the priorities.
3915 (3) It is now possible to give a list of domains (or IP addresses) to be
3916 looked up. The behaviour when one of the lookups defers can be
3917 controlled by a keyword.
3919 (4) It is now possible to specify the separator character for use when
3920 multiple records are returned.
3922 38. The dnslists ACL condition has been extended: it is now possible to supply
3923 a list of IP addresses and/or domains to be looked up in a particular DNS
3926 39. Added log_selector=+queue_time_overall.
3928 40. When running the queue in the test harness, wait just a tad after forking a
3929 delivery process, to get repeatability of debugging output.
3931 41. Include certificate and key file names in error message when GnuTLS fails
3932 to set them up, because the GnuTLS error message doesn't include the name
3933 of the failing file when there is a problem reading it.
3935 42. Allow both -bf and -bF in the same test run.
3937 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3939 44. The "Exiscan patch" is now merged into the mainline Exim source.
3941 45. Sometimes the final signoff response after QUIT could fail to get
3942 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3943 < 0 before doing a fflush(). This bug looks as though it goes back to the
3944 introduction of TLS in release 3.20, but "sometimes" must have been rare
3945 because the tests only now provoked it.
3947 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3948 (this can affect the format of dates).
3950 47. exim_tidydb, when checking for the continued existence of a message for
3951 which it has found a message-specific retry record, was not finding
3952 messages that were in split spool directories. Consequently, it was
3953 deleting retry records that should have stayed in existence.
3955 48. Steve fixed some bugs in eximstats.
3957 49. The SPA authentication driver was not abandoning authentication and moving
3958 on to the next authenticator when an expansion was forced to fail,
3959 contradicting the general specification for all authenticators. Instead it
3960 was generating a temporary error. It now behaves as specified.
3962 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3963 (the order specifies the preference for clients). The order is now AES256,
3964 AES128, 3DES, ARCFOUR128.
3966 51. Small patch to Sieve code - explicitly set From: when generating an
3969 52. Exim crashed if a remote delivery caused a very long error message to be
3970 recorded - for instance if somebody sent an entire SpamAssassin report back
3971 as a large number of 550 error lines. This bug was coincidentally fixed by
3972 increasing the size of one of Exim's internal buffers (big_buffer) that
3973 happened as part of the Exiscan merge. However, to be on the safe side, I
3974 have made the code more robust (and fixed the comments that describe what
3977 53. Now that there can be additional text after "Completed" in log lines (if
3978 the queue_time_overall log selector is set), a one-byte patch to exigrep
3979 was needed to allow it to recognize "Completed" as not the last thing in
3982 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3983 patch that reportedly fixes this has been added. I am not expert enough to
3984 create a test for it. This is what the patch creator wrote:
3986 "I found a little strange behaviour of ldap code when working with
3987 Windows 2003 AD Domain, where users was placed in more than one
3988 Organization Units. When I tried to give exim partial DN, the exit code
3989 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3990 But simultaneously result of request was absolutely normal ldap result,
3991 so I produce this patch..."
3993 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3994 REFERENCE, so I have modified the code to exclude the patch when that macro
3997 55. Some experimental protocols are using DNS PTR records for new purposes. The
3998 keys for these records are domain names, not reversed IP addresses. The
3999 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4000 leaves it alone. Component reversal etc. now happens only for IP addresses.
4003 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4005 57. Double the size of the debug message buffer (to 2048) so that more of very
4006 long debug lines gets shown.
4008 58. The exicyclog utility now does better if the number of log files to keep
4009 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4011 59. Two changes related to the smtp_active_hostname option:
4013 (1) $smtp_active_hostname is now available as a variable.
4014 (2) The default for smtp_banner uses $smtp_active_hostname instead
4015 of $primary_hostname.
4017 60. The host_aton() function is supposed to be passed a string that is known
4018 to be a valid IP address. However, in the case of IPv6 addresses, it was
4019 not checking this. This is a hostage to fortune. Exim now panics and dies
4020 if the condition is not met. A case was found where this could be provoked
4021 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4022 components; fortuitously, this particular loophole had already been fixed
4023 by change 4.50/55 above.
4025 If there are any other similar loopholes, the new check in host_aton()
4026 itself should stop them being exploited. The report I received stated that
4027 data on the command line could provoke the exploit when Exim was running as
4028 exim, but did not say which command line option was involved. All I could
4029 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4030 running as the user.
4033 61. There was a buffer overflow vulnerability in the SPA authentication code
4034 (which came originally from the Samba project). I have added a test to the
4035 spa_base64_to_bits() function which I hope fixes it.
4038 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4039 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4041 63. The daemon start-up calls getloadavg() while still root for those OS that
4042 need the first call to be done as root, but it missed one case: when
4043 deliver_queue_load_max is set with deliver_drop_privilege. This is
4044 necessary for the benefit of the queue runner, because there is no re-exec
4045 when deliver_drop_privilege is set.
4047 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4048 This has been fixed.
4050 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4051 was in use, was not putting the data itself into the right store pool;
4052 consequently, it could be overwritten for a subsequent message in the same
4053 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4056 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4058 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4059 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4060 the function were treating the return as a boolean value, which happened to
4061 work because 0=false and not-0=true, but is not correct code.
4063 68. The host_aton() function was not handling scoped IPv6 addresses (those
4064 with, for example, "%eth0" on the end) correctly.
4066 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4067 negated items (that is, ~something) in unsigned ints. Some compilers
4068 apparently mutter when there is no cast.
4070 70. If an address verification called from an ACL failed, and did not produce a
4071 user-specific message (i.e. there was only a "system" message), nothing was
4072 put in $acl_verify_message. In this situation, it now puts the system
4075 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4076 session; change 4.31/43 added the unwanted input to the log line - except
4077 that it did not do this in the start of session case. It now does.
4079 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4080 This is wrong and can cause the other end to generate a synchronization
4081 error if it is another Exim or anything else that does the synchronization
4082 check. A QUIT command is no longer sent after a timeout.
4084 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4085 during host lookups.
4087 74. The defer_ok option of callout verification was not working if it was used
4088 when verifying addresses in header lines, that is, for this case:
4090 verify = header_sender/callout=defer_ok
4092 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4093 those file descriptors could be used for SMTP connections. If anything
4094 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4095 could be sent to the SMTP client, causing chaos. The daemon now opens
4096 stdin, stdout, and stderr to /dev/null when it puts itself into the
4099 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4100 log by default. The user can override this with suitable Perl magic.
4102 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4103 the log message when discard triggers, was not working for the DATA ACL or
4104 for the non-SMTP ACL.
4106 78. Error message wording change in sieve.c.
4108 79. If smtp_accept_max_per_host was set, the number of connections could be
4109 restricted to fewer than expected, because the daemon was trying to set up
4110 a new connection before checking whether the processes handling previous
4111 connections had finished. The check for completed processes is now done
4112 earlier. On busy systems, this bug wouldn't be noticed because something
4113 else would have woken the daemon, and it would have reaped the completed
4116 80. If a message was submitted locally by a user whose login name contained one
4117 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4118 It caused a spool format error. I have fixed the spool reading code. A
4119 related problem was that the "from" clause in the Received: line became
4120 illegal because of the space(s). It is now covered by ${quote_local_part.
4122 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4125 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4127 83. Patch from Sieve maintainer to fix unterminated string problem in
4128 "vacation" handling.
4130 84. Some minor changes to the Linux configuration files to help with other
4131 OS variants using glibc.
4133 85. One more patch for Sieve to update vacation handling to latest spec.
4136 ----------------------------------------------------
4137 See the note above about the 4.44 and 4.50 releases.
4138 ----------------------------------------------------
4144 1. Change 4.43/35 introduced a bug that caused file counts to be
4145 incorrectly computed when quota_filecount was set in an appendfile
4148 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4149 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4152 3. Give more explanation in the error message when the command for a transport
4153 filter fails to execute.
4155 4. There are several places where Exim runs a non-Exim command in a
4156 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4157 was being done only for the command run by the queryprogram router. It is
4158 now done for all such subprocesses. The other cases are: ${run, transport
4159 filters, and the commands run by the lmtp and pipe transports.
4161 5. Some older OS have a limit of 256 on the maximum number of file
4162 descriptors. Exim was using setrlimit() to set 1000 as a large value
4163 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4164 systems. I've change it so that if it can't get 1000, it tries for 256.
4166 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4167 was an oversight, and furthermore, ever since the addition of extra
4168 controls (e.g. 4.43/32), the checks on when to allow different forms of
4169 "control" were broken. There should now be diagnostics for all cases when a
4170 control that does not make sense is encountered.
4172 7. $recipients is now available in the predata ACL (oversight).
4174 8. Tidy the search cache before the fork to do a delivery from a message
4175 received from the command line. Otherwise the child will trigger a lookup
4176 failure and thereby defer the delivery if it tries to use (for example) a
4177 cached ldap connection that the parent has called unbind on.
4179 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4180 of $address_data from the recipient verification was clobbered by the
4181 sender verification.
4183 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4184 was its contents. (It was OK if the option was not defined at all.)
4186 11. A "Completed" log line is now written for messages that are removed from
4187 the spool by the -Mrm option.
4189 12. $host_address is now set to the target address during the checking of
4190 ignore_target_hosts.
4192 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4193 being passed; this would have caused $sender_host_name to have been used if
4194 matching the list had actually called for a host name (not very likely,
4195 since this list is usually IP addresses). A host name is now passed as
4198 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4199 code that specifies a non-restarting handler (typically sigaction() in
4200 modern systems) in an attempt to fix a rare and obscure crash bug.
4202 15. Narrowed the window for a race in the daemon that could cause it to ignore
4203 SIGCHLD signals. This is not a major problem, because they are used only to
4204 wake it up if nothing else does.
4206 16. A malformed maildirsize file could cause Exim to calculate negative values
4207 for the mailbox size or file count. Odd effects could occur as a result.
4208 The maildirsize information is now recalculated if the size or filecount
4211 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4212 support for a long time. Removed HAVE_SYS_VFS_H.
4214 18. Updated exipick to current release from John Jetmore.
4216 19. Allow an empty sender to be matched against a lookup in an address list.
4217 Previously the only cases considered were a regular expression, or an
4220 20. Exim went into a mad DNS lookup loop when doing a callout where the
4221 host was specified on the transport, if the DNS lookup yielded more than
4224 21. The RFC2047 encoding function was originally intended for short strings
4225 such as real names; it was not keeping to the 75-character limit for
4226 encoded words that the RFC imposes. It now respects the limit, and
4227 generates multiple encoded words if necessary. To be on the safe side, I
4228 have increased the buffer size for the ${rfc2047: expansion operator from
4231 22. Failure to deliver a bounce message always caused it to be frozen, even if
4232 there was an errors_to setting on the router. The errors_to setting is now
4235 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4236 canonical form (fully expanded) before being placed in
4237 $sender_host_address.
4239 24. Updated eximstats to version 1.33
4241 25. Include certificate and key file names in error message when GnuTLS fails
4242 to set them up, because the GnuTLS error message doesn't include the name
4243 of the failing file when there is a problem reading it.
4245 26. Expand error message when OpenSSL has problems setting up cert/key files.
4248 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4249 (this can affect the format of dates).
4251 28. exim_tidydb, when checking for the continued existence of a message for
4252 which it has found a message-specific retry record, was not finding
4253 messages that were in split spool directories. Consequently, it was
4254 deleting retry records that should have stayed in existence.
4256 29. eximstats updated to version 1.35
4257 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4258 1.35 - bugfix such that pie charts by volume are generated correctly
4260 30. The SPA authentication driver was not abandoning authentication and moving
4261 on to the next authenticator when an expansion was forced to fail,
4262 contradicting the general specification for all authenticators. Instead it
4263 was generating a temporary error. It now behaves as specified.
4265 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4266 (the order specifies the preference for clients). The order is now AES256,
4267 AES128, 3DES, ARCFOUR128.
4269 31. Small patch to Sieve code - explicitly set From: when generating an
4272 32. Exim crashed if a remote delivery caused a very long error message to be
4273 recorded - for instance if somebody sent an entire SpamAssassin report back
4274 as a large number of 550 error lines. This bug was coincidentally fixed by
4275 increasing the size of one of Exim's internal buffers (big_buffer) that
4276 happened as part of the Exiscan merge. However, to be on the safe side, I
4277 have made the code more robust (and fixed the comments that describe what
4280 33. Some experimental protocols are using DNS PTR records for new purposes. The
4281 keys for these records are domain names, not reversed IP addresses. The
4282 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4283 leaves it alone. Component reversal etc. now happens only for IP addresses.
4286 34. The host_aton() function is supposed to be passed a string that is known
4287 to be a valid IP address. However, in the case of IPv6 addresses, it was
4288 not checking this. This is a hostage to fortune. Exim now panics and dies
4289 if the condition is not met. A case was found where this could be provoked
4290 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4291 components; fortuitously, this particular loophole had already been fixed
4292 by change 4.50/55 or 4.44/33 above.
4294 If there are any other similar loopholes, the new check in host_aton()
4295 itself should stop them being exploited. The report I received stated that
4296 data on the command line could provoke the exploit when Exim was running as
4297 exim, but did not say which command line option was involved. All I could
4298 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4299 running as the user.
4302 35. There was a buffer overflow vulnerability in the SPA authentication code
4303 (which came originally from the Samba project). I have added a test to the
4304 spa_base64_to_bits() function which I hope fixes it.
4307 36. The daemon start-up calls getloadavg() while still root for those OS that
4308 need the first call to be done as root, but it missed one case: when
4309 deliver_queue_load_max is set with deliver_drop_privilege. This is
4310 necessary for the benefit of the queue runner, because there is no re-exec
4311 when deliver_drop_privilege is set.
4313 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4314 was in use, was not putting the data itself into the right store pool;
4315 consequently, it could be overwritten for a subsequent message in the same
4316 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4319 38. Sometimes the final signoff response after QUIT could fail to get
4320 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4321 < 0 before doing a fflush(). This bug looks as though it goes back to the
4322 introduction of TLS in release 3.20, but "sometimes" must have been rare
4323 because the tests only now provoked it.
4329 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4330 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4331 Now it returns an integer. A number of calls were still expecting a T/F
4332 return. Fortuitously, in all cases, the tests worked in OK situations,
4333 which is the norm. However, things would have gone wrong on any write
4334 failures on the smtp file descriptor. This function is used when sending
4335 messages over SMTP and also when doing verify callouts.
4337 2. When Exim is called to do synchronous delivery of a locally submitted
4338 message (the -odf or -odi options), it no longer closes stderr before doing
4341 3. Implemented the mua_wrapper option.
4343 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4345 5. Implemented the functions header_remove(), header_testname(),
4346 header_add_at_position(), and receive_remove_recipient(), and exported them
4349 6. If an ACL "warn" statement specified the addition of headers, Exim already
4350 inserted X-ACL-Warn: at the start if there was no header name. However, it
4351 was not making this test for the second and subsequent header lines if
4352 there were newlines in the string. This meant that an invalid header could
4353 be inserted if Exim was badly configured.
4355 7. Allow an ACL "warn" statement to add header lines at the start or after all
4356 the Received: headers, as well as at the end.
4358 8. Added the rcpt_4xx retry error code.
4360 9. Added postmaster_mailfrom=xxx to callout verification option.
4362 10. Added mailfrom=xxxx to the callout verification option, for verify=
4365 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4366 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4368 12. Inserted some casts to stop certain compilers warning when using pointer
4369 differences as field lengths or precisions in printf-type calls (mostly
4370 affecting debugging statements).
4372 13. Added optional readline() support for -be (dynamically loaded).
4374 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4375 same clock tick as a message's arrival, so that its received time was the
4376 same as the "first fail" time on the retry record, and that message
4377 remained on the queue past the ultimate address timeout, every queue runner
4378 would try a delivery (because it was past the ultimate address timeout) but
4379 after another failure, the ultimate address timeout, which should have then
4380 bounced the address, did not kick in. This was a "< instead of <=" error;
4381 in most cases the first failure would have been in the next clock tick
4382 after the received time, and all would be well.
4384 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4385 being recognized when the domain list was tested by the match_domain
4386 condition in an expansion string.
4388 16. Added the ${str2b64: operator.
4390 17. Exim was always calling setrlimit() to set a large limit for the number of
4391 processes, without checking whether the existing limit was already
4392 adequate. (It did check for the limit on file descriptors.) Furthermore,
4393 errors from getrlimit() and setrlimit() were being ignored. Now they are
4394 logged to the main and panic logs, but Exim does carry on, to try to do its
4395 job under whatever limits there are.
4397 18. Imported PCRE 5.0.
4399 19. Trivial typo in log message " temporarily refused connection" (the leading
4402 20. If the log selector return_path_on_delivery was set and an address was
4403 redirected to /dev/null, the delivery process crashed because it assumed
4404 that a return path would always be set for a "successful" delivery. In this
4405 case, the whole delivery is bypassed as an optimization, and therefore no
4408 21. Internal re-arrangement: the function for sending a challenge and reading
4409 a response while authentication was assuming a zero-terminated challenge
4410 string. It's now changed to take a pointer and a length, to allow for
4411 binary data in such strings.
4413 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4415 23. Exim was not respecting finduser_retries when seeking the login of the
4416 uid under which it was called; it was always trying 10 times. (The default
4417 setting of finduser_retries is zero.) Also, it was sleeping after the final
4418 failure, which is pointless.
4420 24. Implemented tls_on_connect_ports.
4422 25. Implemented acl_smtp_predata.
4424 26. If the domain in control=submission is set empty, Exim assumes that the
4425 authenticated id is a complete email address when it generates From: or
4426 Sender: header lines.
4428 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4429 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4430 chown and chgrp in /bin and hostname in /usr/bin.
4432 28. Exim was keeping the "process log" file open after each use, just as it
4433 does for the main log. This opens the possibility of it remaining open for
4434 long periods when the USR1 signal hits a daemon. Occasional processlog
4435 errors were reported, that could have been caused by this. Anyway, it seems
4436 much more sensible not to leave this file open at all, so that is what now
4439 29. The long-running daemon process does not normally write to the log once it
4440 has entered its main loop, and it closes the log before doing so. This is
4441 so that log files can straightforwardly be renamed and moved. However,
4442 there are a couple of unusual error situations where the daemon does write
4443 log entries, and I had neglected to close the log afterwards.
4445 30. The text of an SMTP error response that was received during a remote
4446 delivery was being truncated at 512 bytes. This is too short for some of
4447 the long messages that one sometimes sees. I've increased the limit to
4450 31. It is now possible to make retry rules that apply only when a message has a
4451 specific sender, in particular, an empty sender.
4453 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4454 it possible to be selective about when SMTP synchronization is enforced.
4456 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4458 32. Implemented hosts_connection_nolog.
4460 33. Added an ACL for QUIT.
4462 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4465 35. Added mailbox_size and mailbox_filecount to appendfile.
4467 36. Added control = no_multiline_responses to ACLs.
4469 37. There was a bug in the logic of the code that waits for the clock to tick
4470 in the case where the clock went backwards by a substantial amount such
4471 that the microsecond fraction of "now" was more than the microsecond
4472 fraction of "then" (but the whole seconds number was less).
4474 38. Added support for the libradius Radius client library this is found on
4475 FreeBSD (previously only the radiusclient library was supported).
4481 1. When certain lookups returned multiple values in the form name=value, the
4482 quoting of the values was not always being done properly. Specifically:
4483 (a) If the value started with a double quote, but contained no whitespace,
4485 (b) If the value contained whitespace other than a space character (i.e.
4486 tabs or newlines or carriage returns) it was not quoted.
4487 This fix has been applied to the mysql and pgsql lookups by writing a
4488 separate quoting function and calling it from the lookup code. The fix
4489 should probably also be applied to nisplus, ibase and oracle lookups, but
4490 since I cannot test any of those, I have not disturbed their existing code.
4492 2. A hit in the callout cache for a specific address caused a log line with no
4493 reason for rejecting RCPT. Now it says "Previous (cached) callout
4494 verification failure".
4496 3. There was an off-by-one bug in the queryprogram router. An over-long
4497 return line was truncated at 256 instead of 255 characters, thereby
4498 overflowing its buffer with the terminating zero. As well as fixing this, I
4499 have increased the buffer size to 1024 (and made a note to document this).
4501 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4502 when Exim is waiting for an SMTP response from a remote server, Exim
4503 restarts its select() call on the socket, thereby resetting its timeout.
4504 This is not a problem when such interrupts are rare. Somebody set up a cron
4505 job to run exiwhat every 2 minutes, which is less than the normal select()
4506 timeout (5 or 10 minutes). This meant that the select() timeout never
4507 kicked in because it was always reset. I have fixed this by comparing the
4508 time when an interrupt arrives with the time at the start of the first call
4509 to select(). If more time than the timeout has elapsed, the interrupt is
4510 treated as a timeout.
4512 5. Some internal re-factoring in preparation for the addition of Sieve
4513 extensions (by MH). In particular, the "personal" test is moved to a
4514 separate function, and given an option for scanning Cc: and Bcc: (which is
4515 not set for Exim filters).
4517 6. When Exim created an email address using the login of the caller as the
4518 local part (e.g. when creating a From: or Sender: header line), it was not
4519 quoting the local part when it contained special characters such as @.
4521 7. Installed new OpenBSD configuration files.
4523 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4524 try to make them clearer.
4526 9. Callout options, other than the timeout value, were being ignored when
4527 verifying sender addresses in header lines. For example, when using
4529 verify = header_sender/callout=no_cache
4531 the cache was (incorrectly) being used.
4533 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4534 only the Cygwin environment.
4536 11. When return_path_on_delivery was set as a log selector, if different remote
4537 addresses in the same message used different return paths and parallel
4538 remote delivery occurred, the wrong values would sometimes be logged.
4539 (Whenever a remote delivery process finished, the return path value from
4540 the most recently started remote delivery process was logged.)
4542 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4543 lines when AUTH and/or TLS are in use. This is the "received protocol"
4544 field. Exim used to use "asmtp" for authenticated SMTP, without any
4545 indication (in the protocol name) for TLS use. Now it follows the RFC and
4546 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4547 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4548 names appear in log lines as well as in Received: header lines.
4550 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4551 extensions, and comparison tests, and to fix some bugs.
4553 14. Changes to the "personal" filter test:
4555 (1) The test was buggy in that it was just doing the equivalent of
4556 "contains" tests on header lines. For example, if a user's address was
4557 anne@some.where, the "personal" test would incorrectly be true for
4559 To: susanne@some.where
4561 This test is now done by extracting each address from the header in turn,
4562 and checking the entire address. Other tests that are part of "personal"
4563 are now done using regular expressions (for example, to check local parts
4564 of addresses in From: header lines).
4566 (2) The list of non-personal local parts in From: addresses has been
4567 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4568 taken from the Sieve specification recommendations.
4570 (3) If the message contains any header line starting with "List-" it is
4571 treated as non-personal.
4573 (4) The test for "circular" in the Subject: header line has been removed
4574 because it now seems ill-conceived.
4576 15. Minor typos in src/EDITME comments corrected.
4578 16. Installed latest exipick from John Jetmore.
4580 17. If headers_add on a router specified a text string that was too long for
4581 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4582 of string_sprintf() is now avoided.
4584 18. $message_body_size was not set (it was always zero) when running the DATA
4585 ACL and the local_scan() function.
4587 19. For the "mail" command in an Exim filter, no default was being set for
4588 the once_repeat time, causing a random time value to be used if "once" was
4589 specified. (If the value happened to be <= 0, no repeat happened.) The
4590 default is now 0s, meaning "never repeat". The "vacation" command was OK
4591 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4592 (I found it when inspecting the code).
4594 20. There is now an overall timeout for performing a callout verification. It
4595 defaults to 4 times the callout timeout, which applies to individual SMTP
4596 commands during the callout. The overall timeout applies when there is more
4597 than one host that can be tried. The timeout is checked before trying the
4598 next host. This prevents very long delays if there are a large number of
4599 hosts and all are timing out (e.g. when the network connections are timing
4600 out). The value of the overall timeout can be changed by specifying an
4601 additional sub-option for "callout", called "maxwait". For example:
4603 verify = sender/callout=5s,maxwait=20s
4605 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4606 to the end before writing, but this should make it even safer).
4608 22. Exim was forgetting that it had advertised PIPELINING for the second and
4609 subsequent messages on an SMTP connection. It was also not resetting its
4610 memory on STARTTLS and an internal HELO.
4612 23. When Exim logs an SMTP synchronization error within a session, it now
4613 records whether PIPELINING has been advertised or not.
4615 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4616 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4617 rather than long int.
4619 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4621 26. Added the never_mail option to autoreply.
4627 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4628 crash if the getsockname() call failed; this can happen if a connection is
4629 closed very soon after it is established. The problem was simply in the
4630 order in which certain operations were done, causing Exim to try to write
4631 to the SMTP stream before it had set up the file descriptor. The bug has
4632 been fixed by making things happen in the correct order.
4638 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4639 before the connection was closed, thus losing the rejection response.
4641 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4642 some early Solaris releases, but causes trouble in current releases where
4643 socklen_t is defined.
4645 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4648 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4651 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4652 that releasing the top of it at the end releases what was used for sub-
4653 expansions (unless the block got too big). However, discard this block if
4654 the first thing is a variable or header, so that we can use its block when
4655 it is dynamic (useful for very large $message_headers, for example).
4657 6. Lookups now cache *every* query, not just the most recent. A new, separate
4658 store pool is used for this. It can be recovered when all lookup caches are
4659 flushed. Lookups now release memory at the end of their result strings.
4660 This has involved some general refactoring of the lookup sources.
4662 7. Some code has been added to the store_xxx() functions to reduce the amount
4663 of flapping under certain conditions.
4665 8. log_incoming_interface used to affect only the <= reception log lines. Now
4666 it causes the local interface and port to be added to several more SMTP log
4667 lines, for example "SMTP connection from", and rejection lines.
4669 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4671 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4673 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4674 could be overwritten at the end of the current message (or the start of a
4675 new message if it was set in a HELO ACL). The value is now preserved for
4676 the duration of the SMTP connection.
4678 12. If a transport had a headers_rewrite setting, and a matching header line
4679 contained an unqualified address, that address was qualified, even if it
4680 did not match any rewriting rules. The underlying bug was that the values
4681 of the flags that permit the existence of unqualified sender and recipient
4682 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4683 non-local messages, and by -bnq for local messages) were not being
4684 preserved with the message after it was received.
4686 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4687 "next input=" as part of the text comprising the host identity instead of
4688 the correct text. The code was using the same buffer for two different
4689 strings. However, depending on which order the printing function evaluated
4690 its arguments, the bug did not always show up. Under Linux, for example, my
4691 test suite worked just fine.
4693 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4694 doesn't work with some older versions of Perl. It has been changed to "my",
4695 which in any case is probably the better facility to use.
4697 15. A really picky compiler found some instances of statements for creating
4698 error messages that either had too many or two few arguments for the format
4701 16. The size of the buffer for calls to the DNS resolver has been increased
4702 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4703 for addresses that have a lot of PTR records. This alleviates a problem; it
4704 does not fully solve it.
4706 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4707 buffer now truncates the list and logs the incident, which is the same
4708 action as happens when Exim is looking up a host name and its aliases.
4709 Previously in this situation something unpredictable would happen;
4710 sometimes it was "internal error: store_reset failed".
4712 18. If a server dropped the connection unexpectedly when an Exim client was
4713 using GnuTLS and trying to read a response, the client delivery process
4714 crashed while trying to generate an error log message.
4716 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4717 string, for example:
4719 warn message = H1: something\nH2: something
4721 the text was added as a single header line from Exim's point of view
4722 though it ended up OK in the delivered message. However, searching for the
4723 second and subsequent header lines using $h_h2: did not work. This has been
4724 fixed. Similarly, if a system filter added multiple headers in this way,
4725 the routers could not see them.
4727 20. Expanded the error message when iplsearch is called with an invalid key to
4728 suggest using net-iplsearch in a host list.
4730 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4731 ACLs are no longer actually imposed (and a message to that effect is
4734 22. If a "gecos" field in a passwd entry contained escaped characters, in
4735 particular, if it contained a \" sequence, Exim got it wrong when building
4736 a From: or a Sender: header from that name. A second bug also caused
4737 incorrect handling when an unquoted " was present following a character
4738 that needed quoting.
4740 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4741 was not being matched caselessly.
4743 24. Arranged for all hyphens in the exim.8 source to be escaped with
4746 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4747 properly. Recipient callout cache records were still being keyed to include
4748 the sender, even when use_sender was set false. This led to far more
4749 callouts that were necessary. The sender is no longer included in the key
4750 when use_sender is false.
4752 26. Added "control = submission" modifier to ACLs.
4754 27. Added the ${base62d: operator to decode base 62 numbers.
4756 28. dnsdb lookups can now access SRV records.
4758 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4759 the configuration file.
4761 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4762 (-v) mode. This makes the output for a verbose queue run more intelligible.
4764 31. Added a use_postmaster feature to recipient callouts.
4766 32. Added the $body_zerocount variable, containing the number of binary zero
4767 bytes in the message body.
4769 33. The time of last modification of the "new" subdirectory is now used as the
4770 "mailbox time last read" when there is a quota error for a maildir
4773 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4775 35. Added +ignore_unknown as a special item in host lists.
4777 36. Code for decoding IPv6 addresses in host lists is now included, even if
4778 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4779 address was recognized as an IP address, but was then not correctly decoded
4780 into binary, causing unexpected and incorrect effects when compared with
4787 1. Very minor rewording of debugging text in manualroute to say "list of
4788 hosts" instead of "hostlist".
4790 2. If verify=header_syntax was set, and a header line with an unqualified
4791 address (no domain) and a large number of spaces between the end of the
4792 name and the colon was received, the reception process suffered a buffer
4793 overflow, and (when I tested it) crashed. This was caused by some obsolete
4794 code that should have been removed. The fix is to remove it!
4796 3. When running in the test harness, delay a bit after writing a bounce
4797 message to get a bit more predictability in the log output.
4799 4. Added a call to search_tidyup() just before forking a reception process. In
4800 theory, someone could use a lookup in the expansion of smtp_accept_max_
4801 per_host which, without the tidyup, could leave open a database connection.
4803 5. Added the variables $recipient_data and $sender_data which get set from a
4804 lookup success in an ACL "recipients" or "senders" condition, or a router
4805 "senders" option, similar to $domain_data and $local_part_data.
4807 6. Moved the writing of debug_print from before to after the "senders" test
4810 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4811 problems for message scanning, either using a data ACL, or using
4812 local_scan() because the Received: header was not generated till after they
4813 were called (in order to set the time as the time of reception completion).
4814 I have revised the way this works. The header is now generated after the
4815 body is received, but before the ACL or local_scan() are called. After they
4816 are run, the timestamp in the header is updated.
4822 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4823 before starting a queue runner without re-exec. This happened only when
4824 deliver_drop_privilege was set or when the Exim user was set to root. The
4825 effect of the bug was that timeouts during subsequent deliveries caused
4826 crashes instead of being properly handled. The handler is now left at its
4827 default (and expected) setting.
4829 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4830 message, again when deliver_drop_privilege is set or Exim is run as root.
4831 The bug described in (1) was not present in this case, but the tidying up
4832 of the other signals was missing. I have made the two cases consistent.
4834 3. The ignore_target_hosts setting on a manualroute router was being ignored
4835 for hosts that were looked up using the /MX notation.
4837 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4840 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4841 operated on the sender address. After changing the $sender_address to <>
4842 for the sender address verify, Exim was re-instated it as the original
4843 (before rewriting) address, but remembering that it had rewritten it, so it
4844 wasn't rewriting it again. This bug also had the effect of breaking the
4845 sender address verification caching when the sender address was rewritten.
4847 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4848 This has been changed so that if the ip literal address matches
4849 ignore_target_hosts, the router declines.
4851 7. Added expansion conditions match_domain, match_address, and match_local_
4852 part (NOT match_host).
4854 8. The placeholder for the Received: header didn't have a length field set.
4856 9. Added code to Exim itself and to exim_lock to test for a specific race
4857 condition that could lead to file corruption when using MBX delivery. The
4858 issue is with the lockfile that is created in /tmp. If this file is removed
4859 after a process has opened it but before that process has acquired a lock,
4860 there is the potential for a second process to recreate the file and also
4861 acquire a lock. This could lead to two Exim processes writing to the file
4862 at the same time. The added code performs the same test as UW imapd; it
4863 checks after acquiring the lock that its file descriptor still refers to
4864 the same named file.
4866 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4867 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4870 11. Added the smtp_active_hostname option. If used, this will typically be made
4871 to depend on the incoming interface address. Because $interface_address is
4872 not set up until the daemon has forked a reception process, error responses
4873 that can happen earlier (such as "too many connections") no longer contain
4876 12. If an expansion in a condition on a "warn" statement fails because a lookup
4877 defers, the "warn" statement is abandoned, and the next ACL statement is
4878 processed. Previously this caused the whole ACL to be aborted.
4880 13. Added the iplsearch lookup type.
4882 14. Added ident_timeout as a log selector.
4884 15. Added tls_certificate_verified as a log selector.
4886 16. Added a global option tls_require_ciphers (compare the smtp transport
4887 option of the same name). This controls incoming TLS connections.
4889 17. I finally figured out how to make tls_require_ciphers do a similar thing
4890 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4891 before starting the TLS session.
4893 18. Tabs are now shown as \t in -bP output.
4895 19. If the log selector return_path_on_delivery was set, Exim crashed when
4896 bouncing a message because it had too many Received: header lines.
4898 20. If two routers both had headers_remove settings, and the first one included
4899 a superfluous trailing colon, the final name in the first list and the
4900 first name in the second list were incorrectly joined into one item (with a
4901 colon in the middle).
4907 1. Added -C and -D options to the exinext utility, mainly to make it easier
4908 to include in the automated testing, but these could be helpful when
4909 multiple configurations are in use.
4911 2. The exinext utility was not formatting the output nicely when there was
4912 an alternate port involved in the retry record key, nor when there was a
4913 message id as well (for retries that were specific to a specific message
4914 and a specific host). It was also confused by IPv6 addresses, because of
4915 the additional colons they contain. I have fixed the IPv4 problem, and
4916 patched it up to do a reasonable job for IPv6.
4918 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4919 delivery, the log line now contains "pipelined" if PIPELINING was used.
4921 4. An SMTP transport process used to panic and die if the bind() call to set
4922 an explicit outgoing interface failed. This has been changed; it is now
4923 treated in the same way as a connect() failure.
4925 5. A reference to $sender_host_name in the part of a conditional expansion
4926 that was being skipped was still causing a DNS lookup. This no longer
4929 6. The def: expansion condition was not recognizing references to header lines
4930 that used bh_ and bheader_.
4932 7. Added the _cache feature to named lists.
4934 8. The code for checking quota_filecount in the appendfile transport was
4935 allowing one more file than it should have been.
4937 9. For compatibility with Sendmail, the command line option
4945 and sets the incoming protocol and host name (for trusted callers). The
4946 host name and its colon can be omitted when only the protocol is to be set.
4947 Note the Exim already has two private options, -pd and -ps, that refer to
4948 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4949 "s", but I don't think that's a major issue.
4951 10. A number of refactoring changes to the code, none of which should affect
4954 (a) The number of logging options was getting close to filling up the
4955 32-bit word that was used as a bit map. I have split them into two classes:
4956 those that are passed in the argument to log_write(), and those that are
4957 only ever tested independently outside of that function. These are now in
4958 separate 32-bit words, so there is plenty of room for expansion again.
4959 There is no change in the user interface or the logging behaviour.
4961 (b) When building, for example, log lines, the code previously used a
4962 macro that called string_cat() twice, in order to add two strings. This is
4963 not really sufficiently general. Furthermore, there was one instance where
4964 it was actually wrong because one of the argument was used twice, and in
4965 one call a function was used. (As it happened, calling the function twice
4966 did not affect the overall behaviour.) The macro has been replaced by a
4967 function that can join an arbitrary number of extra strings onto a growing
4970 (c) The code for expansion conditions now uses a table and a binary chop
4971 instead of a serial search (which was left over from when there were very
4972 few conditions). Also, it now recognizes conditions like "pam" even when
4973 the relevant support is not compiled in: a suitably worded error message is
4974 given if an attempt is made to use such a condition.
4976 11. Added ${time_interval:xxxxx}.
4978 12. A bug was causing one of the ddress fields not to be passed back correctly
4979 from remote delivery subprocesses. The field in question was not being
4980 subsequently used, so this caused to problems in practice.
4982 13. Added new log selectors queue_time and deliver_time.
4984 14. Might have fixed a bug in maildirsizefile handling that threw up
4985 "unexpected character" debug warnings, and recalculated the data
4986 unnecessarily. In any case, I expanded the warning message to give more
4989 15. Added the message "Restricted characters in address" to the statements in
4990 the default ACL that block characters like @ and % in local parts.
4992 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4993 Three changes have been made:
4995 (a) There was a serious bug; a negative response to MAIL caused the whole
4996 recipient domain to be cached as invalid, thereby blocking all messages
4997 to all local parts at the same domain, from all senders. This bug has
4998 been fixed. The domain is no longer cached after a negative response to
4999 MAIL if the sender used is not empty.
5001 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5004 (c) A new callout option, "use_sender" has been added for people who want
5005 the modified behaviour.
5011 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5014 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5015 indeed breaks things for older releases.
5017 3. Added additional logging to the case where there is a problem reading data
5018 from a filter that is running in a subprocess using a pipe, in order to
5019 try to track down a specific problem.
5021 4. Testing facility fudge: when running in the test harness and attempting
5022 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5023 getting "No route to host". Convert this to a timeout.
5025 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5028 6. Some OS don't have socklen_t but use size_t instead. This affects the
5029 fifth argument of getsockopt() amongst other things. This is now
5030 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5031 can be set for individual OS. I have set it for SunOS5, OSF1, and
5032 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5033 some earlier ones do not.
5035 7. Change 4.30/15 was not doing the test caselessly.
5037 8. The standard form for an IPv6 address literal was being rejected by address
5038 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5039 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5040 this, as well as the form without the "IPv6" on the front (but only when
5041 address literals are enabled, of course).
5043 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5045 10. Exim crashed if a message with an empty sender address specified by -f
5046 encountered a router with an errors_to setting. This could be provoked only
5047 by a command such as
5051 where an empty string was supplied; "<>" did not hit this bug.
5053 11. Installed PCRE release 4.5.
5055 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5056 remained set. It is now erased.
5058 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5059 times from message ids (which are base 36 rather than the normal 62).
5061 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5062 were being counted as actual protocol errors, and logged if the log
5063 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5064 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5065 and DATA following a set of rejected RCPTs do not count as protocol errors.
5066 In other words, Exim assumes they were pipelined, though this may not
5067 actually be the case. Of course, in all cases the client gets an
5068 appropriate error code.
5070 15. If a lookup fails in an ACL condition, a message about the failure may
5071 be available; it is used if testing the ACL cannot continue, because most
5072 such messages specify what the cause of the deferral is. However, some
5073 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5074 that caused an old message to be retained and used if a later statement
5075 caused a defer, replacing the real cause of the deferral.
5077 16. If an IP address had so many PTR records that the DNS lookup buffer
5078 was not large enough to hold them, Exim could crash while trying to process
5079 the truncated data. It now detects and logs this case.
5081 17. Further to 4.21/58, another change has been made: if (and only if) the
5082 first line of a message (the first header line) ends with CRLF, a bare LF
5083 in a subsequent header line has a space inserted after it, so as not to
5084 terminate the header.
5086 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5087 unnecessarily, used atoi() instead of strtol(), and didn't check the
5088 termination when getting file sizes from file names by regex.
5090 19. Completely re-implemented the support for maildirsize files, in the light
5091 of a number of problems with the previous contributed implementation
5092 (4.30/29). In particular:
5094 . If the quota is zero, the maildirsize file is maintained, but no quota is
5097 . If the maildir directory does not exist, it is created before any attempt
5098 to write a maildirsize file.
5100 . The quota value in the file is just a cache; if the quota is changed in
5101 the transport, the new value overrides.
5103 . A regular expression is available for excluding directories from the
5106 20. The autoreply transport checks the characters in options that define the
5107 message's headers; it allows continued headers, but it was checking with
5108 isspace() after an embedded newline instead of explicitly looking for a
5111 21. If all the "regular" hosts to which an address was routed had passed their
5112 expiry times, and had not reached their retry times, the address was
5113 bounced, even if fallback hosts were defined. Now Exim should go on to try
5116 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5117 equivalent code in the SMTP transport. Some hosts send humungous responses
5118 to HELO/EHLO, more than 1024 it seems.
5120 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5121 gives compiler warnings in some environments. I've now done it "properly",
5124 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5125 (because of problems with the built-in one) was declared to return uschar *
5126 instead of char *, causing compiler failure.
5128 25. Fixed a file descriptor leak when processing alias/forward files.
5130 26. Fixed a minor format string issue in dbfn.c.
5132 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5134 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5135 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5138 29. When a sender address is verified, it is cached, to save repeating the test
5139 when there is more than one recipient in a message. However, when the
5140 verification involves a callout, it is possible for different callout
5141 options to be set for different recipients. It is too complicated to keep
5142 track of this in the cache, so now Exim always runs a verification when a
5143 callout is required, relying on the callout cache for the optimization.
5144 The overhead is duplication of the address routing, but this should not be
5147 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5148 to be verified with callout=postmaster, and the main callout worked but the
5149 postmaster check failed, the verification correctly failed. However, if a
5150 subsequent RCPT command asked for sender verification *without* the
5151 postmaster check, incorrect caching caused this verification also to fail,
5154 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5155 it was not caching the DNS options (qualify_single, search_parents) that
5156 were used when the lookup failed. A subsequent lookup with different
5157 options therefore always gave the same answer, though there were cases
5158 where it should not have. (Example: a "domains = !$mx_any" option on a
5159 dnslookup router: the "domains" option is always processed without any
5160 widening, but the router might have qualify_single set.) Now Exim uses the
5161 cached value only when the same options are set.
5163 32. Added John Jetmore's "exipick" utility to the distribution.
5165 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5166 than a timeout (e.g. a certificate is required, and is not provided), an
5167 Exim server now closes the connection immediately. Previously it waited for
5168 the client to close - but if the client is SSL, it seems that they each
5169 wait for each other, leading to a delay before one of them times out.
5171 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5172 maintained 0.8.x compatibility because I don't think many are using it, and
5173 it is clearly obsolete.
5175 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5178 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5179 client certificate was expired. A simple patch fixes this, though I don't
5180 understand the full logic of why the verify callback is called multiple
5183 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5184 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5185 which causes problems with some clients (such as the Certicom SSL Plus
5186 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5187 disables the coutermeasure allowing Eudora to connect."
5189 38. Exim was not checking that a write() to a log file succeeded. This could
5190 lead to Bad Things if a log got too big, in particular if it hit a file
5191 size limit. Exim now panics and dies if it cannot write to a log file, just
5192 as it does if it cannot open a log file.
5194 39. Modified OS/Makefile-Linux so that it now contains
5196 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5198 The two -D definitions ensure that Exim is compiled with large file
5199 support, which makes it possible to handle log files that are bigger than
5202 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5203 instance) a domain was checked against a named list that involved a lookup,
5204 causing $domain_data to be set, then another domain was checked against the
5205 same list, then the first domain was re-checked, the value of $domain_data
5206 after the final check could be wrong. In particular, if the second check
5207 failed, it could be set empty. This bug probably also applied to
5210 41. The strip_trailing_dot option was not being applied to the address given
5211 with the -f command-line option.
5213 42. The code for reading a message's header from the spool was incrementing
5214 $received_count, but never initializing it. This meant that the value was
5215 incorrect (doubled) while delivering a message in the same process in which
5216 it was received. In the most common configuration of Exim, this never
5217 happens - a fresh exec is done - but it can happen when
5218 deliver_drop_privilege is set.
5220 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5221 it now includes up to 150 characters of the unexpected data in the log
5224 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5225 and building data strings. The size of both of these buffers was 10 000
5226 bytes - far larger than anybody would *ever* want, thought I. Needless to
5227 say, somebody hit the limit. I have increased the maximum line length to
5228 20 000 and the maximum data length of concatenated lines to 100 000. I have
5229 also fixed two bugs, because there was no checking on these buffers. Tsk,
5230 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5231 buffer is too small.
5233 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5234 lsearch lookups. Now it does.
5236 46. When parsing a route_list item in a manualroute router, a fixed-length
5237 buffer was used for the list of hosts. I made this 1024 bytes long,
5238 thinking that nobody would ever have a list of hosts that long. Wrong.
5239 Somebody had a whole pile of complicated expansion conditions, and the
5240 string was silently truncated, leading to an expansion error. It turns out
5241 that it is easier to change to an unlimited length (owing to other changes
5242 that have happened since this code was originally written) than to build
5243 structure for giving a limitation error. The length of the item that
5244 expands into the list of hosts is now unlimited.
5246 47. The lsearch lookup could not handle data where the length of text line was
5247 more than 4095 characters. Such lines were truncated, leading to shortened
5248 data being returned. It should now handle lines of any length.
5250 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5251 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5254 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5255 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5256 no longer happens in comment lines. A list of replacements is now placed
5257 at the head of all of the source files, except those whose only change is
5258 to replace PERL_COMMAND in the very first #! line.
5260 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5261 messages on the queue, with a bottom-up merge sort, using code contributed
5262 by Michael Haardt. This should make operations like -bp somewhat faster on
5263 large queues. It won't affect queue runners, except when queue_run_in_order
5266 51. Installed eximstats 1.31 in the distribution.
5268 52. Added support for SRV lookups to the dnslookup router.
5270 53. If an ACL referred to $message_body or $message_body_end, the value was not
5271 reset for any messages that followed in the same SMTP session.
5273 54. The store-handling optimization for building very long strings was not
5274 differentiating between the different store pools. I don't think this
5275 actually made any difference in practice, but I've tidied it.
5277 55. While running the routers to verify a sender address, $sender_address
5278 was still set to the sender address. This is wrong, because when routing to
5279 send a bounce to the sender, it would be empty. Therefore, I have changed
5280 it so that, while verifying a sender address, $sender_address is set to <>.
5281 (There is no change to what happens when verifying a recipient address.)
5283 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5284 target A or AAAA records (if not already returned) without resetting the
5285 qualify_single or search_parents options of the DNS resolver. These are
5286 inappropriate in this case because the targets of MX and SRV records must
5287 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5288 target that, when qualified, matched something in the local domain. These
5289 two options are now turned off when doing these lookups.
5291 57. It seems that at least some releases of Reiserfs (which does not have the
5292 concept of a fixed number of inodes) returns zero and not -1 for the
5293 number of available inodes. This interacted badly with check_spool_inodes,
5294 which assumed that -1 was the "no such thing" setting. What I have done is
5295 to check that the total number of inodes is greater than zero before doing
5296 the test of how many are available.
5298 58. When a "warn" ACL statement has a log_message modifier, the message is
5299 remembered, and not repeated. This is to avoid a lot of repetition when a
5300 message has many recipients that cause the same warning to be written.
5301 Howewer, Exim was preserving the list of already written lines for an
5302 entire SMTP session, which doesn't seem right. The memory is now reset if a
5303 new message is started.
5305 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5306 debugging output unless log_rewrite was also set.
5308 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5309 of (int)(handle) when we know that handle contains (void *)(-1).
5311 61. The Exim daemon panic-logs an error return when it closes the incoming
5312 connection. However "connection reset by peer" seems to be common, and
5313 isn't really an error worthy of noting specially, so that particular error
5316 62. When Exim is trying to find all the local interfaces, it used to panic and
5317 die if the ioctl to get the interface flags failed. However, it seems that
5318 on at least one OS (Solaris 9) it is possible to have an interface that is
5319 included in the list of interfaces, but for which you get a failure error
5320 for this call. This happens when the interface is not "plumbed" into a
5321 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5322 failure of the "get flags" call assumes that the interface is down.
5324 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5325 makes life easier for people who are doing arithmetic on fields extracted
5326 from dates, where you often get leading zeros that should not be
5327 interpreted as octal.
5329 64. Added qualify_domain to the redirect router, to override the global
5332 65. If a pathologically long header line contained very many addresses (the
5333 report of this problem mentioned 10 000) and each of them was rewritten,
5334 Exim could use up a very large amount of memory. (It kept on making new
5335 copies of the header line as it rewrote, and never released the old ones.)
5336 At the expense of a bit more processing, the header rewriting function has
5337 been changed so that it no longer eats memory in this way.
5339 66. The generation of the Received: header has been moved from the time that a
5340 message starts to be received, to the time that it finishes. The timestamp
5341 in the Received: header should now be very close to that of the <= log
5342 line. There are two side-effects of this change:
5344 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5345 logged header lines no longer include the local Received: line, because
5346 it has not yet been created. The same applies to a copy of the message
5347 that is returned to a non-SMTP sender when a message is rejected.
5349 (b) When a filter file is tested using -bf, no additional Received: header
5350 is added to the test message. After some thought, I decided that this
5353 This change does not affect the value of $received_for. It is still set
5354 after address rewriting, but before local_scan() is called.
5356 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5358 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5359 gave an unhelpful panic error message, and a defer error. I have managed to
5360 change this behaviour so that it now rejects any supplied certificate,
5361 which seems right, as the list of acceptable certificates is empty.
5363 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5364 gave an unhelpful defer error. I have not managed to make this reject any
5365 supplied certificates, but the error message it gives is "no certificate
5366 supplied", which is not helpful.
5368 70. exigrep's output now also includes lines that are not associated with any
5369 message, but which match the given pattern. Implemented by a patch from
5370 Martin Sluka, which also tidied up the Perl a bit.
5372 71. Recipient callout verification, like sender verification, was using <> in
5373 the MAIL FROM command. This isn't really the right thing, since the actual
5374 sender may affect whether the remote host accepts the recipient or not. I
5375 have changed it to use the actual sender in the callout; this means that
5376 the cache record is now keyed on a recipient/sender pair, not just the
5377 recipient address. There doesn't seem to be a real danger of callout loops,
5378 since a callout by the remote host to check the sender would use <>.
5379 [SEE ABOVE: changed after hitting problems.]
5381 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5382 temporary errors. However, in the case of such a code being given after
5383 the end of a data transmission (i.e. after ".") Exim was failing to write
5384 a retry record for the message. (Yes, there was some broken host that was
5385 actually sending 8xx at this point.)
5387 73. An unknown lookup type in a host list could cause Exim to panic-die when
5388 the list was checked. (An example that provoked this was putting <; in the
5389 middle of a list instead of at the start.) If this happened during a DATA
5390 ACL check, a -D file could be left lying around. This kind of configuration
5391 error no longer causes Exim to die; instead it causes a defer errror. The
5392 incident is still logged to the main and panic logs.
5394 74. Buglet left over from Exim 3 conversion. The message "too many messages
5395 in one connection" was written to the rejectlog but not the mainlog, except
5396 when address rewriting (yes!) was being logged.
5398 75. Added write_rejectlog option.
5400 76. When a system filter was run not as root (that is, when system_filter_user
5401 was set), the values of the $n variables were not being returned to the
5402 main process; thus, they were not subsequently available in the $sn
5405 77. Added +return_path_on_delivery log selector.
5407 78. A connection timeout was being treated differently from recipients deferred
5408 when testing hosts_max_try with a message that was older than the host's
5409 retry timeout. (The host should not be counted, thus allowing all hosts to
5410 be tried at least once before bouncing.) This may have been the cause of an
5411 occasionally reported bug whereby a message would remain on the queue
5412 longer than the retry timeout, but would be bounced if a delivery was
5413 forced. I say "may" because I never totally pinned down the problem;
5414 setting up timeout/retry tests is difficult. See also the next item.
5416 79. The ultimate address timeout was not being applied to errors that involved
5417 a combination of host plus message (for example, a timeout on a MAIL
5418 command). When an address resolved to a number of possible hosts, and they
5419 were not all tried for each delivery (e.g. because of hosts_max_try), a
5420 message could remain on the queue longer than the retry timeout.
5422 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5425 81. Fixed an obscure SMTP outgoing bug which required at least the following
5426 conditions: (a) there was another message waiting for the same server;
5427 (b) the server returned 5xx to all RCPT commands in the first message so
5428 that the message was not completed; (c) the server dropped the connection
5429 or gave a negative response to the RSET that Exim sends to abort the
5430 transaction. The observed case was a dropped connection after DATA that had
5431 been sent in pipelining mode. That is, the server had advertised PIPELINING
5432 but was not implementing it correctly. The effect of the bug was incorrect
5433 behaviour, such as trying another host, and this could lead to a crash.
5439 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5440 and daemon.c were passed as pointers to ints; they should have been
5441 pointers to socklen_t variables (which are typically unsigned ints).
5443 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5446 3. Fixed a really odd bug that affected only the testing scheme; patching a
5447 certain fixed string in the binary changed the value of another string that
5448 happened to be identical to the end of the original first string.
5450 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5451 name", it returns that address as the IP address. On some operating
5452 systems (e.g. Solaris), it also passes back the IP address string as the
5453 "host name". However, on others (e.g. Linux), it passes back an empty
5454 string. Exim wasn't checking for this, and was changing the host name to an
5455 empty string, assuming it had been canonicized.
5457 5. Although rare, it is permitted to have more than one PTR record for a given
5458 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5459 all the names associated with an address, because they do in Solaris.
5460 However, it seems that they do not in Linux for data that comes from the
5461 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5462 I found this out when I moved to a new Linux workstation and tried to run
5463 the Exim test suite.
5465 To get round this problem I have changed the code so that it now does its
5466 own call to the DNS to look up PTR records when searching for a host name.
5467 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5468 addresses that are only in /etc/hosts are still found.
5470 This behaviour is, however, controlled by an option called host_lookup_
5471 order, which defaults to "bydns:byaddr". If people want to use the other
5472 order, or indeed, just use one or the other means of lookup, they can
5473 specify it in this variable.
5475 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5476 some operating systems, this comes back from gethostbyaddr() as an empty
5477 string, and this is what Exim used to test for. However, it seems that in
5478 other systems, "." is yielded. Exim now tests for this case too.
5480 7. The values of check_spool_space and check_log_space are now held internally
5481 as a number of kilobytes instead of an absolute number of bytes. If a
5482 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5483 kilobyte. This means that much larger values can be stored.
5485 8. Exim monitor: an attempt to get the action menu when not actually pointing
5486 at a message produces an empty menu entitled "No message selected". This
5487 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5488 no entries in it ("Shell widget menu has zero width and/or height"). So I
5489 have added a single, blank menu entry in this case.
5491 9. Added ${quote_local_part.
5493 10. MIME decoding is now applied to the contents of Subject: header lines when
5496 11. Now that a reference to $sender_host_address automatically causes a reverse
5497 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5498 host lookup before query-style lookups in lists that might use this
5499 variable. This has therefore been abolished, and the "net-" prefix is no
5500 longer necessary for query-style lookups.
5502 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5503 have been a typo for LFLAGS, so it has been changed.
5505 13. The install script calls Exim with "-C /dev/null" in order to find the
5506 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5507 to be output. Howeve, since Exim outputs its version number before the
5508 error, it didn't break the script. It just looked ugly. I fixed this by
5509 always allowing "-C /dev/null" if the caller is root.
5511 14. Ignore overlarge ACL variable number when reading spool file - insurance
5512 against a later release with more variables having written the file.
5514 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5515 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5516 this, as well as the form without the "IPv6" on the front.
5518 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5519 OS/Makefile-Darwin file.
5521 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5522 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5524 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5525 variable. However, some OS use ints for inodes. I've added cast to long int
5526 to get rid of the compiler warning.
5528 19. I had forgotten to lock out "/../" in configuration file names when
5529 ALT_CONFIG_PREFIX was set.
5531 20. Routers used for verification do not need to specify transports. However,
5532 if such a router generated a host list, and callout was configured, Exim
5533 crashed, because it could not find a port number from the (non-existent)
5534 transport. It now assumes port 25 in this circumstance.
5536 21. Added the -t option to exigrep.
5538 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5539 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5540 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5541 them is set, LOOKUP_LSEARCH is forced.
5543 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5545 24. Added sender and host information to the "rejected by local_scan()" log
5546 line; previously there was no indication of these.
5548 25. Added .include_if_exists.
5550 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5551 Linux. It turns out that not all file systems support this. Apparently some
5552 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5553 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5554 not supported on the file descriptor, is now ignored when Exim is trying to
5555 sync a directory. This applies only to Linux.
5557 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5559 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5560 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5563 29. Added support for maildirsize files from supplied patch (modified a bit).
5565 30. The use of :fail: followed by an empty string could lead Exim to respond to
5566 sender verification failures with (e.g.):
5568 550 Verification failed for <xxx>
5569 550 Sender verify failed
5571 where the first response line was missing the '-' that indicates it is not
5572 the final line of the response.
5574 31. The loop for finding the name of the user that called Exim had a hardwired
5575 limit of 10; it now uses the value of finduser_retries, which is used for
5576 all other user lookups.
5578 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5581 33. Exim was neglecting to zero errno before one call of strtol() when
5582 expanding a string and expecting an integer value. On some systems this
5583 resulted in spurious "integer overflow" errors. Also, it was casting the
5584 result into an int without checking.
5586 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5587 did not work. The code looks as if it has *never* worked, though it appears
5588 to have been documented since at least releast 1.62. I have made it work.
5590 35. The "timeout_DNS" error in retry rules, also documented since at least
5591 1.62, also never worked. As it isn't clear exactly what this means, and
5592 clearly it isn't a major issue, I have abolished the feature by treating it
5593 as "timeout", and writing a warning to the main and panic logs.
5595 36. The display of retry rules for -brt wasn't always showing the error code
5598 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5599 timeout_connect_A, timeout_connect_MX.
5601 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5602 to the empty sender.
5604 39. The daemon was not analysing the content of -oX till after it had closed
5605 stderr and disconnected from the controlling terminal. This meant that any
5606 syntax errors were only noted on the panic log, and the return code from
5607 the command was 0. By re-arranging the code a little, I've made the
5608 decoding happen first, so such errors now appear on stderr, and the return
5609 code is 1. However, the actual setting up of the sockets still happens in
5610 the disconnected process, so errors there are still only recorded on the
5613 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5614 connections (as happens on some IP stacks) was logged at start up time as
5615 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5616 it from "IPv6 and IPv4", which means that two separate sockets are being
5619 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5620 says whether AF_INET or AF_INET6 was passed as an argument.
5622 42. Exiwhat output was messed up when time zones were included in log
5625 43. Exiwhat now gives more information about the daemon's listening ports,
5626 and whether -tls-on-connect was used.
5628 44. The "port" option of the smtp transport is now expanded.
5630 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5631 silently ignored. Now an error message is written to the main and panic
5634 46. There's a new ACL modifier called "logwrite" which writes to a log file
5635 as soon as it is encountered.
5637 47. Added $local_user_uid and $local_user_gid at routing time.
5639 48. Exim crashed when trying to verify a sender address that was being
5642 49. Exim was recognizing only a space character after ".include". It now also
5643 recognizes a tab character.
5645 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5646 extracting the relevant information from the specification. The man page no
5647 longer contains scrambled data for the -d option, and I've added a section
5648 at the front about calling Exim under different names.
5650 51. Added "extra_headers" argument to the "mail" command in filter files.
5652 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5655 53. Installed eximstats 1.29.
5657 54. Added transport_filter_timeout as a generic transport option.
5659 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5660 Cc: header lines. This was required by RFC 822, but it not required by RFC
5663 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5664 incoming messages that did not have them. Now it does so only if the
5665 message originates locally, that is, if there is no associated remote host
5666 address. When Resent- header lines are present, this applies to the Resent-
5667 lines rather than the non-Resent- lines.
5669 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5670 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5672 58. Messages for configuration errors now include the name of the main
5673 configuration file - useful now that there may be more than one file in a
5674 list (.included file names were always shown).
5676 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5677 for those rare installations that do not start the daemon as root or run it
5678 setuid root. I've cut out the call to initgroups() if the daemon is not
5681 60. The Exim user and group can now be bound into the binary as text strings
5682 that are looked up at the start of Exim's processing.
5684 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5686 62. Added $mailstore_basename variable.
5688 63. Installed patch to sieve.c from Michael Haardt.
5690 64. When Exim failed to open the panic log after failing to open the main log,
5691 the original message it was trying to log was written to stderr and debug
5692 output, but if they were not available (the usual case in production), it
5693 was lost. Now it is written to syslog before the two lines that record the
5694 failures to open the logs.
5696 65. Users' Exim filters run in subprocesses under the user's uid. It is
5697 possible for a "deliver" command or an alias in a "personal" command to
5698 provoke an address rewrite. If logging of address rewriting is configured,
5699 this fails because the process is not running as root or exim. There may be
5700 a better way of dealing with this, but for the moment (because 4.30 needs
5701 to be released), I have disabled address rewrite logging when running a
5702 filter in a non-root, non-exim process.
5708 1. The buildconfig auxiliary program wasn't quoting the value set for
5709 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5710 not defined. This bug was masked in 4.22 by the effect that was fixed in
5713 2. Some messages that were rejected after a message id was allocated were
5714 shown as "incomplete" by exigrep. It no longer does this for messages that
5715 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5717 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5718 have allow_domain_literals set, the ID did not get logged in the <= line.
5719 Domain literals are now always recognized in Message-ID: header lines.
5721 4. The first argument for a ${extract expansion item is the key name or field
5722 number. Leading and trailing spaces in this item were not being ignored,
5723 causing some misleading effects.
5725 5. When deliver_drop_privilege was set, single queue runner processes started
5726 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5727 same command in the process it spins off) were not dropping privilege.
5729 6. When the daemon running as "exim" started a queue runner, it always
5730 re-executed Exim in the spun-off process. This is a waste of effort when
5731 deliver_drop_privilege is set. The new process now just calls the
5732 queue-runner function directly.
5738 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5741 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5742 ignored. Though the use of -oP was forcing the writing of a pid file, it
5743 was always written to the default place.
5745 3. If the message "no IP address found for host xxxx" is generated during
5746 incoming verification, it is now followed by identification of the incoming
5747 connection (so you can more easily find what provoked it).
5749 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5751 5. Added some features to "harden" Exim a bit more against certain attacks:
5753 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5754 be put in Local/Makefile. This is like the never_users runtime option,
5755 but it cannot be overridden. The default setting is "root".
5757 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5758 prefix string with which any file named in a -C command line option
5761 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5762 is retained for -C and -D only if the caller of Exim is root. Without
5763 it, the exim user may also use -C and -D and retain privilege.
5765 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5766 command line option is disabled.
5768 6. Macro names set by the -D option must start with an upper case letter, just
5769 like macro names defined in the configuration file.
5771 7. Added "dereference=" facility to LDAP.
5773 8. Two instances of the typo "uknown" in the source files are fixed.
5775 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5776 the Configure-Makefile script screwed up while processing it.
5778 10. Incorporated PCRE 4.4.
5780 11. The SMTP synchronization check was not operating right at the start of an
5781 SMTP session. For example, it could not catch a HELO sent before the client
5782 waited for the greeting. There is now a check for outstanding input at the
5783 point when the greeting is written. Because of the duplex, asynchronous
5784 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5785 way, but not yet received, when the check is performed.
5787 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5788 on TCP/IP sockets, because this apparently causes some broken clients to
5791 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5792 unchanged) from the Cygwin maintainer.
5794 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5795 was supported instead of testing for mbx. Effectively a typo.
5797 15. The spa authenticator server code was not checking that the input it
5798 received was valid base64.
5800 16. The debug output line for the "set" modifier in ACLs was not showing the
5801 name of the variable that was being set.
5803 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5805 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5806 lookup on its own. Something else was needed to trigger the lookup. For
5807 example, a match in host_lookup or the need for a host name in a host list.
5808 Now, if $sender_host_name is referenced and the host name has not yet been
5809 looked up, a lookup is performed. If the lookup fails, the variable remains
5810 empty, and $host_lookup_failed is set to "1".
5812 19. Added "eqi" as a case-independent comparison operator.
5814 20. The saslauthd authentication condition could segfault if neither service
5815 nor realm was specified.
5817 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5818 error message that was logged was misleading, and incoming SMTP
5819 connections were dropped. The message is now more accurate, and temporary
5820 errors are given to SMTP connections.
5822 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5823 (except RSET) with a 421 error, until QUIT is received. However, it was
5824 failing to send a response to QUIT.
5826 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5827 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5828 ACL is now run after the helo_try_verify_hosts code.
5830 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5831 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5832 case-independent, so other case variants are also recognized). Apparently
5833 some systems use these upper case variants.
5835 25. If more than two messages were waiting for the same host, and a transport
5836 filter was specified for the transport, Exim sent two messages over the
5837 same TCP/IP connection, and then failed with "socket operation on non-
5838 socket" when it tried to send the third.
5840 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5842 27. The extern definition of crypt16() in expand.c was not being excluded when
5843 the OS had its own crypt16() function.
5845 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5846 as a preferred synonym for return_size_limit, both as an option and as an
5849 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5851 30. Changed the default configuration ACL to relax the local part checking rule
5852 for addresses that are not in any local domains. For these addresses,
5853 slashes and pipe symbols are allowed within local parts, but the sequence
5854 /../ is explicitly forbidden.
5856 31. SPA server authentication was not clearing the challenge buffer before
5859 32. log_message in a "warn" ACL statement was writing to the reject log as
5860 well as to the main log, which contradicts the documentation and doesn't
5861 seem right (because no rejection is happening). So I have stopped it.
5863 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5864 However, I am unable to do any testing of this.
5866 34. Fixed an infelicity in the appendfile transport. When checking directories
5867 for a mailbox, to see if any needed to be created, it was accidentally
5868 using path names with one or more superfluous leading slashes; tracing
5869 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5871 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5872 contents are added to the log line that is written for every discarded
5873 recipient. (Previously a log_message setting was ignored.)
5875 36. The ${quote: operator now quotes the string if it is empty.
5877 37. The install script runs exim in order to find its version number. If for
5878 some reason other than non-existence or emptiness, which it checks, it
5879 could not run './exim', it was installing it with an empty version number,
5880 i.e. as "exim-". This error state is now caught, and the installation is
5883 38. An argument was missing from the function that creates an error message
5884 when Exim fails to connect to the socket for saslauthd authentication.
5885 This could cause Exim to crash, or give a corrupted message.
5887 39. Added isip, isip4, and isip6 to ${if conditions.
5889 40. The ACL variables $acl_xx are now saved with the message, and can be
5890 accessed later in routers, transports, and filters.
5892 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5893 strings in the file are not string-expanded.
5895 42. If a MAIL command specified a SIZE value that was too large to fit into an
5896 int variable, the check against message_size_limit failed. Such values are
5897 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5898 day this will have to be increased, but I don't think I want to be around
5899 when emails are that large.
5906 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5907 iconv() is not standard in FreeBSD.
5909 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5910 IPv6 enabled. The observed symptom was a segmentation fault on return from
5911 the function os_common_find_running_interfaces() in src/os.c.
5913 3. In the check_special_case() function in daemon.c I had used "errno" as an
5914 argument name, which causes warnings on some systems. This was basically a
5915 typo, since it was named "eno" in the comments!
5917 4. The code that waits for the clock to tick (at a resolution of some fraction
5918 of a second) so as to ensure message-id uniqueness was always waiting for
5919 at least one whole tick, when it could have waited for less. [This is
5920 almost certainly not relevant at current processor speeds, where it is
5921 unlikely to ever wait at all. But we try to future-proof.]
5923 5. The function that sleeps for a time interval that includes fractions of a
5924 second contained a race. It did not block SIGALRM between setting the
5925 timer, and suspending (a couple of lines later). If the interval was short
5926 and the sigsuspend() was delayed until after it had expired, the suspension
5927 never ended. On busy systems this could lead to processes getting stuck for
5930 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5931 process, before it forks any delivery processes. The open lookup caching
5932 mechanism meant that the open file or database connection was passed into
5933 the delivery process. The problem was that delivery processes always tidy
5934 up cached lookup data. This could cause a problem for the next delivery
5935 process started by the queue runner, because the external queue runner
5936 process does not know about the closure. So the next delivery process
5937 still has data in the lookup cache. In the case of a file lookup, there was
5938 no problem because closing a file descriptor in a subprocess doesn't affect
5939 the parent. However, if the lookup was caching a connection to a database,
5940 the connection was closed, and the second delivery process was likely to
5941 see errors such as "PGSQL: query failed: server closed the connection
5942 unexpectedly". The problem has been fixed by closing all cached lookups
5943 in a queue runner before running a delivery process.
5945 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5946 seem to have the "const" qualifier which it has on other OS. I've
5949 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5950 *of the same type* (client or server) with the same public name that an
5951 error should be diagnosed.
5953 9. When Exim looked up a host name for an IP address, but failed to find the
5954 original IP address when looking up the host name (a safety check), it
5955 output the message "<ip address> does not match any IP for NULL", which was
5956 confusing, to say the least. The bug was that the host name should have
5957 appeared instead of "NULL".
5959 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5960 user that is built into the binary, and the -C or -D options is used, root
5961 privilege is dropped before the configuration file is read. In addition,
5962 logging is switched to stderr instead of the normal log files. If the
5963 configuration then re-defines the Exim user, the unprivileged environment
5964 is probably not what is expected, so Exim logs a panic warning message (but
5967 However, if deliver_drop_privilege is set, the unprivileged state may well
5968 be exactly what is intended, so the warning has been cut out in that case,
5969 and Exim is allowed to try to write to its normal log files.
5975 1. smtp_return_error_details was not giving details for temporary sender
5976 or receiver verification errors.
5978 2. Diagnose a configuration error if two authenticators have the same public
5981 3. Exim used not to create the message log file for a message until the first
5982 delivery attempt. This could be confusing when incoming messages were held
5983 for policy or load reasons. The message log file is now created at the time
5984 the message is received, and an initial "Received" line is written to it.
5986 4. The automatically generated man page for command line options had a minor
5987 bug that caused no ill effects; however, a more serious problem was that
5988 the procedure for building the man page automatically didn't always
5989 operate. Consequently, release 4.20 contains an out-of-date version. This
5990 shouldn't happen again.
5992 5. When building Exim with embedded Perl support, the script that builds the
5993 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5994 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5996 6. The freeze_tell option was not being used for messages that were frozen on
5997 arrival, either by an ACL or by local_scan().
5999 7. Added the smtp_incomplete_transaction log selector.
6001 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6002 was accepting AUTH without a new EHLO.
6004 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6005 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6008 10. Logging of TCP/IP connections (when configured) now happens in the main
6009 daemon process instead of the child process, so that the TCP/IP connection
6010 count is more accurate (but it can never be perfect).
6012 11. The use of "drop" in a nested ACL was not being handled correctly in the
6013 outer ACL. Now, if condition failure induced by the nested "drop" causes
6014 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6015 or "require"), the connection is dropped.
6017 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6018 that yield "discard" can now be used with an "accept" or a "discard" verb,
6019 but an error is generated for any others (because I can't see a useful way
6020 to define what should happen).
6022 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6023 are now processed in the same way as lines in the Exim configuration file.
6024 In particular, continuation lines are supported.
6026 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6028 15. Added -ti meaning -t -i.
6030 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6031 domains, and warn by logging if others are found.
6033 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6034 returned from GIFCONF when Exim is trying to find the list of interfaces on
6035 a host. The code in os.c has been modified to copy each ifreq to an aligned
6036 structure in all cases.
6038 Also, in some cases, the returned ifreq's were being copied to a 'struct
6039 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6040 means the last couple of bytes of an IPv6 address could be chopped if the
6041 ifreq contained only a normal sockaddr (14 bytes storage).
6043 18. Named domain lists were not supported in the hosts_treat_as_local option.
6044 An entry such as +xxxx was not recognized, and was treated as a literal
6047 19. Ensure that header lines added by a DATA ACL are included in the reject log
6048 if the ACL subsequently rejects the message.
6050 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6051 MD5 (which is deprecated).
6053 21. When testing a filter file using -bf, Exim was writing a message when it
6054 took the sender from a "From " line in the message, but it was not doing so
6055 when it took $return_path from a Return-Path: header line. It now does.
6057 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6058 with a valid header line field name (a series of printing characters
6059 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6061 23. Changed "disc" in the source to "disk" to conform to the documentation and
6062 the book and for uniformity.
6064 24. Ignore Sendmail's -Ooption=value command line item.
6066 25. When execve() failed while trying to run a command in a pipe transport,
6067 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6068 could be confused with a return value of 69 from the command itself. This
6069 has been changed to 127, the value the shell returns if it is asked to run
6070 a non-existent command. The wording for the related log line suggests a
6071 non-existent command as the problem.
6073 26. If received_header_text expands to an empty string, do not add a Received:
6074 header line to the message. (Well, it adds a token one on the spool, but
6075 marks it "old" so that it doesn't get used or transmitted.)
6077 27. Installed eximstats 1.28 (addition of -nt option).
6079 28. There was no check for failure on the call to getsockname() in the daemon
6080 code. This can fail if there is a shortage of resources on the system, with
6081 ENOMEM, for example. A temporary error is now given on failure.
6083 29. Contrary to the C standard, it seems that in some environments, the
6084 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6085 program. Exim now does this explicitly; it affects the formatting of
6086 timestamps using strftime().
6088 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6089 complaints. I've now initialized all the variables, to avoid this.
6091 32. Header lines added by a system filter were not being "seen" during
6092 transport-time rewrites.
6094 33. The info_callback() function passed to OpenSSL is set up with type void
6095 (*)(SSL *, int, int), as described somewhere. However, when calling the
6096 function (actually a macro) that sets it up, the type void(*)() is
6097 expected. I've put in a cast to prevent warnings from picky compilers.
6099 34. If a DNS black list lookup found a CNAME record, but there were no A
6100 records associated with the domain it pointed at, Exim crashed.
6102 35. If a DNS black list lookup returned more than one A record, Exim ignored
6103 all but the first. It now scans all returned addresses if a particular IP
6104 value is being sought. In this situation, the contents of the
6105 $dnslist_value variable are a list of all the addresses, separated by a
6108 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6109 to accept a host name and all its aliases if the forward lookup for any of
6110 them yielded the IP address of the incoming connection. Now it accepts only
6111 those names whose forward lookup yields the correct IP address. Any other
6112 names are discarded. This closes a loophole whereby a rogue DNS
6113 administrator could create reverse DNS records to break through a
6114 wildcarded host restriction in an ACL.
6116 37. If a user filter or a system filter that ran in a subprocess used any of
6117 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6118 the wrong values were passed to the pipe command ($thisaddress had the
6119 value of $0, $0 had the value of $1, etc). This bug was introduced by
6120 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6122 38. Improved the line breaking for long SMTP error messages from ACLs.
6123 Previously, if there was no break point between 40 and 75 characters, Exim
6124 left the rest of the message alone. Two changes have been made: (a) I've
6125 reduced the minimum length to 35 characters; (b) if it can't find a break
6126 point between 35 and 75 characters, it looks ahead and uses the first one
6127 that it finds. This may give the occasional overlong line, but at least the
6128 remaining text gets split now.
6130 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6131 file descriptors might be low, and that setting 1000 would always raise it.
6132 It turns out that in some environments, the limit is already over 1000 and
6133 that lowering it causes trouble. So now Exim takes care not to decrease it.
6135 40. When delivering a message, the value of $return_path is set to $sender_
6136 address at the start of routing (routers may change the value). By an
6137 oversight, this default was not being set up when an address was tested by
6138 -bt or -bv, which affected the outcome if any router or filter referred to
6141 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6142 log only when "message" or "log_message" are set. However, if one of the
6143 conditions was an address verification, or a call to a nested ACL, the
6144 messages generated by the underlying test were being passed through. This
6145 no longer happens. The underlying message is available in $acl_verify_
6146 message for both "message" and "log_message" expansions, so it can be
6147 passed through if needed.
6149 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6150 new expansion $bh_ to give the encoded byte string without charset
6151 translation. Translation happens only if iconv() is available; HAVE_ICONV
6152 indicates this at build time. HEADERS_CHARSET gives the charset to
6153 translate to; headers_charset can change it in the configuration, and
6154 "headers charset" can change it in an individual filter file.
6156 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6157 that creates RFC 2047 encoded "words" labels them as that charset instead
6158 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6159 expansion operator; (ii) when Exim creates a From: line for a local
6160 message; (iii) when a header line is rewritten to include a "phrase" part.
6162 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6163 buggy, causing it to skip the first lines of messages whose message ID
6164 ended in 'D'. This would not have bitten before Exim release 4.14, because
6165 message IDs were unlikely to end in 'D' before then. The effect was to have
6166 incorrect size information for certain domains.
6168 45. #include "config.h" was missing at the start of the crypt16.c module. This
6169 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6172 46. If there was a timeout during a "random" callout check, Exim treated it as
6173 a failure of the random address, and carried on sending RSET and the real
6174 address. If the delay was just some slowness somewhere, the response to the
6175 original RCPT would be taken as a response to RSET and so on, causing
6176 mayhem of various kinds.
6178 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6179 when I implemented it. It didn't allow for the fact that some option values
6180 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6181 the right test for positive values.
6183 48. Domain names in DNS records are case-independent. Exim always looks them up
6184 in lower case. Some resolvers return domain names in exactly the case they
6185 appear in the zone file, that is, they may contain uppercase letters. Not
6186 all resolvers do this - some return always lower case. Exim was treating a
6187 change of case by a resolver as a change of domain, similar to a widening
6188 of a domain abbreviation. This triggered its re-routing code and so it was
6189 trying to route what was effectively the same domain again. This normally
6190 caused routing to fail (because the router wouldn't handle the domain
6191 twice). Now Exim checks for this case specially, and just changes the
6192 casing of the domain that it ultimately uses when it transmits the message
6195 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6198 50. If a filter generated a file delivery with a non-absolute name (possible if
6199 no home directory exists for the router), the forbid_file option was not
6202 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6203 the existing equality matching.
6205 52. Exim was using ints instead of ino_t variables in some places where it was
6206 dealing with inode numbers.
6208 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6209 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6210 called TMPDIR, and if it finds it is different, it changes its value.
6212 54. The smtp_printf() function is now made available to local_scan() so
6213 additional output lines can be written before returning. There is also an
6214 smtp_fflush() function to enable the detection of a dropped connection.
6215 The variables smtp_input and smtp_batched_input are exported to
6218 55. Changed the default runtime configuration: the message "Unknown user"
6219 has been removed from the ACL, and instead placed on the localuser router,
6220 using the cannot_route_message feature. This means that any verification
6221 failures that generate their own messages won't get overridden. Similarly,
6222 the "Unrouteable address" message that was in the ACL for unverifiable
6223 relay addresses has also been removed.
6225 56. Added hosts_avoid_esmtp to the smtp transport.
6227 57. The exicyclog script was not checking for the esoteric option
6228 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6229 will work only if exicyclog is run under the appropriate euid.
6231 58. Following a discussion on the list, the rules by which Exim recognises line
6232 endings on incoming messages have been changed. The -dropcr and drop_cr
6233 options are now no-ops, retained only for backwards compatibility. The
6234 following line terminators are recognized: LF CRLF CR. However, special
6235 processing applies to CR:
6237 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6238 nor a local message in the state where . is a terminator.
6240 (ii) If a bare CR is encountered in a header line, an extra space is added
6241 after the line terminator so as not to end the header. The reasoning
6242 behind this is that bare CRs in header lines are most likely either
6243 to be mistakes, or people trying to play silly games.
6245 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6246 was being incorrectly given as 18 bytes larger than it should have been.
6247 This is a VOB (very old bug).
6249 60. This may never have affected anything current, but just in case it has:
6250 When the local host is found other than at the start of a list of hosts,
6251 the local host, those with the same MX, and any that follow, are discarded.
6252 When the list in question was part of a longer list of hosts, the following
6253 hosts (not currently being processed) were also being discarded. This no
6254 longer happens. I'm not sure if this situation could ever has previously
6257 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6260 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6261 header. This is something that is recommended in a new Internet Draft, and
6262 is something that is documented as being done by Sendmail. There are two
6263 possible values. For messages generated by the autoreply transport, Exim
6266 Auto-Submitted: auto-replied
6268 whereas for all other generated messages (e.g. bounces) it adds
6270 Auto-Submitted: auto-generated
6272 63. The "personal" condition in filters now includes a test for the
6273 Auto-Submitted: header. If it contains the string "auto-" the message it
6274 not considered personal.
6276 64. Added rcpt_include_affixes as a generic transport option.
6278 65. Added queue_only_override (default true).
6280 66. Added the syslog_duplication option.
6282 67. If what should have been the first header line of a message consisted of
6283 a space followed by a colon, Exim was mis-interpreting it as a header line.
6284 It isn't of course - it is syntactically invalid and should therefore be
6285 treated as the start of the message body. The misbehaviour could have
6286 caused a number of strange effects, including loss of data in subsequent
6287 header lines, and spool format errors.
6289 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6290 client host had authenticated. This control can now be exercised by an ACL
6291 for more flexibility.
6293 69. By default, callouts do not happen when testing with -bh. There is now a
6294 variant, -bhc, which does actually run the callout code, including
6295 consulting and updating the callout cache.
6297 70. Added support for saslauthd authentication, courtesy of Alexander
6300 71. If statvfs() failed on the spool or log directories while checking their
6301 size for availability, Exim confusingly gave the error "space shortage".
6302 Furthermore, in debugging mode it crashed with a floating point exception.
6303 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6304 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6305 serious problem, Exim now writes to the main and panic logs when this
6306 happens, with details of the failure. It then refuses to accept the
6307 incoming message, giving the message "spool directory problem" or "log
6308 directory problem" with a 421 code for SMTP messages.
6310 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6311 0, 1, and 2 exist, because some OS complain for execs without them (see
6312 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6313 descriptors. However, the code omitted to check that the open succeeded,
6314 causing mysterious errors if for some reason the permissions on /dev/null
6315 got screwed. Now Exim writes a message to the main and panic logs, and
6316 bombs out if it can't open /dev/null.
6318 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6319 interact so that it is all more flexible. It is supposed to remain
6320 backwards compatible. Also added extra_local_interfaces.
6322 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6323 to bomb out with an assertion failure - to the client this appears as a
6324 connection drop. This problem occurs in the part of the code that was taken
6325 from the Samba project. Fortunately, the assertion is in a very simple
6326 function, so I have fixed this by reproducing the function inline in the
6327 one place where it is called, and arranging for authentication to fail
6328 instead of killing the process with assert().
6330 75. The SPA client code was not working when the server requested OEM rather
6331 than Unicode encoding.
6333 76. Added code to make require_files with a specific uid setting more usable in
6334 the case where statting the file as root fails - usually a non-root-mounted
6335 NFS file system. When this happens and the failure is EACCES, Exim now
6336 forks a subprocess and does the per-uid checking as the relevant uid.
6338 77. Added process_log_path.
6340 78. If log_file_path was not explicitly set, a setting of check_log_space or
6341 check_log_inodes was ignored.
6343 79. If a space check for the spool or log partitions fails, the incident is now
6344 logged. Of course, in the latter case the data may get lost...
6346 80. Added the %p formatting code to string_format() so that it can be used to
6347 print addresses in debug_print(). Adjusted all the address printing in the
6348 debugging in store.c to use %p rather than %d.
6350 81. There was a concern that a line of code in smtp_in.c could overflow a
6351 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6352 initially expressed, the concern was not well-founded, because trailing
6353 spaces are removed early. However, if the trailing spaces were followed by
6354 a NULL, they did not get removed, so the overflow was possible. Two fixes
6357 (a) I re-wrote the offending code in a cleaner fashion.
6358 (b) If an incoming SMTP command contains a NULL character, it is rejected
6361 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6362 runs initgroups(), so that if the Exim user is in any additional groups,
6363 they will be used during message reception.
6369 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob