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.
69 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
72 JH/02 The smtp transport option "multi_domain" is now expanded.
74 JH/03 The smtp transport now requests PRDR by default, if the server offers
77 JH/04 Certificate name checking on server certificates, when exim is a client,
78 is now done by default. The transport option tls_verify_cert_hostnames
79 can be used to disable this per-host. The build option
80 EXPERIMENTAL_CERTNAMES is withdrawn.
82 JH/05 The value of the tls_verify_certificates smtp transport and main options
83 default to the word "system" to access the system default CA bundle.
84 For GnuTLS, only version 3.0.20 or later.
86 JH/06 Verification of the server certificate for a TLS connection is now tried
87 (but not required) by default. The verification status is now logged by
88 default, for both outbound TLS and client-certificate supplying inbound
91 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
94 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
95 Status Notification (bounce) messages are now MIME format per RFC 3464.
96 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
97 under the control of the dsn_advertise_hosts option, and routers may
98 have a dsn_lasthop option.
100 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
101 default, modifiable by a malware= option. The list separator for
102 the options can now be changed in the usual way. Bug 68.
104 JH/10 The smtp_receive_timeout main option is now expanded before use.
106 JH/11 The incoming_interface log option now also enables logging of the
107 local interface on delivery outgoing connections.
109 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
110 if the interface and destination host and port all match.
112 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
115 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
116 Patch from Andrew Lewis.
118 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
119 now supports optional time-restrictions, weighting, and priority
120 modifiers per server. Patch originally by <rommer@active.by>.
122 JH/16 The spamd_address main option now supports a mixed list of local
123 and remote servers. Remote servers can be IPv6 addresses, and
124 specify a port-range.
126 JH/17 Bug 68: The spamd_address main option now supports an optional
127 timeout value per server.
129 JH/18 Bug 1581: Router and transport options headers_add/remove can
130 now have the list separator specified.
132 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
135 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
138 JH/21 Support for the A6 type of dns record is withdrawn.
140 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
141 rather than the verbs used.
143 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
144 from 255 to 1024 chars.
146 JH/24 Verification callouts now attempt to use TLS by default.
148 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
149 are generic router options now. The defaults didn't change.
151 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
152 Original patch from Alexander Shikoff, worked over by JH.
154 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
155 files if ZCAT_COMMAND is not executable.
157 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
159 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
161 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
162 Normally benign, it bites when the pair was led to by a CNAME;
163 modern usage is to not canoicalize the domain to a CNAME target
164 (and we were inconsistent anyway for A-only vs AAAA+A).
166 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
168 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
169 when evaluating $sender_host_dnssec.
171 JH/31 Check the HELO verification lookup for DNSSEC, adding new
172 $sender_helo_dnssec variable.
174 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
176 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
178 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
180 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
181 documented as working, but never had. Support all but $spam_report.
183 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
184 added for tls authenticator.
189 TL/01 When running the test suite, the README says that variables such as
190 no_msglog_check are global and can be placed anywhere in a specific
191 test's script, however it was observed that placement needed to be near
192 the beginning for it to behave that way. Changed the runtest perl
193 script to read through the entire script once to detect and set these
194 variables, reset to the beginning of the script, and then run through
195 the script parsing/test process like normal.
197 TL/02 The BSD's have an arc4random API. One of the functions to induce
198 adding randomness was arc4random_stir(), but it has been removed in
199 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
200 function when detected.
202 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
203 cause callback expansion.
205 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
206 syntax errors in an expansion can be treated as a string instead of
207 logging or causing an error, due to the internal use of bool_lax
208 instead of bool when processing it.
210 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
211 server certificates when making smtp deliveries.
213 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
215 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
217 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
219 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
220 Merged patch from Sebastian Wiedenroth.
222 JH/05 Fix results-pipe from transport process. Several recipients, combined
223 with certificate use, exposed issues where response data items split
224 over buffer boundaries were not parsed properly. This eventually
225 resulted in duplicates being sent. This issue only became common enough
226 to notice due to the introduction of conection certificate information,
227 the item size being so much larger. Found and fixed by Wolfgang Breyha.
229 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
230 size buffer was used, resulting in syntax errors when an expansion
233 JH/07 Add support for directories of certificates when compiled with a GnuTLS
234 version 3.3.6 or later.
236 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
237 is EXPERIMENTAL_EVENT, the main-configuration and transport options
238 both become "event_action", the variables become $event_name, $event_data
239 and $event_defer_errno. There is a new variable $verify_mode, usable in
240 routers, transports and related events. The tls:cert event is now also
241 raised for inbound connections, if the main configuration event_action
244 TL/06 In test suite, disable OCSP for old versions of openssl which contained
245 early OCSP support, but no stapling (appears to be less than 1.0.0).
247 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
248 server certificate names available under the smtp transport option
249 "tls_verify_cert_hostname" now do not permit multi-component wildcard
252 JH/10 Time-related extraction expansions from certificates now use the main
253 option "timezone" setting for output formatting, and are consistent
254 between OpenSSL and GnuTLS compilations. Bug 1541.
256 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
257 encoded parameter in the incoming message. Bug 1558.
259 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
260 include certificate info, eximon was claiming there were spoolfile
263 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
265 JH/14 Log delivery-related information more consistently, using the sequence
266 "H=<name> [<ip>]" wherever possible.
268 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
269 are problematic for Debian distribution, omit them from the release
272 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
274 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
277 JH/17 Fix a null-indirection in certextract expansions when a nondefault
278 output list separator was used.
283 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
284 checkers that were complaining about end of non-void function with no
287 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
288 This was a regression intruduced in 4.83 by another bugfix.
290 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
292 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
293 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
299 TF/01 Correctly close the server side of TLS when forking for delivery.
301 When a message was received over SMTP with TLS, Exim failed to clear up
302 the incoming connection properly after forking off the child process to
303 deliver the message. In some situations the subsequent outgoing
304 delivery connection happened to have the same fd number as the incoming
305 connection previously had. Exim would try to use TLS and fail, logging
306 a "Bad file descriptor" error.
308 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
309 utilities have not been installed.
311 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
312 temporary space as the ACL may create new global variables.
314 TL/01 LDAP support uses per connection or global context settings, depending
315 upon the detected version of the libraries at build time.
317 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
318 to extract and use the src ip:port in logging and expansions as if it
319 were a direct connection from the outside internet. PPv2 support was
320 updated based on HAProxy spec change in May 2014.
322 JH/02 Add ${listextract {number}{list}{success}{fail}}.
324 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
325 Properly escape header and check for NULL return.
327 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
330 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
332 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
333 characters in header names, implemented as a verify condition.
334 Contributed by Michael Fischer v. Mollard.
336 TL/05 Rename SPF condition results err_perm and err_temp to standardized
337 results permerror and temperror. Previous values are deprecated but
338 still accepted. In a future release, err_perm and err_temp will be
339 completely removed, which will be a backward incompatibility if the
340 ACL tests for either of these two old results. Patch contributed by
341 user bes-internal on the mailing list.
343 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
345 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
346 selectors, in both main and reject logs.
348 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
351 JH/07 Add malware type "sock" for talking to simple daemon.
353 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
355 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
356 routers/transports under cutthrough routing.
358 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
359 numbers. Touch up "bool" conditional to keep the same definition.
361 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
363 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
365 JH/12 Expand items in router/transport headers_add or headers_remove lists
366 individually rather than the list as a whole. Bug 1452.
368 Required for reasonable handling of multiple headers_ options when
369 they may be empty; requires that headers_remove items with embedded
370 colons must have them doubled (or the list-separator changed).
372 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
373 view the policy declared in the DMARC record. Currently, $dmarc_status
374 is a combined value of both the record presence and the result of the
377 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
379 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
380 dnslookup router and the smtp transport (applying to the forward
383 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
384 of ldap servers used for a specific lookup. Patch provided by Heiko
387 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
388 New variable $lookup_dnssec_authenticated for observability.
390 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
391 Patch submitted by Lars Timman.
393 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
395 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
396 Requires trusted mode and valid format message id, aborts otherwise.
397 Patch contributed by Heiko Schlichting.
399 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
400 certextract with support for various fields. Bug 1358.
402 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
403 is requested by default, modifiable by smtp transport option
406 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
407 operate on certificate variables to give certificate fingerprints
408 Also new ${sha256:cert_variable}.
410 JH/23 The PRDR feature is moved from being Experimental into the mainline.
412 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
413 Christian Aistleitner.
415 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
417 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
418 file. Patch from Wolfgang Breyha.
420 JH/25 Expand the coverage of the delivery $host and $host_address to
421 client authenticators run in verify callout. Bug 1476.
423 JH/26 Port service names are now accepted for tls_on_connect_ports, to
424 align with daemon_smtp_ports. Bug 72.
426 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
427 support and error reporting did not work properly.
429 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
430 and is readable. Patch from Andrew Colin Kissa.
432 TL/14 Enhance documentation of ${run expansion and how it parses the
433 commandline after expansion, particularly in the case when an
434 unquoted variable expansion results in an empty value.
436 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
438 PP/02 Fix internal collision of T_APL on systems which support RFC3123
439 by renaming away from it. Addresses GH issue 15, reported by
442 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
444 TL/15 SECURITY: prevent double expansion in math comparison functions
445 (can expand unsanitized data). Not remotely exploitable.
452 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
454 PP/02 Make -n do something, by making it not do something.
455 When combined with -bP, the name of an option is not output.
457 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
460 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
461 $sender_host_name and config options to manage this, and basic check
464 PP/05 DSCP support for outbound connections and control modifier for inbound.
466 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
467 (Only plugin which currently uses this is kerberos4, which nobody should
468 be using, but we should make it available and other future plugins might
469 conceivably use it, even though it would break NAT; stuff *should* be
470 using channel bindings instead).
472 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
473 name; added for Sendmail compatibility; requires admin caller.
474 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
475 just ignore it); requires trusted caller.
476 Also parse but ignore: -Ac -Am -X<logfile>
479 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
481 TL/02 Add +smtp_confirmation as a default logging option.
483 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
484 Patch by Magnus Holmgren from 2007-02-20.
486 TL/04 Bugzilla 1281 - Spec typo.
487 Bugzilla 1283 - Spec typo.
488 Bugzilla 1290 - Spec grammar fixes.
490 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
492 TL/06 Add Experimental DMARC support using libopendmarc libraries.
494 TL/07 Fix an out of order global option causing a segfault. Reported to dev
495 mailing list by by Dmitry Isaikin.
497 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
499 JH/02 Support "G" suffix to numbers in ${if comparisons.
501 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
503 NM/01 Bugzilla 1197 - Spec typo
504 Bugzilla 1196 - Spec examples corrections
506 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
508 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
509 gnutls_enable_pkcs11, but renamed to more accurately indicate its
512 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
513 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
515 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
516 "acl {{name}{arg}...}", and optional args on acl condition
519 JH/05 Permit multiple router/transport headers_add/remove lines.
521 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
523 JH/07 Avoid using a waiting database for a single-message-only transport.
524 Performance patch from Paul Fisher. Bugzilla 1262.
526 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
529 JH/09 Add $headers_added variable, with content from use of ACL modifier
530 add_header (but not yet added to the message). Bugzilla 199.
532 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
533 Pulled from Bugzilla 817 by Wolfgang Breyha.
535 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
537 (nb: this is the same fix as in Exim 4.80.1)
539 JH/11 Add A= logging on delivery lines, and a client_set_id option on
542 JH/12 Add optional authenticated_sender logging to A= and a log_selector
545 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
547 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
548 advertise SMTP AUTH mechanism to us, instead of a generic
549 protocol violation error. Also, make Exim more robust to bad
550 data from the Dovecot auth socket.
552 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
554 When a queue runner is handling a message, Exim first routes the
555 recipient addresses, during which it prunes them based on the retry
556 hints database. After that it attempts to deliver the message to
557 any remaining recipients. It then updates the hints database using
560 So if a recipient address works intermittently, it can get repeatedly
561 deferred at routing time. The retry hints record remains fresh so the
562 address never reaches the final cutoff time.
564 This is a fairly common occurrence when a user is bumping up against
565 their storage quota. Exim had some logic in its local delivery code
566 to deal with this. However it did not apply to per-recipient defers
567 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
569 This change adds a proper retry rule check during routing so that the
570 final cutoff time is checked against the message's age. We only do
571 this check if there is an address retry record and there is not a
572 domain retry record; this implies that previous attempts to handle
573 the address had the retry_use_local_parts option turned on. We use
574 this as an approximation for the destination being like a local
575 delivery, as in LMTP.
577 I suspect this new check makes the old local delivery cutoff check
578 redundant, but I have not verified this so I left the code in place.
580 TF/02 Correct gecos expansion when From: is a prefix of the username.
582 Test 0254 submits a message to Exim with the header
586 When I ran the test suite under the user fanf2, Exim expanded
587 the header to contain my full name, whereas it should have added
588 a Resent-Sender: header. It erroneously treats any prefix of the
589 username as equal to the username.
591 This change corrects that bug.
593 GF/01 DCC debug and logging tidyup
594 Error conditions log to paniclog rather than rejectlog.
595 Debug lines prefixed by "DCC: " to remove any ambiguity.
597 TF/03 Avoid unnecessary rebuilds of lookup-related code.
599 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
600 Bug spotted by Jeremy Harris; was flawed since initial commit.
601 Would have resulted in OCSP responses post-SNI triggering an Exim
602 NULL dereference and crash.
604 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
606 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
607 Bug detection, analysis and fix by Samuel Thibault.
608 Bugzilla 1331, Debian bug #698092.
610 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
612 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
613 Server implementation by Todd Lyons, client by JH.
614 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
615 config variable "prdr_enable" controls whether the server
616 advertises the facility. If the client requests PRDR a new
617 acl_data_smtp_prdr ACL is called once for each recipient, after
618 the body content is received and before the acl_smtp_data ACL.
619 The client is controlled by bolth of: a hosts_try_prdr option
620 on the smtp transport, and the server advertisement.
621 Default client logging of deliveries and rejections involving
622 PRDR are flagged with the string "PRDR".
624 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
625 fclose(). Diagnosis by Todd Lyons.
627 PP/17 Update configure.default to handle IPv6 localhost better.
628 Patch by Alain Williams (plus minor tweaks).
631 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
632 This is now consistent with GnuTLS, and is now documented: the
633 previous undocumented portable approach to treating the option as
634 unset was to force an expansion failure. That still works, and
635 an empty string is now equivalent.
637 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
638 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
639 not performing validation itself.
641 PP/20 Added force_command boolean option to pipe transport.
642 Patch from Nick Koston, of cPanel Inc.
644 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
647 TF/04 Added udpsend ACL modifer and hexquote expansion operator
649 PP/21 Fix eximon continuous updating with timestamped log-files.
650 Broken in a format-string cleanup in 4.80, missed when I repaired the
651 other false fix of the same issue.
652 Report and fix from Heiko Schlichting.
655 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
656 Report from Prashanth Katuri.
658 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
659 It's SecureTransport, so affects any MacOS clients which use the
660 system-integrated TLS libraries, including email clients.
662 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
663 using a MIME ACL for non-SMTP local injection.
664 Report and assistance in diagnosis by Warren Baker.
666 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
668 JH/16 Fix comparisons for 64b. Bugzilla 1385.
670 TL/09 Add expansion variable $authenticated_fail_id to keep track of
671 last id that failed so it may be referenced in subsequent ACL's.
673 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
676 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
677 ldap library initialization, allowing self-signed CA's to be
678 used. Also properly sets require_cert option later in code by
679 using NULL (global ldap config) instead of ldap handle (per
680 session). Bug diagnosis and testing by alxgomz.
682 TL/12 Enhanced documentation in the ratelimit.pl script provided in
683 the src/util/ subdirectory.
685 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
686 renamed to Transport Post Delivery Action by Jeremy Harris, as
689 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
690 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
691 redis_servers = needs to be configured which will be used by the redis
692 lookup. Patch from Warren Baker, of The Packet Hub.
694 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
696 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
697 hostname or reverse DNS when processing a host list. Used suggestions
698 from multiple comments on this bug.
700 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
702 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
703 Missed a few lines, added it to make the runtest require no keyboard
706 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
707 contains upper case chars. Make router use caseful_local_part.
709 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
710 support when GnuTLS has been built with p11-kit.
716 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
718 This, or similar/improved, will also be change PP/11 of 4.82.
724 PP/01 Handle short writes when writing local log-files.
725 In practice, only affects FreeBSD (8 onwards).
726 Bugzilla 1053, with thanks to Dmitry Isaikin.
728 NM/01 Bugzilla 949 - Documentation tweak
730 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
733 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
735 PP/02 Implemented gsasl authenticator.
737 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
739 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
740 `pkg-config foo` for cflags/libs.
742 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
743 with rest of GSASL and with heimdal_gssapi.
745 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
746 `pkg-config foo` for cflags/libs for the TLS implementation.
748 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
749 properties get this fed in as external SSF. A number of robustness
750 and debugging improvements to the cyrus_sasl authenticator.
752 PP/08 cyrus_sasl server now expands the server_realm option.
754 PP/09 Bugzilla 1214 - Log authentication information in reject log.
755 Patch by Jeremy Harris.
757 PP/10 Added dbmjz lookup type.
759 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
761 PP/12 MAIL args handles TAB as well as SP, for better interop with
762 non-compliant senders.
763 Analysis and variant patch by Todd Lyons.
765 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
766 Bug report from Lars Müller <lars@samba.org> (via SUSE),
767 Patch from Dirk Mueller <dmueller@suse.com>
769 PP/13 tls_peerdn now print-escaped for spool files.
770 Observed some $tls_peerdn in wild which contained \n, which resulted
771 in spool file corruption.
773 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
774 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
775 or write after TLS renegotiation, which otherwise led to messages
778 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
779 as a tracking header (ie: a signed header comes before the signature).
780 Patch from Wolfgang Breyha.
782 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
783 comma-sep list; embedded commas doubled.
785 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
787 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
789 Report and patch from Dmitry Banschikov.
791 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
792 Removed SSL_clear() after SSL_new() which led to protocol negotiation
793 failures. We appear to now support TLS1.1+ with Exim.
795 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
796 lets Exim select keys and certificates based upon TLS SNI from client.
797 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
798 before an outbound SMTP session. New log_selector, +tls_sni.
800 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
801 NULL dereference. Report and patch from Alun Jones.
803 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
804 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
805 Not seeing resolver debug output on NetBSD, but suspect this is a
806 resolver implementation change.
808 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
809 Left warnings. Added "eximon gdb" invocation mode.
811 PP/21 Defaulting "accept_8bitmime" to true, not false.
813 PP/22 Added -bw for inetd wait mode support.
815 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
816 locate the relevant includes and libraries. Made this the default.
818 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
819 Bugzilla 1246, report and most of solution from Tomasz Kusy.
821 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
822 This may cause build issues on older platforms.
824 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
825 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
826 gnutls_require_mac & gnutls_require_protocols (no longer supported).
827 Added SNI support via GnuTLS too.
828 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
830 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
832 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
833 Applied second patch from Janne, implementing suggestion to default
834 multiple-strings-in-record handling to match SPF spec.
836 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
838 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
839 read-only, out of scope).
840 Patch from Wolfgang Breyha, report from Stuart Northfield.
842 PP/29 Fix three issues highlighted by clang analyser static analysis.
843 Only crash-plausible issue would require the Cambridge-specific
844 iplookup router and a misconfiguration.
845 Report from Marcin Mirosław.
847 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
849 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
850 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
851 As part of this, removing so much warning spew let me fix some minor
852 real issues in debug logging.
854 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
855 assignment on my part. Fixed.
857 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
858 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
859 Janne Snabb (who went above and beyond: thank you).
861 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
862 string otherwise requires a connection and a bunch more work and it's
863 relatively easy to get wrong. Should also expose TLS library linkage
866 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
867 64-bit ${eval} (JH/03).
869 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
870 GNU libc to support some of the 64-bit stuff, should not lead to
871 conflicts. Defined before os.h is pulled in, so if a given platform
872 needs to override this, it can.
874 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
875 protection layer was required, which is not implemented.
876 Bugzilla 1254, patch from Wolfgang Breyha.
878 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
879 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
880 tls_dhparam take prime identifiers. Also unbreak combination of
881 OpenSSL+DH_params+TLSSNI.
883 PP/39 Disable SSLv2 by default in OpenSSL support.
889 PP/01 Solaris build fix for Oracle's LDAP libraries.
890 Bugzilla 1109, patch from Stephen Usher.
892 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
894 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
897 TF/02 Fix a couple more cases where we did not log the error message
898 when unlink() failed. See also change 4.74-TF/03.
900 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
901 lock up or crash if it happened to be inside a call to libc when it
902 got a SIGUSR1 from exiwhat.
904 The SIGUSR1 handler appends the current process status to the process
905 log which is later printed by exiwhat. It used to use the general
906 purpose logging code to do this, but several functions it calls are
907 not safe for signals.
909 The new output code in the SIGUSR1 handler is specific to the process
910 log, and simple enough that it's easy to inspect for signal safety.
911 Removing some special cases also simplifies the general logging code.
912 Removing the spurious timestamps from the process log simplifies
915 TF/04 Improved ratelimit ACL condition.
917 The /noupdate option has been deprecated in favour of /readonly which
918 has clearer semantics. The /leaky, /strict, and /readonly update modes
919 are mutually exclusive. The update mode is no longer included in the
920 database key; it just determines when the database is updated. (This
921 means that when you upgrde Exim will forget old rate measurements.)
923 Exim now checks that the per_* options are used with an update mode that
924 makes sense for the current ACL. For example, when Exim is processing a
925 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
926 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
927 must specify per_mail/readonly. If you omit the update mode it defaults to
928 /leaky where that makes sense (as before) or /readonly where required.
930 The /noupdate option is now undocumented but still supported for
931 backwards compatibility. It is equivalent to /readonly except that in
932 ACLs where /readonly is required you may specify /leaky/noupdate or
933 /strict/noupdate which are treated the same as /readonly.
935 A useful new feature is the /count= option. This is a generalization
936 of the per_byte option, so that you can measure the throughput of other
937 aggregate values. For example, the per_byte option is now equivalent
938 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
940 The per_rcpt option has been generalized using the /count= mechanism
941 (though it's more complicated than the per_byte equivalence). When it is
942 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
943 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
944 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
945 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
946 Note that using per_rcpt with a non-readonly update mode in more than
947 one ACL will cause the recipients to be double-counted. (The per_mail
948 and per_byte options don't have this problem.)
950 The handling of very low rates has changed slightly. If the computed rate
951 is less than the event's count (usually one) then this event is the first
952 after a long gap. In this case the rate is set to the same as this event's
953 count, so that the first message of a spam run is counted properly.
955 The major new feature is a mechanism for counting the rate of unique
956 events. The new per_addr option counts the number of different
957 recipients that someone has sent messages to in the last time period. It
958 behaves like per_rcpt if all the recipient addresses are different, but
959 duplicate recipient addresses do not increase the measured rate. Like
960 the /count= option this is a general mechanism, so the per_addr option
961 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
962 example, measure the rate that a client uses different sender addresses
963 with the options per_mail/unique=$sender_address. There are further
964 details in the main documentation.
966 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
968 TF/06 Removed a few PCRE remnants.
970 TF/07 Automatically extract Exim's version number from tags in the git
971 repository when doing development or release builds.
973 PP/02 Raise smtp_cmd_buffer_size to 16kB.
974 Bugzilla 879. Patch from Paul Fisher.
976 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
977 Heavily based on revision 40f9a89a from Simon Arlott's tree.
980 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
982 PP/05 Variable $av_failed, true if the AV scanner deferred.
983 Bugzilla 1078. Patch from John Horne.
985 PP/06 Stop make process more reliably on build failure.
986 Bugzilla 1087. Patch from Heiko Schlittermann.
988 PP/07 Make maildir_use_size_file an _expandable_ boolean.
989 Bugzilla 1089. Patch from Heiko Schlittermann.
991 PP/08 Handle ${run} returning more data than OS pipe buffer size.
992 Bugzilla 1131. Patch from Holger Weiß.
994 PP/09 Handle IPv6 addresses with SPF.
995 Bugzilla 860. Patch from Wolfgang Breyha.
997 PP/10 GnuTLS: support TLS 1.2 & 1.1.
999 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1002 PP/11 match_* no longer expand right-hand-side by default.
1003 New compile-time build option, EXPAND_LISTMATCH_RHS.
1004 New expansion conditions, "inlist", "inlisti".
1006 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1008 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1010 PP/14 fix log_write() format string regression from TF/03.
1011 Bugzilla 1152. Patch from Dmitry Isaikin.
1017 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1019 PP/02 Harmonised TLS library version reporting; only show if debugging.
1020 Layout now matches that introduced for other libraries in 4.74 PP/03.
1022 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1024 PP/04 New "dns_use_edns0" global option.
1026 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1029 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1030 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1032 TK/01 Updated PolarSSL code to 0.14.2.
1033 Bugzilla 1097. Patch from Andreas Metzler.
1035 PP/07 Catch divide-by-zero in ${eval:...}.
1036 Fixes bugzilla 1102.
1038 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1041 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1042 format-string attack -- SECURITY: remote arbitrary code execution.
1044 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1045 time unintentionally subject to list matching rules, letting the header
1046 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1047 arbitrary string expansion). This allowed for information disclosure.
1049 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1050 INT_MIN/-1 -- value coerced to INT_MAX.
1056 NM/01 Workround for PCRE version dependency in version reporting
1059 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1060 This fixes portability to compilers other than gcc, notably
1061 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1063 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1064 makefiles for portability to HP-UX and POSIX correctness.
1066 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1067 Also via indented variable definition in the Makefile.
1068 (Debugging by Oliver Heesakkers).
1070 PP/02 Restore caching of spamd results with expanded spamd_address.
1071 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1073 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1074 Improves build reliability. Fix from: Frank Elsner
1076 NM/02 Fix wide character breakage in the rfc2047 coding
1077 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1079 NM/03 Allow underscore in dnslist lookups
1080 Fixes bug 1026. Patch from Graeme Fowler
1082 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1083 Code patches from Adam Ciarcinski of NetBSD.
1085 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1088 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1089 is logged, to avoid truncation. Patch from John Horne.
1091 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1092 Patch from Jakob Hirsch.
1094 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1095 SQL string expansion failure details.
1096 Patch from Andrey Oktyabrski.
1098 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1099 Patch from Simon Arlott.
1101 PP/09 New lookups functionality failed to compile on old gcc which rejects
1102 extern declarations in function scope.
1103 Patch from Oliver Fleischmann
1105 PP/10 Use sig_atomic_t for flags set from signal handlers.
1106 Check getgroups() return and improve debugging.
1107 Fixed developed for diagnosis in bug 927 (which turned out to be
1110 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1111 Patch from Mark Zealey.
1113 PP/12 Bugzilla 1056: Improved spamd server selection.
1114 Patch from Mark Zealey.
1116 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1117 Based on patch from Heiko Schlittermann.
1119 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1120 Patch from Uwe Doering, sign-off by Michael Haardt.
1122 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1123 variable declaration deep within a block. Bug and patch from
1126 PP/15 lookups-Makefile IRIX compatibilty coercion.
1128 PP/16 Make DISABLE_DKIM build knob functional.
1130 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1131 Patch by Simon Arlott
1133 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1134 variable argument macros. Our copy now differs from upstream.
1140 TF/01 Failure to get a lock on a hints database can have serious
1141 consequences so log it to the panic log.
1143 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1144 controlled using the smtp_confirmation log selector.
1146 TF/03 Include the error message when we fail to unlink a spool file.
1148 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1149 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1150 for maintaining out-of-tree patches for some time.
1152 PP/01 Bugzilla 139: Documentation and portability issues.
1153 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1154 Handle per-OS dynamic-module compilation flags.
1156 PP/02 Let /dev/null have normal permissions.
1157 The 4.73 fixes were a little too stringent and complained about the
1158 permissions on /dev/null. Exempt it from some checks.
1159 Reported by Andreas M. Kirchwitz.
1161 PP/03 Report version information for many libraries, including
1162 Exim version information for dynamically loaded libraries. Created
1163 version.h, now support a version extension string for distributors
1164 who patch heavily. Dynamic module ABI change.
1166 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1167 privilege escalation vulnerability whereby the Exim run-time user
1168 can cause root to append content of the attacker's choosing to
1171 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1174 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1175 If dropping privileges for untrusted macros, we disabled normal logging
1176 on the basis that it would fail; for the Exim run-time user, this is not
1177 the case, and it resulted in successful deliveries going unlogged.
1178 Fixed. Reported by Andreas Metzler.
1184 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1185 only prepend for the Resent-* case. Fixes regression introduced in
1186 Exim 4.70 by NM/22 for Bugzilla 607.
1188 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1189 increasing numbers of administrators be bitten by this.
1191 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1193 PP/03 Bugzilla 994: added openssl_options main configuration option.
1195 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1197 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1199 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1201 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1202 without a peer certificate, leading to a segfault because of an
1203 assumption that peers always have certificates. Be a little more
1204 paranoid. Problem reported by Martin Tscholak.
1206 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1207 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1208 NB: ClamAV planning to remove STREAM in "middle of 2010".
1209 CL also introduces -bmalware, various -d+acl logging additions and
1210 more caution in buffer sizes.
1212 PP/09 Implemented reverse_ip expansion operator.
1214 PP/10 Bugzilla 937: provide a "debug" ACL control.
1216 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1218 PP/12 Bugzilla 973: Implement --version.
1220 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1222 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1224 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1226 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1227 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1228 ignore trailing whitespace.
1230 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1232 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1235 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1236 Notification from Dr Andrew Aitchison.
1238 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1239 ExtendedDetectionInfo response format.
1240 Notification from John Horne.
1242 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1245 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1246 XSL and documented dependency on system catalogs, with examples of how
1249 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1252 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1253 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1254 configuration file which is writeable by the Exim user or group.
1256 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1257 of configuration files to cover files specified with the -C option if
1258 they are going to be used with root privileges, not just the default
1261 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1262 option (effectively making it always true).
1264 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1265 files to be used while preserving root privileges.
1267 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1268 that rogue child processes cannot use them.
1270 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1271 run-time user, instead of root.
1273 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1274 Exim run-time user without dropping privileges.
1276 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1277 result string, instead of calling string_vformat() twice with the same
1280 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1281 for other users. Others should always drop root privileges if they use
1282 -C on the command line, even for a whitelisted configure file.
1284 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1286 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1292 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1293 $data_path, and $header_path variables; fixed documentation bugs and
1296 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1297 exipick to access non-standard spools, including the "frozen" queue
1300 NM/01 Bugzilla 965: Support mysql stored procedures.
1301 Patch from Alain Williams
1303 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1305 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1306 Patch from Andreas Metzler
1308 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1309 Patch from Kirill Miazine
1311 NM/05 Bugzilla 671: Added umask to procmail example.
1313 JJ/03 installed exipick 20100323.0, fixing doc bug
1315 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1316 directory. Notification and patch from Dan Rosenberg.
1318 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1320 TK/02 Improve log output when DKIM signing operation fails.
1322 MH/01 Treat the transport option dkim_domain as a colon separated
1323 list, not as a single string, and sign the message with each element,
1324 omitting multiple occurences of the same signer.
1326 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1327 Bugzilla 985, 986. Patch by Simon Arlott
1329 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1330 Patch by Simon Arlott
1332 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1333 MBX locking. Notification from Dan Rosenberg.
1339 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1341 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1343 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1345 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1347 NM/04 Bugzilla 915: Flag broken perl installation during build.
1353 TK/01 Added patch by Johannes Berg that expands the main option
1354 "spamd_address" if it starts with a dollar sign.
1356 TK/02 Write list of recipients to X-Envelope-Sender header when building
1357 the mbox-format spool file for content scanning (suggested by Jakob
1360 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1361 (http://www.dcc-servers.net/) support via dccifd. Activated by
1362 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1364 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1365 by Mark Daniel Reidel <mr@df.eu>.
1367 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1368 When building exim an external PCRE library is now needed -
1369 PCRE is a system library on the majority of modern systems.
1370 See entry on PCRE_LIBS in EDITME file.
1372 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1373 conversation. Added nologin parameter to request.
1374 Patch contributed by Kirill Miazine.
1376 TF/01 Do not log submission mode rewrites if they do not change the address.
1378 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1380 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1381 log files in place. Contributed by Roberto Lima.
1383 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1385 TF/03 Bugzilla 615: When checking the local_parts router precondition
1386 after a local_part_suffix or local_part_prefix option, Exim now
1387 does not use the address's named list lookup cache, since this
1388 contains cached lookups for the whole local part.
1390 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1391 Robert Millan. Documentation is in experimental-spec.txt.
1393 TF/04 Bugzilla 668: Fix parallel build (make -j).
1395 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1397 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1398 Patch provided by Jan Srzednicki.
1400 TF/05 Leading white space used to be stripped from $spam_report which
1401 wrecked the formatting. Now it is preserved.
1403 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1404 that they are available at delivery time.
1406 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1408 TF/08 TLS error reporting now respects the incoming_interface and
1409 incoming_port log selectors.
1411 TF/09 Produce a more useful error message if an SMTP transport's hosts
1412 setting expands to an empty string.
1414 NM/06 Bugzilla 744: EXPN did not work under TLS.
1415 Patch provided by Phil Pennock.
1417 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1418 Patch provided by Richard Godbee.
1420 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1421 acl_smtp_notquit, added index entry.
1423 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1424 Patch provided by Eugene Bujak.
1426 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1427 accept(). Patch provided by Maxim Dounin.
1429 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1430 Patch provided by Phil Pennock.
1432 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1434 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1435 Patch provided by Brad "anomie" Jorsch.
1437 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1438 Patch provided by Dean Brooks.
1440 TK/05 Add native DKIM support (does not depend on external libraries).
1442 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1443 Patch provided by Graeme Fowler.
1445 NM/16 Bugzilla 851: Documentation example syntax fix.
1447 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1449 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1452 NM/19 Bugzilla 745: TLS version reporting.
1453 Patch provided by Phil Pennock.
1455 NM/20 Bugzilla 167: bool: condition support.
1456 Patch provided by Phil Pennock.
1458 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1459 clients. Patch provided by Phil Pennock.
1461 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1462 Patch provided by Brad "anomie" Jorsch.
1464 NM/23 Bugzilla 687: Fix misparses in eximstats.
1465 Patch provided by Heiko Schlittermann.
1467 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1468 Patch provided by Heiko Schlittermann.
1470 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1471 plus update to original patch.
1473 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1475 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1476 Patch provided by David Brownlee.
1478 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1480 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1482 NM/30 Bugzilla 888: TLS documentation bugfixes.
1484 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1486 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1487 Unlike the original bugzilla I have changed all shell scripts in src tree.
1489 NM/33 Bugzilla 898: Transport filter timeout fix.
1490 Patch by Todd Rinaldo.
1492 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1493 Patch by Serge Demonchaux.
1495 NM/35 Bugzilla 39: Base64 decode bug fixes.
1496 Patch by Jakob Hirsch.
1498 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1500 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1502 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1504 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1510 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1511 ALT-N's libdkim that I have put here:
1512 http://duncanthrax.net/exim-experimental/
1514 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1515 were called 'true' and it seems that C99 defines that as a reserved
1516 keyword to be used with 'bool' variable types. That means you could
1517 not include C99-style headers which use bools without triggering
1518 build errors in sieve.c.
1520 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1521 as mailq or other aliases. Changed the --help handling significantly
1522 to do whats expected. exim_usage() emits usage/help information.
1524 SC/01 Added the -bylocaldomain option to eximstats.
1526 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1528 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1530 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1536 PH/01 Another patch from the Sieve maintainer.
1538 PH/02 When an IPv6 address is converted to a string for single-key lookup
1539 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1540 dots are used instead of colons so that keys in lsearch files need not
1541 contain colons. This was done some time before quoting was made available
1542 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1543 (notated using the quote facility) so as to distinguish them from IPv4
1544 keys. This meant that lookups for IP addresses in host lists did not work
1545 for iplsearch lookups.
1547 This has been fixed by arranging for IPv6 addresses to be expressed with
1548 colons if the lookup type is iplsearch. This is not incompatible, because
1549 previously such lookups could never work.
1551 The situation is now rather anomolous, since one *can* have colons in
1552 ordinary lsearch keys. However, making the change in all cases is
1553 incompatible and would probably break a number of configurations.
1555 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1558 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1559 conversion specification without a maximum field width, thereby enabling
1560 a rogue spamd server to cause a buffer overflow. While nobody in their
1561 right mind would setup Exim to query an untrusted spamd server, an
1562 attacker that gains access to a server running spamd could potentially
1563 exploit this vulnerability to run arbitrary code as the Exim user.
1565 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1566 $primary_hostname instead of what libspf2 thinks the hosts name is.
1568 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1569 a directory entry by the name of the lookup key. Previously, if a
1570 symlink pointed to a non-existing file or a file in a directory that
1571 Exim lacked permissions to read, a lookup for a key matching that
1572 symlink would fail. Now it is enough that a matching directory entry
1573 exists, symlink or not. (Bugzilla 503.)
1575 PH/03 The body_linecount and body_zerocount variables are now exported in the
1578 PH/04 Added the $dnslist_matched variable.
1580 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1581 This means they are set thereafter only if the connection becomes
1584 PH/06 Added the client_condition to authenticators so that some can be skipped
1585 by clients under certain conditions.
1587 PH/07 The error message for a badly-placed control=no_multiline_responses left
1588 "_responses" off the end of the name.
1590 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1592 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1593 (without spaces) instead of just copying the configuration text.
1595 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1597 PH/11 Added $max_received_linelength.
1599 PH/12 Added +ignore_defer and +include_defer to host lists.
1601 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1602 way in which PCRE > 7.0 is built.
1604 PH/14 Implemented queue_only_load_latch.
1606 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1607 MAIL command. The effect was to mangle the value on 64-bit systems.
1609 PH/16 Another patch from the Sieve maintainer.
1611 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1613 PH/18 If a system quota error occurred while trying to create the file for
1614 a maildir delivery, the message "Mailbox is full" was not appended to the
1615 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1616 only to a quota excession during the actual writing of the file.
1618 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1619 characters?) which causes problems in log lines. The DN values are now
1620 passed through string_printing() before being added to log lines.
1622 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1623 and InterBase are left for another time.)
1625 PH/21 Added message_body_newlines option.
1627 PH/22 Guard against possible overflow in moan_check_errorcopy().
1629 PH/23 POSIX allows open() to be a macro; guard against that.
1631 PH/24 If the recipient of an error message contained an @ in the local part
1632 (suitably quoted, of course), incorrect values were put in $domain and
1633 $local_part during the evaluation of errors_copy.
1639 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1640 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1643 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1644 issue a MAIL command.
1646 PH/02 In an ACL statement such as
1648 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1650 if a client was not listed at all, or was listed with a value other than
1651 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1652 the condition was not true (as it should be), so access was not denied.
1653 The bug was that the ! inversion was incorrectly passed on to the second
1654 item. This has been fixed.
1656 PH/03 Added additional dnslists conditions == and =& which are different from
1657 = and & when the dns lookup returns more than one IP address.
1659 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1660 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1662 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1663 FSYNC, which compiles an option called disable_fsync that allows for
1664 bypassing fsync(). The documentation is heavily laced with warnings.
1666 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1668 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1669 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1670 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1671 including adding "make clean"; (3) Added -fPIC when compiling the test
1672 dynamically loaded module, to get rid of a warning.
1674 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1675 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1676 = 0s. The bug is otherwise harmless.
1678 PH/07 There was a bug in the dovecot authenticator such that the value of
1679 $auth1 could be overwritten, and so not correctly preserved, after a
1680 successful authentication. This usually meant that the value preserved by
1681 the server_setid option was incorrect.
1683 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1685 PH/09 Installed PCRE release 7.0.
1687 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1688 run for batched SMTP input. It is now run at the start of every message
1689 in the batch. While fixing this I discovered that the process information
1690 (output by running exiwhat) was not always getting set for -bs and -bS
1691 input. This is fixed, and it now also says "batched" for BSMTP.
1693 PH/11 Added control=no_pipelining.
1695 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1696 patch, slightly modified), and move the expansion of helo_data till after
1697 the connection is made in the smtp transport (so it can use these
1700 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1702 PH/14 Added log_selector = +pid.
1704 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1706 PH/16 Add ${if forany and ${if forall.
1708 PH/17 Added dsn_from option to vary the From: line in DSNs.
1710 PH/18 Flush SMTP output before performing a callout, unless control =
1711 no_callout_flush is set.
1713 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1714 was true (the default) a successful delivery failed to delete the retry
1715 item, thus causing premature timeout of the address. The bug is now
1718 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1720 PH/21 Long custom messages for fakedefer and fakereject are now split up
1721 into multiline reponses in the same way that messages for "deny" and
1722 other ACL rejections are.
1724 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1725 with slight modification.
1727 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1728 draft, changing the syntax and factoring some duplicate code".
1730 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1731 for deliveries of the second and subsequent messages over the same SMTP
1734 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1735 ${reduce, with only minor "tidies".
1737 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1739 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1740 expansion side effects.
1742 PH/27 When a message times out after an over-quota error from an Exim-imposed
1743 quota, the bounce message says "mailbox is full". This message was not
1744 being given when it was a system quota that was exceeded. It now should
1747 MH/03 Made $recipients available in local_scan(). local_scan() already has
1748 better access to the recipient list through recipients_list[], but
1749 $recipients can be useful in postmaster-provided expansion strings.
1751 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1752 in the case of a MAIL command with additional options following the
1753 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1754 were accidentally chopped off.
1756 PH/29 SMTP synchronization checks are implemented when a command is read -
1757 there is a check that no more input is waiting when there shouldn't be
1758 any. However, for some commands, a delay in an ACL can mean that it is
1759 some time before the response is written. In this time, more input might
1760 arrive, invalidly. So now there are extra checks after an ACL has run for
1761 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1762 pipelining has not been advertised.
1764 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1766 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1767 returned in the SMTP response when smtp_return_error_details was false.
1768 This has been fixed.
1770 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1771 instead of the C I/O that was originally supplied, because problems were
1772 reported on Solaris.
1774 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1775 Exim which did not show up earlier: it was assuming that a call to
1776 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1777 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1778 was a macro that became an assignment, so it seemed to work. This has
1779 changed to a proper function call with a void return, hence the compile
1780 error. Exim's code has been fixed.
1782 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1785 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1787 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1789 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1790 because it turns out to be too big or there is a timeout) there is no
1791 "Completed" line in the log. When some messages of this type were
1792 selected by exigrep, they were listed as "not completed". Others were
1793 picked up by some special patterns. I have improved the selection
1794 criteria to be more general.
1796 PH/38 The host_find_failed option in the manualroute router can now be set
1797 to "ignore", to completely ignore a host whose IP address cannot be
1798 found. If all hosts are ignored, the behaviour is controlled by the new
1799 host_all_ignored option.
1801 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1802 homing or because of multiple MX records with /mx) generated more than
1803 one IP address, and the following item turned out to be the local host,
1804 all the secondary addresses of the first item were incorrectly removed
1805 from the list, along with the local host and any following hosts (which
1806 is what is supposed to happen).
1808 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1809 whoever called Exim into the -H file. In the case of the daemon it was
1810 behaving confusingly. When first started, it used values for whoever
1811 started the daemon, but after a SIGHUP it used the Exim user (because it
1812 calls itself on a restart). I have changed the code so that it now always
1815 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1816 message are rejected with the same error (e.g. no authentication or bad
1817 sender address), and a DATA command is nevertheless sent (as can happen
1818 with PIPELINING or a stupid MUA), the error message that was given to the
1819 RCPT commands is included in the rejection of the DATA command. This is
1820 intended to be helpful for MUAs that show only the final error to their
1823 PH/42 Another patch from the Sieve maintainer.
1825 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1826 Eximstats - Fixed some broken HTML links and added missing column headers
1828 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1829 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1831 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1833 PH/43 Yet another patch from the Sieve maintainer.
1835 PH/44 I found a way to check for a TCP/IP connection going away before sending
1836 the response to the final '.' that terminates a message, but only in the
1837 case where the client has not sent further data following the '.'
1838 (unfortunately, this is allowed). However, in many cases there won't be
1839 any further data because there won't be any more messages to send. A call
1840 to select() can be used: if it shows that the input is "ready", there is
1841 either input waiting, or the socket has been closed. An attempt to read
1842 the next input character can distinguish the two cases. Previously, Exim
1843 would have sent an OK response which the client would never have see.
1844 This could lead to message repetition. This fix should cure that, at
1845 least in a lot of common cases.
1847 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1848 advertised in response to EHLO.
1854 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1855 fixed by 4.65/MH/01 (is this a record?) are fixed:
1857 (i) An empty string was always treated as zero by the numeric comparison
1858 operators. This behaviour has been restored.
1860 (ii) It is documented that the numeric comparison operators always treat
1861 their arguments as decimal numbers. This was broken in that numbers
1862 starting with 0 were being interpreted as octal.
1864 While fixing these problems I realized that there was another issue that
1865 hadn't been noticed. Values of message_size_limit (both the global option
1866 and the transport option) were treated as octal if they started with 0.
1867 The documentation was vague. These values are now always treated as
1868 decimal, and I will make that clear in the documentation.
1874 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1875 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1878 MH/01 Don't check that the operands of numeric comparison operators are
1879 integers when their expansion is in "skipping" mode (fixes bug
1880 introduced by 4.64-PH/07).
1882 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1883 child addresses, Exim now panics and dies. Previously, because the count
1884 is held in a short int, deliveries were likely to be lost. As such a
1885 large number of recipients for a single message is ridiculous
1886 (performance will be very, very poor), I have chosen to impose a limit
1887 rather than extend the field.
1893 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1894 leftover -K file (the existence of which was triggered by #402).
1895 While we were at it, introduced process PID as part of the -K
1896 filename. This should rule out race conditions when creating
1899 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1900 processing considerably. Previous code took too long for large mails,
1901 triggering a timeout which in turn triggers #401.
1903 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1904 in the DK code in transports.c. sendfile() is not really portable,
1905 hence the _LINUX specificness.
1907 TF/01 In the add_headers option to the mail command in an Exim filter,
1908 there was a bug that Exim would claim a syntax error in any
1909 header after the first one which had an odd number of characters
1912 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1913 callout verification, Exim cached a "reject" for the entire domain. This
1914 is correct for most verifications, but it is not correct for a recipient
1915 verification with use_sender or use_postmaster set, because in that case
1916 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1917 case of MAIL FROM:<> rejection from other early rejections (e.g.
1918 rejection of HELO). When verifying a recipient using a non-null MAIL
1919 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1920 Whatever the result of the callout, the value of the domain cache is
1921 left unchanged (for any other kind of callout, getting as far as trying
1922 RCPT means that the domain itself is ok).
1924 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1927 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1928 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1929 session, the daemon ignores ECONNECT errors and logs others; it now
1930 ignores EPIPE as well.
1932 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1933 (quoted-printable decoding).
1935 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1936 later the small subsequent patch to fix an introduced bug.
1938 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1940 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1942 PH/08 An error is now given if message_size_limit is specified negative.
1944 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1945 to be given (somewhat) arbitrary names.
1947 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1950 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1951 miscellaneous code fixes
1953 PH/10 Added the log_reject_target ACL modifier to specify where to log
1956 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1957 hostname. This is wrong, because it relates to the incoming message (and
1958 probably the interface on which it is arriving) and not to the outgoing
1959 callout (which could be using a different interface). This has been
1960 changed to use the value of the helo_data option from the smtp transport
1961 instead - this is what is used when a message is actually being sent. If
1962 there is no remote transport (possible with a router that sets up host
1963 addresses), $smtp_active_hostname is used.
1965 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1966 tweaks were necessary in order to get it to work (see also 21 below):
1967 (a) The code assumed that strncpy() returns a negative number on buffer
1968 overflow, which isn't the case. Replaced with Exim's string_format()
1970 (b) There were several signed/unsigned issues. I just did the minimum
1971 hacking in of casts. There is scope for a larger refactoring.
1972 (c) The code used strcasecmp() which is not a standard C function.
1973 Replaced with Exim's strcmpic() function.
1974 (d) The code set only $1; it now sets $auth1 as well.
1975 (e) A simple test gave the error "authentication client didn't specify
1976 service in request". It would seem that Dovecot has changed its
1977 interface. Fortunately there's a specification; I followed it and
1978 changed what the client sends and it appears to be working now.
1980 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1983 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1984 address A is aliased to B and C, where B exists and C does not. Without
1985 -v the output is "A verified" because verification stops after a
1986 successful redirection if more than one address is generated. However,
1987 with -v the child addresses are also verified. Exim was outputting "A
1988 failed to verify" and then showing the successful verification for C,
1989 with its parentage. It now outputs "B failed to verify", showing B's
1990 parentage before showing the successful verification of C.
1992 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1993 look up a TXT record in a specific list after matching in a combined
1996 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1997 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1998 they consult the DNS. I had assumed they would set it the way they
1999 wanted; and indeed my experiments on Linux seem to show that in some
2000 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2001 To be on the safe side, however, I have now made the interface to
2002 host_find_byname() similar to host_find_bydns(), with an argument
2003 containing the DNS resolver options. The host_find_byname() function now
2004 sets these options at its start, just as host_find_bydns() does. The smtp
2005 transport options dns_qualify_single and dns_search_parents are passed to
2006 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2007 of host_find_byname() use the default settings of RES_DEFNAMES
2008 (qualify_single) but not RES_DNSRCH (search_parents).
2010 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2011 spool_read_header() do less string testing, by means of a preliminary
2012 switch on the second character of optional "-foo" lines. (This is
2013 overdue, caused by the large number of possibilities that now exist.
2014 Originally there were few.) While I was there, I also converted the
2015 str(n)cmp tests so they don't re-test the leading "-" and the first
2016 character, in the hope this might squeeze out yet more improvement.
2018 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2019 flag allowing group syntax was set by the header_syntax check but not
2020 turned off, possible causing trouble later; (2) The flag was not being
2021 set at all for the header_verify test, causing "group"-style headers to
2022 be rejected. I have now set it in this case, and also caused header_
2023 verify to ignore an empty address taken from a group. While doing this, I
2024 came across some other cases where the code for allowing group syntax
2025 while scanning a header line wasn't quite right (mostly, not resetting
2026 the flag correctly in the right place). These bugs could have caused
2027 trouble for malformed header lines. I hope it is now all correct.
2029 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2030 with the "reply" argument non-NULL. The code, however (which originally
2031 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2032 but it didn't always do it. This confused somebody who was copying the
2033 code for some other use. I have removed all the tests.
2035 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2036 feature that was used to support insecure browsers during the U.S. crypto
2037 embargo. It requires special client support, and Exim is probably the
2038 only MTA that supported it -- and would never use it because real RSA is
2039 always available. This code has been removed, because it had the bad
2040 effect of slowing Exim down by computing (never used) parameters for the
2041 RSA_EXPORT functionality.
2043 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2044 authenticator to fail if there's a tab character in the incoming data
2045 (there should never be unless someone is messing about, as it's supposed
2046 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2047 if the connection is using TLS or if the remote IP is the same as the
2048 local IP, and the "valid-client-cert option" if a client certificate has
2051 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2052 authenticators. This can be used for authorization after authentication
2053 succeeds. (In the case of plaintext, it servers for both authentication
2056 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2057 if any retry times were supplied.
2059 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2060 connection, where there is no client IP address to check. In this
2061 situation, the verify now always succeeds.
2063 PH/25 Applied John Jetmore's -Mset patch.
2065 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2067 PH/27 In a string expansion for a processed (not raw) header when multiple
2068 headers of the same name were present, leading whitespace was being
2069 removed from all of them, but trailing whitespace was being removed only
2070 from the last one. Now trailing whitespace is removed from each header
2071 before concatenation. Completely empty headers in a concatenation (as
2072 before) are ignored.
2074 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2075 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2077 PH/29 [Removed. This was a change that I later backed out, and forgot to
2078 correct the ChangeLog entry (that I had efficiently created) before
2079 committing the later change.]
2081 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2082 address errors (4xx response to RCPT) over the same connection as other
2083 messages routed to the same hosts. Such deliveries are always "forced",
2084 so retry times are not inspected. This resulted in far too many retries
2085 for the affected addresses. The effect occurred only when there were more
2086 hosts than the hosts_max_try setting in the smtp transport when it had
2087 the 4xx errors. Those hosts that it had tried were not added to the list
2088 of hosts for which the message was waiting, so if all were tried, there
2089 was no problem. Two fixes have been applied:
2091 (i) If there are any address or message errors in an SMTP delivery, none
2092 of the hosts (tried or untried) are now added to the list of hosts
2093 for which the message is waiting, so the message should not be a
2094 candidate for sending over the same connection that was used for a
2095 successful delivery of some other message. This seems entirely
2096 reasonable: after all the message is NOT "waiting for some host".
2097 This is so "obvious" that I'm not sure why it wasn't done
2098 previously. Hope I haven't missed anything, but it can't do any
2099 harm, as the worst effect is to miss an optimization.
2101 (ii) If, despite (i), such a delivery is accidentally attempted, the
2102 routing retry time is respected, so at least it doesn't keep
2103 hammering the server.
2105 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2106 in ${readsocket because some servers need this prod.
2108 PH/32 Added some extra debug output when updating a wait-xxx database.
2110 PH/33 The hint "could be header name not terminated by colon", which has been
2111 given for certain expansion errors for a long time, was not being given
2112 for the ${if def:h_colon_omitted{... case.
2114 PH/34 The spec says: "With one important exception, whenever a domain list is
2115 being scanned, $domain contains the subject domain." There was at least
2116 one case where this was not true.
2118 PH/35 The error "getsockname() failed: connection reset by peer" was being
2119 written to the panic log as well as the main log, but it isn't really
2120 panic-worthy as it just means the connection died rather early on. I have
2121 removed the panic log writing for the ECONNRESET error when getsockname()
2124 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2125 runs only) independently of the message's sender address. This meant
2126 that, if the 4xx error was in fact related to the sender, a different
2127 message to the same recipient with a different sender could confuse
2128 things. In particualar, this can happen when sending to a greylisting
2129 server, but other circumstances could also provoke similar problems.
2130 I have changed the default so that the retry time for these errors is now
2131 based a combination of the sender and recipient addresses. This change
2132 can be overridden by setting address_retry_include_sender=false in the
2135 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2136 remote server are returned as part of bounce messages. This was not
2137 happening for LMTP over a pipe (the lmtp transport), but now it is the
2138 same for both kinds of LMTP.
2140 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2141 in header lines that were in fact CNAMEs. This is no longer the case.
2143 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2144 and queue runs started by the daemon processed all messages. This has
2145 been fixed so that -R and -S can now usefully be given with -q<time>.
2147 PH/40 Import PCRE release 6.7 (fixes some bugs).
2149 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2151 PH/42 Give an error if -q is specified more than once.
2153 PH/43 Renamed the variables $interface_address and $interface_port as
2154 $received_ip_address and $received_port, to make it clear that these
2155 values apply to message reception, and not to the outgoing interface when
2156 a message is delivered. (The old names remain recognized, of course.)
2158 PH/44 There was no timeout on the connect() call when using a Unix domain
2159 socket in the ${readsocket expansion. There now is.
2161 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2162 be meaningful with "accept".
2164 SC/01 Eximstats V1.43
2165 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2167 SC/02 Eximstats V1.44
2168 Use a glob alias rather than an array ref in the generated
2169 parser. This improves both readability and performance.
2171 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2172 Collect SpamAssassin and rejection statistics.
2173 Don't display local sender or destination tables unless
2174 there is data to show.
2175 Added average volumes into the top table text output.
2177 SC/04 Eximstats V1.46
2178 Collect data on the number of addresses (recipients)
2179 as well as the number of messages.
2181 SC/05 Eximstats V1.47
2182 Added 'Message too big' to the list of mail rejection
2183 reasons (thanks to Marco Gaiarin).
2185 SC/06 Eximstats V1.48
2186 Mainlog lines which have GMT offsets and are too short to
2187 have a flag are now skipped.
2189 SC/07 Eximstats V1.49 (Alain Williams)
2190 Added the -emptyok flag.
2192 SC/08 Eximstats V1.50
2193 Fixes for obtaining the IP address from reject messages.
2195 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2196 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2197 whitesspace changes from 4.64-PH/27
2199 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2202 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2203 are found, allow negative numbers in numeric criteria)
2205 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2207 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2208 to match changes made in 4.64-PH/43
2210 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2212 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2213 "domain" is actually an IP address, and force a failure. This locks out
2214 those revolvers/nameservers that support "A-for-A" lookups, in
2215 contravention of the specifications.
2217 PH/48 When a host name was looked up from an IP address, and the subsequent
2218 forward lookup of the name timed out, the host name was left in
2219 $sender_host_name, contrary to the specification.
2221 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2222 restricted to single-key lookups, Exim was not diagnosing an error if
2223 * or *@ was used with a query-style lookup.
2225 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2227 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2228 long ago, but noone interested enough thought of it. Let's just say that
2229 the "1.1" means that there are some new functions that weren't there at
2230 some point in the past.
2232 PH/51 Error processing for expansion failure of helo_data from an smtp
2233 transport during callout processing was broken.
2235 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2236 tested/used via the -bh/-bhc/-bs options.
2238 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2239 bug, fixed in subsequent PCRE releases).
2241 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2242 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2244 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2250 SC/01 Use a glob alias rather than an array ref in eximstats generated
2251 parser. This improves both readability and performance.
2253 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2254 Don't display local sender or destination tables in eximstats unless
2255 there is data to show.
2256 Added average volumes into the eximstats top table text output.
2258 SC/03 Collect data on the number of addresses (recipients) as well
2259 as the number of messages in eximstats.
2261 TF/01 Correct an error in the documentation for the redirect router. Exim
2262 does (usually) call initgroups() when daemonizing.
2264 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2265 with consistent privilege compared to when running as a daemon.
2267 TF/03 Note in the spec that $authenticated_id is not set for local
2268 submissions from trusted users.
2270 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2271 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2273 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2274 by adding some example configuration directives to the default
2275 configuration file. A little bit of work is required to uncomment the
2276 directives and define how usernames and passwords are checked, but
2277 there is now a framework to start from.
2279 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2280 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2281 without this. I don't know how relevant this is to other LDAP libraries.
2283 PH/02 Add the verb name to the "unknown ACL verb" error.
2285 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2287 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2289 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2290 directory not expanded when it should be if an expanded home directory
2291 was set for the address (which is overridden by the transport).
2293 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2296 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2297 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2298 because it is too late at that time, and has no effect.
2300 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2301 security issue with \' (bugzilla #107). I could not use the
2302 PQescapeStringConn() function, because it needs a PGconn value as one of
2305 PH/08 When testing addresses using -bt, indicate those final addresses that
2306 are duplicates that would not cause an additional delivery. At least one
2307 person was confused, thinking that -bt output corresponded to deliveries.
2308 (Suppressing duplicates isn't a good idea as you lose the information
2309 about possibly different redirections that led to the duplicates.)
2311 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2312 systems where poll() doesn't work, in particular OS X.
2314 PH/10 Added more information to debugging output for retry time not reached.
2316 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2317 operations in malware.c.
2319 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2322 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2323 syslog_duplication set false, log lines that would normally be written
2324 both the the main log and to the reject log were not written to syslog at
2327 PH/14 In the default configuration, change the use of "message" in ACL warn
2328 statements to "add_header".
2330 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2331 not followed by a command (e.g. "seen endif").
2333 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2334 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2337 PH/17 Added extra conditions to the default value of delay_warning_condition
2341 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2342 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2343 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2346 The Auto-Submitted: and various List- headers are standardised, whereas I
2347 don't think Precedence: ever was.
2349 PH/18 Refactored debugging code in route_finduser() to show more information,
2350 in particular, the error code if getpwnam() issues one.
2352 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2353 This is apparently needed in addition to the PH/07 change above to avoid
2354 any possible encoding problems.
2356 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2357 but not after initializing Perl.
2359 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2360 output them only if debugging. By default they are written stderr,
2361 apparently, which is not desirable.
2363 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2366 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2369 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2371 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2372 authenticated or an ident call has been made. Suppress the default
2373 values for $authenticated_id and $authenticated_sender (but permit -oMai
2374 and -oMas) when testing with -bh.
2376 PH/24 Re-jigged the order of the tests in the default configuration so that the
2377 tests for valid domains and recipients precede the DNS black list and CSA
2378 tests, on the grounds that those ones are more expensive.
2380 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2381 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2382 This bug exists in every version of Exim that I still have, right back to
2385 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2386 However, an attempt to turn on case-sensitivity in a regex key by
2387 including (?-i) didn't work because the subject string was already
2388 lowercased, and the effects were non-intuitive. It turns out that a
2389 one-line patch can be used to allow (?-i) to work as expected.
2395 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2396 other effects) broke the use of negated acl sub-conditions.
2398 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2401 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2402 "Deny" causes Exim to reject the incoming connection with a 554 error.
2403 Unfortunately, if there is a major crisis, such as a disk failure,
2404 tcp-wrappers gives "deny", whereas what one would like would be some
2405 kind of temporary error. A kludge has been added to help with this.
2406 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2407 554 error is used if errno is still zero or contains ENOENT (which occurs
2408 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2411 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2413 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2414 errors. Otherwise a message that provokes a temporary error (when other
2415 messages do not) can cause a whole host to time out.
2417 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2418 addresses were routed directly to files or pipes from a redirect router.
2419 File deliveries just didn't batch; pipe deliveries might have suffered
2422 PH/06 A failure to get a lock for a hints database would erroneously always say
2423 "Failed to get write lock", even when it was really a read lock.
2425 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2426 of 0600. This has been changed to use the value of the lockfile_mode
2427 option (which defaults to 0600).
2429 PH/08 Applied small patch from the Sieve maintainer.
2431 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2432 folder from quota calculations, a direct delivery into this folder messed
2433 up the contents of the maildirsize file. This was because the regex was
2434 used only to exclude .Trash (or whatever) when the size of the mailbox
2435 was calculated. There was no check that a delivery was happening into an
2436 excluded directory. This bug has been fixed by ignoring all quota
2437 processing for deliveries into excluded directories.
2439 PH/10 Added the maildirfolder_create_regex option to appendfile.
2445 PH/01 The code for finding all the local interface addresses on a FreeBSD
2446 system running IPv6 was broken. This may well have applied to all BSD
2447 systems, as well as to others that have similar system calls. The broken
2448 code found IPv4 interfaces correctly, but gave incorrect values for the
2449 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2450 that it would not match correctly against @[] and not recognize the IPv6
2453 PH/02 The ipliteral router was not recognizing addresses of the form user@
2454 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2456 PH/03 Added disable_ipv6.
2458 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2459 decoded form, because it is most often used to construct To: headers
2460 lines in autoreplies, and the decoded form may well be syntactically
2461 invalid. However, $reply_address has leading white space removed, and all
2462 newlines turned into spaces so that the autoreply transport does not
2465 PH/05 If group was specified without a user on a router, and no group or user
2466 was specified on a transport, the group from the router was ignored.
2468 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2469 for visible compile-time settings that can be used to change these
2470 numbers, for those that want even more. Backwards compatibility with old
2471 spool files has been maintained. However, going back to a previous Exim
2472 release will lost any variables that are in spool files.
2474 PH/07 Two small changes when running in the test harness: increase delay when
2475 passing a TCP/IP connection to a new process, in case the original
2476 process has to generate a bounce, and remove special handling of
2477 127.0.0.2 (sic), which is no longer necessary.
2479 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2480 be the same on different OS.
2482 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2485 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2486 whether --show-vars was specified or not
2488 JJ/02 exipick: Added support for new ACL variable spool format introduced
2491 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2492 syntactically invalid From: or Reply-to: line, and a filter used this to
2493 generate an autoreply, and therefore failed to obtain an address for the
2494 autoreply, Exim could try to deliver to a non-existent relative file
2495 name, causing unrelated and misleading errors. What now happens is that
2496 it logs this as a hard delivery error, but does not attempt to create a
2499 PH/11 The exinext utility has a -C option for testing purposes, but although
2500 the given file was scanned by exinext itself; it wasn't being passed on
2501 when Exim was called.
2503 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2504 an end-of-file indication when reading a command response.
2506 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2507 compiled. In many other places in Exim, IPv6 addresses are always
2508 recognized, so I have changed this. It also means that IPv4 domain
2509 literals of the form [IPV4:n.n.n.n] are now always recognized.
2511 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2512 used if the router is not running as root, for example, when verifying at
2513 ACL time, or when using -bh. The debugging output from this situation was
2514 non-existent - all you got was a failure to exec. I have made two
2517 (a) Failures to set uid/gid, the current directory, or a process leader
2518 in a subprocess such as that created by queryprogram now generate
2519 suitable debugging ouput when -d is set.
2521 (b) The queryprogram router detects when it is not running as root,
2522 outputs suitable debugging information if -d is set, and then runs
2523 the subprocess without attempting to change uid/gid.
2525 PH/15 Minor change to Makefile for building test_host (undocumented testing
2528 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2529 additional section of a DNS packet that returns MX or SRV records.
2530 Instead, it always explicitly searches for A/AAAA records. This avoids
2531 major problems that occur when a DNS server includes only records of one
2532 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2533 fixed another bug: if SRV records were looked up and the corresponding
2534 address records were *not* found in the additional section, the port
2535 values from the SRV records were lost.
2537 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2538 using the correct key (the original address) when searching the retry
2539 rules in order to find which one to use for generating the retry hint.
2541 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2542 adding the default one. Similarly, if it contains a Reply-To: header, the
2543 errors_reply_to option, if set, is not used.
2545 PH/19 When calculating a retry time, Exim used to measure the "time since
2546 failure" by looking at the "first failed" field in the retry record. Now
2547 it does not use this if it is later than than the arrival time of the
2548 message. Instead it uses the arrival time. This makes for better
2549 behaviour in cases where some deliveries succeed, thus re-setting the
2550 "first failed" field. An example is a quota failure for a huge message
2551 when small messages continue to be delivered. Without this change, the
2552 "time since failure" will always be short, possible causing more frequent
2553 delivery attempts for the huge message than are intended.
2554 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2556 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2557 $1, $2, $3) because the numerical variables can be reset during some
2558 expansion items (e.g. "match"), thereby losing the authentication data.
2560 PH/21 Make -bV show the size of off_t variables so that the test suite can
2561 decide whether to run tests for quotas > 2G.
2563 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2564 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2565 filecount value is greater than 2G or if a quota value is greater than 2G
2566 on a system where the size of off_t is not greater than 4, a panic error
2569 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2570 never match. The debug and -bh output now contains an explicit error
2571 message indicating a malformed IPv4 address or mask.
2573 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2574 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2575 PH/23 above applies.
2577 PH/25 Do not write to syslog when running in the test harness. The only
2578 occasion when this arises is a failure to open the main or panic logs
2579 (for which there is an explicit test).
2581 PH/26 Added the /no_tell option to "control=freeze".
2583 PH/27 If a host name lookup failed very early in a connection, for example, if
2584 the IP address matched host_lookup and the reverse lookup yielded a name
2585 that did not have a forward lookup, an error message of the form "no IP
2586 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2587 could be logged. Now it outputs the IP address instead of "NULL".
2589 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2590 allows the sender and the authenticated sender to be set when
2591 submitting a message from within Exim. Since child_open_exim() is
2592 documented for local_scan(), the new function should be too.
2594 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2595 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2596 results in an empty string is now treated as unset.
2598 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2600 PH/31 Added sender_verify_fail logging option.
2602 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2603 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2604 tidied the source and removed it altogether.
2606 PH/33 When a queue run was abandoned because the load average was too high, a
2607 log line was always written; now it is written only if the queue_run log
2608 selector is set. In addition, the log line for abandonment now contains
2609 information about the queue run such as the pid. This is always present
2610 in "start" and "stop" lines but was omitted from the "abandon" line.
2612 PH/34 Omit spaces between a header name and the colon in the error message that
2613 is given when verify = headers_syntax fails (if there are lots of them,
2614 the message gets confusing).
2616 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2617 names, as there are now some PTR records that contain slashes. This check
2618 is only to protect against broken name servers that fall over on strange
2619 characters, so the fact that it applies to all lookups doesn't matter.
2621 PH/36 Now that the new test suite is complete, we can remove some of the
2622 special code in Exim that was needed for the old test suite. For example,
2623 sorting DNS records because real resolvers return them in an arbitrary
2624 order. The new test suite's fake resolver always returns records in the
2627 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2628 bounces) except when queue_only is set, to avoid logging races between
2629 the different processes.
2631 PH/38 Panic-die if .include specifies a non-absolute path.
2633 PH/39 A tweak to the "H" retry rule from its user.
2635 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2636 a label. They prevented compilation on older perls.
2638 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2639 a warning to be raised on newish perls.
2641 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2642 on queue. Changes to match documented behaviour of showing count of
2643 messages matching specified criteria.
2645 PH/40 Changed the default ident timeout from 30s to 5s.
2647 PH/41 Added support for the use of login_cap features, on those BSD systems
2648 that have them, for controlling the resources used by pipe deliveries.
2650 PH/42 The content-scanning code uses fopen() to create files in which to put
2651 message data. Previously it was not paying any attention to the mode of
2652 the files. Exim runs with umask(0) because the rest of the code creates
2653 files with open(), and sets the required mode explicitly. Thus, these
2654 files were ending up world-writeable. This was not a big issue, because,
2655 being within the spool directory, they were not world-accessible. I have
2656 created a function called modefopen, which takes an additional mode
2657 argument. It sets umask(777), creates the file, chmods it to the required
2658 mode, then resets the umask. All the relevant calls to fopen() in the
2659 content scanning code have been changed to use this function.
2661 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2662 to 24 hours. This avoids potential overflow problems when processing G
2663 and H retry rules. I suspect nobody ever tinkers with this value.
2665 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2667 PH/45 When the plaintext authenticator is running as a client, the server's
2668 challenges are checked to ensure they are valid base64 strings. By
2669 default, the authentication attempt is cancelled if an invalid string is
2670 received. Setting client_ignore_invalid_base64 true ignores these errors.
2671 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2672 they are received. Thus, the responses can be made to depend on the
2673 challenges. If an invalid string is ignored, an empty string is placed in
2676 PH/46 Messages that are created by the autoreply transport now contains a
2677 References: header, in accordance with RFCs 2822 and 3834.
2679 PH/47 Added authenticated_sender_force to the smtp transport.
2681 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2683 PH/49 Installed latest patch from the Sieve maintainer.
2685 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2686 was also set, the appendfile transport was unnecessarily scanning a
2687 directory of message files (e.g. for maildir delivery) to find the count
2688 of files (along with the size), even though it did not need this
2689 information. It now does the scan only if it needs to find either the
2690 size of the count of files.
2692 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2694 PH/52 Two bugs concerned with error handling when the smtp transport is
2697 (i) Exim was not creating retry information for temporary errors given
2698 for individual recipients after the DATA command when the smtp transport
2699 was used in LMTP mode. This meant that they could be retried too
2700 frequently, and not timed out correctly.
2702 (ii) Exim was setting the flag that allows error details to be returned
2703 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2704 recipients that were returned after the DATA command.
2706 PH/53 This is related to PH/52, but is more general: for any failing address,
2707 when detailed error information was permitted to be returned to the
2708 sender, but the error was temporary, then after the final timeout, only
2709 "retry timeout exceeded" was returned. Now it returns the full error as
2710 well as "retry timeout exceeded".
2712 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2713 do this, and (what is worse) MTAs that accept it.
2715 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2716 will now be deprecated.
2718 PH/56 New os.c-cygwin from the Cygwin maintainer.
2720 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2721 formats (previously only available in exim formats via -bpr, -bpru,
2722 and -bpra. Now also available in native and exiqgrep formats)
2724 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2725 with very large, slow to parse queues
2727 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2729 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2731 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2732 responses to authentication challenges, though it was showing the
2733 challenges; (ii) I've removed the CR characters from the debug output for
2736 PH/58 Allow for the insertion of a newline as well as a space when a string
2737 is turned into more than one encoded-word during RFC 2047 encoding. The
2738 Sieve code now uses this.
2740 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2741 data_4xx, lost_connection, tls_required.
2743 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2744 message was being sent as an SMTP response.
2746 PH/61 Add -l and -k options to exicyclog.
2748 PH/62 When verifying, if an address was redirected to one new address, so that
2749 verification continued, and the new address failed or deferred after
2750 having set something in $address_data, the value of $address_data was not
2751 passed back to the ACL. This was different to the case when no
2752 redirection occurred. The value is now passed back in both cases.
2754 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2755 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2756 use login_cap.h, so on its own it isn't the distinguishing feature. The
2757 new name refers directly to the setclassresources() function.
2759 PH/65 Added configuration files for NetBSD3.
2761 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2763 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2764 is preferred over IPv4.
2766 PH/68 The bounce_return_message and bounce_return_body options were not being
2767 honoured for bounces generated during the reception of non-SMTP messages.
2768 In particular, this applied to messages rejected by the ACL. This bug has
2769 been fixed. However, if bounce_return_message is true and bounce_return_
2770 body is false, the headers that are returned for a non-SMTP message
2771 include only those that have been read before the error was detected.
2772 (In the case of an ACL rejection, they have all been read.)
2774 PH/69 The HTML version of the specification is now built in a directory called
2775 spec_html instead of spec.html, because the latter looks like a path with
2776 a MIME-type, and this confuses some software.
2778 PH/70 Catch two compiler warnings in sieve.c.
2780 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2781 function verify_get_ident() calls ip_connect() to connect a socket, but
2782 if the "connect()" function timed out, ip_connect() used to close the
2783 socket. However, verify_get_ident() also closes the socket later, and in
2784 between Exim writes to the log, which may get opened at this point. When
2785 the socket was closed in ip_connect(), the log could get the same file
2786 descriptor number as the socket. This naturally causes chaos. The fix is
2787 not to close the socket in ip_connect(); the socket should be closed by
2788 the function that creates it. There was only one place in the code where
2789 this was missing, in the iplookup router, which I don't think anybody now
2790 uses, but I've fixed it anyway.
2792 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2793 well as to direct DNS lookups. Otherwise the handling of names in host
2794 lists is inconsistent and therefore confusing.
2800 PH/01 Two changes to the default runtime configuration:
2802 (1) Move the checks for relay_from_hosts and authenticated clients from
2803 after to before the (commented out) DNS black list checks.
2805 (2) Add control=submission to the relay_from_hosts and authenticated
2806 clients checks, on the grounds that messages accepted by these
2807 statements are most likely to be submissions.
2809 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2811 (1) Generate an error if the third argument for the ${prvs expansion is
2814 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2817 (3) Reset the variables that are obtained from the first argument of
2818 ${prvscheck and used in the second argument before leaving the code,
2819 because their memory is reclaimed, so using them afterwards may do
2822 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2823 one (it's much easier than Tom thought :-).
2825 (5) Because of (4), we can now allow for the use of $prvscheck_result
2826 inside the third argument.
2828 PH/03 For some reason, the default setting of PATH when running a command from
2829 a pipe transport was just "/usr/bin". I have changed it to
2832 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2833 anything to be listed in the output from -bV.
2835 PH/05 When a filter generated an autoreply, the entire To: header line was
2836 quoted in the delivery log line, like this:
2838 => >A.N.Other <ano@some.domain> <original@ddress> ...
2840 This has been changed so that it extracts the operative address. There
2841 may be more than one such address. If so, they are comma-separated, like
2844 => >ano@some.domain,ona@other.domain <original@ddress> ...
2846 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2847 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2848 not being looked up in the rDNS to get a host name, Exim was showing the
2849 IP address twice in Received: lines, even though the IP addresses were
2850 identical. For example:
2852 Received: from [1.2.3.4] (helo=[1.2.3.4])
2854 However, if the real host name was known, it was omitting the HELO data
2855 if it matched the actual IP address. This has been tidied up so that it
2856 doesn't show the same IP address twice.
2858 PH/07 When both +timestamp and +memory debugging was on, the value given by
2859 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2860 called by the debug printing, thereby overwriting the timestamp buffer.
2861 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2863 PH/08 When the original message was included in an autoreply transport, it
2864 always said "this is a copy of the message, including all the headers",
2865 even if body_only or headers_only was set. It now gives an appropriate
2868 PH/09 Applied a patch from the Sieve maintainer which:
2870 o fixes some comments
2871 o adds the (disabled) notify extension core
2872 o adds some debug output for the result of if/elsif tests
2873 o points to the current vacation draft in the documentation
2874 and documents the missing references header update
2878 o fixes a bug in processing the envelope test (when testing
2879 multiple envelope elements, the last element determinted the
2882 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2883 Electronic Mail") by including:
2885 Auto-submitted: auto-generated
2887 in the messages that it generates (bounce messages and others, such as
2888 warnings). In the case of bounce messages for non-SMTP mesages, there was
2889 also a typo: it was using "Auto_submitted" (underscore instead of
2890 hyphen). Since every message generated by Exim is necessarily in response
2891 to another message, thes have all been changed to:
2893 Auto-Submitted: auto-replied
2895 in accordance with these statements in the RFC:
2897 The auto-replied keyword:
2899 - SHOULD be used on messages sent in direct response to another
2900 message by an automatic process,
2902 - MUST NOT be used on manually-generated messages,
2904 - MAY be used on Delivery Status Notifications (DSNs) and Message
2905 Disposition Notifications (MDNs),
2907 - MUST NOT be used on messages generated by automatic or periodic
2908 processes, except for messages which are automatic responses to
2911 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2912 to the default Received: header definition.
2914 PH/12 Added log selector acl_warn_skipped (default on).
2916 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2917 variables because (a) they are in the wrong storage pool and (b) even if
2918 they were copied, it wouldn't work properly because of the caching.
2920 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2921 checking when decoding. Apparently there are clients that generate
2922 overlong encoded strings. Why am I not surprised?
2924 PH/15 If the first argument of "${if match_address" was not empty, but did not
2925 contain an "@" character, Exim crashed. Now it writes a panic log message
2926 and treats the condition as false.
2928 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2930 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2931 extension "envelope-auth". The code is finished and in agreement with
2932 other implementations, but there is no documentation so far and in fact,
2933 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2934 not changing the active code.
2936 Print executed "if" and "elsif" statements when debugging is used. This
2937 helps a great deal to understand what a filter does.
2939 Document more things not specified clearly in RFC3028. I had all this
2940 sorted out, when out of a sudden new issues came to my mind. Oops."
2942 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2945 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2946 canonical form (as documented). However, after a host name lookup from
2947 the IP address, check_host() was doing a simple string comparison with
2948 addresses acquired from the DNS when checking that the found name did
2949 have the original IP as one of its addresses. Since any found IPv6
2950 addresses are likely to be in abbreviated form, the comparison could
2951 fail. Luckily, there already exists a function for doing the comparison
2952 by converting both addresses to binary, so now that is used instead of
2953 the text comparison.
2955 PH/20 There was another similar case to PH/19, when a complete host name was
2956 given in a host list; looking up its IP address could give an abbreviated
2957 form, whereas the current host's name might or might not be abbreviated.
2958 The same fix has been applied.
2964 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2965 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2968 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2969 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2971 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2973 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2974 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2975 Also, for "Auto-Submitted", treat anything other than "no" as
2976 non-personal, in accordance with RFC 3834. (Previously it treated
2977 anything starting "auto-" as non-personal.)
2979 TF/01 The control=submission/name=... option had a problem with syntax
2980 errors if the name included a slash character. The /name= option
2981 now slurps the rest of the string, so it can include any characters
2982 but it must come last in the list of options (after /sender_retain
2985 PH/05 Some modifications to the interface to the fake nameserver for the new
2993 TK/01 Added the "success_on_redirect" address verification option. See
2994 NewStuff for rationale and an example.
2996 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2998 PH/02 Patch to exigrep to allow it to work on syslog lines.
3000 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3001 fread() to skip over the body file's header line, because in Cygwin the
3002 header line is locked and is inaccessible.
3004 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3005 co-exist for some time) to make it clear that it is the Exim ID that is
3006 referenced, not the Message-ID: header line.
3008 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3009 string_format() function, because snprintf() does not exist on all
3012 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3013 ${stat: expansion item.
3015 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3016 protocol synchronization error", to keep the pedants happy.
3018 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3019 well as for IRIX systems, when gcc is being used. See the host.c source
3022 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3024 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3027 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3028 transport and to the smtp transport in LMTP mode.
3030 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3032 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3033 run a filter in a subprocess. This could lead to confusion in subsequent
3034 lookups in the parent process. There should also be a search_tidyup() at
3035 the end of the subprocess.
3037 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3038 only if the host matched helo_try_verify_hosts, which caused the
3039 verification to occur when the EHLO/HELO command was issued. The ACL just
3040 tested the remembered result. Now, if a previous verification attempt has
3041 not happened, "verify = helo" does it there and then.
3043 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3045 TK/03 Fix log output including CR from clamd.
3047 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3048 exist provoked a memory error which could cause a segfault.
3050 PH/15 Installed PCRE 6.2
3052 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3054 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3055 of the problem. Specifically, suggested +O2 rather than +O1 for the
3058 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3060 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3061 record, the port was not correctly logged when the outgoing_port log
3062 selector was set (it logged the transort's default port).
3064 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3065 fallback_hosts, and "hosts" in the smtp transport.
3067 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3068 host errors such as "Connection refused".
3070 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3071 authentication with radiusclient 0.4.9:
3073 - Error returned from rc_read_config was caught wrongly
3074 - Username/password not passed on to radius server due to wrong length.
3076 The presumption is that some radiusclient API changes for 4.51/PH/17
3077 were not taken care of correctly. The code is still untested by me (my
3078 Linux distribution still has 0.3.2 of radiusclient), but it was
3079 contributed by a Radius user.
3081 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3082 expanding the "port" option of the smtp transport.
3084 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3085 while reading a MIME header. Thanks to Tom Hughes for a patch.
3087 PH/24 Include config.h inside local_scan.h so that configuration settings are
3090 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3091 that in an ACL for RCPT (for example), you can examine exactly what was
3094 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3095 commands, but it was not correctly comparing the address with the actual
3096 client host address. Thus, it would show the EHLO address in Received:
3097 header lines when this was not necessary.
3099 PH/27 Added the % operator to ${eval:}.
3101 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3102 it should be ignoring failures (because with -C, for example, it has lost
3103 privilege). It wasn't ignoring creation failures other than "already
3106 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3109 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3110 because an input line was too long, either on its own, or by virtue of
3111 too many continuations, the temporary file was not being removed, and the
3112 return code was incorrect.
3114 PH/31 Missing "BOOL" in function definition in filtertest.c.
3116 PH/32 Applied Sieve patches from the maintainer.
3118 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3120 PH/33 Added "verify = not_blind".
3122 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3123 Local/Makefile (with some defaults set). These are used in built scripts
3124 such as exicyclog, but they have never been used in the exim_install
3125 script (though there are many overriding facilities there). I have
3126 arranged that the exim_install script now takes note of these two
3129 PH/35 Installed configuration files for Dragonfly.
3131 PH/36 When a locally submitted message by a trusted user did not contain a
3132 From: header, and the sender address was obtained from -f or from an SMTP
3133 MAIL command, and the trusted user did not use -F to supply a sender
3134 name, $originator_name was incorrectly used when constructing a From:
3135 header. Furthermore, $originator_name was used for submission mode
3136 messages from external hosts without From: headers in a similar way,
3137 which is clearly wrong.
3139 PH/37 Added control=suppress_local_fixups.
3141 PH/38 When log_selector = +received_sender was set, and the addition of the
3142 sender made the log line's construction buffer exactly full, or one byte
3143 less than full, an overflow happened when the terminating "\n" was
3146 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3147 when the result of a list match is failure because a DNS lookup failed.
3149 PH/40 RM_COMMAND is now used in the building process.
3151 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3152 the "build-* directories that it finds.
3154 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3155 address in a domain literal was a prefix of an interface address.
3157 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3158 when verifying a sender address, unless rewrite_headers is false.
3160 PH/44 Wrote a long comment about why errors_to addresses are verified as
3161 recipients, not senders.
3163 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3164 the ratelimit ACL was added.
3166 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3168 PH/46 Added extra information about PostgreSQL errors to the error string.
3170 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3171 avoiding the need to install special zones in a real server. This is
3172 backwards compatible; if it can't find the fake resolver, it drops back.
3173 Thus, both old and new test suites can be run.
3175 TF/02 Added util/ratelimit.pl
3177 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3178 clock is set back in time.
3180 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3181 Candler <B.Candler@pobox.com>.
3183 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3184 OK for addresses that are the result of redirections.
3186 PH/48 A number of further additions for the benefit of the new test suite,
3187 including a fake gethostbyname() that interfaces to the fake DNS resolver
3190 TF/06 The fix for widen_domains has also been applied to qualify_single and
3191 search_parents which are the other dnslookup options that can cause
3194 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3197 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3199 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3200 service" if it encountered an MX record with an empty target hostname.
3201 The message is now "an MX or SRV record indicated no SMTP service".
3203 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3204 if the DNS of the sending site is misconfigured. This is quite a
3205 common situation. This change restores the behaviour of treating a
3206 helo verification defer as a failure.
3208 PH/51 If self=fail was set on a router, the bounce message did not include the
3209 actual error message.
3215 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3217 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3218 command itself ended in error, the underlying message about the transport
3219 filter timeout was being overwritten with the pipe command error. Now the
3220 underlying error message should be appended to the second error message.
3222 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3224 PH/02 Reduce the amount of output that "make" produces by default. Full output
3225 can still be requested.
3227 PH/03 The warning log line about a condition test deferring for a "warn" verb
3228 was being output only once per connection, rather than after each
3229 occurrence (because it was using the same function as for successful
3230 "warn" verbs). This seems wrong, so I have changed it.
3232 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3233 it should not have, which might have caused a crash in the right
3234 circumstances, but probably never did.
3236 PH/04 Installed a modified version of Tony Finch's patch to make submission
3237 mode fix the return path as well as the Sender: header line, and to
3238 add a /name= option so that you can make the user's friendly name appear
3241 TF/03 Added the control = fakedefer ACL modifier.
3243 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3244 Mark Lowes for thorough testing.
3246 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3248 TK/03 Merged latest SRS patch from Miles Wilton.
3250 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3251 with the definition in sysexits.h (which is #included earlier).
3252 Fortunately, Exim does not actually use EX_OK. The code used to try to
3253 preserve the sysexits.h value, by assumimg that macro definitions were
3254 scanned for macro replacements. I have been disabused of this notion,
3255 so now the code just undefines EX_OK before #including unistd.h.
3257 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3258 in the smtp transport. When a block could not be written in a single
3259 write() function, the timeout was being re-applied to each part-write.
3260 This seems wrong - if the receiver was accepting one byte at a time it
3261 would take for ever. The timeout is now adjusted when this happens. It
3262 doesn't have to be particularly precise.
3264 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3265 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3267 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3268 without a domain if the check to <postmaster@domain> fails.
3270 SC/01 Eximstats: added -xls and the ability to specify output files
3271 (patch written by Frank Heydlauf).
3273 SC/02 Eximstats: use FileHandles for outputing results.
3275 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3277 SC/04 Eximstats: fixed display of large numbers with -nvr option
3279 SC/05 Eximstats: fixed merging of reports with empty tables.
3281 SC/06 Eximstats: added the -include_original_destination flag
3283 SC/07 Eximstats: removed tabs and trailing whitespace.
3285 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3287 TK/06 MBOX spool code: Add real "From " MBOX separator line
3288 so the .eml file is really in mbox format (even though
3289 most programs do not really care). Patch from Alex Miller.
3291 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3292 The latter is generated from $received_to and is only set if the
3293 message has one envelope recipient. SA can use these headers,
3294 obviously out-of-the-box. Patch from Alex Miller.
3296 PH/08 The ${def test on a variable was returning false if the variable's
3297 value was "0", contrary to what the specification has always said!
3298 The result should be true unless the variable is empty.
3300 PH/09 The syntax error of a character other than { following "${if
3301 def:variable_name" (after optional whitespace) was not being diagnosed.
3302 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3303 accidental colon was present, for example, could give incorrect results.
3305 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3306 result is used (not including appendfile, where other changes are about
3309 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3310 This involved changing a lot of size variables from int to off_t. It
3311 should work with maildirs and everything.
3313 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3314 spamd dying while we are connected to it.
3316 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3319 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3320 function for BDB 4.3.
3322 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3324 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3325 the sources for PCRE have been re-arranged and more files are now
3328 PH/15 The code I had for printing potentially long long variables in PH/11
3329 above was not the best (it lost precision). The length of off_t variables
3330 is now inspected at build time, and an appropriate printing format (%ld
3331 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3332 to be "long long int" or "long int". This is needed for the internal
3333 formatting function string_vformat().
3335 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3336 the configuration file to be ":syslog", then the script "guesses" where
3337 the logs files are, rather than using the compiled in default. In our
3338 case the guess is not the same as the compiled default, so the script
3339 suddenly stopped working when I started to use syslog. The patch checks
3340 to see if log_file_path is "". If so, it attempts to read it from exim
3341 with no configuration file to get the compiled in version, before it
3342 falls back to the previous guessing code."
3344 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3345 implementing BATV in an Exim configuration. See NewStuff for the gory
3348 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3349 Makefile that are specific to HP-UX.
3351 PH/18 If the "use_postmaster" option was set for a recipient callout together
3352 with the "random" option, the postmaster address was used as the MAIL
3353 FROM address for the random test, but not for the subsequent recipient
3354 test. It is now used for both.
3356 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3357 patch removes a few documentation additions to RFC 3028, because the
3358 latest draft now contains them. It adds the new en;ascii-case comparator
3359 and a new error check for 8bit text in MIME parts. Comparator and
3360 require names are now matched exactly. I enabled the subaddress
3361 extension, but it is not well tested yet (read: it works for me)."
3363 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3364 rework some of the code of TK/09 above to avoid the hardwired use of
3365 "%lld" and "long long". Replaced the call to snprintf() with a call to
3368 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3369 records point to non-existent hosts", "retry timeout exceeded", and
3370 "retry time not reached for any host after a long failure period".
3372 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3373 experimental DomainKeys support:
3375 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3376 (2) On an error such as an illegally used "control", the wrong name for
3377 the control was given.
3379 These problems did NOT occur unless DomainKeys support was compiled.
3381 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3383 PH/24 Added ${if match_ip condition.
3385 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3386 databases so that it will be absolutely obvious if a crash occurs in the
3387 DB library. This is a regular occurrence (often caused by mis-matched
3390 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3391 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3392 picked up on a user's system that detects such things. There doesn't seem
3393 to be a gcc warning option for this - only an attribute that has to be
3394 put on the function's prototype. It seems that in Fedora Core 4 they have
3395 set this on a number of new functions. No doubt there will be more in due
3398 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3399 specify a transport. However, if an address that was verified by such a
3400 router was the subject of a callout, Exim crashed because it tried to
3401 read the rcpt_include_affixes from the non-existent transport. Now it
3402 just assumes that the setting of that option is false. This bug was
3403 introduced by 4.51/PH/31.
3405 PH/28 Changed -d+all to exclude +memory, because that information is very
3406 rarely of interest, but it makes the output a lot bigger. People tend to
3407 do -d+all out of habit.
3409 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3410 code in os-type was giving problems when libc.so lives in lib64, like on
3413 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3414 aren't the modern standard, and it seems that some systems' include files
3415 don't always have them. Exim was already checking for some of the newer
3416 ones like T_AAAA, and defining it itself. I've added checks for all the
3417 record types that Exim uses.
3419 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3420 not automatically generating a new one, as it is supposed to. This
3421 prevented TLS from working. If the file did exist, but contained invalid
3422 data, a new version was generated, as expected. It was only the case of a
3423 non-existent file that was broken.
3425 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3426 with a change in libdomainkeys > 0.64.
3428 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3429 from DNS. If the selector record carries the flag, it now has
3430 precedence over the domain-wide flag.
3432 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3434 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3435 the use of a port name that isn't defined in /etc/services) occurred, the
3436 message was deferred as in a normal delivery, and thus remained on the
3437 spool, instead of being failed because of the mua_wrapper setting. This
3438 is now fixed, and I tidied up some of the mua_wrapper messages at the
3441 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3442 the messages in a hash of arrays rather than using individual hashes.
3443 This is a bit cleaner and results in dramatic memory savings, albeit
3444 at a slight CPU cost.
3446 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3447 as requested by Marc Sherman.
3449 SC/10 Eximstats: added histograms for user specified patterns as requested
3452 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3454 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3455 fopen() in the content-scanning modules that did not already have it.
3461 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3462 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3464 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3466 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3468 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3469 to test Sieve filters that use "vacation".
3471 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3472 that changes the way the GnuTLS parameters are stored in the cache file.
3473 The new format can be generated externally. For backward compatibility,
3474 if the data in the cache doesn't make sense, Exim assumes it has read an
3475 old-format file, and it generates new data and writes a new file. This
3476 means that you can't go back to an older release without removing the
3479 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3480 work if there are any delivery delays because "one_time" forces the
3481 parent to be marked "delivered", so its unseen clone is never tried
3482 again. For this reason, Exim now forbids the simultaneous setting of
3485 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3486 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3487 entry if you want to know the details. The fix, however, neglected to
3488 consider the case when local delivery batching is involved. The test for
3489 "previously delivered" was not happening when checking to see if an
3490 address could be batched with a previous (undelivered) one; under
3491 certain circumstances this could lead to multiple deliveries to the same
3494 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3495 in its include files, and this causes problems building Exim.
3497 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3498 header_syntax) but Exim was just ignoring anything given after a slash.
3499 In particular, this caused confusion with an attempt to use "verify =
3500 reverse_host_lookup/defer_ok". An error is now given when options are
3501 supplied for verify items that do not have them. (Maybe reverse_host_
3502 lookup should have a defer_ok option, but that's a different point.)
3504 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3505 defined by RFC 821) to 2048, because there were problems with some AUTH
3506 commands, and RFC 1869 says the size should be increased for extended
3507 SMTP commands that take arguments.
3509 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3512 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3513 "unknown" error; now it says that the functionality isn't in the binary.
3515 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3516 an address' error message when a string expansion fails (syntax or
3517 whatever). Otherwise the password may appear in the log. Following change
3518 PH/42 below, there is no longer a chance of it appearing in a bounce
3521 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3523 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3524 of its addresses was ever tried. (Bugzilla bug #2.)
3526 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3527 the result incorrectly in the debug output. (It correctly added a newline
3528 to what was transported.)
3530 TF/01 Added $received_time.
3532 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3533 commented out examples of how to interface to a virus scanner and to
3534 SpamAssassin. Also added commented examples of av_scanner and
3535 spamd_address settings.
3537 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3538 and controls are allowed in which ACLs. There were a couple of minor
3539 errors. Some of the entries in the conditions table (which is a table of
3540 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3541 negation of where the condition IS allowed.
3543 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3545 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3546 header file does not have a version number, so I've had to invent a new
3547 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3548 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3549 radiusclient), but it was contributed by a Radius user.
3551 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3552 files or directories, for OpenSSL.
3554 PH/19 When an Exim process that is running as root has to create an Exim log
3555 file, it does so in a subprocess that runs as exim:exim so as to get the
3556 ownership right at creation (otherwise, other Exim processes might see
3557 the file with the wrong ownership). There was no test for failure of this
3558 fork() call, which would lead to the process getting stuck as it waited
3559 for a non-existent subprocess. Forks do occasionally fail when resources
3560 run out. I reviewed all the other calls to fork(); they all seem to check
3563 PH/20 When checking for unexpected SMTP input at connect time (before writing
3564 the banner), Exim was not dealing correctly with a non-positive return
3565 from the read() function. If the client had disconnected by this time,
3566 the result was a log entry for a synchronization error with an empty
3567 string after "input=" when read() returned zero. If read() returned -1
3568 (an event I could not check), uninitialized data bytes were printed.
3569 There were reports of junk text (parts of files, etc) appearing after
3572 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3574 PH/22 Added support for macro redefinition, and (re)definition in between
3575 driver and ACL definitions.
3577 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3578 forgetting to use the resulting value; it was using the unexpanded value.
3580 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3581 hadn't been configured. The fix is from Juergen Kreileder, who
3582 understands it better than I do:
3584 "Here's what I see happening with three configured cyrus_sasl
3585 authenticators configured (plain, login, cram-md5):
3587 On startup auth_cyrus_sasl_init() gets called for each of these.
3588 This means three calls to sasl_listmech() without a specified mech_list.
3589 => SASL tests which mechs of all available mechs actually work
3590 => three warnings about OTP not working
3591 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3593 With the patch, sasl_listmech() also gets called three times. But now
3594 SASL's mech_list option is set to the server_mech specified in the the
3595 authenticator. Or in other words, the answer from sasl_listmech()
3596 gets limited to just the mech you're testing for (which is different
3598 => the return list contains just 'plain' or 'login', 'cram-md5' or
3599 nothing depending on the value of ob->server_mech.
3601 I've just tested the patch: Authentication still works fine,
3602 unavailable mechs specified in the exim configuration are still
3603 caught, and the auth.log warnings about OTP are gone."
3605 PH/25 When debugging is enabled, the contents of the command line are added
3606 to the debugging output, even when log_selector=+arguments is not
3609 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3610 answer is "GNU", and only if the return is "GNU/something" is the answer
3613 PH/27 $acl_verify_message is now set immediately after the failure of a
3614 verification in an ACL, and so is available in subsequent modifiers. In
3615 particular, the message can be preserved by coding like this:
3617 warn !verify = sender
3618 set acl_m0 = $acl_verify_message
3620 Previously, $acl_verify_message was set only while expanding "message"
3621 and "log_message" when a very denied access.
3623 PH/28 Modified OS/os.c-Linux with
3625 -#ifndef OS_LOAD_AVERAGE
3626 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3628 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3629 nomenclature these days.)
3631 PH/29 Installed patch from the Sieve maintainer that adds the options
3632 sieve_useraddress and sieve_subaddress to the redirect router.
3634 PH/30 In these circumstances:
3635 . Two addresses routed to the same list of hosts;
3636 . First host does not offer TLS;
3637 . First host accepts first address;
3638 . First host gives temporary error to second address;
3639 . Second host offers TLS and a TLS session is established;
3640 . Second host accepts second address.
3641 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3642 and peerdn, if requested) that were in fact used only for the second
3645 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3646 attention to any local part prefix or suffix that was matched by the
3647 router that accepted the address. It now behaves in the same way as it
3648 does for delivery: the affixes are removed from the local part unless
3649 rcpt_include_affixes is set on the transport.
3651 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3652 timeout during the DATA phase of an incoming message.
3654 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3655 applied a patch sanctioned by the Sieve maintainer.
3657 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3658 the uid or gid is negative. A case of a negative gid caused this to be
3659 noticed. The fix allows for either to be negative.
3661 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3662 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3665 PH/36 The change PH/12 above was broken. Fixed it.
3667 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3668 the grounds that routing the same address twice would always produce the
3669 same answer. This might have been true once, but it is certainly no
3670 longer true now. Routing a child address may depend on the previous
3671 routing that produced that child. Some complicated redirection strategies
3672 went wrong when messages had multiple recipients, and made Exim's
3673 behaviour dependent on the order in which the addresses were given.
3675 I have moved the duplicate checking until after the routing is complete.
3676 Exim scans the addresses that are assigned to local and remote
3677 transports, and removes any duplicates. This means that more work will be
3678 done, as duplicates will always all be routed, but duplicates are
3679 presumably rare, so I don't expect this is of any significance.
3681 For deliveries to pipes, files, and autoreplies, the duplicate checking
3682 still happens during the routing process, since they are not going to be
3685 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3686 It corrects a timeout issue with spamd. This is Ian's comment: "The
3687 background is that sometimes spamd either never reads data from a
3688 connection it has accepted, or it never writes response data. The exiscan
3689 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3690 blindly assumes that writes won't block so it may never time out."
3692 PH/39 Allow G after quota size as well as K and M.
3694 PH/40 The value set for $authenticated_id in an authenticator may not contain
3695 binary zeroes or newlines because the value is written to log lines and
3696 to spool files. There was no check on this. Now the value is run through
3697 the string_printing() function so that such characters are converted to
3698 printable escape sequences.
3700 PH/41 $message_linecount is a new variable that contains the total number of
3701 lines in the message. Compare $body_linecount, which is the count for the
3704 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3705 bounce and delay warning messages, except in certain special cases, which
3708 (a) An SMTP error message from a remote host;
3709 (b) A message specified in a :fail: redirection;
3710 (c) A message specified in a "fail" command in a system filter;
3711 (d) A message specified in a FAIL return from the queryprogram router;
3712 (e) A message specified by the cannot_route_message router option.
3714 In these cases only, Exim does include the error details in bounce and
3715 warning messages. There are also a few cases where bland messages such
3716 as "unrouteable address" or "local delivery error" are given.
3718 PH/43 $value is now also set for the "else" part of a ${run expansion.
3720 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3721 being worked on, but at least Exim now implements the latest version to
3724 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3725 process to complete was treated as a delivery failure, a timeout while
3726 writing the message to the pipe was logged, but erroneously treated as a
3727 successful delivery. Such timeouts include transport filter timeouts. For
3728 consistency with the overall process timeout, these timeouts are now
3729 treated as errors, giving rise to delivery failures by default. However,
3730 there is now a new Boolean option for the pipe transport called
3731 timeout_defer, which, if set TRUE, converts the failures into defers for
3732 both kinds of timeout. A transport filter timeout is now identified in
3735 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3736 systems where "make" and "gmake" are different, calling "gmake" at top
3737 level broke things. I've arranged for the value of $(MAKE) to be passed
3738 from the Makefile to this script so that it can call the same version of
3742 A note about Exim versions 4.44 and 4.50
3743 ----------------------------------------
3745 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3746 changes of various kinds. As a consequence, a big documentation update was
3747 needed. This delayed the release for rather longer than seemed good, especially
3748 in the light of a couple of (minor) security issues. Therefore, the changes
3749 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3750 release. So 4.44 and 4.50 are in effect two different branches that both start
3753 I have left the 4.50 change log unchanged; it contains all the changes since
3754 4.43. The change log for 4.44 is below; many of its items are identical to
3755 those for 4.50. This seems to be the most sensible way to preserve the
3756 historical information.
3762 1. Minor wording change to the doc/README.SIEVE file.
3764 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3765 computation of the current number of files was incorrect.
3767 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3768 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3771 4. Give more explanation in the error message when the command for a transport
3772 filter fails to execute.
3774 5. There are several places where Exim runs a non-Exim command in a
3775 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3776 was being done only for the command run by the queryprogram router. It is
3777 now done for all such subprocesses. The other cases are: ${run, transport
3778 filters, and the commands run by the lmtp and pipe transports.
3780 6. Added CONFIGURE_GROUP build-time option.
3782 7. Some older OS have a limit of 256 on the maximum number of file
3783 descriptors. Exim was using setrlimit() to set 1000 as a large value
3784 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3785 systems. I've change it so that if it can't get 1000, it tries for 256.
3787 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3788 was an oversight, and furthermore, ever since the addition of extra
3789 controls (e.g. 4.43/32), the checks on when to allow different forms of
3790 "control" were broken. There should now be diagnostics for all cases when a
3791 control that does not make sense is encountered.
3793 9. Added the /retain_sender option to "control=submission".
3795 10. $recipients is now available in the predata ACL (oversight).
3797 11. Tidy the search cache before the fork to do a delivery from a message
3798 received from the command line. Otherwise the child will trigger a lookup
3799 failure and thereby defer the delivery if it tries to use (for example) a
3800 cached ldap connection that the parent has called unbind on.
3802 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3803 of $address_data from the recipient verification was clobbered by the
3804 sender verification.
3806 13. The value of address_data from a sender verification is now available in
3807 $sender_address_data in subsequent conditions in the ACL statement.
3809 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3811 15. Added a new option "connect=<time>" to callout options, to set a different
3814 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3815 was its contents. (It was OK if the option was not defined at all.)
3817 17. A "Completed" log line is now written for messages that are removed from
3818 the spool by the -Mrm option.
3820 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3821 information about exactly what failed.
3823 19. Added -dd to debug only the daemon process.
3825 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3826 handles timeouts, both on the server side and network timeouts. Renamed the
3827 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3829 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3830 It is now set to "smtps".
3832 22. $host_address is now set to the target address during the checking of
3833 ignore_target_hosts.
3835 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3836 being passed; this would have caused $sender_host_name to have been used if
3837 matching the list had actually called for a host name (not very likely,
3838 since this list is usually IP addresses). A host name is now passed as
3841 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3842 code that specifies a non-restarting handler (typically sigaction() in
3843 modern systems) in an attempt to fix a rare and obscure crash bug.
3845 25. Narrowed the window for a race in the daemon that could cause it to ignore
3846 SIGCHLD signals. This is not a major problem, because they are used only to
3847 wake it up if nothing else does.
3849 26. A malformed maildirsize file could cause Exim to calculate negative values
3850 for the mailbox size or file count. Odd effects could occur as a result.
3851 The maildirsize information is now recalculated if the size or filecount
3854 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3855 support for a long time. Removed HAVE_SYS_VFS_H.
3857 28. Installed the latest version of exipick from John Jetmore.
3859 29. In an address list, if the pattern was not a regular expression, an empty
3860 subject address (from a bounce message) matched only if the pattern was an
3861 empty string. Non-empty patterns were not even tested. This was the wrong
3862 because it is perfectly reasonable to use an empty address as part of a
3863 database query. An empty address is now tested by patterns that are
3864 lookups. However, all the other forms of pattern expect the subject to
3865 contain a local part and a domain, and therefore, for them, an empty
3866 address still always fails if the pattern is not itself empty.
3868 30. Exim went into a mad DNS loop when attempting to do a callout where the
3869 host was specified on an smtp transport, and looking it up yielded more
3870 than one IP address.
3872 31. Re-factored the code for checking spool and log partition space into a
3873 function that finds that data and another that does the check. The former
3874 is then used to implement four new variables: $spool_space, $log_space,
3875 $spool_inodes, and $log_inodes.
3877 32. The RFC2047 encoding function was originally intended for short strings
3878 such as real names; it was not keeping to the 75-character limit for
3879 encoded words that the RFC imposes. It now respects the limit, and
3880 generates multiple encoded words if necessary. To be on the safe side, I
3881 have increased the buffer size for the ${rfc2047: expansion operator from
3884 33. It is now permitted to omit both strings after an "if" condition; if the
3885 condition is true, the result is "true". As before, when the second string
3886 is omitted, a false condition yields an empty string. This makes it less
3887 cumbersome to write custom ACL and router conditions.
3889 34. Failure to deliver a bounce message always caused it to be frozen, even if
3890 there was an errors_to setting on the router. The errors_to setting is now
3893 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3894 canonical form (fully expanded) before being placed in
3895 $sender_host_address.
3897 36. The table in the code that translates DNS record types into text (T_A to
3898 "A" for instance) was missing entries for NS and CNAME. It is just possible
3899 that this could have caused confusion if both these types were looked up
3900 for the same domain, because the text type is used as part of Exim's
3901 per-process caching. But the chance of anyone hitting this buglet seems
3904 37. The dnsdb lookup has been extended in a number of ways.
3906 (1) There is a new type, "zns", which walks up the domain tree until it
3907 finds some nameserver records. It should be used with care.
3909 (2) There is a new type, "mxh", which is like "mx" except that it returns
3910 just the host names, not the priorities.
3912 (3) It is now possible to give a list of domains (or IP addresses) to be
3913 looked up. The behaviour when one of the lookups defers can be
3914 controlled by a keyword.
3916 (4) It is now possible to specify the separator character for use when
3917 multiple records are returned.
3919 38. The dnslists ACL condition has been extended: it is now possible to supply
3920 a list of IP addresses and/or domains to be looked up in a particular DNS
3923 39. Added log_selector=+queue_time_overall.
3925 40. When running the queue in the test harness, wait just a tad after forking a
3926 delivery process, to get repeatability of debugging output.
3928 41. Include certificate and key file names in error message when GnuTLS fails
3929 to set them up, because the GnuTLS error message doesn't include the name
3930 of the failing file when there is a problem reading it.
3932 42. Allow both -bf and -bF in the same test run.
3934 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3936 44. The "Exiscan patch" is now merged into the mainline Exim source.
3938 45. Sometimes the final signoff response after QUIT could fail to get
3939 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3940 < 0 before doing a fflush(). This bug looks as though it goes back to the
3941 introduction of TLS in release 3.20, but "sometimes" must have been rare
3942 because the tests only now provoked it.
3944 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3945 (this can affect the format of dates).
3947 47. exim_tidydb, when checking for the continued existence of a message for
3948 which it has found a message-specific retry record, was not finding
3949 messages that were in split spool directories. Consequently, it was
3950 deleting retry records that should have stayed in existence.
3952 48. Steve fixed some bugs in eximstats.
3954 49. The SPA authentication driver was not abandoning authentication and moving
3955 on to the next authenticator when an expansion was forced to fail,
3956 contradicting the general specification for all authenticators. Instead it
3957 was generating a temporary error. It now behaves as specified.
3959 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3960 (the order specifies the preference for clients). The order is now AES256,
3961 AES128, 3DES, ARCFOUR128.
3963 51. Small patch to Sieve code - explicitly set From: when generating an
3966 52. Exim crashed if a remote delivery caused a very long error message to be
3967 recorded - for instance if somebody sent an entire SpamAssassin report back
3968 as a large number of 550 error lines. This bug was coincidentally fixed by
3969 increasing the size of one of Exim's internal buffers (big_buffer) that
3970 happened as part of the Exiscan merge. However, to be on the safe side, I
3971 have made the code more robust (and fixed the comments that describe what
3974 53. Now that there can be additional text after "Completed" in log lines (if
3975 the queue_time_overall log selector is set), a one-byte patch to exigrep
3976 was needed to allow it to recognize "Completed" as not the last thing in
3979 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3980 patch that reportedly fixes this has been added. I am not expert enough to
3981 create a test for it. This is what the patch creator wrote:
3983 "I found a little strange behaviour of ldap code when working with
3984 Windows 2003 AD Domain, where users was placed in more than one
3985 Organization Units. When I tried to give exim partial DN, the exit code
3986 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3987 But simultaneously result of request was absolutely normal ldap result,
3988 so I produce this patch..."
3990 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3991 REFERENCE, so I have modified the code to exclude the patch when that macro
3994 55. Some experimental protocols are using DNS PTR records for new purposes. The
3995 keys for these records are domain names, not reversed IP addresses. The
3996 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3997 leaves it alone. Component reversal etc. now happens only for IP addresses.
4000 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4002 57. Double the size of the debug message buffer (to 2048) so that more of very
4003 long debug lines gets shown.
4005 58. The exicyclog utility now does better if the number of log files to keep
4006 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4008 59. Two changes related to the smtp_active_hostname option:
4010 (1) $smtp_active_hostname is now available as a variable.
4011 (2) The default for smtp_banner uses $smtp_active_hostname instead
4012 of $primary_hostname.
4014 60. The host_aton() function is supposed to be passed a string that is known
4015 to be a valid IP address. However, in the case of IPv6 addresses, it was
4016 not checking this. This is a hostage to fortune. Exim now panics and dies
4017 if the condition is not met. A case was found where this could be provoked
4018 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4019 components; fortuitously, this particular loophole had already been fixed
4020 by change 4.50/55 above.
4022 If there are any other similar loopholes, the new check in host_aton()
4023 itself should stop them being exploited. The report I received stated that
4024 data on the command line could provoke the exploit when Exim was running as
4025 exim, but did not say which command line option was involved. All I could
4026 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4027 running as the user.
4030 61. There was a buffer overflow vulnerability in the SPA authentication code
4031 (which came originally from the Samba project). I have added a test to the
4032 spa_base64_to_bits() function which I hope fixes it.
4035 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4036 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4038 63. The daemon start-up calls getloadavg() while still root for those OS that
4039 need the first call to be done as root, but it missed one case: when
4040 deliver_queue_load_max is set with deliver_drop_privilege. This is
4041 necessary for the benefit of the queue runner, because there is no re-exec
4042 when deliver_drop_privilege is set.
4044 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4045 This has been fixed.
4047 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4048 was in use, was not putting the data itself into the right store pool;
4049 consequently, it could be overwritten for a subsequent message in the same
4050 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4053 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4055 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4056 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4057 the function were treating the return as a boolean value, which happened to
4058 work because 0=false and not-0=true, but is not correct code.
4060 68. The host_aton() function was not handling scoped IPv6 addresses (those
4061 with, for example, "%eth0" on the end) correctly.
4063 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4064 negated items (that is, ~something) in unsigned ints. Some compilers
4065 apparently mutter when there is no cast.
4067 70. If an address verification called from an ACL failed, and did not produce a
4068 user-specific message (i.e. there was only a "system" message), nothing was
4069 put in $acl_verify_message. In this situation, it now puts the system
4072 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4073 session; change 4.31/43 added the unwanted input to the log line - except
4074 that it did not do this in the start of session case. It now does.
4076 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4077 This is wrong and can cause the other end to generate a synchronization
4078 error if it is another Exim or anything else that does the synchronization
4079 check. A QUIT command is no longer sent after a timeout.
4081 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4082 during host lookups.
4084 74. The defer_ok option of callout verification was not working if it was used
4085 when verifying addresses in header lines, that is, for this case:
4087 verify = header_sender/callout=defer_ok
4089 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4090 those file descriptors could be used for SMTP connections. If anything
4091 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4092 could be sent to the SMTP client, causing chaos. The daemon now opens
4093 stdin, stdout, and stderr to /dev/null when it puts itself into the
4096 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4097 log by default. The user can override this with suitable Perl magic.
4099 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4100 the log message when discard triggers, was not working for the DATA ACL or
4101 for the non-SMTP ACL.
4103 78. Error message wording change in sieve.c.
4105 79. If smtp_accept_max_per_host was set, the number of connections could be
4106 restricted to fewer than expected, because the daemon was trying to set up
4107 a new connection before checking whether the processes handling previous
4108 connections had finished. The check for completed processes is now done
4109 earlier. On busy systems, this bug wouldn't be noticed because something
4110 else would have woken the daemon, and it would have reaped the completed
4113 80. If a message was submitted locally by a user whose login name contained one
4114 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4115 It caused a spool format error. I have fixed the spool reading code. A
4116 related problem was that the "from" clause in the Received: line became
4117 illegal because of the space(s). It is now covered by ${quote_local_part.
4119 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4122 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4124 83. Patch from Sieve maintainer to fix unterminated string problem in
4125 "vacation" handling.
4127 84. Some minor changes to the Linux configuration files to help with other
4128 OS variants using glibc.
4130 85. One more patch for Sieve to update vacation handling to latest spec.
4133 ----------------------------------------------------
4134 See the note above about the 4.44 and 4.50 releases.
4135 ----------------------------------------------------
4141 1. Change 4.43/35 introduced a bug that caused file counts to be
4142 incorrectly computed when quota_filecount was set in an appendfile
4145 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4146 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4149 3. Give more explanation in the error message when the command for a transport
4150 filter fails to execute.
4152 4. There are several places where Exim runs a non-Exim command in a
4153 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4154 was being done only for the command run by the queryprogram router. It is
4155 now done for all such subprocesses. The other cases are: ${run, transport
4156 filters, and the commands run by the lmtp and pipe transports.
4158 5. Some older OS have a limit of 256 on the maximum number of file
4159 descriptors. Exim was using setrlimit() to set 1000 as a large value
4160 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4161 systems. I've change it so that if it can't get 1000, it tries for 256.
4163 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4164 was an oversight, and furthermore, ever since the addition of extra
4165 controls (e.g. 4.43/32), the checks on when to allow different forms of
4166 "control" were broken. There should now be diagnostics for all cases when a
4167 control that does not make sense is encountered.
4169 7. $recipients is now available in the predata ACL (oversight).
4171 8. Tidy the search cache before the fork to do a delivery from a message
4172 received from the command line. Otherwise the child will trigger a lookup
4173 failure and thereby defer the delivery if it tries to use (for example) a
4174 cached ldap connection that the parent has called unbind on.
4176 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4177 of $address_data from the recipient verification was clobbered by the
4178 sender verification.
4180 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4181 was its contents. (It was OK if the option was not defined at all.)
4183 11. A "Completed" log line is now written for messages that are removed from
4184 the spool by the -Mrm option.
4186 12. $host_address is now set to the target address during the checking of
4187 ignore_target_hosts.
4189 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4190 being passed; this would have caused $sender_host_name to have been used if
4191 matching the list had actually called for a host name (not very likely,
4192 since this list is usually IP addresses). A host name is now passed as
4195 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4196 code that specifies a non-restarting handler (typically sigaction() in
4197 modern systems) in an attempt to fix a rare and obscure crash bug.
4199 15. Narrowed the window for a race in the daemon that could cause it to ignore
4200 SIGCHLD signals. This is not a major problem, because they are used only to
4201 wake it up if nothing else does.
4203 16. A malformed maildirsize file could cause Exim to calculate negative values
4204 for the mailbox size or file count. Odd effects could occur as a result.
4205 The maildirsize information is now recalculated if the size or filecount
4208 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4209 support for a long time. Removed HAVE_SYS_VFS_H.
4211 18. Updated exipick to current release from John Jetmore.
4213 19. Allow an empty sender to be matched against a lookup in an address list.
4214 Previously the only cases considered were a regular expression, or an
4217 20. Exim went into a mad DNS lookup loop when doing a callout where the
4218 host was specified on the transport, if the DNS lookup yielded more than
4221 21. The RFC2047 encoding function was originally intended for short strings
4222 such as real names; it was not keeping to the 75-character limit for
4223 encoded words that the RFC imposes. It now respects the limit, and
4224 generates multiple encoded words if necessary. To be on the safe side, I
4225 have increased the buffer size for the ${rfc2047: expansion operator from
4228 22. Failure to deliver a bounce message always caused it to be frozen, even if
4229 there was an errors_to setting on the router. The errors_to setting is now
4232 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4233 canonical form (fully expanded) before being placed in
4234 $sender_host_address.
4236 24. Updated eximstats to version 1.33
4238 25. Include certificate and key file names in error message when GnuTLS fails
4239 to set them up, because the GnuTLS error message doesn't include the name
4240 of the failing file when there is a problem reading it.
4242 26. Expand error message when OpenSSL has problems setting up cert/key files.
4245 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4246 (this can affect the format of dates).
4248 28. exim_tidydb, when checking for the continued existence of a message for
4249 which it has found a message-specific retry record, was not finding
4250 messages that were in split spool directories. Consequently, it was
4251 deleting retry records that should have stayed in existence.
4253 29. eximstats updated to version 1.35
4254 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4255 1.35 - bugfix such that pie charts by volume are generated correctly
4257 30. The SPA authentication driver was not abandoning authentication and moving
4258 on to the next authenticator when an expansion was forced to fail,
4259 contradicting the general specification for all authenticators. Instead it
4260 was generating a temporary error. It now behaves as specified.
4262 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4263 (the order specifies the preference for clients). The order is now AES256,
4264 AES128, 3DES, ARCFOUR128.
4266 31. Small patch to Sieve code - explicitly set From: when generating an
4269 32. Exim crashed if a remote delivery caused a very long error message to be
4270 recorded - for instance if somebody sent an entire SpamAssassin report back
4271 as a large number of 550 error lines. This bug was coincidentally fixed by
4272 increasing the size of one of Exim's internal buffers (big_buffer) that
4273 happened as part of the Exiscan merge. However, to be on the safe side, I
4274 have made the code more robust (and fixed the comments that describe what
4277 33. Some experimental protocols are using DNS PTR records for new purposes. The
4278 keys for these records are domain names, not reversed IP addresses. The
4279 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4280 leaves it alone. Component reversal etc. now happens only for IP addresses.
4283 34. The host_aton() function is supposed to be passed a string that is known
4284 to be a valid IP address. However, in the case of IPv6 addresses, it was
4285 not checking this. This is a hostage to fortune. Exim now panics and dies
4286 if the condition is not met. A case was found where this could be provoked
4287 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4288 components; fortuitously, this particular loophole had already been fixed
4289 by change 4.50/55 or 4.44/33 above.
4291 If there are any other similar loopholes, the new check in host_aton()
4292 itself should stop them being exploited. The report I received stated that
4293 data on the command line could provoke the exploit when Exim was running as
4294 exim, but did not say which command line option was involved. All I could
4295 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4296 running as the user.
4299 35. There was a buffer overflow vulnerability in the SPA authentication code
4300 (which came originally from the Samba project). I have added a test to the
4301 spa_base64_to_bits() function which I hope fixes it.
4304 36. The daemon start-up calls getloadavg() while still root for those OS that
4305 need the first call to be done as root, but it missed one case: when
4306 deliver_queue_load_max is set with deliver_drop_privilege. This is
4307 necessary for the benefit of the queue runner, because there is no re-exec
4308 when deliver_drop_privilege is set.
4310 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4311 was in use, was not putting the data itself into the right store pool;
4312 consequently, it could be overwritten for a subsequent message in the same
4313 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4316 38. Sometimes the final signoff response after QUIT could fail to get
4317 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4318 < 0 before doing a fflush(). This bug looks as though it goes back to the
4319 introduction of TLS in release 3.20, but "sometimes" must have been rare
4320 because the tests only now provoked it.
4326 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4327 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4328 Now it returns an integer. A number of calls were still expecting a T/F
4329 return. Fortuitously, in all cases, the tests worked in OK situations,
4330 which is the norm. However, things would have gone wrong on any write
4331 failures on the smtp file descriptor. This function is used when sending
4332 messages over SMTP and also when doing verify callouts.
4334 2. When Exim is called to do synchronous delivery of a locally submitted
4335 message (the -odf or -odi options), it no longer closes stderr before doing
4338 3. Implemented the mua_wrapper option.
4340 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4342 5. Implemented the functions header_remove(), header_testname(),
4343 header_add_at_position(), and receive_remove_recipient(), and exported them
4346 6. If an ACL "warn" statement specified the addition of headers, Exim already
4347 inserted X-ACL-Warn: at the start if there was no header name. However, it
4348 was not making this test for the second and subsequent header lines if
4349 there were newlines in the string. This meant that an invalid header could
4350 be inserted if Exim was badly configured.
4352 7. Allow an ACL "warn" statement to add header lines at the start or after all
4353 the Received: headers, as well as at the end.
4355 8. Added the rcpt_4xx retry error code.
4357 9. Added postmaster_mailfrom=xxx to callout verification option.
4359 10. Added mailfrom=xxxx to the callout verification option, for verify=
4362 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4363 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4365 12. Inserted some casts to stop certain compilers warning when using pointer
4366 differences as field lengths or precisions in printf-type calls (mostly
4367 affecting debugging statements).
4369 13. Added optional readline() support for -be (dynamically loaded).
4371 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4372 same clock tick as a message's arrival, so that its received time was the
4373 same as the "first fail" time on the retry record, and that message
4374 remained on the queue past the ultimate address timeout, every queue runner
4375 would try a delivery (because it was past the ultimate address timeout) but
4376 after another failure, the ultimate address timeout, which should have then
4377 bounced the address, did not kick in. This was a "< instead of <=" error;
4378 in most cases the first failure would have been in the next clock tick
4379 after the received time, and all would be well.
4381 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4382 being recognized when the domain list was tested by the match_domain
4383 condition in an expansion string.
4385 16. Added the ${str2b64: operator.
4387 17. Exim was always calling setrlimit() to set a large limit for the number of
4388 processes, without checking whether the existing limit was already
4389 adequate. (It did check for the limit on file descriptors.) Furthermore,
4390 errors from getrlimit() and setrlimit() were being ignored. Now they are
4391 logged to the main and panic logs, but Exim does carry on, to try to do its
4392 job under whatever limits there are.
4394 18. Imported PCRE 5.0.
4396 19. Trivial typo in log message " temporarily refused connection" (the leading
4399 20. If the log selector return_path_on_delivery was set and an address was
4400 redirected to /dev/null, the delivery process crashed because it assumed
4401 that a return path would always be set for a "successful" delivery. In this
4402 case, the whole delivery is bypassed as an optimization, and therefore no
4405 21. Internal re-arrangement: the function for sending a challenge and reading
4406 a response while authentication was assuming a zero-terminated challenge
4407 string. It's now changed to take a pointer and a length, to allow for
4408 binary data in such strings.
4410 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4412 23. Exim was not respecting finduser_retries when seeking the login of the
4413 uid under which it was called; it was always trying 10 times. (The default
4414 setting of finduser_retries is zero.) Also, it was sleeping after the final
4415 failure, which is pointless.
4417 24. Implemented tls_on_connect_ports.
4419 25. Implemented acl_smtp_predata.
4421 26. If the domain in control=submission is set empty, Exim assumes that the
4422 authenticated id is a complete email address when it generates From: or
4423 Sender: header lines.
4425 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4426 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4427 chown and chgrp in /bin and hostname in /usr/bin.
4429 28. Exim was keeping the "process log" file open after each use, just as it
4430 does for the main log. This opens the possibility of it remaining open for
4431 long periods when the USR1 signal hits a daemon. Occasional processlog
4432 errors were reported, that could have been caused by this. Anyway, it seems
4433 much more sensible not to leave this file open at all, so that is what now
4436 29. The long-running daemon process does not normally write to the log once it
4437 has entered its main loop, and it closes the log before doing so. This is
4438 so that log files can straightforwardly be renamed and moved. However,
4439 there are a couple of unusual error situations where the daemon does write
4440 log entries, and I had neglected to close the log afterwards.
4442 30. The text of an SMTP error response that was received during a remote
4443 delivery was being truncated at 512 bytes. This is too short for some of
4444 the long messages that one sometimes sees. I've increased the limit to
4447 31. It is now possible to make retry rules that apply only when a message has a
4448 specific sender, in particular, an empty sender.
4450 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4451 it possible to be selective about when SMTP synchronization is enforced.
4453 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4455 32. Implemented hosts_connection_nolog.
4457 33. Added an ACL for QUIT.
4459 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4462 35. Added mailbox_size and mailbox_filecount to appendfile.
4464 36. Added control = no_multiline_responses to ACLs.
4466 37. There was a bug in the logic of the code that waits for the clock to tick
4467 in the case where the clock went backwards by a substantial amount such
4468 that the microsecond fraction of "now" was more than the microsecond
4469 fraction of "then" (but the whole seconds number was less).
4471 38. Added support for the libradius Radius client library this is found on
4472 FreeBSD (previously only the radiusclient library was supported).
4478 1. When certain lookups returned multiple values in the form name=value, the
4479 quoting of the values was not always being done properly. Specifically:
4480 (a) If the value started with a double quote, but contained no whitespace,
4482 (b) If the value contained whitespace other than a space character (i.e.
4483 tabs or newlines or carriage returns) it was not quoted.
4484 This fix has been applied to the mysql and pgsql lookups by writing a
4485 separate quoting function and calling it from the lookup code. The fix
4486 should probably also be applied to nisplus, ibase and oracle lookups, but
4487 since I cannot test any of those, I have not disturbed their existing code.
4489 2. A hit in the callout cache for a specific address caused a log line with no
4490 reason for rejecting RCPT. Now it says "Previous (cached) callout
4491 verification failure".
4493 3. There was an off-by-one bug in the queryprogram router. An over-long
4494 return line was truncated at 256 instead of 255 characters, thereby
4495 overflowing its buffer with the terminating zero. As well as fixing this, I
4496 have increased the buffer size to 1024 (and made a note to document this).
4498 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4499 when Exim is waiting for an SMTP response from a remote server, Exim
4500 restarts its select() call on the socket, thereby resetting its timeout.
4501 This is not a problem when such interrupts are rare. Somebody set up a cron
4502 job to run exiwhat every 2 minutes, which is less than the normal select()
4503 timeout (5 or 10 minutes). This meant that the select() timeout never
4504 kicked in because it was always reset. I have fixed this by comparing the
4505 time when an interrupt arrives with the time at the start of the first call
4506 to select(). If more time than the timeout has elapsed, the interrupt is
4507 treated as a timeout.
4509 5. Some internal re-factoring in preparation for the addition of Sieve
4510 extensions (by MH). In particular, the "personal" test is moved to a
4511 separate function, and given an option for scanning Cc: and Bcc: (which is
4512 not set for Exim filters).
4514 6. When Exim created an email address using the login of the caller as the
4515 local part (e.g. when creating a From: or Sender: header line), it was not
4516 quoting the local part when it contained special characters such as @.
4518 7. Installed new OpenBSD configuration files.
4520 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4521 try to make them clearer.
4523 9. Callout options, other than the timeout value, were being ignored when
4524 verifying sender addresses in header lines. For example, when using
4526 verify = header_sender/callout=no_cache
4528 the cache was (incorrectly) being used.
4530 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4531 only the Cygwin environment.
4533 11. When return_path_on_delivery was set as a log selector, if different remote
4534 addresses in the same message used different return paths and parallel
4535 remote delivery occurred, the wrong values would sometimes be logged.
4536 (Whenever a remote delivery process finished, the return path value from
4537 the most recently started remote delivery process was logged.)
4539 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4540 lines when AUTH and/or TLS are in use. This is the "received protocol"
4541 field. Exim used to use "asmtp" for authenticated SMTP, without any
4542 indication (in the protocol name) for TLS use. Now it follows the RFC and
4543 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4544 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4545 names appear in log lines as well as in Received: header lines.
4547 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4548 extensions, and comparison tests, and to fix some bugs.
4550 14. Changes to the "personal" filter test:
4552 (1) The test was buggy in that it was just doing the equivalent of
4553 "contains" tests on header lines. For example, if a user's address was
4554 anne@some.where, the "personal" test would incorrectly be true for
4556 To: susanne@some.where
4558 This test is now done by extracting each address from the header in turn,
4559 and checking the entire address. Other tests that are part of "personal"
4560 are now done using regular expressions (for example, to check local parts
4561 of addresses in From: header lines).
4563 (2) The list of non-personal local parts in From: addresses has been
4564 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4565 taken from the Sieve specification recommendations.
4567 (3) If the message contains any header line starting with "List-" it is
4568 treated as non-personal.
4570 (4) The test for "circular" in the Subject: header line has been removed
4571 because it now seems ill-conceived.
4573 15. Minor typos in src/EDITME comments corrected.
4575 16. Installed latest exipick from John Jetmore.
4577 17. If headers_add on a router specified a text string that was too long for
4578 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4579 of string_sprintf() is now avoided.
4581 18. $message_body_size was not set (it was always zero) when running the DATA
4582 ACL and the local_scan() function.
4584 19. For the "mail" command in an Exim filter, no default was being set for
4585 the once_repeat time, causing a random time value to be used if "once" was
4586 specified. (If the value happened to be <= 0, no repeat happened.) The
4587 default is now 0s, meaning "never repeat". The "vacation" command was OK
4588 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4589 (I found it when inspecting the code).
4591 20. There is now an overall timeout for performing a callout verification. It
4592 defaults to 4 times the callout timeout, which applies to individual SMTP
4593 commands during the callout. The overall timeout applies when there is more
4594 than one host that can be tried. The timeout is checked before trying the
4595 next host. This prevents very long delays if there are a large number of
4596 hosts and all are timing out (e.g. when the network connections are timing
4597 out). The value of the overall timeout can be changed by specifying an
4598 additional sub-option for "callout", called "maxwait". For example:
4600 verify = sender/callout=5s,maxwait=20s
4602 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4603 to the end before writing, but this should make it even safer).
4605 22. Exim was forgetting that it had advertised PIPELINING for the second and
4606 subsequent messages on an SMTP connection. It was also not resetting its
4607 memory on STARTTLS and an internal HELO.
4609 23. When Exim logs an SMTP synchronization error within a session, it now
4610 records whether PIPELINING has been advertised or not.
4612 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4613 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4614 rather than long int.
4616 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4618 26. Added the never_mail option to autoreply.
4624 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4625 crash if the getsockname() call failed; this can happen if a connection is
4626 closed very soon after it is established. The problem was simply in the
4627 order in which certain operations were done, causing Exim to try to write
4628 to the SMTP stream before it had set up the file descriptor. The bug has
4629 been fixed by making things happen in the correct order.
4635 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4636 before the connection was closed, thus losing the rejection response.
4638 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4639 some early Solaris releases, but causes trouble in current releases where
4640 socklen_t is defined.
4642 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4645 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4648 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4649 that releasing the top of it at the end releases what was used for sub-
4650 expansions (unless the block got too big). However, discard this block if
4651 the first thing is a variable or header, so that we can use its block when
4652 it is dynamic (useful for very large $message_headers, for example).
4654 6. Lookups now cache *every* query, not just the most recent. A new, separate
4655 store pool is used for this. It can be recovered when all lookup caches are
4656 flushed. Lookups now release memory at the end of their result strings.
4657 This has involved some general refactoring of the lookup sources.
4659 7. Some code has been added to the store_xxx() functions to reduce the amount
4660 of flapping under certain conditions.
4662 8. log_incoming_interface used to affect only the <= reception log lines. Now
4663 it causes the local interface and port to be added to several more SMTP log
4664 lines, for example "SMTP connection from", and rejection lines.
4666 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4668 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4670 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4671 could be overwritten at the end of the current message (or the start of a
4672 new message if it was set in a HELO ACL). The value is now preserved for
4673 the duration of the SMTP connection.
4675 12. If a transport had a headers_rewrite setting, and a matching header line
4676 contained an unqualified address, that address was qualified, even if it
4677 did not match any rewriting rules. The underlying bug was that the values
4678 of the flags that permit the existence of unqualified sender and recipient
4679 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4680 non-local messages, and by -bnq for local messages) were not being
4681 preserved with the message after it was received.
4683 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4684 "next input=" as part of the text comprising the host identity instead of
4685 the correct text. The code was using the same buffer for two different
4686 strings. However, depending on which order the printing function evaluated
4687 its arguments, the bug did not always show up. Under Linux, for example, my
4688 test suite worked just fine.
4690 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4691 doesn't work with some older versions of Perl. It has been changed to "my",
4692 which in any case is probably the better facility to use.
4694 15. A really picky compiler found some instances of statements for creating
4695 error messages that either had too many or two few arguments for the format
4698 16. The size of the buffer for calls to the DNS resolver has been increased
4699 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4700 for addresses that have a lot of PTR records. This alleviates a problem; it
4701 does not fully solve it.
4703 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4704 buffer now truncates the list and logs the incident, which is the same
4705 action as happens when Exim is looking up a host name and its aliases.
4706 Previously in this situation something unpredictable would happen;
4707 sometimes it was "internal error: store_reset failed".
4709 18. If a server dropped the connection unexpectedly when an Exim client was
4710 using GnuTLS and trying to read a response, the client delivery process
4711 crashed while trying to generate an error log message.
4713 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4714 string, for example:
4716 warn message = H1: something\nH2: something
4718 the text was added as a single header line from Exim's point of view
4719 though it ended up OK in the delivered message. However, searching for the
4720 second and subsequent header lines using $h_h2: did not work. This has been
4721 fixed. Similarly, if a system filter added multiple headers in this way,
4722 the routers could not see them.
4724 20. Expanded the error message when iplsearch is called with an invalid key to
4725 suggest using net-iplsearch in a host list.
4727 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4728 ACLs are no longer actually imposed (and a message to that effect is
4731 22. If a "gecos" field in a passwd entry contained escaped characters, in
4732 particular, if it contained a \" sequence, Exim got it wrong when building
4733 a From: or a Sender: header from that name. A second bug also caused
4734 incorrect handling when an unquoted " was present following a character
4735 that needed quoting.
4737 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4738 was not being matched caselessly.
4740 24. Arranged for all hyphens in the exim.8 source to be escaped with
4743 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4744 properly. Recipient callout cache records were still being keyed to include
4745 the sender, even when use_sender was set false. This led to far more
4746 callouts that were necessary. The sender is no longer included in the key
4747 when use_sender is false.
4749 26. Added "control = submission" modifier to ACLs.
4751 27. Added the ${base62d: operator to decode base 62 numbers.
4753 28. dnsdb lookups can now access SRV records.
4755 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4756 the configuration file.
4758 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4759 (-v) mode. This makes the output for a verbose queue run more intelligible.
4761 31. Added a use_postmaster feature to recipient callouts.
4763 32. Added the $body_zerocount variable, containing the number of binary zero
4764 bytes in the message body.
4766 33. The time of last modification of the "new" subdirectory is now used as the
4767 "mailbox time last read" when there is a quota error for a maildir
4770 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4772 35. Added +ignore_unknown as a special item in host lists.
4774 36. Code for decoding IPv6 addresses in host lists is now included, even if
4775 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4776 address was recognized as an IP address, but was then not correctly decoded
4777 into binary, causing unexpected and incorrect effects when compared with
4784 1. Very minor rewording of debugging text in manualroute to say "list of
4785 hosts" instead of "hostlist".
4787 2. If verify=header_syntax was set, and a header line with an unqualified
4788 address (no domain) and a large number of spaces between the end of the
4789 name and the colon was received, the reception process suffered a buffer
4790 overflow, and (when I tested it) crashed. This was caused by some obsolete
4791 code that should have been removed. The fix is to remove it!
4793 3. When running in the test harness, delay a bit after writing a bounce
4794 message to get a bit more predictability in the log output.
4796 4. Added a call to search_tidyup() just before forking a reception process. In
4797 theory, someone could use a lookup in the expansion of smtp_accept_max_
4798 per_host which, without the tidyup, could leave open a database connection.
4800 5. Added the variables $recipient_data and $sender_data which get set from a
4801 lookup success in an ACL "recipients" or "senders" condition, or a router
4802 "senders" option, similar to $domain_data and $local_part_data.
4804 6. Moved the writing of debug_print from before to after the "senders" test
4807 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4808 problems for message scanning, either using a data ACL, or using
4809 local_scan() because the Received: header was not generated till after they
4810 were called (in order to set the time as the time of reception completion).
4811 I have revised the way this works. The header is now generated after the
4812 body is received, but before the ACL or local_scan() are called. After they
4813 are run, the timestamp in the header is updated.
4819 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4820 before starting a queue runner without re-exec. This happened only when
4821 deliver_drop_privilege was set or when the Exim user was set to root. The
4822 effect of the bug was that timeouts during subsequent deliveries caused
4823 crashes instead of being properly handled. The handler is now left at its
4824 default (and expected) setting.
4826 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4827 message, again when deliver_drop_privilege is set or Exim is run as root.
4828 The bug described in (1) was not present in this case, but the tidying up
4829 of the other signals was missing. I have made the two cases consistent.
4831 3. The ignore_target_hosts setting on a manualroute router was being ignored
4832 for hosts that were looked up using the /MX notation.
4834 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4837 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4838 operated on the sender address. After changing the $sender_address to <>
4839 for the sender address verify, Exim was re-instated it as the original
4840 (before rewriting) address, but remembering that it had rewritten it, so it
4841 wasn't rewriting it again. This bug also had the effect of breaking the
4842 sender address verification caching when the sender address was rewritten.
4844 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4845 This has been changed so that if the ip literal address matches
4846 ignore_target_hosts, the router declines.
4848 7. Added expansion conditions match_domain, match_address, and match_local_
4849 part (NOT match_host).
4851 8. The placeholder for the Received: header didn't have a length field set.
4853 9. Added code to Exim itself and to exim_lock to test for a specific race
4854 condition that could lead to file corruption when using MBX delivery. The
4855 issue is with the lockfile that is created in /tmp. If this file is removed
4856 after a process has opened it but before that process has acquired a lock,
4857 there is the potential for a second process to recreate the file and also
4858 acquire a lock. This could lead to two Exim processes writing to the file
4859 at the same time. The added code performs the same test as UW imapd; it
4860 checks after acquiring the lock that its file descriptor still refers to
4861 the same named file.
4863 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4864 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4867 11. Added the smtp_active_hostname option. If used, this will typically be made
4868 to depend on the incoming interface address. Because $interface_address is
4869 not set up until the daemon has forked a reception process, error responses
4870 that can happen earlier (such as "too many connections") no longer contain
4873 12. If an expansion in a condition on a "warn" statement fails because a lookup
4874 defers, the "warn" statement is abandoned, and the next ACL statement is
4875 processed. Previously this caused the whole ACL to be aborted.
4877 13. Added the iplsearch lookup type.
4879 14. Added ident_timeout as a log selector.
4881 15. Added tls_certificate_verified as a log selector.
4883 16. Added a global option tls_require_ciphers (compare the smtp transport
4884 option of the same name). This controls incoming TLS connections.
4886 17. I finally figured out how to make tls_require_ciphers do a similar thing
4887 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4888 before starting the TLS session.
4890 18. Tabs are now shown as \t in -bP output.
4892 19. If the log selector return_path_on_delivery was set, Exim crashed when
4893 bouncing a message because it had too many Received: header lines.
4895 20. If two routers both had headers_remove settings, and the first one included
4896 a superfluous trailing colon, the final name in the first list and the
4897 first name in the second list were incorrectly joined into one item (with a
4898 colon in the middle).
4904 1. Added -C and -D options to the exinext utility, mainly to make it easier
4905 to include in the automated testing, but these could be helpful when
4906 multiple configurations are in use.
4908 2. The exinext utility was not formatting the output nicely when there was
4909 an alternate port involved in the retry record key, nor when there was a
4910 message id as well (for retries that were specific to a specific message
4911 and a specific host). It was also confused by IPv6 addresses, because of
4912 the additional colons they contain. I have fixed the IPv4 problem, and
4913 patched it up to do a reasonable job for IPv6.
4915 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4916 delivery, the log line now contains "pipelined" if PIPELINING was used.
4918 4. An SMTP transport process used to panic and die if the bind() call to set
4919 an explicit outgoing interface failed. This has been changed; it is now
4920 treated in the same way as a connect() failure.
4922 5. A reference to $sender_host_name in the part of a conditional expansion
4923 that was being skipped was still causing a DNS lookup. This no longer
4926 6. The def: expansion condition was not recognizing references to header lines
4927 that used bh_ and bheader_.
4929 7. Added the _cache feature to named lists.
4931 8. The code for checking quota_filecount in the appendfile transport was
4932 allowing one more file than it should have been.
4934 9. For compatibility with Sendmail, the command line option
4942 and sets the incoming protocol and host name (for trusted callers). The
4943 host name and its colon can be omitted when only the protocol is to be set.
4944 Note the Exim already has two private options, -pd and -ps, that refer to
4945 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4946 "s", but I don't think that's a major issue.
4948 10. A number of refactoring changes to the code, none of which should affect
4951 (a) The number of logging options was getting close to filling up the
4952 32-bit word that was used as a bit map. I have split them into two classes:
4953 those that are passed in the argument to log_write(), and those that are
4954 only ever tested independently outside of that function. These are now in
4955 separate 32-bit words, so there is plenty of room for expansion again.
4956 There is no change in the user interface or the logging behaviour.
4958 (b) When building, for example, log lines, the code previously used a
4959 macro that called string_cat() twice, in order to add two strings. This is
4960 not really sufficiently general. Furthermore, there was one instance where
4961 it was actually wrong because one of the argument was used twice, and in
4962 one call a function was used. (As it happened, calling the function twice
4963 did not affect the overall behaviour.) The macro has been replaced by a
4964 function that can join an arbitrary number of extra strings onto a growing
4967 (c) The code for expansion conditions now uses a table and a binary chop
4968 instead of a serial search (which was left over from when there were very
4969 few conditions). Also, it now recognizes conditions like "pam" even when
4970 the relevant support is not compiled in: a suitably worded error message is
4971 given if an attempt is made to use such a condition.
4973 11. Added ${time_interval:xxxxx}.
4975 12. A bug was causing one of the ddress fields not to be passed back correctly
4976 from remote delivery subprocesses. The field in question was not being
4977 subsequently used, so this caused to problems in practice.
4979 13. Added new log selectors queue_time and deliver_time.
4981 14. Might have fixed a bug in maildirsizefile handling that threw up
4982 "unexpected character" debug warnings, and recalculated the data
4983 unnecessarily. In any case, I expanded the warning message to give more
4986 15. Added the message "Restricted characters in address" to the statements in
4987 the default ACL that block characters like @ and % in local parts.
4989 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4990 Three changes have been made:
4992 (a) There was a serious bug; a negative response to MAIL caused the whole
4993 recipient domain to be cached as invalid, thereby blocking all messages
4994 to all local parts at the same domain, from all senders. This bug has
4995 been fixed. The domain is no longer cached after a negative response to
4996 MAIL if the sender used is not empty.
4998 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5001 (c) A new callout option, "use_sender" has been added for people who want
5002 the modified behaviour.
5008 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5011 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5012 indeed breaks things for older releases.
5014 3. Added additional logging to the case where there is a problem reading data
5015 from a filter that is running in a subprocess using a pipe, in order to
5016 try to track down a specific problem.
5018 4. Testing facility fudge: when running in the test harness and attempting
5019 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5020 getting "No route to host". Convert this to a timeout.
5022 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5025 6. Some OS don't have socklen_t but use size_t instead. This affects the
5026 fifth argument of getsockopt() amongst other things. This is now
5027 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5028 can be set for individual OS. I have set it for SunOS5, OSF1, and
5029 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5030 some earlier ones do not.
5032 7. Change 4.30/15 was not doing the test caselessly.
5034 8. The standard form for an IPv6 address literal was being rejected by address
5035 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5036 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5037 this, as well as the form without the "IPv6" on the front (but only when
5038 address literals are enabled, of course).
5040 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5042 10. Exim crashed if a message with an empty sender address specified by -f
5043 encountered a router with an errors_to setting. This could be provoked only
5044 by a command such as
5048 where an empty string was supplied; "<>" did not hit this bug.
5050 11. Installed PCRE release 4.5.
5052 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5053 remained set. It is now erased.
5055 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5056 times from message ids (which are base 36 rather than the normal 62).
5058 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5059 were being counted as actual protocol errors, and logged if the log
5060 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5061 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5062 and DATA following a set of rejected RCPTs do not count as protocol errors.
5063 In other words, Exim assumes they were pipelined, though this may not
5064 actually be the case. Of course, in all cases the client gets an
5065 appropriate error code.
5067 15. If a lookup fails in an ACL condition, a message about the failure may
5068 be available; it is used if testing the ACL cannot continue, because most
5069 such messages specify what the cause of the deferral is. However, some
5070 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5071 that caused an old message to be retained and used if a later statement
5072 caused a defer, replacing the real cause of the deferral.
5074 16. If an IP address had so many PTR records that the DNS lookup buffer
5075 was not large enough to hold them, Exim could crash while trying to process
5076 the truncated data. It now detects and logs this case.
5078 17. Further to 4.21/58, another change has been made: if (and only if) the
5079 first line of a message (the first header line) ends with CRLF, a bare LF
5080 in a subsequent header line has a space inserted after it, so as not to
5081 terminate the header.
5083 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5084 unnecessarily, used atoi() instead of strtol(), and didn't check the
5085 termination when getting file sizes from file names by regex.
5087 19. Completely re-implemented the support for maildirsize files, in the light
5088 of a number of problems with the previous contributed implementation
5089 (4.30/29). In particular:
5091 . If the quota is zero, the maildirsize file is maintained, but no quota is
5094 . If the maildir directory does not exist, it is created before any attempt
5095 to write a maildirsize file.
5097 . The quota value in the file is just a cache; if the quota is changed in
5098 the transport, the new value overrides.
5100 . A regular expression is available for excluding directories from the
5103 20. The autoreply transport checks the characters in options that define the
5104 message's headers; it allows continued headers, but it was checking with
5105 isspace() after an embedded newline instead of explicitly looking for a
5108 21. If all the "regular" hosts to which an address was routed had passed their
5109 expiry times, and had not reached their retry times, the address was
5110 bounced, even if fallback hosts were defined. Now Exim should go on to try
5113 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5114 equivalent code in the SMTP transport. Some hosts send humungous responses
5115 to HELO/EHLO, more than 1024 it seems.
5117 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5118 gives compiler warnings in some environments. I've now done it "properly",
5121 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5122 (because of problems with the built-in one) was declared to return uschar *
5123 instead of char *, causing compiler failure.
5125 25. Fixed a file descriptor leak when processing alias/forward files.
5127 26. Fixed a minor format string issue in dbfn.c.
5129 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5131 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5132 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5135 29. When a sender address is verified, it is cached, to save repeating the test
5136 when there is more than one recipient in a message. However, when the
5137 verification involves a callout, it is possible for different callout
5138 options to be set for different recipients. It is too complicated to keep
5139 track of this in the cache, so now Exim always runs a verification when a
5140 callout is required, relying on the callout cache for the optimization.
5141 The overhead is duplication of the address routing, but this should not be
5144 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5145 to be verified with callout=postmaster, and the main callout worked but the
5146 postmaster check failed, the verification correctly failed. However, if a
5147 subsequent RCPT command asked for sender verification *without* the
5148 postmaster check, incorrect caching caused this verification also to fail,
5151 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5152 it was not caching the DNS options (qualify_single, search_parents) that
5153 were used when the lookup failed. A subsequent lookup with different
5154 options therefore always gave the same answer, though there were cases
5155 where it should not have. (Example: a "domains = !$mx_any" option on a
5156 dnslookup router: the "domains" option is always processed without any
5157 widening, but the router might have qualify_single set.) Now Exim uses the
5158 cached value only when the same options are set.
5160 32. Added John Jetmore's "exipick" utility to the distribution.
5162 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5163 than a timeout (e.g. a certificate is required, and is not provided), an
5164 Exim server now closes the connection immediately. Previously it waited for
5165 the client to close - but if the client is SSL, it seems that they each
5166 wait for each other, leading to a delay before one of them times out.
5168 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5169 maintained 0.8.x compatibility because I don't think many are using it, and
5170 it is clearly obsolete.
5172 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5175 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5176 client certificate was expired. A simple patch fixes this, though I don't
5177 understand the full logic of why the verify callback is called multiple
5180 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5181 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5182 which causes problems with some clients (such as the Certicom SSL Plus
5183 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5184 disables the coutermeasure allowing Eudora to connect."
5186 38. Exim was not checking that a write() to a log file succeeded. This could
5187 lead to Bad Things if a log got too big, in particular if it hit a file
5188 size limit. Exim now panics and dies if it cannot write to a log file, just
5189 as it does if it cannot open a log file.
5191 39. Modified OS/Makefile-Linux so that it now contains
5193 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5195 The two -D definitions ensure that Exim is compiled with large file
5196 support, which makes it possible to handle log files that are bigger than
5199 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5200 instance) a domain was checked against a named list that involved a lookup,
5201 causing $domain_data to be set, then another domain was checked against the
5202 same list, then the first domain was re-checked, the value of $domain_data
5203 after the final check could be wrong. In particular, if the second check
5204 failed, it could be set empty. This bug probably also applied to
5207 41. The strip_trailing_dot option was not being applied to the address given
5208 with the -f command-line option.
5210 42. The code for reading a message's header from the spool was incrementing
5211 $received_count, but never initializing it. This meant that the value was
5212 incorrect (doubled) while delivering a message in the same process in which
5213 it was received. In the most common configuration of Exim, this never
5214 happens - a fresh exec is done - but it can happen when
5215 deliver_drop_privilege is set.
5217 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5218 it now includes up to 150 characters of the unexpected data in the log
5221 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5222 and building data strings. The size of both of these buffers was 10 000
5223 bytes - far larger than anybody would *ever* want, thought I. Needless to
5224 say, somebody hit the limit. I have increased the maximum line length to
5225 20 000 and the maximum data length of concatenated lines to 100 000. I have
5226 also fixed two bugs, because there was no checking on these buffers. Tsk,
5227 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5228 buffer is too small.
5230 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5231 lsearch lookups. Now it does.
5233 46. When parsing a route_list item in a manualroute router, a fixed-length
5234 buffer was used for the list of hosts. I made this 1024 bytes long,
5235 thinking that nobody would ever have a list of hosts that long. Wrong.
5236 Somebody had a whole pile of complicated expansion conditions, and the
5237 string was silently truncated, leading to an expansion error. It turns out
5238 that it is easier to change to an unlimited length (owing to other changes
5239 that have happened since this code was originally written) than to build
5240 structure for giving a limitation error. The length of the item that
5241 expands into the list of hosts is now unlimited.
5243 47. The lsearch lookup could not handle data where the length of text line was
5244 more than 4095 characters. Such lines were truncated, leading to shortened
5245 data being returned. It should now handle lines of any length.
5247 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5248 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5251 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5252 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5253 no longer happens in comment lines. A list of replacements is now placed
5254 at the head of all of the source files, except those whose only change is
5255 to replace PERL_COMMAND in the very first #! line.
5257 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5258 messages on the queue, with a bottom-up merge sort, using code contributed
5259 by Michael Haardt. This should make operations like -bp somewhat faster on
5260 large queues. It won't affect queue runners, except when queue_run_in_order
5263 51. Installed eximstats 1.31 in the distribution.
5265 52. Added support for SRV lookups to the dnslookup router.
5267 53. If an ACL referred to $message_body or $message_body_end, the value was not
5268 reset for any messages that followed in the same SMTP session.
5270 54. The store-handling optimization for building very long strings was not
5271 differentiating between the different store pools. I don't think this
5272 actually made any difference in practice, but I've tidied it.
5274 55. While running the routers to verify a sender address, $sender_address
5275 was still set to the sender address. This is wrong, because when routing to
5276 send a bounce to the sender, it would be empty. Therefore, I have changed
5277 it so that, while verifying a sender address, $sender_address is set to <>.
5278 (There is no change to what happens when verifying a recipient address.)
5280 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5281 target A or AAAA records (if not already returned) without resetting the
5282 qualify_single or search_parents options of the DNS resolver. These are
5283 inappropriate in this case because the targets of MX and SRV records must
5284 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5285 target that, when qualified, matched something in the local domain. These
5286 two options are now turned off when doing these lookups.
5288 57. It seems that at least some releases of Reiserfs (which does not have the
5289 concept of a fixed number of inodes) returns zero and not -1 for the
5290 number of available inodes. This interacted badly with check_spool_inodes,
5291 which assumed that -1 was the "no such thing" setting. What I have done is
5292 to check that the total number of inodes is greater than zero before doing
5293 the test of how many are available.
5295 58. When a "warn" ACL statement has a log_message modifier, the message is
5296 remembered, and not repeated. This is to avoid a lot of repetition when a
5297 message has many recipients that cause the same warning to be written.
5298 Howewer, Exim was preserving the list of already written lines for an
5299 entire SMTP session, which doesn't seem right. The memory is now reset if a
5300 new message is started.
5302 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5303 debugging output unless log_rewrite was also set.
5305 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5306 of (int)(handle) when we know that handle contains (void *)(-1).
5308 61. The Exim daemon panic-logs an error return when it closes the incoming
5309 connection. However "connection reset by peer" seems to be common, and
5310 isn't really an error worthy of noting specially, so that particular error
5313 62. When Exim is trying to find all the local interfaces, it used to panic and
5314 die if the ioctl to get the interface flags failed. However, it seems that
5315 on at least one OS (Solaris 9) it is possible to have an interface that is
5316 included in the list of interfaces, but for which you get a failure error
5317 for this call. This happens when the interface is not "plumbed" into a
5318 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5319 failure of the "get flags" call assumes that the interface is down.
5321 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5322 makes life easier for people who are doing arithmetic on fields extracted
5323 from dates, where you often get leading zeros that should not be
5324 interpreted as octal.
5326 64. Added qualify_domain to the redirect router, to override the global
5329 65. If a pathologically long header line contained very many addresses (the
5330 report of this problem mentioned 10 000) and each of them was rewritten,
5331 Exim could use up a very large amount of memory. (It kept on making new
5332 copies of the header line as it rewrote, and never released the old ones.)
5333 At the expense of a bit more processing, the header rewriting function has
5334 been changed so that it no longer eats memory in this way.
5336 66. The generation of the Received: header has been moved from the time that a
5337 message starts to be received, to the time that it finishes. The timestamp
5338 in the Received: header should now be very close to that of the <= log
5339 line. There are two side-effects of this change:
5341 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5342 logged header lines no longer include the local Received: line, because
5343 it has not yet been created. The same applies to a copy of the message
5344 that is returned to a non-SMTP sender when a message is rejected.
5346 (b) When a filter file is tested using -bf, no additional Received: header
5347 is added to the test message. After some thought, I decided that this
5350 This change does not affect the value of $received_for. It is still set
5351 after address rewriting, but before local_scan() is called.
5353 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5355 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5356 gave an unhelpful panic error message, and a defer error. I have managed to
5357 change this behaviour so that it now rejects any supplied certificate,
5358 which seems right, as the list of acceptable certificates is empty.
5360 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5361 gave an unhelpful defer error. I have not managed to make this reject any
5362 supplied certificates, but the error message it gives is "no certificate
5363 supplied", which is not helpful.
5365 70. exigrep's output now also includes lines that are not associated with any
5366 message, but which match the given pattern. Implemented by a patch from
5367 Martin Sluka, which also tidied up the Perl a bit.
5369 71. Recipient callout verification, like sender verification, was using <> in
5370 the MAIL FROM command. This isn't really the right thing, since the actual
5371 sender may affect whether the remote host accepts the recipient or not. I
5372 have changed it to use the actual sender in the callout; this means that
5373 the cache record is now keyed on a recipient/sender pair, not just the
5374 recipient address. There doesn't seem to be a real danger of callout loops,
5375 since a callout by the remote host to check the sender would use <>.
5376 [SEE ABOVE: changed after hitting problems.]
5378 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5379 temporary errors. However, in the case of such a code being given after
5380 the end of a data transmission (i.e. after ".") Exim was failing to write
5381 a retry record for the message. (Yes, there was some broken host that was
5382 actually sending 8xx at this point.)
5384 73. An unknown lookup type in a host list could cause Exim to panic-die when
5385 the list was checked. (An example that provoked this was putting <; in the
5386 middle of a list instead of at the start.) If this happened during a DATA
5387 ACL check, a -D file could be left lying around. This kind of configuration
5388 error no longer causes Exim to die; instead it causes a defer errror. The
5389 incident is still logged to the main and panic logs.
5391 74. Buglet left over from Exim 3 conversion. The message "too many messages
5392 in one connection" was written to the rejectlog but not the mainlog, except
5393 when address rewriting (yes!) was being logged.
5395 75. Added write_rejectlog option.
5397 76. When a system filter was run not as root (that is, when system_filter_user
5398 was set), the values of the $n variables were not being returned to the
5399 main process; thus, they were not subsequently available in the $sn
5402 77. Added +return_path_on_delivery log selector.
5404 78. A connection timeout was being treated differently from recipients deferred
5405 when testing hosts_max_try with a message that was older than the host's
5406 retry timeout. (The host should not be counted, thus allowing all hosts to
5407 be tried at least once before bouncing.) This may have been the cause of an
5408 occasionally reported bug whereby a message would remain on the queue
5409 longer than the retry timeout, but would be bounced if a delivery was
5410 forced. I say "may" because I never totally pinned down the problem;
5411 setting up timeout/retry tests is difficult. See also the next item.
5413 79. The ultimate address timeout was not being applied to errors that involved
5414 a combination of host plus message (for example, a timeout on a MAIL
5415 command). When an address resolved to a number of possible hosts, and they
5416 were not all tried for each delivery (e.g. because of hosts_max_try), a
5417 message could remain on the queue longer than the retry timeout.
5419 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5422 81. Fixed an obscure SMTP outgoing bug which required at least the following
5423 conditions: (a) there was another message waiting for the same server;
5424 (b) the server returned 5xx to all RCPT commands in the first message so
5425 that the message was not completed; (c) the server dropped the connection
5426 or gave a negative response to the RSET that Exim sends to abort the
5427 transaction. The observed case was a dropped connection after DATA that had
5428 been sent in pipelining mode. That is, the server had advertised PIPELINING
5429 but was not implementing it correctly. The effect of the bug was incorrect
5430 behaviour, such as trying another host, and this could lead to a crash.
5436 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5437 and daemon.c were passed as pointers to ints; they should have been
5438 pointers to socklen_t variables (which are typically unsigned ints).
5440 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5443 3. Fixed a really odd bug that affected only the testing scheme; patching a
5444 certain fixed string in the binary changed the value of another string that
5445 happened to be identical to the end of the original first string.
5447 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5448 name", it returns that address as the IP address. On some operating
5449 systems (e.g. Solaris), it also passes back the IP address string as the
5450 "host name". However, on others (e.g. Linux), it passes back an empty
5451 string. Exim wasn't checking for this, and was changing the host name to an
5452 empty string, assuming it had been canonicized.
5454 5. Although rare, it is permitted to have more than one PTR record for a given
5455 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5456 all the names associated with an address, because they do in Solaris.
5457 However, it seems that they do not in Linux for data that comes from the
5458 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5459 I found this out when I moved to a new Linux workstation and tried to run
5460 the Exim test suite.
5462 To get round this problem I have changed the code so that it now does its
5463 own call to the DNS to look up PTR records when searching for a host name.
5464 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5465 addresses that are only in /etc/hosts are still found.
5467 This behaviour is, however, controlled by an option called host_lookup_
5468 order, which defaults to "bydns:byaddr". If people want to use the other
5469 order, or indeed, just use one or the other means of lookup, they can
5470 specify it in this variable.
5472 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5473 some operating systems, this comes back from gethostbyaddr() as an empty
5474 string, and this is what Exim used to test for. However, it seems that in
5475 other systems, "." is yielded. Exim now tests for this case too.
5477 7. The values of check_spool_space and check_log_space are now held internally
5478 as a number of kilobytes instead of an absolute number of bytes. If a
5479 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5480 kilobyte. This means that much larger values can be stored.
5482 8. Exim monitor: an attempt to get the action menu when not actually pointing
5483 at a message produces an empty menu entitled "No message selected". This
5484 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5485 no entries in it ("Shell widget menu has zero width and/or height"). So I
5486 have added a single, blank menu entry in this case.
5488 9. Added ${quote_local_part.
5490 10. MIME decoding is now applied to the contents of Subject: header lines when
5493 11. Now that a reference to $sender_host_address automatically causes a reverse
5494 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5495 host lookup before query-style lookups in lists that might use this
5496 variable. This has therefore been abolished, and the "net-" prefix is no
5497 longer necessary for query-style lookups.
5499 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5500 have been a typo for LFLAGS, so it has been changed.
5502 13. The install script calls Exim with "-C /dev/null" in order to find the
5503 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5504 to be output. Howeve, since Exim outputs its version number before the
5505 error, it didn't break the script. It just looked ugly. I fixed this by
5506 always allowing "-C /dev/null" if the caller is root.
5508 14. Ignore overlarge ACL variable number when reading spool file - insurance
5509 against a later release with more variables having written the file.
5511 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5512 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5513 this, as well as the form without the "IPv6" on the front.
5515 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5516 OS/Makefile-Darwin file.
5518 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5519 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5521 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5522 variable. However, some OS use ints for inodes. I've added cast to long int
5523 to get rid of the compiler warning.
5525 19. I had forgotten to lock out "/../" in configuration file names when
5526 ALT_CONFIG_PREFIX was set.
5528 20. Routers used for verification do not need to specify transports. However,
5529 if such a router generated a host list, and callout was configured, Exim
5530 crashed, because it could not find a port number from the (non-existent)
5531 transport. It now assumes port 25 in this circumstance.
5533 21. Added the -t option to exigrep.
5535 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5536 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5537 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5538 them is set, LOOKUP_LSEARCH is forced.
5540 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5542 24. Added sender and host information to the "rejected by local_scan()" log
5543 line; previously there was no indication of these.
5545 25. Added .include_if_exists.
5547 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5548 Linux. It turns out that not all file systems support this. Apparently some
5549 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5550 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5551 not supported on the file descriptor, is now ignored when Exim is trying to
5552 sync a directory. This applies only to Linux.
5554 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5556 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5557 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5560 29. Added support for maildirsize files from supplied patch (modified a bit).
5562 30. The use of :fail: followed by an empty string could lead Exim to respond to
5563 sender verification failures with (e.g.):
5565 550 Verification failed for <xxx>
5566 550 Sender verify failed
5568 where the first response line was missing the '-' that indicates it is not
5569 the final line of the response.
5571 31. The loop for finding the name of the user that called Exim had a hardwired
5572 limit of 10; it now uses the value of finduser_retries, which is used for
5573 all other user lookups.
5575 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5578 33. Exim was neglecting to zero errno before one call of strtol() when
5579 expanding a string and expecting an integer value. On some systems this
5580 resulted in spurious "integer overflow" errors. Also, it was casting the
5581 result into an int without checking.
5583 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5584 did not work. The code looks as if it has *never* worked, though it appears
5585 to have been documented since at least releast 1.62. I have made it work.
5587 35. The "timeout_DNS" error in retry rules, also documented since at least
5588 1.62, also never worked. As it isn't clear exactly what this means, and
5589 clearly it isn't a major issue, I have abolished the feature by treating it
5590 as "timeout", and writing a warning to the main and panic logs.
5592 36. The display of retry rules for -brt wasn't always showing the error code
5595 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5596 timeout_connect_A, timeout_connect_MX.
5598 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5599 to the empty sender.
5601 39. The daemon was not analysing the content of -oX till after it had closed
5602 stderr and disconnected from the controlling terminal. This meant that any
5603 syntax errors were only noted on the panic log, and the return code from
5604 the command was 0. By re-arranging the code a little, I've made the
5605 decoding happen first, so such errors now appear on stderr, and the return
5606 code is 1. However, the actual setting up of the sockets still happens in
5607 the disconnected process, so errors there are still only recorded on the
5610 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5611 connections (as happens on some IP stacks) was logged at start up time as
5612 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5613 it from "IPv6 and IPv4", which means that two separate sockets are being
5616 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5617 says whether AF_INET or AF_INET6 was passed as an argument.
5619 42. Exiwhat output was messed up when time zones were included in log
5622 43. Exiwhat now gives more information about the daemon's listening ports,
5623 and whether -tls-on-connect was used.
5625 44. The "port" option of the smtp transport is now expanded.
5627 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5628 silently ignored. Now an error message is written to the main and panic
5631 46. There's a new ACL modifier called "logwrite" which writes to a log file
5632 as soon as it is encountered.
5634 47. Added $local_user_uid and $local_user_gid at routing time.
5636 48. Exim crashed when trying to verify a sender address that was being
5639 49. Exim was recognizing only a space character after ".include". It now also
5640 recognizes a tab character.
5642 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5643 extracting the relevant information from the specification. The man page no
5644 longer contains scrambled data for the -d option, and I've added a section
5645 at the front about calling Exim under different names.
5647 51. Added "extra_headers" argument to the "mail" command in filter files.
5649 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5652 53. Installed eximstats 1.29.
5654 54. Added transport_filter_timeout as a generic transport option.
5656 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5657 Cc: header lines. This was required by RFC 822, but it not required by RFC
5660 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5661 incoming messages that did not have them. Now it does so only if the
5662 message originates locally, that is, if there is no associated remote host
5663 address. When Resent- header lines are present, this applies to the Resent-
5664 lines rather than the non-Resent- lines.
5666 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5667 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5669 58. Messages for configuration errors now include the name of the main
5670 configuration file - useful now that there may be more than one file in a
5671 list (.included file names were always shown).
5673 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5674 for those rare installations that do not start the daemon as root or run it
5675 setuid root. I've cut out the call to initgroups() if the daemon is not
5678 60. The Exim user and group can now be bound into the binary as text strings
5679 that are looked up at the start of Exim's processing.
5681 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5683 62. Added $mailstore_basename variable.
5685 63. Installed patch to sieve.c from Michael Haardt.
5687 64. When Exim failed to open the panic log after failing to open the main log,
5688 the original message it was trying to log was written to stderr and debug
5689 output, but if they were not available (the usual case in production), it
5690 was lost. Now it is written to syslog before the two lines that record the
5691 failures to open the logs.
5693 65. Users' Exim filters run in subprocesses under the user's uid. It is
5694 possible for a "deliver" command or an alias in a "personal" command to
5695 provoke an address rewrite. If logging of address rewriting is configured,
5696 this fails because the process is not running as root or exim. There may be
5697 a better way of dealing with this, but for the moment (because 4.30 needs
5698 to be released), I have disabled address rewrite logging when running a
5699 filter in a non-root, non-exim process.
5705 1. The buildconfig auxiliary program wasn't quoting the value set for
5706 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5707 not defined. This bug was masked in 4.22 by the effect that was fixed in
5710 2. Some messages that were rejected after a message id was allocated were
5711 shown as "incomplete" by exigrep. It no longer does this for messages that
5712 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5714 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5715 have allow_domain_literals set, the ID did not get logged in the <= line.
5716 Domain literals are now always recognized in Message-ID: header lines.
5718 4. The first argument for a ${extract expansion item is the key name or field
5719 number. Leading and trailing spaces in this item were not being ignored,
5720 causing some misleading effects.
5722 5. When deliver_drop_privilege was set, single queue runner processes started
5723 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5724 same command in the process it spins off) were not dropping privilege.
5726 6. When the daemon running as "exim" started a queue runner, it always
5727 re-executed Exim in the spun-off process. This is a waste of effort when
5728 deliver_drop_privilege is set. The new process now just calls the
5729 queue-runner function directly.
5735 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5738 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5739 ignored. Though the use of -oP was forcing the writing of a pid file, it
5740 was always written to the default place.
5742 3. If the message "no IP address found for host xxxx" is generated during
5743 incoming verification, it is now followed by identification of the incoming
5744 connection (so you can more easily find what provoked it).
5746 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5748 5. Added some features to "harden" Exim a bit more against certain attacks:
5750 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5751 be put in Local/Makefile. This is like the never_users runtime option,
5752 but it cannot be overridden. The default setting is "root".
5754 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5755 prefix string with which any file named in a -C command line option
5758 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5759 is retained for -C and -D only if the caller of Exim is root. Without
5760 it, the exim user may also use -C and -D and retain privilege.
5762 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5763 command line option is disabled.
5765 6. Macro names set by the -D option must start with an upper case letter, just
5766 like macro names defined in the configuration file.
5768 7. Added "dereference=" facility to LDAP.
5770 8. Two instances of the typo "uknown" in the source files are fixed.
5772 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5773 the Configure-Makefile script screwed up while processing it.
5775 10. Incorporated PCRE 4.4.
5777 11. The SMTP synchronization check was not operating right at the start of an
5778 SMTP session. For example, it could not catch a HELO sent before the client
5779 waited for the greeting. There is now a check for outstanding input at the
5780 point when the greeting is written. Because of the duplex, asynchronous
5781 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5782 way, but not yet received, when the check is performed.
5784 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5785 on TCP/IP sockets, because this apparently causes some broken clients to
5788 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5789 unchanged) from the Cygwin maintainer.
5791 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5792 was supported instead of testing for mbx. Effectively a typo.
5794 15. The spa authenticator server code was not checking that the input it
5795 received was valid base64.
5797 16. The debug output line for the "set" modifier in ACLs was not showing the
5798 name of the variable that was being set.
5800 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5802 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5803 lookup on its own. Something else was needed to trigger the lookup. For
5804 example, a match in host_lookup or the need for a host name in a host list.
5805 Now, if $sender_host_name is referenced and the host name has not yet been
5806 looked up, a lookup is performed. If the lookup fails, the variable remains
5807 empty, and $host_lookup_failed is set to "1".
5809 19. Added "eqi" as a case-independent comparison operator.
5811 20. The saslauthd authentication condition could segfault if neither service
5812 nor realm was specified.
5814 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5815 error message that was logged was misleading, and incoming SMTP
5816 connections were dropped. The message is now more accurate, and temporary
5817 errors are given to SMTP connections.
5819 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5820 (except RSET) with a 421 error, until QUIT is received. However, it was
5821 failing to send a response to QUIT.
5823 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5824 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5825 ACL is now run after the helo_try_verify_hosts code.
5827 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5828 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5829 case-independent, so other case variants are also recognized). Apparently
5830 some systems use these upper case variants.
5832 25. If more than two messages were waiting for the same host, and a transport
5833 filter was specified for the transport, Exim sent two messages over the
5834 same TCP/IP connection, and then failed with "socket operation on non-
5835 socket" when it tried to send the third.
5837 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5839 27. The extern definition of crypt16() in expand.c was not being excluded when
5840 the OS had its own crypt16() function.
5842 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5843 as a preferred synonym for return_size_limit, both as an option and as an
5846 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5848 30. Changed the default configuration ACL to relax the local part checking rule
5849 for addresses that are not in any local domains. For these addresses,
5850 slashes and pipe symbols are allowed within local parts, but the sequence
5851 /../ is explicitly forbidden.
5853 31. SPA server authentication was not clearing the challenge buffer before
5856 32. log_message in a "warn" ACL statement was writing to the reject log as
5857 well as to the main log, which contradicts the documentation and doesn't
5858 seem right (because no rejection is happening). So I have stopped it.
5860 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5861 However, I am unable to do any testing of this.
5863 34. Fixed an infelicity in the appendfile transport. When checking directories
5864 for a mailbox, to see if any needed to be created, it was accidentally
5865 using path names with one or more superfluous leading slashes; tracing
5866 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5868 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5869 contents are added to the log line that is written for every discarded
5870 recipient. (Previously a log_message setting was ignored.)
5872 36. The ${quote: operator now quotes the string if it is empty.
5874 37. The install script runs exim in order to find its version number. If for
5875 some reason other than non-existence or emptiness, which it checks, it
5876 could not run './exim', it was installing it with an empty version number,
5877 i.e. as "exim-". This error state is now caught, and the installation is
5880 38. An argument was missing from the function that creates an error message
5881 when Exim fails to connect to the socket for saslauthd authentication.
5882 This could cause Exim to crash, or give a corrupted message.
5884 39. Added isip, isip4, and isip6 to ${if conditions.
5886 40. The ACL variables $acl_xx are now saved with the message, and can be
5887 accessed later in routers, transports, and filters.
5889 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5890 strings in the file are not string-expanded.
5892 42. If a MAIL command specified a SIZE value that was too large to fit into an
5893 int variable, the check against message_size_limit failed. Such values are
5894 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5895 day this will have to be increased, but I don't think I want to be around
5896 when emails are that large.
5903 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5904 iconv() is not standard in FreeBSD.
5906 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5907 IPv6 enabled. The observed symptom was a segmentation fault on return from
5908 the function os_common_find_running_interfaces() in src/os.c.
5910 3. In the check_special_case() function in daemon.c I had used "errno" as an
5911 argument name, which causes warnings on some systems. This was basically a
5912 typo, since it was named "eno" in the comments!
5914 4. The code that waits for the clock to tick (at a resolution of some fraction
5915 of a second) so as to ensure message-id uniqueness was always waiting for
5916 at least one whole tick, when it could have waited for less. [This is
5917 almost certainly not relevant at current processor speeds, where it is
5918 unlikely to ever wait at all. But we try to future-proof.]
5920 5. The function that sleeps for a time interval that includes fractions of a
5921 second contained a race. It did not block SIGALRM between setting the
5922 timer, and suspending (a couple of lines later). If the interval was short
5923 and the sigsuspend() was delayed until after it had expired, the suspension
5924 never ended. On busy systems this could lead to processes getting stuck for
5927 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5928 process, before it forks any delivery processes. The open lookup caching
5929 mechanism meant that the open file or database connection was passed into
5930 the delivery process. The problem was that delivery processes always tidy
5931 up cached lookup data. This could cause a problem for the next delivery
5932 process started by the queue runner, because the external queue runner
5933 process does not know about the closure. So the next delivery process
5934 still has data in the lookup cache. In the case of a file lookup, there was
5935 no problem because closing a file descriptor in a subprocess doesn't affect
5936 the parent. However, if the lookup was caching a connection to a database,
5937 the connection was closed, and the second delivery process was likely to
5938 see errors such as "PGSQL: query failed: server closed the connection
5939 unexpectedly". The problem has been fixed by closing all cached lookups
5940 in a queue runner before running a delivery process.
5942 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5943 seem to have the "const" qualifier which it has on other OS. I've
5946 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5947 *of the same type* (client or server) with the same public name that an
5948 error should be diagnosed.
5950 9. When Exim looked up a host name for an IP address, but failed to find the
5951 original IP address when looking up the host name (a safety check), it
5952 output the message "<ip address> does not match any IP for NULL", which was
5953 confusing, to say the least. The bug was that the host name should have
5954 appeared instead of "NULL".
5956 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5957 user that is built into the binary, and the -C or -D options is used, root
5958 privilege is dropped before the configuration file is read. In addition,
5959 logging is switched to stderr instead of the normal log files. If the
5960 configuration then re-defines the Exim user, the unprivileged environment
5961 is probably not what is expected, so Exim logs a panic warning message (but
5964 However, if deliver_drop_privilege is set, the unprivileged state may well
5965 be exactly what is intended, so the warning has been cut out in that case,
5966 and Exim is allowed to try to write to its normal log files.
5972 1. smtp_return_error_details was not giving details for temporary sender
5973 or receiver verification errors.
5975 2. Diagnose a configuration error if two authenticators have the same public
5978 3. Exim used not to create the message log file for a message until the first
5979 delivery attempt. This could be confusing when incoming messages were held
5980 for policy or load reasons. The message log file is now created at the time
5981 the message is received, and an initial "Received" line is written to it.
5983 4. The automatically generated man page for command line options had a minor
5984 bug that caused no ill effects; however, a more serious problem was that
5985 the procedure for building the man page automatically didn't always
5986 operate. Consequently, release 4.20 contains an out-of-date version. This
5987 shouldn't happen again.
5989 5. When building Exim with embedded Perl support, the script that builds the
5990 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5991 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5993 6. The freeze_tell option was not being used for messages that were frozen on
5994 arrival, either by an ACL or by local_scan().
5996 7. Added the smtp_incomplete_transaction log selector.
5998 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5999 was accepting AUTH without a new EHLO.
6001 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6002 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6005 10. Logging of TCP/IP connections (when configured) now happens in the main
6006 daemon process instead of the child process, so that the TCP/IP connection
6007 count is more accurate (but it can never be perfect).
6009 11. The use of "drop" in a nested ACL was not being handled correctly in the
6010 outer ACL. Now, if condition failure induced by the nested "drop" causes
6011 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6012 or "require"), the connection is dropped.
6014 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6015 that yield "discard" can now be used with an "accept" or a "discard" verb,
6016 but an error is generated for any others (because I can't see a useful way
6017 to define what should happen).
6019 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6020 are now processed in the same way as lines in the Exim configuration file.
6021 In particular, continuation lines are supported.
6023 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6025 15. Added -ti meaning -t -i.
6027 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6028 domains, and warn by logging if others are found.
6030 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6031 returned from GIFCONF when Exim is trying to find the list of interfaces on
6032 a host. The code in os.c has been modified to copy each ifreq to an aligned
6033 structure in all cases.
6035 Also, in some cases, the returned ifreq's were being copied to a 'struct
6036 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6037 means the last couple of bytes of an IPv6 address could be chopped if the
6038 ifreq contained only a normal sockaddr (14 bytes storage).
6040 18. Named domain lists were not supported in the hosts_treat_as_local option.
6041 An entry such as +xxxx was not recognized, and was treated as a literal
6044 19. Ensure that header lines added by a DATA ACL are included in the reject log
6045 if the ACL subsequently rejects the message.
6047 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6048 MD5 (which is deprecated).
6050 21. When testing a filter file using -bf, Exim was writing a message when it
6051 took the sender from a "From " line in the message, but it was not doing so
6052 when it took $return_path from a Return-Path: header line. It now does.
6054 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6055 with a valid header line field name (a series of printing characters
6056 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6058 23. Changed "disc" in the source to "disk" to conform to the documentation and
6059 the book and for uniformity.
6061 24. Ignore Sendmail's -Ooption=value command line item.
6063 25. When execve() failed while trying to run a command in a pipe transport,
6064 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6065 could be confused with a return value of 69 from the command itself. This
6066 has been changed to 127, the value the shell returns if it is asked to run
6067 a non-existent command. The wording for the related log line suggests a
6068 non-existent command as the problem.
6070 26. If received_header_text expands to an empty string, do not add a Received:
6071 header line to the message. (Well, it adds a token one on the spool, but
6072 marks it "old" so that it doesn't get used or transmitted.)
6074 27. Installed eximstats 1.28 (addition of -nt option).
6076 28. There was no check for failure on the call to getsockname() in the daemon
6077 code. This can fail if there is a shortage of resources on the system, with
6078 ENOMEM, for example. A temporary error is now given on failure.
6080 29. Contrary to the C standard, it seems that in some environments, the
6081 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6082 program. Exim now does this explicitly; it affects the formatting of
6083 timestamps using strftime().
6085 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6086 complaints. I've now initialized all the variables, to avoid this.
6088 32. Header lines added by a system filter were not being "seen" during
6089 transport-time rewrites.
6091 33. The info_callback() function passed to OpenSSL is set up with type void
6092 (*)(SSL *, int, int), as described somewhere. However, when calling the
6093 function (actually a macro) that sets it up, the type void(*)() is
6094 expected. I've put in a cast to prevent warnings from picky compilers.
6096 34. If a DNS black list lookup found a CNAME record, but there were no A
6097 records associated with the domain it pointed at, Exim crashed.
6099 35. If a DNS black list lookup returned more than one A record, Exim ignored
6100 all but the first. It now scans all returned addresses if a particular IP
6101 value is being sought. In this situation, the contents of the
6102 $dnslist_value variable are a list of all the addresses, separated by a
6105 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6106 to accept a host name and all its aliases if the forward lookup for any of
6107 them yielded the IP address of the incoming connection. Now it accepts only
6108 those names whose forward lookup yields the correct IP address. Any other
6109 names are discarded. This closes a loophole whereby a rogue DNS
6110 administrator could create reverse DNS records to break through a
6111 wildcarded host restriction in an ACL.
6113 37. If a user filter or a system filter that ran in a subprocess used any of
6114 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6115 the wrong values were passed to the pipe command ($thisaddress had the
6116 value of $0, $0 had the value of $1, etc). This bug was introduced by
6117 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6119 38. Improved the line breaking for long SMTP error messages from ACLs.
6120 Previously, if there was no break point between 40 and 75 characters, Exim
6121 left the rest of the message alone. Two changes have been made: (a) I've
6122 reduced the minimum length to 35 characters; (b) if it can't find a break
6123 point between 35 and 75 characters, it looks ahead and uses the first one
6124 that it finds. This may give the occasional overlong line, but at least the
6125 remaining text gets split now.
6127 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6128 file descriptors might be low, and that setting 1000 would always raise it.
6129 It turns out that in some environments, the limit is already over 1000 and
6130 that lowering it causes trouble. So now Exim takes care not to decrease it.
6132 40. When delivering a message, the value of $return_path is set to $sender_
6133 address at the start of routing (routers may change the value). By an
6134 oversight, this default was not being set up when an address was tested by
6135 -bt or -bv, which affected the outcome if any router or filter referred to
6138 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6139 log only when "message" or "log_message" are set. However, if one of the
6140 conditions was an address verification, or a call to a nested ACL, the
6141 messages generated by the underlying test were being passed through. This
6142 no longer happens. The underlying message is available in $acl_verify_
6143 message for both "message" and "log_message" expansions, so it can be
6144 passed through if needed.
6146 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6147 new expansion $bh_ to give the encoded byte string without charset
6148 translation. Translation happens only if iconv() is available; HAVE_ICONV
6149 indicates this at build time. HEADERS_CHARSET gives the charset to
6150 translate to; headers_charset can change it in the configuration, and
6151 "headers charset" can change it in an individual filter file.
6153 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6154 that creates RFC 2047 encoded "words" labels them as that charset instead
6155 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6156 expansion operator; (ii) when Exim creates a From: line for a local
6157 message; (iii) when a header line is rewritten to include a "phrase" part.
6159 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6160 buggy, causing it to skip the first lines of messages whose message ID
6161 ended in 'D'. This would not have bitten before Exim release 4.14, because
6162 message IDs were unlikely to end in 'D' before then. The effect was to have
6163 incorrect size information for certain domains.
6165 45. #include "config.h" was missing at the start of the crypt16.c module. This
6166 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6169 46. If there was a timeout during a "random" callout check, Exim treated it as
6170 a failure of the random address, and carried on sending RSET and the real
6171 address. If the delay was just some slowness somewhere, the response to the
6172 original RCPT would be taken as a response to RSET and so on, causing
6173 mayhem of various kinds.
6175 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6176 when I implemented it. It didn't allow for the fact that some option values
6177 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6178 the right test for positive values.
6180 48. Domain names in DNS records are case-independent. Exim always looks them up
6181 in lower case. Some resolvers return domain names in exactly the case they
6182 appear in the zone file, that is, they may contain uppercase letters. Not
6183 all resolvers do this - some return always lower case. Exim was treating a
6184 change of case by a resolver as a change of domain, similar to a widening
6185 of a domain abbreviation. This triggered its re-routing code and so it was
6186 trying to route what was effectively the same domain again. This normally
6187 caused routing to fail (because the router wouldn't handle the domain
6188 twice). Now Exim checks for this case specially, and just changes the
6189 casing of the domain that it ultimately uses when it transmits the message
6192 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6195 50. If a filter generated a file delivery with a non-absolute name (possible if
6196 no home directory exists for the router), the forbid_file option was not
6199 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6200 the existing equality matching.
6202 52. Exim was using ints instead of ino_t variables in some places where it was
6203 dealing with inode numbers.
6205 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6206 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6207 called TMPDIR, and if it finds it is different, it changes its value.
6209 54. The smtp_printf() function is now made available to local_scan() so
6210 additional output lines can be written before returning. There is also an
6211 smtp_fflush() function to enable the detection of a dropped connection.
6212 The variables smtp_input and smtp_batched_input are exported to
6215 55. Changed the default runtime configuration: the message "Unknown user"
6216 has been removed from the ACL, and instead placed on the localuser router,
6217 using the cannot_route_message feature. This means that any verification
6218 failures that generate their own messages won't get overridden. Similarly,
6219 the "Unrouteable address" message that was in the ACL for unverifiable
6220 relay addresses has also been removed.
6222 56. Added hosts_avoid_esmtp to the smtp transport.
6224 57. The exicyclog script was not checking for the esoteric option
6225 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6226 will work only if exicyclog is run under the appropriate euid.
6228 58. Following a discussion on the list, the rules by which Exim recognises line
6229 endings on incoming messages have been changed. The -dropcr and drop_cr
6230 options are now no-ops, retained only for backwards compatibility. The
6231 following line terminators are recognized: LF CRLF CR. However, special
6232 processing applies to CR:
6234 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6235 nor a local message in the state where . is a terminator.
6237 (ii) If a bare CR is encountered in a header line, an extra space is added
6238 after the line terminator so as not to end the header. The reasoning
6239 behind this is that bare CRs in header lines are most likely either
6240 to be mistakes, or people trying to play silly games.
6242 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6243 was being incorrectly given as 18 bytes larger than it should have been.
6244 This is a VOB (very old bug).
6246 60. This may never have affected anything current, but just in case it has:
6247 When the local host is found other than at the start of a list of hosts,
6248 the local host, those with the same MX, and any that follow, are discarded.
6249 When the list in question was part of a longer list of hosts, the following
6250 hosts (not currently being processed) were also being discarded. This no
6251 longer happens. I'm not sure if this situation could ever has previously
6254 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6257 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6258 header. This is something that is recommended in a new Internet Draft, and
6259 is something that is documented as being done by Sendmail. There are two
6260 possible values. For messages generated by the autoreply transport, Exim
6263 Auto-Submitted: auto-replied
6265 whereas for all other generated messages (e.g. bounces) it adds
6267 Auto-Submitted: auto-generated
6269 63. The "personal" condition in filters now includes a test for the
6270 Auto-Submitted: header. If it contains the string "auto-" the message it
6271 not considered personal.
6273 64. Added rcpt_include_affixes as a generic transport option.
6275 65. Added queue_only_override (default true).
6277 66. Added the syslog_duplication option.
6279 67. If what should have been the first header line of a message consisted of
6280 a space followed by a colon, Exim was mis-interpreting it as a header line.
6281 It isn't of course - it is syntactically invalid and should therefore be
6282 treated as the start of the message body. The misbehaviour could have
6283 caused a number of strange effects, including loss of data in subsequent
6284 header lines, and spool format errors.
6286 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6287 client host had authenticated. This control can now be exercised by an ACL
6288 for more flexibility.
6290 69. By default, callouts do not happen when testing with -bh. There is now a
6291 variant, -bhc, which does actually run the callout code, including
6292 consulting and updating the callout cache.
6294 70. Added support for saslauthd authentication, courtesy of Alexander
6297 71. If statvfs() failed on the spool or log directories while checking their
6298 size for availability, Exim confusingly gave the error "space shortage".
6299 Furthermore, in debugging mode it crashed with a floating point exception.
6300 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6301 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6302 serious problem, Exim now writes to the main and panic logs when this
6303 happens, with details of the failure. It then refuses to accept the
6304 incoming message, giving the message "spool directory problem" or "log
6305 directory problem" with a 421 code for SMTP messages.
6307 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6308 0, 1, and 2 exist, because some OS complain for execs without them (see
6309 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6310 descriptors. However, the code omitted to check that the open succeeded,
6311 causing mysterious errors if for some reason the permissions on /dev/null
6312 got screwed. Now Exim writes a message to the main and panic logs, and
6313 bombs out if it can't open /dev/null.
6315 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6316 interact so that it is all more flexible. It is supposed to remain
6317 backwards compatible. Also added extra_local_interfaces.
6319 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6320 to bomb out with an assertion failure - to the client this appears as a
6321 connection drop. This problem occurs in the part of the code that was taken
6322 from the Samba project. Fortunately, the assertion is in a very simple
6323 function, so I have fixed this by reproducing the function inline in the
6324 one place where it is called, and arranging for authentication to fail
6325 instead of killing the process with assert().
6327 75. The SPA client code was not working when the server requested OEM rather
6328 than Unicode encoding.
6330 76. Added code to make require_files with a specific uid setting more usable in
6331 the case where statting the file as root fails - usually a non-root-mounted
6332 NFS file system. When this happens and the failure is EACCES, Exim now
6333 forks a subprocess and does the per-uid checking as the relevant uid.
6335 77. Added process_log_path.
6337 78. If log_file_path was not explicitly set, a setting of check_log_space or
6338 check_log_inodes was ignored.
6340 79. If a space check for the spool or log partitions fails, the incident is now
6341 logged. Of course, in the latter case the data may get lost...
6343 80. Added the %p formatting code to string_format() so that it can be used to
6344 print addresses in debug_print(). Adjusted all the address printing in the
6345 debugging in store.c to use %p rather than %d.
6347 81. There was a concern that a line of code in smtp_in.c could overflow a
6348 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6349 initially expressed, the concern was not well-founded, because trailing
6350 spaces are removed early. However, if the trailing spaces were followed by
6351 a NULL, they did not get removed, so the overflow was possible. Two fixes
6354 (a) I re-wrote the offending code in a cleaner fashion.
6355 (b) If an incoming SMTP command contains a NULL character, it is rejected
6358 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6359 runs initgroups(), so that if the Exim user is in any additional groups,
6360 they will be used during message reception.
6366 The change log for 4.20 and earlier releases has been archived.