1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
8 and 3.4.4 - once the server is enabled to respond to an OCSP request
9 it does even when not requested, resulting in a stapling non-aware
10 client dropping the TLS connection.
12 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
13 support variable-length bit vectors. No functional change.
15 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
16 The I= interface field on outgoing lines is now after the H= remote
17 host field, same as incoming lines. There is a separate
18 outgoing_interface log selector which allows you to disable the
21 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
22 If not running log_selector +smtp_connection the mainlog would be held
23 open indefinitely after a "too many connections" event, including to a
24 deleted file after a log rotate. Leave the per net connection logging
25 leaving it open for efficiency as that will be quickly detected by the
26 check on the next write.
28 HS/01 Bug 1671: Fix post transport crash.
29 Processing the wait-<transport> messages could crash the delivery
30 process if the message IDs didn't exist for some reason. When
31 using 'split_spool_directory=yes' the construction of the spool
32 file name failed already, exposing the same netto behaviour.
34 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
35 mime_regex ACL conditions.
37 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
38 to DSN fail messages (bounces): remote IP, remote greeting, remote response
39 to HELO, local diagnostic string.
41 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
42 log line to debug. Even when configured with a tls authenticator many
43 client connections are expected to not authenticate in this way, so
44 an authenticate fail is not an error.
46 HS/02 Add the Exim version string to the process info. This way exiwhat
47 gives some more detail about the running daemon.
49 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
50 matter for fast-change records such as DNSBLs.
52 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
53 retry record, even if constant. There may be multiple transports with
54 different interface settings and the retry behaviour needs to be kept
57 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
59 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
61 JH/10 Bug 840: fix log_defer_output option of pipe transport
63 JH/11 Bug 830: use same host for all RCPTS of a message, even under
64 hosts_randomize. This matters a lot when combined with mua_wrapper.
66 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
67 ${quote_pgsql:<string>} operator.
69 JH/13 Bug 1708: avoid misaligned access in cached lookup.
71 JH/14 Change header file name for freeradius-client. Relevant if compiling
72 with Radius support; from the Gentoo tree and checked under Fedora.
74 JH/15 Bug 1712: Introduce $prdr_requested flag variable
79 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
82 JH/02 The smtp transport option "multi_domain" is now expanded.
84 JH/03 The smtp transport now requests PRDR by default, if the server offers
87 JH/04 Certificate name checking on server certificates, when exim is a client,
88 is now done by default. The transport option tls_verify_cert_hostnames
89 can be used to disable this per-host. The build option
90 EXPERIMENTAL_CERTNAMES is withdrawn.
92 JH/05 The value of the tls_verify_certificates smtp transport and main options
93 default to the word "system" to access the system default CA bundle.
94 For GnuTLS, only version 3.0.20 or later.
96 JH/06 Verification of the server certificate for a TLS connection is now tried
97 (but not required) by default. The verification status is now logged by
98 default, for both outbound TLS and client-certificate supplying inbound
101 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
104 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
105 Status Notification (bounce) messages are now MIME format per RFC 3464.
106 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
107 under the control of the dsn_advertise_hosts option, and routers may
108 have a dsn_lasthop option.
110 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
111 default, modifiable by a malware= option. The list separator for
112 the options can now be changed in the usual way. Bug 68.
114 JH/10 The smtp_receive_timeout main option is now expanded before use.
116 JH/11 The incoming_interface log option now also enables logging of the
117 local interface on delivery outgoing connections.
119 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
120 if the interface and destination host and port all match.
122 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
125 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
126 Patch from Andrew Lewis.
128 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
129 now supports optional time-restrictions, weighting, and priority
130 modifiers per server. Patch originally by <rommer@active.by>.
132 JH/16 The spamd_address main option now supports a mixed list of local
133 and remote servers. Remote servers can be IPv6 addresses, and
134 specify a port-range.
136 JH/17 Bug 68: The spamd_address main option now supports an optional
137 timeout value per server.
139 JH/18 Bug 1581: Router and transport options headers_add/remove can
140 now have the list separator specified.
142 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
145 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
148 JH/21 Support for the A6 type of dns record is withdrawn.
150 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
151 rather than the verbs used.
153 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
154 from 255 to 1024 chars.
156 JH/24 Verification callouts now attempt to use TLS by default.
158 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
159 are generic router options now. The defaults didn't change.
161 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
162 Original patch from Alexander Shikoff, worked over by JH.
164 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
165 files if ZCAT_COMMAND is not executable.
167 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
169 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
171 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
172 Normally benign, it bites when the pair was led to by a CNAME;
173 modern usage is to not canoicalize the domain to a CNAME target
174 (and we were inconsistent anyway for A-only vs AAAA+A).
176 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
178 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
179 when evaluating $sender_host_dnssec.
181 JH/31 Check the HELO verification lookup for DNSSEC, adding new
182 $sender_helo_dnssec variable.
184 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
186 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
188 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
190 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
191 documented as working, but never had. Support all but $spam_report.
193 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
194 added for tls authenticator.
199 TL/01 When running the test suite, the README says that variables such as
200 no_msglog_check are global and can be placed anywhere in a specific
201 test's script, however it was observed that placement needed to be near
202 the beginning for it to behave that way. Changed the runtest perl
203 script to read through the entire script once to detect and set these
204 variables, reset to the beginning of the script, and then run through
205 the script parsing/test process like normal.
207 TL/02 The BSD's have an arc4random API. One of the functions to induce
208 adding randomness was arc4random_stir(), but it has been removed in
209 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
210 function when detected.
212 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
213 cause callback expansion.
215 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
216 syntax errors in an expansion can be treated as a string instead of
217 logging or causing an error, due to the internal use of bool_lax
218 instead of bool when processing it.
220 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
221 server certificates when making smtp deliveries.
223 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
225 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
227 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
229 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
230 Merged patch from Sebastian Wiedenroth.
232 JH/05 Fix results-pipe from transport process. Several recipients, combined
233 with certificate use, exposed issues where response data items split
234 over buffer boundaries were not parsed properly. This eventually
235 resulted in duplicates being sent. This issue only became common enough
236 to notice due to the introduction of conection certificate information,
237 the item size being so much larger. Found and fixed by Wolfgang Breyha.
239 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
240 size buffer was used, resulting in syntax errors when an expansion
243 JH/07 Add support for directories of certificates when compiled with a GnuTLS
244 version 3.3.6 or later.
246 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
247 is EXPERIMENTAL_EVENT, the main-configuration and transport options
248 both become "event_action", the variables become $event_name, $event_data
249 and $event_defer_errno. There is a new variable $verify_mode, usable in
250 routers, transports and related events. The tls:cert event is now also
251 raised for inbound connections, if the main configuration event_action
254 TL/06 In test suite, disable OCSP for old versions of openssl which contained
255 early OCSP support, but no stapling (appears to be less than 1.0.0).
257 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
258 server certificate names available under the smtp transport option
259 "tls_verify_cert_hostname" now do not permit multi-component wildcard
262 JH/10 Time-related extraction expansions from certificates now use the main
263 option "timezone" setting for output formatting, and are consistent
264 between OpenSSL and GnuTLS compilations. Bug 1541.
266 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
267 encoded parameter in the incoming message. Bug 1558.
269 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
270 include certificate info, eximon was claiming there were spoolfile
273 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
275 JH/14 Log delivery-related information more consistently, using the sequence
276 "H=<name> [<ip>]" wherever possible.
278 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
279 are problematic for Debian distribution, omit them from the release
282 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
284 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
287 JH/17 Fix a null-indirection in certextract expansions when a nondefault
288 output list separator was used.
293 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
294 checkers that were complaining about end of non-void function with no
297 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
298 This was a regression intruduced in 4.83 by another bugfix.
300 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
302 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
303 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
309 TF/01 Correctly close the server side of TLS when forking for delivery.
311 When a message was received over SMTP with TLS, Exim failed to clear up
312 the incoming connection properly after forking off the child process to
313 deliver the message. In some situations the subsequent outgoing
314 delivery connection happened to have the same fd number as the incoming
315 connection previously had. Exim would try to use TLS and fail, logging
316 a "Bad file descriptor" error.
318 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
319 utilities have not been installed.
321 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
322 temporary space as the ACL may create new global variables.
324 TL/01 LDAP support uses per connection or global context settings, depending
325 upon the detected version of the libraries at build time.
327 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
328 to extract and use the src ip:port in logging and expansions as if it
329 were a direct connection from the outside internet. PPv2 support was
330 updated based on HAProxy spec change in May 2014.
332 JH/02 Add ${listextract {number}{list}{success}{fail}}.
334 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
335 Properly escape header and check for NULL return.
337 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
340 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
342 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
343 characters in header names, implemented as a verify condition.
344 Contributed by Michael Fischer v. Mollard.
346 TL/05 Rename SPF condition results err_perm and err_temp to standardized
347 results permerror and temperror. Previous values are deprecated but
348 still accepted. In a future release, err_perm and err_temp will be
349 completely removed, which will be a backward incompatibility if the
350 ACL tests for either of these two old results. Patch contributed by
351 user bes-internal on the mailing list.
353 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
355 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
356 selectors, in both main and reject logs.
358 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
361 JH/07 Add malware type "sock" for talking to simple daemon.
363 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
365 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
366 routers/transports under cutthrough routing.
368 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
369 numbers. Touch up "bool" conditional to keep the same definition.
371 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
373 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
375 JH/12 Expand items in router/transport headers_add or headers_remove lists
376 individually rather than the list as a whole. Bug 1452.
378 Required for reasonable handling of multiple headers_ options when
379 they may be empty; requires that headers_remove items with embedded
380 colons must have them doubled (or the list-separator changed).
382 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
383 view the policy declared in the DMARC record. Currently, $dmarc_status
384 is a combined value of both the record presence and the result of the
387 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
389 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
390 dnslookup router and the smtp transport (applying to the forward
393 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
394 of ldap servers used for a specific lookup. Patch provided by Heiko
397 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
398 New variable $lookup_dnssec_authenticated for observability.
400 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
401 Patch submitted by Lars Timman.
403 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
405 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
406 Requires trusted mode and valid format message id, aborts otherwise.
407 Patch contributed by Heiko Schlichting.
409 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
410 certextract with support for various fields. Bug 1358.
412 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
413 is requested by default, modifiable by smtp transport option
416 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
417 operate on certificate variables to give certificate fingerprints
418 Also new ${sha256:cert_variable}.
420 JH/23 The PRDR feature is moved from being Experimental into the mainline.
422 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
423 Christian Aistleitner.
425 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
427 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
428 file. Patch from Wolfgang Breyha.
430 JH/25 Expand the coverage of the delivery $host and $host_address to
431 client authenticators run in verify callout. Bug 1476.
433 JH/26 Port service names are now accepted for tls_on_connect_ports, to
434 align with daemon_smtp_ports. Bug 72.
436 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
437 support and error reporting did not work properly.
439 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
440 and is readable. Patch from Andrew Colin Kissa.
442 TL/14 Enhance documentation of ${run expansion and how it parses the
443 commandline after expansion, particularly in the case when an
444 unquoted variable expansion results in an empty value.
446 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
448 PP/02 Fix internal collision of T_APL on systems which support RFC3123
449 by renaming away from it. Addresses GH issue 15, reported by
452 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
454 TL/15 SECURITY: prevent double expansion in math comparison functions
455 (can expand unsanitized data). Not remotely exploitable.
462 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
464 PP/02 Make -n do something, by making it not do something.
465 When combined with -bP, the name of an option is not output.
467 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
470 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
471 $sender_host_name and config options to manage this, and basic check
474 PP/05 DSCP support for outbound connections and control modifier for inbound.
476 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
477 (Only plugin which currently uses this is kerberos4, which nobody should
478 be using, but we should make it available and other future plugins might
479 conceivably use it, even though it would break NAT; stuff *should* be
480 using channel bindings instead).
482 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
483 name; added for Sendmail compatibility; requires admin caller.
484 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
485 just ignore it); requires trusted caller.
486 Also parse but ignore: -Ac -Am -X<logfile>
489 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
491 TL/02 Add +smtp_confirmation as a default logging option.
493 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
494 Patch by Magnus Holmgren from 2007-02-20.
496 TL/04 Bugzilla 1281 - Spec typo.
497 Bugzilla 1283 - Spec typo.
498 Bugzilla 1290 - Spec grammar fixes.
500 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
502 TL/06 Add Experimental DMARC support using libopendmarc libraries.
504 TL/07 Fix an out of order global option causing a segfault. Reported to dev
505 mailing list by by Dmitry Isaikin.
507 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
509 JH/02 Support "G" suffix to numbers in ${if comparisons.
511 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
513 NM/01 Bugzilla 1197 - Spec typo
514 Bugzilla 1196 - Spec examples corrections
516 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
518 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
519 gnutls_enable_pkcs11, but renamed to more accurately indicate its
522 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
523 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
525 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
526 "acl {{name}{arg}...}", and optional args on acl condition
529 JH/05 Permit multiple router/transport headers_add/remove lines.
531 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
533 JH/07 Avoid using a waiting database for a single-message-only transport.
534 Performance patch from Paul Fisher. Bugzilla 1262.
536 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
539 JH/09 Add $headers_added variable, with content from use of ACL modifier
540 add_header (but not yet added to the message). Bugzilla 199.
542 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
543 Pulled from Bugzilla 817 by Wolfgang Breyha.
545 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
547 (nb: this is the same fix as in Exim 4.80.1)
549 JH/11 Add A= logging on delivery lines, and a client_set_id option on
552 JH/12 Add optional authenticated_sender logging to A= and a log_selector
555 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
557 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
558 advertise SMTP AUTH mechanism to us, instead of a generic
559 protocol violation error. Also, make Exim more robust to bad
560 data from the Dovecot auth socket.
562 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
564 When a queue runner is handling a message, Exim first routes the
565 recipient addresses, during which it prunes them based on the retry
566 hints database. After that it attempts to deliver the message to
567 any remaining recipients. It then updates the hints database using
570 So if a recipient address works intermittently, it can get repeatedly
571 deferred at routing time. The retry hints record remains fresh so the
572 address never reaches the final cutoff time.
574 This is a fairly common occurrence when a user is bumping up against
575 their storage quota. Exim had some logic in its local delivery code
576 to deal with this. However it did not apply to per-recipient defers
577 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
579 This change adds a proper retry rule check during routing so that the
580 final cutoff time is checked against the message's age. We only do
581 this check if there is an address retry record and there is not a
582 domain retry record; this implies that previous attempts to handle
583 the address had the retry_use_local_parts option turned on. We use
584 this as an approximation for the destination being like a local
585 delivery, as in LMTP.
587 I suspect this new check makes the old local delivery cutoff check
588 redundant, but I have not verified this so I left the code in place.
590 TF/02 Correct gecos expansion when From: is a prefix of the username.
592 Test 0254 submits a message to Exim with the header
596 When I ran the test suite under the user fanf2, Exim expanded
597 the header to contain my full name, whereas it should have added
598 a Resent-Sender: header. It erroneously treats any prefix of the
599 username as equal to the username.
601 This change corrects that bug.
603 GF/01 DCC debug and logging tidyup
604 Error conditions log to paniclog rather than rejectlog.
605 Debug lines prefixed by "DCC: " to remove any ambiguity.
607 TF/03 Avoid unnecessary rebuilds of lookup-related code.
609 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
610 Bug spotted by Jeremy Harris; was flawed since initial commit.
611 Would have resulted in OCSP responses post-SNI triggering an Exim
612 NULL dereference and crash.
614 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
616 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
617 Bug detection, analysis and fix by Samuel Thibault.
618 Bugzilla 1331, Debian bug #698092.
620 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
622 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
623 Server implementation by Todd Lyons, client by JH.
624 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
625 config variable "prdr_enable" controls whether the server
626 advertises the facility. If the client requests PRDR a new
627 acl_data_smtp_prdr ACL is called once for each recipient, after
628 the body content is received and before the acl_smtp_data ACL.
629 The client is controlled by bolth of: a hosts_try_prdr option
630 on the smtp transport, and the server advertisement.
631 Default client logging of deliveries and rejections involving
632 PRDR are flagged with the string "PRDR".
634 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
635 fclose(). Diagnosis by Todd Lyons.
637 PP/17 Update configure.default to handle IPv6 localhost better.
638 Patch by Alain Williams (plus minor tweaks).
641 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
642 This is now consistent with GnuTLS, and is now documented: the
643 previous undocumented portable approach to treating the option as
644 unset was to force an expansion failure. That still works, and
645 an empty string is now equivalent.
647 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
648 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
649 not performing validation itself.
651 PP/20 Added force_command boolean option to pipe transport.
652 Patch from Nick Koston, of cPanel Inc.
654 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
657 TF/04 Added udpsend ACL modifer and hexquote expansion operator
659 PP/21 Fix eximon continuous updating with timestamped log-files.
660 Broken in a format-string cleanup in 4.80, missed when I repaired the
661 other false fix of the same issue.
662 Report and fix from Heiko Schlichting.
665 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
666 Report from Prashanth Katuri.
668 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
669 It's SecureTransport, so affects any MacOS clients which use the
670 system-integrated TLS libraries, including email clients.
672 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
673 using a MIME ACL for non-SMTP local injection.
674 Report and assistance in diagnosis by Warren Baker.
676 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
678 JH/16 Fix comparisons for 64b. Bugzilla 1385.
680 TL/09 Add expansion variable $authenticated_fail_id to keep track of
681 last id that failed so it may be referenced in subsequent ACL's.
683 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
686 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
687 ldap library initialization, allowing self-signed CA's to be
688 used. Also properly sets require_cert option later in code by
689 using NULL (global ldap config) instead of ldap handle (per
690 session). Bug diagnosis and testing by alxgomz.
692 TL/12 Enhanced documentation in the ratelimit.pl script provided in
693 the src/util/ subdirectory.
695 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
696 renamed to Transport Post Delivery Action by Jeremy Harris, as
699 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
700 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
701 redis_servers = needs to be configured which will be used by the redis
702 lookup. Patch from Warren Baker, of The Packet Hub.
704 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
706 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
707 hostname or reverse DNS when processing a host list. Used suggestions
708 from multiple comments on this bug.
710 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
712 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
713 Missed a few lines, added it to make the runtest require no keyboard
716 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
717 contains upper case chars. Make router use caseful_local_part.
719 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
720 support when GnuTLS has been built with p11-kit.
726 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
728 This, or similar/improved, will also be change PP/11 of 4.82.
734 PP/01 Handle short writes when writing local log-files.
735 In practice, only affects FreeBSD (8 onwards).
736 Bugzilla 1053, with thanks to Dmitry Isaikin.
738 NM/01 Bugzilla 949 - Documentation tweak
740 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
743 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
745 PP/02 Implemented gsasl authenticator.
747 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
749 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
750 `pkg-config foo` for cflags/libs.
752 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
753 with rest of GSASL and with heimdal_gssapi.
755 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
756 `pkg-config foo` for cflags/libs for the TLS implementation.
758 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
759 properties get this fed in as external SSF. A number of robustness
760 and debugging improvements to the cyrus_sasl authenticator.
762 PP/08 cyrus_sasl server now expands the server_realm option.
764 PP/09 Bugzilla 1214 - Log authentication information in reject log.
765 Patch by Jeremy Harris.
767 PP/10 Added dbmjz lookup type.
769 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
771 PP/12 MAIL args handles TAB as well as SP, for better interop with
772 non-compliant senders.
773 Analysis and variant patch by Todd Lyons.
775 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
776 Bug report from Lars Müller <lars@samba.org> (via SUSE),
777 Patch from Dirk Mueller <dmueller@suse.com>
779 PP/13 tls_peerdn now print-escaped for spool files.
780 Observed some $tls_peerdn in wild which contained \n, which resulted
781 in spool file corruption.
783 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
784 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
785 or write after TLS renegotiation, which otherwise led to messages
788 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
789 as a tracking header (ie: a signed header comes before the signature).
790 Patch from Wolfgang Breyha.
792 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
793 comma-sep list; embedded commas doubled.
795 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
797 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
799 Report and patch from Dmitry Banschikov.
801 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
802 Removed SSL_clear() after SSL_new() which led to protocol negotiation
803 failures. We appear to now support TLS1.1+ with Exim.
805 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
806 lets Exim select keys and certificates based upon TLS SNI from client.
807 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
808 before an outbound SMTP session. New log_selector, +tls_sni.
810 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
811 NULL dereference. Report and patch from Alun Jones.
813 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
814 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
815 Not seeing resolver debug output on NetBSD, but suspect this is a
816 resolver implementation change.
818 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
819 Left warnings. Added "eximon gdb" invocation mode.
821 PP/21 Defaulting "accept_8bitmime" to true, not false.
823 PP/22 Added -bw for inetd wait mode support.
825 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
826 locate the relevant includes and libraries. Made this the default.
828 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
829 Bugzilla 1246, report and most of solution from Tomasz Kusy.
831 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
832 This may cause build issues on older platforms.
834 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
835 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
836 gnutls_require_mac & gnutls_require_protocols (no longer supported).
837 Added SNI support via GnuTLS too.
838 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
840 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
842 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
843 Applied second patch from Janne, implementing suggestion to default
844 multiple-strings-in-record handling to match SPF spec.
846 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
848 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
849 read-only, out of scope).
850 Patch from Wolfgang Breyha, report from Stuart Northfield.
852 PP/29 Fix three issues highlighted by clang analyser static analysis.
853 Only crash-plausible issue would require the Cambridge-specific
854 iplookup router and a misconfiguration.
855 Report from Marcin Mirosław.
857 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
859 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
860 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
861 As part of this, removing so much warning spew let me fix some minor
862 real issues in debug logging.
864 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
865 assignment on my part. Fixed.
867 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
868 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
869 Janne Snabb (who went above and beyond: thank you).
871 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
872 string otherwise requires a connection and a bunch more work and it's
873 relatively easy to get wrong. Should also expose TLS library linkage
876 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
877 64-bit ${eval} (JH/03).
879 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
880 GNU libc to support some of the 64-bit stuff, should not lead to
881 conflicts. Defined before os.h is pulled in, so if a given platform
882 needs to override this, it can.
884 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
885 protection layer was required, which is not implemented.
886 Bugzilla 1254, patch from Wolfgang Breyha.
888 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
889 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
890 tls_dhparam take prime identifiers. Also unbreak combination of
891 OpenSSL+DH_params+TLSSNI.
893 PP/39 Disable SSLv2 by default in OpenSSL support.
899 PP/01 Solaris build fix for Oracle's LDAP libraries.
900 Bugzilla 1109, patch from Stephen Usher.
902 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
904 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
907 TF/02 Fix a couple more cases where we did not log the error message
908 when unlink() failed. See also change 4.74-TF/03.
910 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
911 lock up or crash if it happened to be inside a call to libc when it
912 got a SIGUSR1 from exiwhat.
914 The SIGUSR1 handler appends the current process status to the process
915 log which is later printed by exiwhat. It used to use the general
916 purpose logging code to do this, but several functions it calls are
917 not safe for signals.
919 The new output code in the SIGUSR1 handler is specific to the process
920 log, and simple enough that it's easy to inspect for signal safety.
921 Removing some special cases also simplifies the general logging code.
922 Removing the spurious timestamps from the process log simplifies
925 TF/04 Improved ratelimit ACL condition.
927 The /noupdate option has been deprecated in favour of /readonly which
928 has clearer semantics. The /leaky, /strict, and /readonly update modes
929 are mutually exclusive. The update mode is no longer included in the
930 database key; it just determines when the database is updated. (This
931 means that when you upgrde Exim will forget old rate measurements.)
933 Exim now checks that the per_* options are used with an update mode that
934 makes sense for the current ACL. For example, when Exim is processing a
935 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
936 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
937 must specify per_mail/readonly. If you omit the update mode it defaults to
938 /leaky where that makes sense (as before) or /readonly where required.
940 The /noupdate option is now undocumented but still supported for
941 backwards compatibility. It is equivalent to /readonly except that in
942 ACLs where /readonly is required you may specify /leaky/noupdate or
943 /strict/noupdate which are treated the same as /readonly.
945 A useful new feature is the /count= option. This is a generalization
946 of the per_byte option, so that you can measure the throughput of other
947 aggregate values. For example, the per_byte option is now equivalent
948 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
950 The per_rcpt option has been generalized using the /count= mechanism
951 (though it's more complicated than the per_byte equivalence). When it is
952 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
953 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
954 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
955 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
956 Note that using per_rcpt with a non-readonly update mode in more than
957 one ACL will cause the recipients to be double-counted. (The per_mail
958 and per_byte options don't have this problem.)
960 The handling of very low rates has changed slightly. If the computed rate
961 is less than the event's count (usually one) then this event is the first
962 after a long gap. In this case the rate is set to the same as this event's
963 count, so that the first message of a spam run is counted properly.
965 The major new feature is a mechanism for counting the rate of unique
966 events. The new per_addr option counts the number of different
967 recipients that someone has sent messages to in the last time period. It
968 behaves like per_rcpt if all the recipient addresses are different, but
969 duplicate recipient addresses do not increase the measured rate. Like
970 the /count= option this is a general mechanism, so the per_addr option
971 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
972 example, measure the rate that a client uses different sender addresses
973 with the options per_mail/unique=$sender_address. There are further
974 details in the main documentation.
976 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
978 TF/06 Removed a few PCRE remnants.
980 TF/07 Automatically extract Exim's version number from tags in the git
981 repository when doing development or release builds.
983 PP/02 Raise smtp_cmd_buffer_size to 16kB.
984 Bugzilla 879. Patch from Paul Fisher.
986 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
987 Heavily based on revision 40f9a89a from Simon Arlott's tree.
990 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
992 PP/05 Variable $av_failed, true if the AV scanner deferred.
993 Bugzilla 1078. Patch from John Horne.
995 PP/06 Stop make process more reliably on build failure.
996 Bugzilla 1087. Patch from Heiko Schlittermann.
998 PP/07 Make maildir_use_size_file an _expandable_ boolean.
999 Bugzilla 1089. Patch from Heiko Schlittermann.
1001 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1002 Bugzilla 1131. Patch from Holger Weiß.
1004 PP/09 Handle IPv6 addresses with SPF.
1005 Bugzilla 860. Patch from Wolfgang Breyha.
1007 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1009 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1012 PP/11 match_* no longer expand right-hand-side by default.
1013 New compile-time build option, EXPAND_LISTMATCH_RHS.
1014 New expansion conditions, "inlist", "inlisti".
1016 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1018 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1020 PP/14 fix log_write() format string regression from TF/03.
1021 Bugzilla 1152. Patch from Dmitry Isaikin.
1027 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1029 PP/02 Harmonised TLS library version reporting; only show if debugging.
1030 Layout now matches that introduced for other libraries in 4.74 PP/03.
1032 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1034 PP/04 New "dns_use_edns0" global option.
1036 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1039 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1040 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1042 TK/01 Updated PolarSSL code to 0.14.2.
1043 Bugzilla 1097. Patch from Andreas Metzler.
1045 PP/07 Catch divide-by-zero in ${eval:...}.
1046 Fixes bugzilla 1102.
1048 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1051 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1052 format-string attack -- SECURITY: remote arbitrary code execution.
1054 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1055 time unintentionally subject to list matching rules, letting the header
1056 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1057 arbitrary string expansion). This allowed for information disclosure.
1059 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1060 INT_MIN/-1 -- value coerced to INT_MAX.
1066 NM/01 Workround for PCRE version dependency in version reporting
1069 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1070 This fixes portability to compilers other than gcc, notably
1071 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1073 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1074 makefiles for portability to HP-UX and POSIX correctness.
1076 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1077 Also via indented variable definition in the Makefile.
1078 (Debugging by Oliver Heesakkers).
1080 PP/02 Restore caching of spamd results with expanded spamd_address.
1081 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1083 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1084 Improves build reliability. Fix from: Frank Elsner
1086 NM/02 Fix wide character breakage in the rfc2047 coding
1087 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1089 NM/03 Allow underscore in dnslist lookups
1090 Fixes bug 1026. Patch from Graeme Fowler
1092 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1093 Code patches from Adam Ciarcinski of NetBSD.
1095 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1098 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1099 is logged, to avoid truncation. Patch from John Horne.
1101 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1102 Patch from Jakob Hirsch.
1104 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1105 SQL string expansion failure details.
1106 Patch from Andrey Oktyabrski.
1108 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1109 Patch from Simon Arlott.
1111 PP/09 New lookups functionality failed to compile on old gcc which rejects
1112 extern declarations in function scope.
1113 Patch from Oliver Fleischmann
1115 PP/10 Use sig_atomic_t for flags set from signal handlers.
1116 Check getgroups() return and improve debugging.
1117 Fixed developed for diagnosis in bug 927 (which turned out to be
1120 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1121 Patch from Mark Zealey.
1123 PP/12 Bugzilla 1056: Improved spamd server selection.
1124 Patch from Mark Zealey.
1126 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1127 Based on patch from Heiko Schlittermann.
1129 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1130 Patch from Uwe Doering, sign-off by Michael Haardt.
1132 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1133 variable declaration deep within a block. Bug and patch from
1136 PP/15 lookups-Makefile IRIX compatibilty coercion.
1138 PP/16 Make DISABLE_DKIM build knob functional.
1140 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1141 Patch by Simon Arlott
1143 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1144 variable argument macros. Our copy now differs from upstream.
1150 TF/01 Failure to get a lock on a hints database can have serious
1151 consequences so log it to the panic log.
1153 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1154 controlled using the smtp_confirmation log selector.
1156 TF/03 Include the error message when we fail to unlink a spool file.
1158 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1159 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1160 for maintaining out-of-tree patches for some time.
1162 PP/01 Bugzilla 139: Documentation and portability issues.
1163 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1164 Handle per-OS dynamic-module compilation flags.
1166 PP/02 Let /dev/null have normal permissions.
1167 The 4.73 fixes were a little too stringent and complained about the
1168 permissions on /dev/null. Exempt it from some checks.
1169 Reported by Andreas M. Kirchwitz.
1171 PP/03 Report version information for many libraries, including
1172 Exim version information for dynamically loaded libraries. Created
1173 version.h, now support a version extension string for distributors
1174 who patch heavily. Dynamic module ABI change.
1176 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1177 privilege escalation vulnerability whereby the Exim run-time user
1178 can cause root to append content of the attacker's choosing to
1181 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1184 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1185 If dropping privileges for untrusted macros, we disabled normal logging
1186 on the basis that it would fail; for the Exim run-time user, this is not
1187 the case, and it resulted in successful deliveries going unlogged.
1188 Fixed. Reported by Andreas Metzler.
1194 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1195 only prepend for the Resent-* case. Fixes regression introduced in
1196 Exim 4.70 by NM/22 for Bugzilla 607.
1198 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1199 increasing numbers of administrators be bitten by this.
1201 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1203 PP/03 Bugzilla 994: added openssl_options main configuration option.
1205 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1207 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1209 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1211 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1212 without a peer certificate, leading to a segfault because of an
1213 assumption that peers always have certificates. Be a little more
1214 paranoid. Problem reported by Martin Tscholak.
1216 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1217 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1218 NB: ClamAV planning to remove STREAM in "middle of 2010".
1219 CL also introduces -bmalware, various -d+acl logging additions and
1220 more caution in buffer sizes.
1222 PP/09 Implemented reverse_ip expansion operator.
1224 PP/10 Bugzilla 937: provide a "debug" ACL control.
1226 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1228 PP/12 Bugzilla 973: Implement --version.
1230 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1232 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1234 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1236 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1237 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1238 ignore trailing whitespace.
1240 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1242 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1245 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1246 Notification from Dr Andrew Aitchison.
1248 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1249 ExtendedDetectionInfo response format.
1250 Notification from John Horne.
1252 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1255 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1256 XSL and documented dependency on system catalogs, with examples of how
1259 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1262 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1263 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1264 configuration file which is writeable by the Exim user or group.
1266 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1267 of configuration files to cover files specified with the -C option if
1268 they are going to be used with root privileges, not just the default
1271 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1272 option (effectively making it always true).
1274 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1275 files to be used while preserving root privileges.
1277 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1278 that rogue child processes cannot use them.
1280 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1281 run-time user, instead of root.
1283 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1284 Exim run-time user without dropping privileges.
1286 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1287 result string, instead of calling string_vformat() twice with the same
1290 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1291 for other users. Others should always drop root privileges if they use
1292 -C on the command line, even for a whitelisted configure file.
1294 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1296 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1302 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1303 $data_path, and $header_path variables; fixed documentation bugs and
1306 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1307 exipick to access non-standard spools, including the "frozen" queue
1310 NM/01 Bugzilla 965: Support mysql stored procedures.
1311 Patch from Alain Williams
1313 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1315 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1316 Patch from Andreas Metzler
1318 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1319 Patch from Kirill Miazine
1321 NM/05 Bugzilla 671: Added umask to procmail example.
1323 JJ/03 installed exipick 20100323.0, fixing doc bug
1325 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1326 directory. Notification and patch from Dan Rosenberg.
1328 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1330 TK/02 Improve log output when DKIM signing operation fails.
1332 MH/01 Treat the transport option dkim_domain as a colon separated
1333 list, not as a single string, and sign the message with each element,
1334 omitting multiple occurences of the same signer.
1336 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1337 Bugzilla 985, 986. Patch by Simon Arlott
1339 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1340 Patch by Simon Arlott
1342 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1343 MBX locking. Notification from Dan Rosenberg.
1349 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1351 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1353 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1355 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1357 NM/04 Bugzilla 915: Flag broken perl installation during build.
1363 TK/01 Added patch by Johannes Berg that expands the main option
1364 "spamd_address" if it starts with a dollar sign.
1366 TK/02 Write list of recipients to X-Envelope-Sender header when building
1367 the mbox-format spool file for content scanning (suggested by Jakob
1370 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1371 (http://www.dcc-servers.net/) support via dccifd. Activated by
1372 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1374 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1375 by Mark Daniel Reidel <mr@df.eu>.
1377 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1378 When building exim an external PCRE library is now needed -
1379 PCRE is a system library on the majority of modern systems.
1380 See entry on PCRE_LIBS in EDITME file.
1382 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1383 conversation. Added nologin parameter to request.
1384 Patch contributed by Kirill Miazine.
1386 TF/01 Do not log submission mode rewrites if they do not change the address.
1388 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1390 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1391 log files in place. Contributed by Roberto Lima.
1393 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1395 TF/03 Bugzilla 615: When checking the local_parts router precondition
1396 after a local_part_suffix or local_part_prefix option, Exim now
1397 does not use the address's named list lookup cache, since this
1398 contains cached lookups for the whole local part.
1400 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1401 Robert Millan. Documentation is in experimental-spec.txt.
1403 TF/04 Bugzilla 668: Fix parallel build (make -j).
1405 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1407 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1408 Patch provided by Jan Srzednicki.
1410 TF/05 Leading white space used to be stripped from $spam_report which
1411 wrecked the formatting. Now it is preserved.
1413 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1414 that they are available at delivery time.
1416 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1418 TF/08 TLS error reporting now respects the incoming_interface and
1419 incoming_port log selectors.
1421 TF/09 Produce a more useful error message if an SMTP transport's hosts
1422 setting expands to an empty string.
1424 NM/06 Bugzilla 744: EXPN did not work under TLS.
1425 Patch provided by Phil Pennock.
1427 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1428 Patch provided by Richard Godbee.
1430 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1431 acl_smtp_notquit, added index entry.
1433 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1434 Patch provided by Eugene Bujak.
1436 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1437 accept(). Patch provided by Maxim Dounin.
1439 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1440 Patch provided by Phil Pennock.
1442 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1444 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1445 Patch provided by Brad "anomie" Jorsch.
1447 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1448 Patch provided by Dean Brooks.
1450 TK/05 Add native DKIM support (does not depend on external libraries).
1452 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1453 Patch provided by Graeme Fowler.
1455 NM/16 Bugzilla 851: Documentation example syntax fix.
1457 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1459 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1462 NM/19 Bugzilla 745: TLS version reporting.
1463 Patch provided by Phil Pennock.
1465 NM/20 Bugzilla 167: bool: condition support.
1466 Patch provided by Phil Pennock.
1468 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1469 clients. Patch provided by Phil Pennock.
1471 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1472 Patch provided by Brad "anomie" Jorsch.
1474 NM/23 Bugzilla 687: Fix misparses in eximstats.
1475 Patch provided by Heiko Schlittermann.
1477 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1478 Patch provided by Heiko Schlittermann.
1480 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1481 plus update to original patch.
1483 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1485 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1486 Patch provided by David Brownlee.
1488 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1490 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1492 NM/30 Bugzilla 888: TLS documentation bugfixes.
1494 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1496 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1497 Unlike the original bugzilla I have changed all shell scripts in src tree.
1499 NM/33 Bugzilla 898: Transport filter timeout fix.
1500 Patch by Todd Rinaldo.
1502 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1503 Patch by Serge Demonchaux.
1505 NM/35 Bugzilla 39: Base64 decode bug fixes.
1506 Patch by Jakob Hirsch.
1508 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1510 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1512 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1514 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1520 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1521 ALT-N's libdkim that I have put here:
1522 http://duncanthrax.net/exim-experimental/
1524 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1525 were called 'true' and it seems that C99 defines that as a reserved
1526 keyword to be used with 'bool' variable types. That means you could
1527 not include C99-style headers which use bools without triggering
1528 build errors in sieve.c.
1530 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1531 as mailq or other aliases. Changed the --help handling significantly
1532 to do whats expected. exim_usage() emits usage/help information.
1534 SC/01 Added the -bylocaldomain option to eximstats.
1536 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1538 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1540 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1546 PH/01 Another patch from the Sieve maintainer.
1548 PH/02 When an IPv6 address is converted to a string for single-key lookup
1549 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1550 dots are used instead of colons so that keys in lsearch files need not
1551 contain colons. This was done some time before quoting was made available
1552 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1553 (notated using the quote facility) so as to distinguish them from IPv4
1554 keys. This meant that lookups for IP addresses in host lists did not work
1555 for iplsearch lookups.
1557 This has been fixed by arranging for IPv6 addresses to be expressed with
1558 colons if the lookup type is iplsearch. This is not incompatible, because
1559 previously such lookups could never work.
1561 The situation is now rather anomolous, since one *can* have colons in
1562 ordinary lsearch keys. However, making the change in all cases is
1563 incompatible and would probably break a number of configurations.
1565 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1568 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1569 conversion specification without a maximum field width, thereby enabling
1570 a rogue spamd server to cause a buffer overflow. While nobody in their
1571 right mind would setup Exim to query an untrusted spamd server, an
1572 attacker that gains access to a server running spamd could potentially
1573 exploit this vulnerability to run arbitrary code as the Exim user.
1575 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1576 $primary_hostname instead of what libspf2 thinks the hosts name is.
1578 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1579 a directory entry by the name of the lookup key. Previously, if a
1580 symlink pointed to a non-existing file or a file in a directory that
1581 Exim lacked permissions to read, a lookup for a key matching that
1582 symlink would fail. Now it is enough that a matching directory entry
1583 exists, symlink or not. (Bugzilla 503.)
1585 PH/03 The body_linecount and body_zerocount variables are now exported in the
1588 PH/04 Added the $dnslist_matched variable.
1590 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1591 This means they are set thereafter only if the connection becomes
1594 PH/06 Added the client_condition to authenticators so that some can be skipped
1595 by clients under certain conditions.
1597 PH/07 The error message for a badly-placed control=no_multiline_responses left
1598 "_responses" off the end of the name.
1600 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1602 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1603 (without spaces) instead of just copying the configuration text.
1605 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1607 PH/11 Added $max_received_linelength.
1609 PH/12 Added +ignore_defer and +include_defer to host lists.
1611 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1612 way in which PCRE > 7.0 is built.
1614 PH/14 Implemented queue_only_load_latch.
1616 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1617 MAIL command. The effect was to mangle the value on 64-bit systems.
1619 PH/16 Another patch from the Sieve maintainer.
1621 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1623 PH/18 If a system quota error occurred while trying to create the file for
1624 a maildir delivery, the message "Mailbox is full" was not appended to the
1625 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1626 only to a quota excession during the actual writing of the file.
1628 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1629 characters?) which causes problems in log lines. The DN values are now
1630 passed through string_printing() before being added to log lines.
1632 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1633 and InterBase are left for another time.)
1635 PH/21 Added message_body_newlines option.
1637 PH/22 Guard against possible overflow in moan_check_errorcopy().
1639 PH/23 POSIX allows open() to be a macro; guard against that.
1641 PH/24 If the recipient of an error message contained an @ in the local part
1642 (suitably quoted, of course), incorrect values were put in $domain and
1643 $local_part during the evaluation of errors_copy.
1649 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1650 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1653 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1654 issue a MAIL command.
1656 PH/02 In an ACL statement such as
1658 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1660 if a client was not listed at all, or was listed with a value other than
1661 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1662 the condition was not true (as it should be), so access was not denied.
1663 The bug was that the ! inversion was incorrectly passed on to the second
1664 item. This has been fixed.
1666 PH/03 Added additional dnslists conditions == and =& which are different from
1667 = and & when the dns lookup returns more than one IP address.
1669 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1670 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1672 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1673 FSYNC, which compiles an option called disable_fsync that allows for
1674 bypassing fsync(). The documentation is heavily laced with warnings.
1676 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1678 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1679 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1680 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1681 including adding "make clean"; (3) Added -fPIC when compiling the test
1682 dynamically loaded module, to get rid of a warning.
1684 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1685 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1686 = 0s. The bug is otherwise harmless.
1688 PH/07 There was a bug in the dovecot authenticator such that the value of
1689 $auth1 could be overwritten, and so not correctly preserved, after a
1690 successful authentication. This usually meant that the value preserved by
1691 the server_setid option was incorrect.
1693 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1695 PH/09 Installed PCRE release 7.0.
1697 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1698 run for batched SMTP input. It is now run at the start of every message
1699 in the batch. While fixing this I discovered that the process information
1700 (output by running exiwhat) was not always getting set for -bs and -bS
1701 input. This is fixed, and it now also says "batched" for BSMTP.
1703 PH/11 Added control=no_pipelining.
1705 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1706 patch, slightly modified), and move the expansion of helo_data till after
1707 the connection is made in the smtp transport (so it can use these
1710 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1712 PH/14 Added log_selector = +pid.
1714 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1716 PH/16 Add ${if forany and ${if forall.
1718 PH/17 Added dsn_from option to vary the From: line in DSNs.
1720 PH/18 Flush SMTP output before performing a callout, unless control =
1721 no_callout_flush is set.
1723 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1724 was true (the default) a successful delivery failed to delete the retry
1725 item, thus causing premature timeout of the address. The bug is now
1728 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1730 PH/21 Long custom messages for fakedefer and fakereject are now split up
1731 into multiline reponses in the same way that messages for "deny" and
1732 other ACL rejections are.
1734 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1735 with slight modification.
1737 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1738 draft, changing the syntax and factoring some duplicate code".
1740 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1741 for deliveries of the second and subsequent messages over the same SMTP
1744 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1745 ${reduce, with only minor "tidies".
1747 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1749 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1750 expansion side effects.
1752 PH/27 When a message times out after an over-quota error from an Exim-imposed
1753 quota, the bounce message says "mailbox is full". This message was not
1754 being given when it was a system quota that was exceeded. It now should
1757 MH/03 Made $recipients available in local_scan(). local_scan() already has
1758 better access to the recipient list through recipients_list[], but
1759 $recipients can be useful in postmaster-provided expansion strings.
1761 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1762 in the case of a MAIL command with additional options following the
1763 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1764 were accidentally chopped off.
1766 PH/29 SMTP synchronization checks are implemented when a command is read -
1767 there is a check that no more input is waiting when there shouldn't be
1768 any. However, for some commands, a delay in an ACL can mean that it is
1769 some time before the response is written. In this time, more input might
1770 arrive, invalidly. So now there are extra checks after an ACL has run for
1771 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1772 pipelining has not been advertised.
1774 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1776 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1777 returned in the SMTP response when smtp_return_error_details was false.
1778 This has been fixed.
1780 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1781 instead of the C I/O that was originally supplied, because problems were
1782 reported on Solaris.
1784 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1785 Exim which did not show up earlier: it was assuming that a call to
1786 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1787 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1788 was a macro that became an assignment, so it seemed to work. This has
1789 changed to a proper function call with a void return, hence the compile
1790 error. Exim's code has been fixed.
1792 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1795 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1797 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1799 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1800 because it turns out to be too big or there is a timeout) there is no
1801 "Completed" line in the log. When some messages of this type were
1802 selected by exigrep, they were listed as "not completed". Others were
1803 picked up by some special patterns. I have improved the selection
1804 criteria to be more general.
1806 PH/38 The host_find_failed option in the manualroute router can now be set
1807 to "ignore", to completely ignore a host whose IP address cannot be
1808 found. If all hosts are ignored, the behaviour is controlled by the new
1809 host_all_ignored option.
1811 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1812 homing or because of multiple MX records with /mx) generated more than
1813 one IP address, and the following item turned out to be the local host,
1814 all the secondary addresses of the first item were incorrectly removed
1815 from the list, along with the local host and any following hosts (which
1816 is what is supposed to happen).
1818 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1819 whoever called Exim into the -H file. In the case of the daemon it was
1820 behaving confusingly. When first started, it used values for whoever
1821 started the daemon, but after a SIGHUP it used the Exim user (because it
1822 calls itself on a restart). I have changed the code so that it now always
1825 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1826 message are rejected with the same error (e.g. no authentication or bad
1827 sender address), and a DATA command is nevertheless sent (as can happen
1828 with PIPELINING or a stupid MUA), the error message that was given to the
1829 RCPT commands is included in the rejection of the DATA command. This is
1830 intended to be helpful for MUAs that show only the final error to their
1833 PH/42 Another patch from the Sieve maintainer.
1835 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1836 Eximstats - Fixed some broken HTML links and added missing column headers
1838 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1839 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1841 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1843 PH/43 Yet another patch from the Sieve maintainer.
1845 PH/44 I found a way to check for a TCP/IP connection going away before sending
1846 the response to the final '.' that terminates a message, but only in the
1847 case where the client has not sent further data following the '.'
1848 (unfortunately, this is allowed). However, in many cases there won't be
1849 any further data because there won't be any more messages to send. A call
1850 to select() can be used: if it shows that the input is "ready", there is
1851 either input waiting, or the socket has been closed. An attempt to read
1852 the next input character can distinguish the two cases. Previously, Exim
1853 would have sent an OK response which the client would never have see.
1854 This could lead to message repetition. This fix should cure that, at
1855 least in a lot of common cases.
1857 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1858 advertised in response to EHLO.
1864 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1865 fixed by 4.65/MH/01 (is this a record?) are fixed:
1867 (i) An empty string was always treated as zero by the numeric comparison
1868 operators. This behaviour has been restored.
1870 (ii) It is documented that the numeric comparison operators always treat
1871 their arguments as decimal numbers. This was broken in that numbers
1872 starting with 0 were being interpreted as octal.
1874 While fixing these problems I realized that there was another issue that
1875 hadn't been noticed. Values of message_size_limit (both the global option
1876 and the transport option) were treated as octal if they started with 0.
1877 The documentation was vague. These values are now always treated as
1878 decimal, and I will make that clear in the documentation.
1884 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1885 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1888 MH/01 Don't check that the operands of numeric comparison operators are
1889 integers when their expansion is in "skipping" mode (fixes bug
1890 introduced by 4.64-PH/07).
1892 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1893 child addresses, Exim now panics and dies. Previously, because the count
1894 is held in a short int, deliveries were likely to be lost. As such a
1895 large number of recipients for a single message is ridiculous
1896 (performance will be very, very poor), I have chosen to impose a limit
1897 rather than extend the field.
1903 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1904 leftover -K file (the existence of which was triggered by #402).
1905 While we were at it, introduced process PID as part of the -K
1906 filename. This should rule out race conditions when creating
1909 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1910 processing considerably. Previous code took too long for large mails,
1911 triggering a timeout which in turn triggers #401.
1913 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1914 in the DK code in transports.c. sendfile() is not really portable,
1915 hence the _LINUX specificness.
1917 TF/01 In the add_headers option to the mail command in an Exim filter,
1918 there was a bug that Exim would claim a syntax error in any
1919 header after the first one which had an odd number of characters
1922 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1923 callout verification, Exim cached a "reject" for the entire domain. This
1924 is correct for most verifications, but it is not correct for a recipient
1925 verification with use_sender or use_postmaster set, because in that case
1926 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1927 case of MAIL FROM:<> rejection from other early rejections (e.g.
1928 rejection of HELO). When verifying a recipient using a non-null MAIL
1929 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1930 Whatever the result of the callout, the value of the domain cache is
1931 left unchanged (for any other kind of callout, getting as far as trying
1932 RCPT means that the domain itself is ok).
1934 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1937 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1938 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1939 session, the daemon ignores ECONNECT errors and logs others; it now
1940 ignores EPIPE as well.
1942 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1943 (quoted-printable decoding).
1945 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1946 later the small subsequent patch to fix an introduced bug.
1948 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1950 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1952 PH/08 An error is now given if message_size_limit is specified negative.
1954 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1955 to be given (somewhat) arbitrary names.
1957 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1960 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1961 miscellaneous code fixes
1963 PH/10 Added the log_reject_target ACL modifier to specify where to log
1966 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1967 hostname. This is wrong, because it relates to the incoming message (and
1968 probably the interface on which it is arriving) and not to the outgoing
1969 callout (which could be using a different interface). This has been
1970 changed to use the value of the helo_data option from the smtp transport
1971 instead - this is what is used when a message is actually being sent. If
1972 there is no remote transport (possible with a router that sets up host
1973 addresses), $smtp_active_hostname is used.
1975 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1976 tweaks were necessary in order to get it to work (see also 21 below):
1977 (a) The code assumed that strncpy() returns a negative number on buffer
1978 overflow, which isn't the case. Replaced with Exim's string_format()
1980 (b) There were several signed/unsigned issues. I just did the minimum
1981 hacking in of casts. There is scope for a larger refactoring.
1982 (c) The code used strcasecmp() which is not a standard C function.
1983 Replaced with Exim's strcmpic() function.
1984 (d) The code set only $1; it now sets $auth1 as well.
1985 (e) A simple test gave the error "authentication client didn't specify
1986 service in request". It would seem that Dovecot has changed its
1987 interface. Fortunately there's a specification; I followed it and
1988 changed what the client sends and it appears to be working now.
1990 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1993 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1994 address A is aliased to B and C, where B exists and C does not. Without
1995 -v the output is "A verified" because verification stops after a
1996 successful redirection if more than one address is generated. However,
1997 with -v the child addresses are also verified. Exim was outputting "A
1998 failed to verify" and then showing the successful verification for C,
1999 with its parentage. It now outputs "B failed to verify", showing B's
2000 parentage before showing the successful verification of C.
2002 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2003 look up a TXT record in a specific list after matching in a combined
2006 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2007 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2008 they consult the DNS. I had assumed they would set it the way they
2009 wanted; and indeed my experiments on Linux seem to show that in some
2010 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2011 To be on the safe side, however, I have now made the interface to
2012 host_find_byname() similar to host_find_bydns(), with an argument
2013 containing the DNS resolver options. The host_find_byname() function now
2014 sets these options at its start, just as host_find_bydns() does. The smtp
2015 transport options dns_qualify_single and dns_search_parents are passed to
2016 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2017 of host_find_byname() use the default settings of RES_DEFNAMES
2018 (qualify_single) but not RES_DNSRCH (search_parents).
2020 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2021 spool_read_header() do less string testing, by means of a preliminary
2022 switch on the second character of optional "-foo" lines. (This is
2023 overdue, caused by the large number of possibilities that now exist.
2024 Originally there were few.) While I was there, I also converted the
2025 str(n)cmp tests so they don't re-test the leading "-" and the first
2026 character, in the hope this might squeeze out yet more improvement.
2028 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2029 flag allowing group syntax was set by the header_syntax check but not
2030 turned off, possible causing trouble later; (2) The flag was not being
2031 set at all for the header_verify test, causing "group"-style headers to
2032 be rejected. I have now set it in this case, and also caused header_
2033 verify to ignore an empty address taken from a group. While doing this, I
2034 came across some other cases where the code for allowing group syntax
2035 while scanning a header line wasn't quite right (mostly, not resetting
2036 the flag correctly in the right place). These bugs could have caused
2037 trouble for malformed header lines. I hope it is now all correct.
2039 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2040 with the "reply" argument non-NULL. The code, however (which originally
2041 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2042 but it didn't always do it. This confused somebody who was copying the
2043 code for some other use. I have removed all the tests.
2045 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2046 feature that was used to support insecure browsers during the U.S. crypto
2047 embargo. It requires special client support, and Exim is probably the
2048 only MTA that supported it -- and would never use it because real RSA is
2049 always available. This code has been removed, because it had the bad
2050 effect of slowing Exim down by computing (never used) parameters for the
2051 RSA_EXPORT functionality.
2053 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2054 authenticator to fail if there's a tab character in the incoming data
2055 (there should never be unless someone is messing about, as it's supposed
2056 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2057 if the connection is using TLS or if the remote IP is the same as the
2058 local IP, and the "valid-client-cert option" if a client certificate has
2061 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2062 authenticators. This can be used for authorization after authentication
2063 succeeds. (In the case of plaintext, it servers for both authentication
2066 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2067 if any retry times were supplied.
2069 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2070 connection, where there is no client IP address to check. In this
2071 situation, the verify now always succeeds.
2073 PH/25 Applied John Jetmore's -Mset patch.
2075 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2077 PH/27 In a string expansion for a processed (not raw) header when multiple
2078 headers of the same name were present, leading whitespace was being
2079 removed from all of them, but trailing whitespace was being removed only
2080 from the last one. Now trailing whitespace is removed from each header
2081 before concatenation. Completely empty headers in a concatenation (as
2082 before) are ignored.
2084 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2085 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2087 PH/29 [Removed. This was a change that I later backed out, and forgot to
2088 correct the ChangeLog entry (that I had efficiently created) before
2089 committing the later change.]
2091 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2092 address errors (4xx response to RCPT) over the same connection as other
2093 messages routed to the same hosts. Such deliveries are always "forced",
2094 so retry times are not inspected. This resulted in far too many retries
2095 for the affected addresses. The effect occurred only when there were more
2096 hosts than the hosts_max_try setting in the smtp transport when it had
2097 the 4xx errors. Those hosts that it had tried were not added to the list
2098 of hosts for which the message was waiting, so if all were tried, there
2099 was no problem. Two fixes have been applied:
2101 (i) If there are any address or message errors in an SMTP delivery, none
2102 of the hosts (tried or untried) are now added to the list of hosts
2103 for which the message is waiting, so the message should not be a
2104 candidate for sending over the same connection that was used for a
2105 successful delivery of some other message. This seems entirely
2106 reasonable: after all the message is NOT "waiting for some host".
2107 This is so "obvious" that I'm not sure why it wasn't done
2108 previously. Hope I haven't missed anything, but it can't do any
2109 harm, as the worst effect is to miss an optimization.
2111 (ii) If, despite (i), such a delivery is accidentally attempted, the
2112 routing retry time is respected, so at least it doesn't keep
2113 hammering the server.
2115 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2116 in ${readsocket because some servers need this prod.
2118 PH/32 Added some extra debug output when updating a wait-xxx database.
2120 PH/33 The hint "could be header name not terminated by colon", which has been
2121 given for certain expansion errors for a long time, was not being given
2122 for the ${if def:h_colon_omitted{... case.
2124 PH/34 The spec says: "With one important exception, whenever a domain list is
2125 being scanned, $domain contains the subject domain." There was at least
2126 one case where this was not true.
2128 PH/35 The error "getsockname() failed: connection reset by peer" was being
2129 written to the panic log as well as the main log, but it isn't really
2130 panic-worthy as it just means the connection died rather early on. I have
2131 removed the panic log writing for the ECONNRESET error when getsockname()
2134 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2135 runs only) independently of the message's sender address. This meant
2136 that, if the 4xx error was in fact related to the sender, a different
2137 message to the same recipient with a different sender could confuse
2138 things. In particualar, this can happen when sending to a greylisting
2139 server, but other circumstances could also provoke similar problems.
2140 I have changed the default so that the retry time for these errors is now
2141 based a combination of the sender and recipient addresses. This change
2142 can be overridden by setting address_retry_include_sender=false in the
2145 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2146 remote server are returned as part of bounce messages. This was not
2147 happening for LMTP over a pipe (the lmtp transport), but now it is the
2148 same for both kinds of LMTP.
2150 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2151 in header lines that were in fact CNAMEs. This is no longer the case.
2153 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2154 and queue runs started by the daemon processed all messages. This has
2155 been fixed so that -R and -S can now usefully be given with -q<time>.
2157 PH/40 Import PCRE release 6.7 (fixes some bugs).
2159 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2161 PH/42 Give an error if -q is specified more than once.
2163 PH/43 Renamed the variables $interface_address and $interface_port as
2164 $received_ip_address and $received_port, to make it clear that these
2165 values apply to message reception, and not to the outgoing interface when
2166 a message is delivered. (The old names remain recognized, of course.)
2168 PH/44 There was no timeout on the connect() call when using a Unix domain
2169 socket in the ${readsocket expansion. There now is.
2171 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2172 be meaningful with "accept".
2174 SC/01 Eximstats V1.43
2175 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2177 SC/02 Eximstats V1.44
2178 Use a glob alias rather than an array ref in the generated
2179 parser. This improves both readability and performance.
2181 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2182 Collect SpamAssassin and rejection statistics.
2183 Don't display local sender or destination tables unless
2184 there is data to show.
2185 Added average volumes into the top table text output.
2187 SC/04 Eximstats V1.46
2188 Collect data on the number of addresses (recipients)
2189 as well as the number of messages.
2191 SC/05 Eximstats V1.47
2192 Added 'Message too big' to the list of mail rejection
2193 reasons (thanks to Marco Gaiarin).
2195 SC/06 Eximstats V1.48
2196 Mainlog lines which have GMT offsets and are too short to
2197 have a flag are now skipped.
2199 SC/07 Eximstats V1.49 (Alain Williams)
2200 Added the -emptyok flag.
2202 SC/08 Eximstats V1.50
2203 Fixes for obtaining the IP address from reject messages.
2205 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2206 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2207 whitesspace changes from 4.64-PH/27
2209 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2212 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2213 are found, allow negative numbers in numeric criteria)
2215 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2217 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2218 to match changes made in 4.64-PH/43
2220 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2222 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2223 "domain" is actually an IP address, and force a failure. This locks out
2224 those revolvers/nameservers that support "A-for-A" lookups, in
2225 contravention of the specifications.
2227 PH/48 When a host name was looked up from an IP address, and the subsequent
2228 forward lookup of the name timed out, the host name was left in
2229 $sender_host_name, contrary to the specification.
2231 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2232 restricted to single-key lookups, Exim was not diagnosing an error if
2233 * or *@ was used with a query-style lookup.
2235 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2237 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2238 long ago, but noone interested enough thought of it. Let's just say that
2239 the "1.1" means that there are some new functions that weren't there at
2240 some point in the past.
2242 PH/51 Error processing for expansion failure of helo_data from an smtp
2243 transport during callout processing was broken.
2245 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2246 tested/used via the -bh/-bhc/-bs options.
2248 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2249 bug, fixed in subsequent PCRE releases).
2251 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2252 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2254 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2260 SC/01 Use a glob alias rather than an array ref in eximstats generated
2261 parser. This improves both readability and performance.
2263 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2264 Don't display local sender or destination tables in eximstats unless
2265 there is data to show.
2266 Added average volumes into the eximstats top table text output.
2268 SC/03 Collect data on the number of addresses (recipients) as well
2269 as the number of messages in eximstats.
2271 TF/01 Correct an error in the documentation for the redirect router. Exim
2272 does (usually) call initgroups() when daemonizing.
2274 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2275 with consistent privilege compared to when running as a daemon.
2277 TF/03 Note in the spec that $authenticated_id is not set for local
2278 submissions from trusted users.
2280 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2281 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2283 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2284 by adding some example configuration directives to the default
2285 configuration file. A little bit of work is required to uncomment the
2286 directives and define how usernames and passwords are checked, but
2287 there is now a framework to start from.
2289 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2290 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2291 without this. I don't know how relevant this is to other LDAP libraries.
2293 PH/02 Add the verb name to the "unknown ACL verb" error.
2295 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2297 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2299 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2300 directory not expanded when it should be if an expanded home directory
2301 was set for the address (which is overridden by the transport).
2303 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2306 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2307 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2308 because it is too late at that time, and has no effect.
2310 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2311 security issue with \' (bugzilla #107). I could not use the
2312 PQescapeStringConn() function, because it needs a PGconn value as one of
2315 PH/08 When testing addresses using -bt, indicate those final addresses that
2316 are duplicates that would not cause an additional delivery. At least one
2317 person was confused, thinking that -bt output corresponded to deliveries.
2318 (Suppressing duplicates isn't a good idea as you lose the information
2319 about possibly different redirections that led to the duplicates.)
2321 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2322 systems where poll() doesn't work, in particular OS X.
2324 PH/10 Added more information to debugging output for retry time not reached.
2326 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2327 operations in malware.c.
2329 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2332 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2333 syslog_duplication set false, log lines that would normally be written
2334 both the the main log and to the reject log were not written to syslog at
2337 PH/14 In the default configuration, change the use of "message" in ACL warn
2338 statements to "add_header".
2340 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2341 not followed by a command (e.g. "seen endif").
2343 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2344 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2347 PH/17 Added extra conditions to the default value of delay_warning_condition
2351 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2352 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2353 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2356 The Auto-Submitted: and various List- headers are standardised, whereas I
2357 don't think Precedence: ever was.
2359 PH/18 Refactored debugging code in route_finduser() to show more information,
2360 in particular, the error code if getpwnam() issues one.
2362 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2363 This is apparently needed in addition to the PH/07 change above to avoid
2364 any possible encoding problems.
2366 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2367 but not after initializing Perl.
2369 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2370 output them only if debugging. By default they are written stderr,
2371 apparently, which is not desirable.
2373 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2376 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2379 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2381 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2382 authenticated or an ident call has been made. Suppress the default
2383 values for $authenticated_id and $authenticated_sender (but permit -oMai
2384 and -oMas) when testing with -bh.
2386 PH/24 Re-jigged the order of the tests in the default configuration so that the
2387 tests for valid domains and recipients precede the DNS black list and CSA
2388 tests, on the grounds that those ones are more expensive.
2390 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2391 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2392 This bug exists in every version of Exim that I still have, right back to
2395 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2396 However, an attempt to turn on case-sensitivity in a regex key by
2397 including (?-i) didn't work because the subject string was already
2398 lowercased, and the effects were non-intuitive. It turns out that a
2399 one-line patch can be used to allow (?-i) to work as expected.
2405 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2406 other effects) broke the use of negated acl sub-conditions.
2408 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2411 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2412 "Deny" causes Exim to reject the incoming connection with a 554 error.
2413 Unfortunately, if there is a major crisis, such as a disk failure,
2414 tcp-wrappers gives "deny", whereas what one would like would be some
2415 kind of temporary error. A kludge has been added to help with this.
2416 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2417 554 error is used if errno is still zero or contains ENOENT (which occurs
2418 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2421 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2423 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2424 errors. Otherwise a message that provokes a temporary error (when other
2425 messages do not) can cause a whole host to time out.
2427 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2428 addresses were routed directly to files or pipes from a redirect router.
2429 File deliveries just didn't batch; pipe deliveries might have suffered
2432 PH/06 A failure to get a lock for a hints database would erroneously always say
2433 "Failed to get write lock", even when it was really a read lock.
2435 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2436 of 0600. This has been changed to use the value of the lockfile_mode
2437 option (which defaults to 0600).
2439 PH/08 Applied small patch from the Sieve maintainer.
2441 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2442 folder from quota calculations, a direct delivery into this folder messed
2443 up the contents of the maildirsize file. This was because the regex was
2444 used only to exclude .Trash (or whatever) when the size of the mailbox
2445 was calculated. There was no check that a delivery was happening into an
2446 excluded directory. This bug has been fixed by ignoring all quota
2447 processing for deliveries into excluded directories.
2449 PH/10 Added the maildirfolder_create_regex option to appendfile.
2455 PH/01 The code for finding all the local interface addresses on a FreeBSD
2456 system running IPv6 was broken. This may well have applied to all BSD
2457 systems, as well as to others that have similar system calls. The broken
2458 code found IPv4 interfaces correctly, but gave incorrect values for the
2459 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2460 that it would not match correctly against @[] and not recognize the IPv6
2463 PH/02 The ipliteral router was not recognizing addresses of the form user@
2464 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2466 PH/03 Added disable_ipv6.
2468 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2469 decoded form, because it is most often used to construct To: headers
2470 lines in autoreplies, and the decoded form may well be syntactically
2471 invalid. However, $reply_address has leading white space removed, and all
2472 newlines turned into spaces so that the autoreply transport does not
2475 PH/05 If group was specified without a user on a router, and no group or user
2476 was specified on a transport, the group from the router was ignored.
2478 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2479 for visible compile-time settings that can be used to change these
2480 numbers, for those that want even more. Backwards compatibility with old
2481 spool files has been maintained. However, going back to a previous Exim
2482 release will lost any variables that are in spool files.
2484 PH/07 Two small changes when running in the test harness: increase delay when
2485 passing a TCP/IP connection to a new process, in case the original
2486 process has to generate a bounce, and remove special handling of
2487 127.0.0.2 (sic), which is no longer necessary.
2489 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2490 be the same on different OS.
2492 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2495 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2496 whether --show-vars was specified or not
2498 JJ/02 exipick: Added support for new ACL variable spool format introduced
2501 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2502 syntactically invalid From: or Reply-to: line, and a filter used this to
2503 generate an autoreply, and therefore failed to obtain an address for the
2504 autoreply, Exim could try to deliver to a non-existent relative file
2505 name, causing unrelated and misleading errors. What now happens is that
2506 it logs this as a hard delivery error, but does not attempt to create a
2509 PH/11 The exinext utility has a -C option for testing purposes, but although
2510 the given file was scanned by exinext itself; it wasn't being passed on
2511 when Exim was called.
2513 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2514 an end-of-file indication when reading a command response.
2516 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2517 compiled. In many other places in Exim, IPv6 addresses are always
2518 recognized, so I have changed this. It also means that IPv4 domain
2519 literals of the form [IPV4:n.n.n.n] are now always recognized.
2521 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2522 used if the router is not running as root, for example, when verifying at
2523 ACL time, or when using -bh. The debugging output from this situation was
2524 non-existent - all you got was a failure to exec. I have made two
2527 (a) Failures to set uid/gid, the current directory, or a process leader
2528 in a subprocess such as that created by queryprogram now generate
2529 suitable debugging ouput when -d is set.
2531 (b) The queryprogram router detects when it is not running as root,
2532 outputs suitable debugging information if -d is set, and then runs
2533 the subprocess without attempting to change uid/gid.
2535 PH/15 Minor change to Makefile for building test_host (undocumented testing
2538 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2539 additional section of a DNS packet that returns MX or SRV records.
2540 Instead, it always explicitly searches for A/AAAA records. This avoids
2541 major problems that occur when a DNS server includes only records of one
2542 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2543 fixed another bug: if SRV records were looked up and the corresponding
2544 address records were *not* found in the additional section, the port
2545 values from the SRV records were lost.
2547 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2548 using the correct key (the original address) when searching the retry
2549 rules in order to find which one to use for generating the retry hint.
2551 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2552 adding the default one. Similarly, if it contains a Reply-To: header, the
2553 errors_reply_to option, if set, is not used.
2555 PH/19 When calculating a retry time, Exim used to measure the "time since
2556 failure" by looking at the "first failed" field in the retry record. Now
2557 it does not use this if it is later than than the arrival time of the
2558 message. Instead it uses the arrival time. This makes for better
2559 behaviour in cases where some deliveries succeed, thus re-setting the
2560 "first failed" field. An example is a quota failure for a huge message
2561 when small messages continue to be delivered. Without this change, the
2562 "time since failure" will always be short, possible causing more frequent
2563 delivery attempts for the huge message than are intended.
2564 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2566 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2567 $1, $2, $3) because the numerical variables can be reset during some
2568 expansion items (e.g. "match"), thereby losing the authentication data.
2570 PH/21 Make -bV show the size of off_t variables so that the test suite can
2571 decide whether to run tests for quotas > 2G.
2573 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2574 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2575 filecount value is greater than 2G or if a quota value is greater than 2G
2576 on a system where the size of off_t is not greater than 4, a panic error
2579 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2580 never match. The debug and -bh output now contains an explicit error
2581 message indicating a malformed IPv4 address or mask.
2583 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2584 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2585 PH/23 above applies.
2587 PH/25 Do not write to syslog when running in the test harness. The only
2588 occasion when this arises is a failure to open the main or panic logs
2589 (for which there is an explicit test).
2591 PH/26 Added the /no_tell option to "control=freeze".
2593 PH/27 If a host name lookup failed very early in a connection, for example, if
2594 the IP address matched host_lookup and the reverse lookup yielded a name
2595 that did not have a forward lookup, an error message of the form "no IP
2596 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2597 could be logged. Now it outputs the IP address instead of "NULL".
2599 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2600 allows the sender and the authenticated sender to be set when
2601 submitting a message from within Exim. Since child_open_exim() is
2602 documented for local_scan(), the new function should be too.
2604 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2605 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2606 results in an empty string is now treated as unset.
2608 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2610 PH/31 Added sender_verify_fail logging option.
2612 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2613 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2614 tidied the source and removed it altogether.
2616 PH/33 When a queue run was abandoned because the load average was too high, a
2617 log line was always written; now it is written only if the queue_run log
2618 selector is set. In addition, the log line for abandonment now contains
2619 information about the queue run such as the pid. This is always present
2620 in "start" and "stop" lines but was omitted from the "abandon" line.
2622 PH/34 Omit spaces between a header name and the colon in the error message that
2623 is given when verify = headers_syntax fails (if there are lots of them,
2624 the message gets confusing).
2626 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2627 names, as there are now some PTR records that contain slashes. This check
2628 is only to protect against broken name servers that fall over on strange
2629 characters, so the fact that it applies to all lookups doesn't matter.
2631 PH/36 Now that the new test suite is complete, we can remove some of the
2632 special code in Exim that was needed for the old test suite. For example,
2633 sorting DNS records because real resolvers return them in an arbitrary
2634 order. The new test suite's fake resolver always returns records in the
2637 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2638 bounces) except when queue_only is set, to avoid logging races between
2639 the different processes.
2641 PH/38 Panic-die if .include specifies a non-absolute path.
2643 PH/39 A tweak to the "H" retry rule from its user.
2645 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2646 a label. They prevented compilation on older perls.
2648 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2649 a warning to be raised on newish perls.
2651 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2652 on queue. Changes to match documented behaviour of showing count of
2653 messages matching specified criteria.
2655 PH/40 Changed the default ident timeout from 30s to 5s.
2657 PH/41 Added support for the use of login_cap features, on those BSD systems
2658 that have them, for controlling the resources used by pipe deliveries.
2660 PH/42 The content-scanning code uses fopen() to create files in which to put
2661 message data. Previously it was not paying any attention to the mode of
2662 the files. Exim runs with umask(0) because the rest of the code creates
2663 files with open(), and sets the required mode explicitly. Thus, these
2664 files were ending up world-writeable. This was not a big issue, because,
2665 being within the spool directory, they were not world-accessible. I have
2666 created a function called modefopen, which takes an additional mode
2667 argument. It sets umask(777), creates the file, chmods it to the required
2668 mode, then resets the umask. All the relevant calls to fopen() in the
2669 content scanning code have been changed to use this function.
2671 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2672 to 24 hours. This avoids potential overflow problems when processing G
2673 and H retry rules. I suspect nobody ever tinkers with this value.
2675 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2677 PH/45 When the plaintext authenticator is running as a client, the server's
2678 challenges are checked to ensure they are valid base64 strings. By
2679 default, the authentication attempt is cancelled if an invalid string is
2680 received. Setting client_ignore_invalid_base64 true ignores these errors.
2681 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2682 they are received. Thus, the responses can be made to depend on the
2683 challenges. If an invalid string is ignored, an empty string is placed in
2686 PH/46 Messages that are created by the autoreply transport now contains a
2687 References: header, in accordance with RFCs 2822 and 3834.
2689 PH/47 Added authenticated_sender_force to the smtp transport.
2691 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2693 PH/49 Installed latest patch from the Sieve maintainer.
2695 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2696 was also set, the appendfile transport was unnecessarily scanning a
2697 directory of message files (e.g. for maildir delivery) to find the count
2698 of files (along with the size), even though it did not need this
2699 information. It now does the scan only if it needs to find either the
2700 size of the count of files.
2702 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2704 PH/52 Two bugs concerned with error handling when the smtp transport is
2707 (i) Exim was not creating retry information for temporary errors given
2708 for individual recipients after the DATA command when the smtp transport
2709 was used in LMTP mode. This meant that they could be retried too
2710 frequently, and not timed out correctly.
2712 (ii) Exim was setting the flag that allows error details to be returned
2713 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2714 recipients that were returned after the DATA command.
2716 PH/53 This is related to PH/52, but is more general: for any failing address,
2717 when detailed error information was permitted to be returned to the
2718 sender, but the error was temporary, then after the final timeout, only
2719 "retry timeout exceeded" was returned. Now it returns the full error as
2720 well as "retry timeout exceeded".
2722 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2723 do this, and (what is worse) MTAs that accept it.
2725 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2726 will now be deprecated.
2728 PH/56 New os.c-cygwin from the Cygwin maintainer.
2730 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2731 formats (previously only available in exim formats via -bpr, -bpru,
2732 and -bpra. Now also available in native and exiqgrep formats)
2734 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2735 with very large, slow to parse queues
2737 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2739 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2741 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2742 responses to authentication challenges, though it was showing the
2743 challenges; (ii) I've removed the CR characters from the debug output for
2746 PH/58 Allow for the insertion of a newline as well as a space when a string
2747 is turned into more than one encoded-word during RFC 2047 encoding. The
2748 Sieve code now uses this.
2750 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2751 data_4xx, lost_connection, tls_required.
2753 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2754 message was being sent as an SMTP response.
2756 PH/61 Add -l and -k options to exicyclog.
2758 PH/62 When verifying, if an address was redirected to one new address, so that
2759 verification continued, and the new address failed or deferred after
2760 having set something in $address_data, the value of $address_data was not
2761 passed back to the ACL. This was different to the case when no
2762 redirection occurred. The value is now passed back in both cases.
2764 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2765 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2766 use login_cap.h, so on its own it isn't the distinguishing feature. The
2767 new name refers directly to the setclassresources() function.
2769 PH/65 Added configuration files for NetBSD3.
2771 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2773 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2774 is preferred over IPv4.
2776 PH/68 The bounce_return_message and bounce_return_body options were not being
2777 honoured for bounces generated during the reception of non-SMTP messages.
2778 In particular, this applied to messages rejected by the ACL. This bug has
2779 been fixed. However, if bounce_return_message is true and bounce_return_
2780 body is false, the headers that are returned for a non-SMTP message
2781 include only those that have been read before the error was detected.
2782 (In the case of an ACL rejection, they have all been read.)
2784 PH/69 The HTML version of the specification is now built in a directory called
2785 spec_html instead of spec.html, because the latter looks like a path with
2786 a MIME-type, and this confuses some software.
2788 PH/70 Catch two compiler warnings in sieve.c.
2790 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2791 function verify_get_ident() calls ip_connect() to connect a socket, but
2792 if the "connect()" function timed out, ip_connect() used to close the
2793 socket. However, verify_get_ident() also closes the socket later, and in
2794 between Exim writes to the log, which may get opened at this point. When
2795 the socket was closed in ip_connect(), the log could get the same file
2796 descriptor number as the socket. This naturally causes chaos. The fix is
2797 not to close the socket in ip_connect(); the socket should be closed by
2798 the function that creates it. There was only one place in the code where
2799 this was missing, in the iplookup router, which I don't think anybody now
2800 uses, but I've fixed it anyway.
2802 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2803 well as to direct DNS lookups. Otherwise the handling of names in host
2804 lists is inconsistent and therefore confusing.
2810 PH/01 Two changes to the default runtime configuration:
2812 (1) Move the checks for relay_from_hosts and authenticated clients from
2813 after to before the (commented out) DNS black list checks.
2815 (2) Add control=submission to the relay_from_hosts and authenticated
2816 clients checks, on the grounds that messages accepted by these
2817 statements are most likely to be submissions.
2819 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2821 (1) Generate an error if the third argument for the ${prvs expansion is
2824 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2827 (3) Reset the variables that are obtained from the first argument of
2828 ${prvscheck and used in the second argument before leaving the code,
2829 because their memory is reclaimed, so using them afterwards may do
2832 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2833 one (it's much easier than Tom thought :-).
2835 (5) Because of (4), we can now allow for the use of $prvscheck_result
2836 inside the third argument.
2838 PH/03 For some reason, the default setting of PATH when running a command from
2839 a pipe transport was just "/usr/bin". I have changed it to
2842 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2843 anything to be listed in the output from -bV.
2845 PH/05 When a filter generated an autoreply, the entire To: header line was
2846 quoted in the delivery log line, like this:
2848 => >A.N.Other <ano@some.domain> <original@ddress> ...
2850 This has been changed so that it extracts the operative address. There
2851 may be more than one such address. If so, they are comma-separated, like
2854 => >ano@some.domain,ona@other.domain <original@ddress> ...
2856 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2857 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2858 not being looked up in the rDNS to get a host name, Exim was showing the
2859 IP address twice in Received: lines, even though the IP addresses were
2860 identical. For example:
2862 Received: from [1.2.3.4] (helo=[1.2.3.4])
2864 However, if the real host name was known, it was omitting the HELO data
2865 if it matched the actual IP address. This has been tidied up so that it
2866 doesn't show the same IP address twice.
2868 PH/07 When both +timestamp and +memory debugging was on, the value given by
2869 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2870 called by the debug printing, thereby overwriting the timestamp buffer.
2871 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2873 PH/08 When the original message was included in an autoreply transport, it
2874 always said "this is a copy of the message, including all the headers",
2875 even if body_only or headers_only was set. It now gives an appropriate
2878 PH/09 Applied a patch from the Sieve maintainer which:
2880 o fixes some comments
2881 o adds the (disabled) notify extension core
2882 o adds some debug output for the result of if/elsif tests
2883 o points to the current vacation draft in the documentation
2884 and documents the missing references header update
2888 o fixes a bug in processing the envelope test (when testing
2889 multiple envelope elements, the last element determinted the
2892 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2893 Electronic Mail") by including:
2895 Auto-submitted: auto-generated
2897 in the messages that it generates (bounce messages and others, such as
2898 warnings). In the case of bounce messages for non-SMTP mesages, there was
2899 also a typo: it was using "Auto_submitted" (underscore instead of
2900 hyphen). Since every message generated by Exim is necessarily in response
2901 to another message, thes have all been changed to:
2903 Auto-Submitted: auto-replied
2905 in accordance with these statements in the RFC:
2907 The auto-replied keyword:
2909 - SHOULD be used on messages sent in direct response to another
2910 message by an automatic process,
2912 - MUST NOT be used on manually-generated messages,
2914 - MAY be used on Delivery Status Notifications (DSNs) and Message
2915 Disposition Notifications (MDNs),
2917 - MUST NOT be used on messages generated by automatic or periodic
2918 processes, except for messages which are automatic responses to
2921 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2922 to the default Received: header definition.
2924 PH/12 Added log selector acl_warn_skipped (default on).
2926 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2927 variables because (a) they are in the wrong storage pool and (b) even if
2928 they were copied, it wouldn't work properly because of the caching.
2930 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2931 checking when decoding. Apparently there are clients that generate
2932 overlong encoded strings. Why am I not surprised?
2934 PH/15 If the first argument of "${if match_address" was not empty, but did not
2935 contain an "@" character, Exim crashed. Now it writes a panic log message
2936 and treats the condition as false.
2938 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2940 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2941 extension "envelope-auth". The code is finished and in agreement with
2942 other implementations, but there is no documentation so far and in fact,
2943 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2944 not changing the active code.
2946 Print executed "if" and "elsif" statements when debugging is used. This
2947 helps a great deal to understand what a filter does.
2949 Document more things not specified clearly in RFC3028. I had all this
2950 sorted out, when out of a sudden new issues came to my mind. Oops."
2952 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2955 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2956 canonical form (as documented). However, after a host name lookup from
2957 the IP address, check_host() was doing a simple string comparison with
2958 addresses acquired from the DNS when checking that the found name did
2959 have the original IP as one of its addresses. Since any found IPv6
2960 addresses are likely to be in abbreviated form, the comparison could
2961 fail. Luckily, there already exists a function for doing the comparison
2962 by converting both addresses to binary, so now that is used instead of
2963 the text comparison.
2965 PH/20 There was another similar case to PH/19, when a complete host name was
2966 given in a host list; looking up its IP address could give an abbreviated
2967 form, whereas the current host's name might or might not be abbreviated.
2968 The same fix has been applied.
2974 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2975 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2978 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2979 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2981 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2983 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2984 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2985 Also, for "Auto-Submitted", treat anything other than "no" as
2986 non-personal, in accordance with RFC 3834. (Previously it treated
2987 anything starting "auto-" as non-personal.)
2989 TF/01 The control=submission/name=... option had a problem with syntax
2990 errors if the name included a slash character. The /name= option
2991 now slurps the rest of the string, so it can include any characters
2992 but it must come last in the list of options (after /sender_retain
2995 PH/05 Some modifications to the interface to the fake nameserver for the new
3003 TK/01 Added the "success_on_redirect" address verification option. See
3004 NewStuff for rationale and an example.
3006 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3008 PH/02 Patch to exigrep to allow it to work on syslog lines.
3010 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3011 fread() to skip over the body file's header line, because in Cygwin the
3012 header line is locked and is inaccessible.
3014 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3015 co-exist for some time) to make it clear that it is the Exim ID that is
3016 referenced, not the Message-ID: header line.
3018 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3019 string_format() function, because snprintf() does not exist on all
3022 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3023 ${stat: expansion item.
3025 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3026 protocol synchronization error", to keep the pedants happy.
3028 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3029 well as for IRIX systems, when gcc is being used. See the host.c source
3032 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3034 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3037 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3038 transport and to the smtp transport in LMTP mode.
3040 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3042 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3043 run a filter in a subprocess. This could lead to confusion in subsequent
3044 lookups in the parent process. There should also be a search_tidyup() at
3045 the end of the subprocess.
3047 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3048 only if the host matched helo_try_verify_hosts, which caused the
3049 verification to occur when the EHLO/HELO command was issued. The ACL just
3050 tested the remembered result. Now, if a previous verification attempt has
3051 not happened, "verify = helo" does it there and then.
3053 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3055 TK/03 Fix log output including CR from clamd.
3057 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3058 exist provoked a memory error which could cause a segfault.
3060 PH/15 Installed PCRE 6.2
3062 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3064 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3065 of the problem. Specifically, suggested +O2 rather than +O1 for the
3068 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3070 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3071 record, the port was not correctly logged when the outgoing_port log
3072 selector was set (it logged the transort's default port).
3074 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3075 fallback_hosts, and "hosts" in the smtp transport.
3077 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3078 host errors such as "Connection refused".
3080 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3081 authentication with radiusclient 0.4.9:
3083 - Error returned from rc_read_config was caught wrongly
3084 - Username/password not passed on to radius server due to wrong length.
3086 The presumption is that some radiusclient API changes for 4.51/PH/17
3087 were not taken care of correctly. The code is still untested by me (my
3088 Linux distribution still has 0.3.2 of radiusclient), but it was
3089 contributed by a Radius user.
3091 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3092 expanding the "port" option of the smtp transport.
3094 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3095 while reading a MIME header. Thanks to Tom Hughes for a patch.
3097 PH/24 Include config.h inside local_scan.h so that configuration settings are
3100 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3101 that in an ACL for RCPT (for example), you can examine exactly what was
3104 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3105 commands, but it was not correctly comparing the address with the actual
3106 client host address. Thus, it would show the EHLO address in Received:
3107 header lines when this was not necessary.
3109 PH/27 Added the % operator to ${eval:}.
3111 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3112 it should be ignoring failures (because with -C, for example, it has lost
3113 privilege). It wasn't ignoring creation failures other than "already
3116 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3119 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3120 because an input line was too long, either on its own, or by virtue of
3121 too many continuations, the temporary file was not being removed, and the
3122 return code was incorrect.
3124 PH/31 Missing "BOOL" in function definition in filtertest.c.
3126 PH/32 Applied Sieve patches from the maintainer.
3128 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3130 PH/33 Added "verify = not_blind".
3132 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3133 Local/Makefile (with some defaults set). These are used in built scripts
3134 such as exicyclog, but they have never been used in the exim_install
3135 script (though there are many overriding facilities there). I have
3136 arranged that the exim_install script now takes note of these two
3139 PH/35 Installed configuration files for Dragonfly.
3141 PH/36 When a locally submitted message by a trusted user did not contain a
3142 From: header, and the sender address was obtained from -f or from an SMTP
3143 MAIL command, and the trusted user did not use -F to supply a sender
3144 name, $originator_name was incorrectly used when constructing a From:
3145 header. Furthermore, $originator_name was used for submission mode
3146 messages from external hosts without From: headers in a similar way,
3147 which is clearly wrong.
3149 PH/37 Added control=suppress_local_fixups.
3151 PH/38 When log_selector = +received_sender was set, and the addition of the
3152 sender made the log line's construction buffer exactly full, or one byte
3153 less than full, an overflow happened when the terminating "\n" was
3156 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3157 when the result of a list match is failure because a DNS lookup failed.
3159 PH/40 RM_COMMAND is now used in the building process.
3161 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3162 the "build-* directories that it finds.
3164 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3165 address in a domain literal was a prefix of an interface address.
3167 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3168 when verifying a sender address, unless rewrite_headers is false.
3170 PH/44 Wrote a long comment about why errors_to addresses are verified as
3171 recipients, not senders.
3173 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3174 the ratelimit ACL was added.
3176 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3178 PH/46 Added extra information about PostgreSQL errors to the error string.
3180 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3181 avoiding the need to install special zones in a real server. This is
3182 backwards compatible; if it can't find the fake resolver, it drops back.
3183 Thus, both old and new test suites can be run.
3185 TF/02 Added util/ratelimit.pl
3187 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3188 clock is set back in time.
3190 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3191 Candler <B.Candler@pobox.com>.
3193 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3194 OK for addresses that are the result of redirections.
3196 PH/48 A number of further additions for the benefit of the new test suite,
3197 including a fake gethostbyname() that interfaces to the fake DNS resolver
3200 TF/06 The fix for widen_domains has also been applied to qualify_single and
3201 search_parents which are the other dnslookup options that can cause
3204 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3207 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3209 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3210 service" if it encountered an MX record with an empty target hostname.
3211 The message is now "an MX or SRV record indicated no SMTP service".
3213 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3214 if the DNS of the sending site is misconfigured. This is quite a
3215 common situation. This change restores the behaviour of treating a
3216 helo verification defer as a failure.
3218 PH/51 If self=fail was set on a router, the bounce message did not include the
3219 actual error message.
3225 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3227 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3228 command itself ended in error, the underlying message about the transport
3229 filter timeout was being overwritten with the pipe command error. Now the
3230 underlying error message should be appended to the second error message.
3232 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3234 PH/02 Reduce the amount of output that "make" produces by default. Full output
3235 can still be requested.
3237 PH/03 The warning log line about a condition test deferring for a "warn" verb
3238 was being output only once per connection, rather than after each
3239 occurrence (because it was using the same function as for successful
3240 "warn" verbs). This seems wrong, so I have changed it.
3242 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3243 it should not have, which might have caused a crash in the right
3244 circumstances, but probably never did.
3246 PH/04 Installed a modified version of Tony Finch's patch to make submission
3247 mode fix the return path as well as the Sender: header line, and to
3248 add a /name= option so that you can make the user's friendly name appear
3251 TF/03 Added the control = fakedefer ACL modifier.
3253 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3254 Mark Lowes for thorough testing.
3256 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3258 TK/03 Merged latest SRS patch from Miles Wilton.
3260 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3261 with the definition in sysexits.h (which is #included earlier).
3262 Fortunately, Exim does not actually use EX_OK. The code used to try to
3263 preserve the sysexits.h value, by assumimg that macro definitions were
3264 scanned for macro replacements. I have been disabused of this notion,
3265 so now the code just undefines EX_OK before #including unistd.h.
3267 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3268 in the smtp transport. When a block could not be written in a single
3269 write() function, the timeout was being re-applied to each part-write.
3270 This seems wrong - if the receiver was accepting one byte at a time it
3271 would take for ever. The timeout is now adjusted when this happens. It
3272 doesn't have to be particularly precise.
3274 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3275 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3277 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3278 without a domain if the check to <postmaster@domain> fails.
3280 SC/01 Eximstats: added -xls and the ability to specify output files
3281 (patch written by Frank Heydlauf).
3283 SC/02 Eximstats: use FileHandles for outputing results.
3285 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3287 SC/04 Eximstats: fixed display of large numbers with -nvr option
3289 SC/05 Eximstats: fixed merging of reports with empty tables.
3291 SC/06 Eximstats: added the -include_original_destination flag
3293 SC/07 Eximstats: removed tabs and trailing whitespace.
3295 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3297 TK/06 MBOX spool code: Add real "From " MBOX separator line
3298 so the .eml file is really in mbox format (even though
3299 most programs do not really care). Patch from Alex Miller.
3301 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3302 The latter is generated from $received_to and is only set if the
3303 message has one envelope recipient. SA can use these headers,
3304 obviously out-of-the-box. Patch from Alex Miller.
3306 PH/08 The ${def test on a variable was returning false if the variable's
3307 value was "0", contrary to what the specification has always said!
3308 The result should be true unless the variable is empty.
3310 PH/09 The syntax error of a character other than { following "${if
3311 def:variable_name" (after optional whitespace) was not being diagnosed.
3312 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3313 accidental colon was present, for example, could give incorrect results.
3315 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3316 result is used (not including appendfile, where other changes are about
3319 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3320 This involved changing a lot of size variables from int to off_t. It
3321 should work with maildirs and everything.
3323 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3324 spamd dying while we are connected to it.
3326 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3329 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3330 function for BDB 4.3.
3332 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3334 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3335 the sources for PCRE have been re-arranged and more files are now
3338 PH/15 The code I had for printing potentially long long variables in PH/11
3339 above was not the best (it lost precision). The length of off_t variables
3340 is now inspected at build time, and an appropriate printing format (%ld
3341 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3342 to be "long long int" or "long int". This is needed for the internal
3343 formatting function string_vformat().
3345 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3346 the configuration file to be ":syslog", then the script "guesses" where
3347 the logs files are, rather than using the compiled in default. In our
3348 case the guess is not the same as the compiled default, so the script
3349 suddenly stopped working when I started to use syslog. The patch checks
3350 to see if log_file_path is "". If so, it attempts to read it from exim
3351 with no configuration file to get the compiled in version, before it
3352 falls back to the previous guessing code."
3354 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3355 implementing BATV in an Exim configuration. See NewStuff for the gory
3358 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3359 Makefile that are specific to HP-UX.
3361 PH/18 If the "use_postmaster" option was set for a recipient callout together
3362 with the "random" option, the postmaster address was used as the MAIL
3363 FROM address for the random test, but not for the subsequent recipient
3364 test. It is now used for both.
3366 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3367 patch removes a few documentation additions to RFC 3028, because the
3368 latest draft now contains them. It adds the new en;ascii-case comparator
3369 and a new error check for 8bit text in MIME parts. Comparator and
3370 require names are now matched exactly. I enabled the subaddress
3371 extension, but it is not well tested yet (read: it works for me)."
3373 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3374 rework some of the code of TK/09 above to avoid the hardwired use of
3375 "%lld" and "long long". Replaced the call to snprintf() with a call to
3378 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3379 records point to non-existent hosts", "retry timeout exceeded", and
3380 "retry time not reached for any host after a long failure period".
3382 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3383 experimental DomainKeys support:
3385 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3386 (2) On an error such as an illegally used "control", the wrong name for
3387 the control was given.
3389 These problems did NOT occur unless DomainKeys support was compiled.
3391 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3393 PH/24 Added ${if match_ip condition.
3395 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3396 databases so that it will be absolutely obvious if a crash occurs in the
3397 DB library. This is a regular occurrence (often caused by mis-matched
3400 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3401 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3402 picked up on a user's system that detects such things. There doesn't seem
3403 to be a gcc warning option for this - only an attribute that has to be
3404 put on the function's prototype. It seems that in Fedora Core 4 they have
3405 set this on a number of new functions. No doubt there will be more in due
3408 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3409 specify a transport. However, if an address that was verified by such a
3410 router was the subject of a callout, Exim crashed because it tried to
3411 read the rcpt_include_affixes from the non-existent transport. Now it
3412 just assumes that the setting of that option is false. This bug was
3413 introduced by 4.51/PH/31.
3415 PH/28 Changed -d+all to exclude +memory, because that information is very
3416 rarely of interest, but it makes the output a lot bigger. People tend to
3417 do -d+all out of habit.
3419 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3420 code in os-type was giving problems when libc.so lives in lib64, like on
3423 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3424 aren't the modern standard, and it seems that some systems' include files
3425 don't always have them. Exim was already checking for some of the newer
3426 ones like T_AAAA, and defining it itself. I've added checks for all the
3427 record types that Exim uses.
3429 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3430 not automatically generating a new one, as it is supposed to. This
3431 prevented TLS from working. If the file did exist, but contained invalid
3432 data, a new version was generated, as expected. It was only the case of a
3433 non-existent file that was broken.
3435 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3436 with a change in libdomainkeys > 0.64.
3438 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3439 from DNS. If the selector record carries the flag, it now has
3440 precedence over the domain-wide flag.
3442 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3444 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3445 the use of a port name that isn't defined in /etc/services) occurred, the
3446 message was deferred as in a normal delivery, and thus remained on the
3447 spool, instead of being failed because of the mua_wrapper setting. This
3448 is now fixed, and I tidied up some of the mua_wrapper messages at the
3451 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3452 the messages in a hash of arrays rather than using individual hashes.
3453 This is a bit cleaner and results in dramatic memory savings, albeit
3454 at a slight CPU cost.
3456 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3457 as requested by Marc Sherman.
3459 SC/10 Eximstats: added histograms for user specified patterns as requested
3462 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3464 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3465 fopen() in the content-scanning modules that did not already have it.
3471 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3472 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3474 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3476 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3478 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3479 to test Sieve filters that use "vacation".
3481 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3482 that changes the way the GnuTLS parameters are stored in the cache file.
3483 The new format can be generated externally. For backward compatibility,
3484 if the data in the cache doesn't make sense, Exim assumes it has read an
3485 old-format file, and it generates new data and writes a new file. This
3486 means that you can't go back to an older release without removing the
3489 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3490 work if there are any delivery delays because "one_time" forces the
3491 parent to be marked "delivered", so its unseen clone is never tried
3492 again. For this reason, Exim now forbids the simultaneous setting of
3495 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3496 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3497 entry if you want to know the details. The fix, however, neglected to
3498 consider the case when local delivery batching is involved. The test for
3499 "previously delivered" was not happening when checking to see if an
3500 address could be batched with a previous (undelivered) one; under
3501 certain circumstances this could lead to multiple deliveries to the same
3504 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3505 in its include files, and this causes problems building Exim.
3507 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3508 header_syntax) but Exim was just ignoring anything given after a slash.
3509 In particular, this caused confusion with an attempt to use "verify =
3510 reverse_host_lookup/defer_ok". An error is now given when options are
3511 supplied for verify items that do not have them. (Maybe reverse_host_
3512 lookup should have a defer_ok option, but that's a different point.)
3514 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3515 defined by RFC 821) to 2048, because there were problems with some AUTH
3516 commands, and RFC 1869 says the size should be increased for extended
3517 SMTP commands that take arguments.
3519 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3522 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3523 "unknown" error; now it says that the functionality isn't in the binary.
3525 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3526 an address' error message when a string expansion fails (syntax or
3527 whatever). Otherwise the password may appear in the log. Following change
3528 PH/42 below, there is no longer a chance of it appearing in a bounce
3531 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3533 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3534 of its addresses was ever tried. (Bugzilla bug #2.)
3536 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3537 the result incorrectly in the debug output. (It correctly added a newline
3538 to what was transported.)
3540 TF/01 Added $received_time.
3542 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3543 commented out examples of how to interface to a virus scanner and to
3544 SpamAssassin. Also added commented examples of av_scanner and
3545 spamd_address settings.
3547 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3548 and controls are allowed in which ACLs. There were a couple of minor
3549 errors. Some of the entries in the conditions table (which is a table of
3550 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3551 negation of where the condition IS allowed.
3553 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3555 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3556 header file does not have a version number, so I've had to invent a new
3557 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3558 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3559 radiusclient), but it was contributed by a Radius user.
3561 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3562 files or directories, for OpenSSL.
3564 PH/19 When an Exim process that is running as root has to create an Exim log
3565 file, it does so in a subprocess that runs as exim:exim so as to get the
3566 ownership right at creation (otherwise, other Exim processes might see
3567 the file with the wrong ownership). There was no test for failure of this
3568 fork() call, which would lead to the process getting stuck as it waited
3569 for a non-existent subprocess. Forks do occasionally fail when resources
3570 run out. I reviewed all the other calls to fork(); they all seem to check
3573 PH/20 When checking for unexpected SMTP input at connect time (before writing
3574 the banner), Exim was not dealing correctly with a non-positive return
3575 from the read() function. If the client had disconnected by this time,
3576 the result was a log entry for a synchronization error with an empty
3577 string after "input=" when read() returned zero. If read() returned -1
3578 (an event I could not check), uninitialized data bytes were printed.
3579 There were reports of junk text (parts of files, etc) appearing after
3582 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3584 PH/22 Added support for macro redefinition, and (re)definition in between
3585 driver and ACL definitions.
3587 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3588 forgetting to use the resulting value; it was using the unexpanded value.
3590 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3591 hadn't been configured. The fix is from Juergen Kreileder, who
3592 understands it better than I do:
3594 "Here's what I see happening with three configured cyrus_sasl
3595 authenticators configured (plain, login, cram-md5):
3597 On startup auth_cyrus_sasl_init() gets called for each of these.
3598 This means three calls to sasl_listmech() without a specified mech_list.
3599 => SASL tests which mechs of all available mechs actually work
3600 => three warnings about OTP not working
3601 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3603 With the patch, sasl_listmech() also gets called three times. But now
3604 SASL's mech_list option is set to the server_mech specified in the the
3605 authenticator. Or in other words, the answer from sasl_listmech()
3606 gets limited to just the mech you're testing for (which is different
3608 => the return list contains just 'plain' or 'login', 'cram-md5' or
3609 nothing depending on the value of ob->server_mech.
3611 I've just tested the patch: Authentication still works fine,
3612 unavailable mechs specified in the exim configuration are still
3613 caught, and the auth.log warnings about OTP are gone."
3615 PH/25 When debugging is enabled, the contents of the command line are added
3616 to the debugging output, even when log_selector=+arguments is not
3619 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3620 answer is "GNU", and only if the return is "GNU/something" is the answer
3623 PH/27 $acl_verify_message is now set immediately after the failure of a
3624 verification in an ACL, and so is available in subsequent modifiers. In
3625 particular, the message can be preserved by coding like this:
3627 warn !verify = sender
3628 set acl_m0 = $acl_verify_message
3630 Previously, $acl_verify_message was set only while expanding "message"
3631 and "log_message" when a very denied access.
3633 PH/28 Modified OS/os.c-Linux with
3635 -#ifndef OS_LOAD_AVERAGE
3636 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3638 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3639 nomenclature these days.)
3641 PH/29 Installed patch from the Sieve maintainer that adds the options
3642 sieve_useraddress and sieve_subaddress to the redirect router.
3644 PH/30 In these circumstances:
3645 . Two addresses routed to the same list of hosts;
3646 . First host does not offer TLS;
3647 . First host accepts first address;
3648 . First host gives temporary error to second address;
3649 . Second host offers TLS and a TLS session is established;
3650 . Second host accepts second address.
3651 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3652 and peerdn, if requested) that were in fact used only for the second
3655 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3656 attention to any local part prefix or suffix that was matched by the
3657 router that accepted the address. It now behaves in the same way as it
3658 does for delivery: the affixes are removed from the local part unless
3659 rcpt_include_affixes is set on the transport.
3661 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3662 timeout during the DATA phase of an incoming message.
3664 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3665 applied a patch sanctioned by the Sieve maintainer.
3667 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3668 the uid or gid is negative. A case of a negative gid caused this to be
3669 noticed. The fix allows for either to be negative.
3671 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3672 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3675 PH/36 The change PH/12 above was broken. Fixed it.
3677 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3678 the grounds that routing the same address twice would always produce the
3679 same answer. This might have been true once, but it is certainly no
3680 longer true now. Routing a child address may depend on the previous
3681 routing that produced that child. Some complicated redirection strategies
3682 went wrong when messages had multiple recipients, and made Exim's
3683 behaviour dependent on the order in which the addresses were given.
3685 I have moved the duplicate checking until after the routing is complete.
3686 Exim scans the addresses that are assigned to local and remote
3687 transports, and removes any duplicates. This means that more work will be
3688 done, as duplicates will always all be routed, but duplicates are
3689 presumably rare, so I don't expect this is of any significance.
3691 For deliveries to pipes, files, and autoreplies, the duplicate checking
3692 still happens during the routing process, since they are not going to be
3695 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3696 It corrects a timeout issue with spamd. This is Ian's comment: "The
3697 background is that sometimes spamd either never reads data from a
3698 connection it has accepted, or it never writes response data. The exiscan
3699 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3700 blindly assumes that writes won't block so it may never time out."
3702 PH/39 Allow G after quota size as well as K and M.
3704 PH/40 The value set for $authenticated_id in an authenticator may not contain
3705 binary zeroes or newlines because the value is written to log lines and
3706 to spool files. There was no check on this. Now the value is run through
3707 the string_printing() function so that such characters are converted to
3708 printable escape sequences.
3710 PH/41 $message_linecount is a new variable that contains the total number of
3711 lines in the message. Compare $body_linecount, which is the count for the
3714 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3715 bounce and delay warning messages, except in certain special cases, which
3718 (a) An SMTP error message from a remote host;
3719 (b) A message specified in a :fail: redirection;
3720 (c) A message specified in a "fail" command in a system filter;
3721 (d) A message specified in a FAIL return from the queryprogram router;
3722 (e) A message specified by the cannot_route_message router option.
3724 In these cases only, Exim does include the error details in bounce and
3725 warning messages. There are also a few cases where bland messages such
3726 as "unrouteable address" or "local delivery error" are given.
3728 PH/43 $value is now also set for the "else" part of a ${run expansion.
3730 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3731 being worked on, but at least Exim now implements the latest version to
3734 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3735 process to complete was treated as a delivery failure, a timeout while
3736 writing the message to the pipe was logged, but erroneously treated as a
3737 successful delivery. Such timeouts include transport filter timeouts. For
3738 consistency with the overall process timeout, these timeouts are now
3739 treated as errors, giving rise to delivery failures by default. However,
3740 there is now a new Boolean option for the pipe transport called
3741 timeout_defer, which, if set TRUE, converts the failures into defers for
3742 both kinds of timeout. A transport filter timeout is now identified in
3745 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3746 systems where "make" and "gmake" are different, calling "gmake" at top
3747 level broke things. I've arranged for the value of $(MAKE) to be passed
3748 from the Makefile to this script so that it can call the same version of
3752 A note about Exim versions 4.44 and 4.50
3753 ----------------------------------------
3755 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3756 changes of various kinds. As a consequence, a big documentation update was
3757 needed. This delayed the release for rather longer than seemed good, especially
3758 in the light of a couple of (minor) security issues. Therefore, the changes
3759 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3760 release. So 4.44 and 4.50 are in effect two different branches that both start
3763 I have left the 4.50 change log unchanged; it contains all the changes since
3764 4.43. The change log for 4.44 is below; many of its items are identical to
3765 those for 4.50. This seems to be the most sensible way to preserve the
3766 historical information.
3772 1. Minor wording change to the doc/README.SIEVE file.
3774 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3775 computation of the current number of files was incorrect.
3777 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3778 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3781 4. Give more explanation in the error message when the command for a transport
3782 filter fails to execute.
3784 5. There are several places where Exim runs a non-Exim command in a
3785 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3786 was being done only for the command run by the queryprogram router. It is
3787 now done for all such subprocesses. The other cases are: ${run, transport
3788 filters, and the commands run by the lmtp and pipe transports.
3790 6. Added CONFIGURE_GROUP build-time option.
3792 7. Some older OS have a limit of 256 on the maximum number of file
3793 descriptors. Exim was using setrlimit() to set 1000 as a large value
3794 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3795 systems. I've change it so that if it can't get 1000, it tries for 256.
3797 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3798 was an oversight, and furthermore, ever since the addition of extra
3799 controls (e.g. 4.43/32), the checks on when to allow different forms of
3800 "control" were broken. There should now be diagnostics for all cases when a
3801 control that does not make sense is encountered.
3803 9. Added the /retain_sender option to "control=submission".
3805 10. $recipients is now available in the predata ACL (oversight).
3807 11. Tidy the search cache before the fork to do a delivery from a message
3808 received from the command line. Otherwise the child will trigger a lookup
3809 failure and thereby defer the delivery if it tries to use (for example) a
3810 cached ldap connection that the parent has called unbind on.
3812 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3813 of $address_data from the recipient verification was clobbered by the
3814 sender verification.
3816 13. The value of address_data from a sender verification is now available in
3817 $sender_address_data in subsequent conditions in the ACL statement.
3819 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3821 15. Added a new option "connect=<time>" to callout options, to set a different
3824 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3825 was its contents. (It was OK if the option was not defined at all.)
3827 17. A "Completed" log line is now written for messages that are removed from
3828 the spool by the -Mrm option.
3830 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3831 information about exactly what failed.
3833 19. Added -dd to debug only the daemon process.
3835 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3836 handles timeouts, both on the server side and network timeouts. Renamed the
3837 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3839 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3840 It is now set to "smtps".
3842 22. $host_address is now set to the target address during the checking of
3843 ignore_target_hosts.
3845 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3846 being passed; this would have caused $sender_host_name to have been used if
3847 matching the list had actually called for a host name (not very likely,
3848 since this list is usually IP addresses). A host name is now passed as
3851 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3852 code that specifies a non-restarting handler (typically sigaction() in
3853 modern systems) in an attempt to fix a rare and obscure crash bug.
3855 25. Narrowed the window for a race in the daemon that could cause it to ignore
3856 SIGCHLD signals. This is not a major problem, because they are used only to
3857 wake it up if nothing else does.
3859 26. A malformed maildirsize file could cause Exim to calculate negative values
3860 for the mailbox size or file count. Odd effects could occur as a result.
3861 The maildirsize information is now recalculated if the size or filecount
3864 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3865 support for a long time. Removed HAVE_SYS_VFS_H.
3867 28. Installed the latest version of exipick from John Jetmore.
3869 29. In an address list, if the pattern was not a regular expression, an empty
3870 subject address (from a bounce message) matched only if the pattern was an
3871 empty string. Non-empty patterns were not even tested. This was the wrong
3872 because it is perfectly reasonable to use an empty address as part of a
3873 database query. An empty address is now tested by patterns that are
3874 lookups. However, all the other forms of pattern expect the subject to
3875 contain a local part and a domain, and therefore, for them, an empty
3876 address still always fails if the pattern is not itself empty.
3878 30. Exim went into a mad DNS loop when attempting to do a callout where the
3879 host was specified on an smtp transport, and looking it up yielded more
3880 than one IP address.
3882 31. Re-factored the code for checking spool and log partition space into a
3883 function that finds that data and another that does the check. The former
3884 is then used to implement four new variables: $spool_space, $log_space,
3885 $spool_inodes, and $log_inodes.
3887 32. The RFC2047 encoding function was originally intended for short strings
3888 such as real names; it was not keeping to the 75-character limit for
3889 encoded words that the RFC imposes. It now respects the limit, and
3890 generates multiple encoded words if necessary. To be on the safe side, I
3891 have increased the buffer size for the ${rfc2047: expansion operator from
3894 33. It is now permitted to omit both strings after an "if" condition; if the
3895 condition is true, the result is "true". As before, when the second string
3896 is omitted, a false condition yields an empty string. This makes it less
3897 cumbersome to write custom ACL and router conditions.
3899 34. Failure to deliver a bounce message always caused it to be frozen, even if
3900 there was an errors_to setting on the router. The errors_to setting is now
3903 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3904 canonical form (fully expanded) before being placed in
3905 $sender_host_address.
3907 36. The table in the code that translates DNS record types into text (T_A to
3908 "A" for instance) was missing entries for NS and CNAME. It is just possible
3909 that this could have caused confusion if both these types were looked up
3910 for the same domain, because the text type is used as part of Exim's
3911 per-process caching. But the chance of anyone hitting this buglet seems
3914 37. The dnsdb lookup has been extended in a number of ways.
3916 (1) There is a new type, "zns", which walks up the domain tree until it
3917 finds some nameserver records. It should be used with care.
3919 (2) There is a new type, "mxh", which is like "mx" except that it returns
3920 just the host names, not the priorities.
3922 (3) It is now possible to give a list of domains (or IP addresses) to be
3923 looked up. The behaviour when one of the lookups defers can be
3924 controlled by a keyword.
3926 (4) It is now possible to specify the separator character for use when
3927 multiple records are returned.
3929 38. The dnslists ACL condition has been extended: it is now possible to supply
3930 a list of IP addresses and/or domains to be looked up in a particular DNS
3933 39. Added log_selector=+queue_time_overall.
3935 40. When running the queue in the test harness, wait just a tad after forking a
3936 delivery process, to get repeatability of debugging output.
3938 41. Include certificate and key file names in error message when GnuTLS fails
3939 to set them up, because the GnuTLS error message doesn't include the name
3940 of the failing file when there is a problem reading it.
3942 42. Allow both -bf and -bF in the same test run.
3944 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3946 44. The "Exiscan patch" is now merged into the mainline Exim source.
3948 45. Sometimes the final signoff response after QUIT could fail to get
3949 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3950 < 0 before doing a fflush(). This bug looks as though it goes back to the
3951 introduction of TLS in release 3.20, but "sometimes" must have been rare
3952 because the tests only now provoked it.
3954 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3955 (this can affect the format of dates).
3957 47. exim_tidydb, when checking for the continued existence of a message for
3958 which it has found a message-specific retry record, was not finding
3959 messages that were in split spool directories. Consequently, it was
3960 deleting retry records that should have stayed in existence.
3962 48. Steve fixed some bugs in eximstats.
3964 49. The SPA authentication driver was not abandoning authentication and moving
3965 on to the next authenticator when an expansion was forced to fail,
3966 contradicting the general specification for all authenticators. Instead it
3967 was generating a temporary error. It now behaves as specified.
3969 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3970 (the order specifies the preference for clients). The order is now AES256,
3971 AES128, 3DES, ARCFOUR128.
3973 51. Small patch to Sieve code - explicitly set From: when generating an
3976 52. Exim crashed if a remote delivery caused a very long error message to be
3977 recorded - for instance if somebody sent an entire SpamAssassin report back
3978 as a large number of 550 error lines. This bug was coincidentally fixed by
3979 increasing the size of one of Exim's internal buffers (big_buffer) that
3980 happened as part of the Exiscan merge. However, to be on the safe side, I
3981 have made the code more robust (and fixed the comments that describe what
3984 53. Now that there can be additional text after "Completed" in log lines (if
3985 the queue_time_overall log selector is set), a one-byte patch to exigrep
3986 was needed to allow it to recognize "Completed" as not the last thing in
3989 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3990 patch that reportedly fixes this has been added. I am not expert enough to
3991 create a test for it. This is what the patch creator wrote:
3993 "I found a little strange behaviour of ldap code when working with
3994 Windows 2003 AD Domain, where users was placed in more than one
3995 Organization Units. When I tried to give exim partial DN, the exit code
3996 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3997 But simultaneously result of request was absolutely normal ldap result,
3998 so I produce this patch..."
4000 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4001 REFERENCE, so I have modified the code to exclude the patch when that macro
4004 55. Some experimental protocols are using DNS PTR records for new purposes. The
4005 keys for these records are domain names, not reversed IP addresses. The
4006 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4007 leaves it alone. Component reversal etc. now happens only for IP addresses.
4010 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4012 57. Double the size of the debug message buffer (to 2048) so that more of very
4013 long debug lines gets shown.
4015 58. The exicyclog utility now does better if the number of log files to keep
4016 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4018 59. Two changes related to the smtp_active_hostname option:
4020 (1) $smtp_active_hostname is now available as a variable.
4021 (2) The default for smtp_banner uses $smtp_active_hostname instead
4022 of $primary_hostname.
4024 60. The host_aton() function is supposed to be passed a string that is known
4025 to be a valid IP address. However, in the case of IPv6 addresses, it was
4026 not checking this. This is a hostage to fortune. Exim now panics and dies
4027 if the condition is not met. A case was found where this could be provoked
4028 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4029 components; fortuitously, this particular loophole had already been fixed
4030 by change 4.50/55 above.
4032 If there are any other similar loopholes, the new check in host_aton()
4033 itself should stop them being exploited. The report I received stated that
4034 data on the command line could provoke the exploit when Exim was running as
4035 exim, but did not say which command line option was involved. All I could
4036 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4037 running as the user.
4040 61. There was a buffer overflow vulnerability in the SPA authentication code
4041 (which came originally from the Samba project). I have added a test to the
4042 spa_base64_to_bits() function which I hope fixes it.
4045 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4046 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4048 63. The daemon start-up calls getloadavg() while still root for those OS that
4049 need the first call to be done as root, but it missed one case: when
4050 deliver_queue_load_max is set with deliver_drop_privilege. This is
4051 necessary for the benefit of the queue runner, because there is no re-exec
4052 when deliver_drop_privilege is set.
4054 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4055 This has been fixed.
4057 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4058 was in use, was not putting the data itself into the right store pool;
4059 consequently, it could be overwritten for a subsequent message in the same
4060 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4063 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4065 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4066 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4067 the function were treating the return as a boolean value, which happened to
4068 work because 0=false and not-0=true, but is not correct code.
4070 68. The host_aton() function was not handling scoped IPv6 addresses (those
4071 with, for example, "%eth0" on the end) correctly.
4073 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4074 negated items (that is, ~something) in unsigned ints. Some compilers
4075 apparently mutter when there is no cast.
4077 70. If an address verification called from an ACL failed, and did not produce a
4078 user-specific message (i.e. there was only a "system" message), nothing was
4079 put in $acl_verify_message. In this situation, it now puts the system
4082 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4083 session; change 4.31/43 added the unwanted input to the log line - except
4084 that it did not do this in the start of session case. It now does.
4086 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4087 This is wrong and can cause the other end to generate a synchronization
4088 error if it is another Exim or anything else that does the synchronization
4089 check. A QUIT command is no longer sent after a timeout.
4091 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4092 during host lookups.
4094 74. The defer_ok option of callout verification was not working if it was used
4095 when verifying addresses in header lines, that is, for this case:
4097 verify = header_sender/callout=defer_ok
4099 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4100 those file descriptors could be used for SMTP connections. If anything
4101 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4102 could be sent to the SMTP client, causing chaos. The daemon now opens
4103 stdin, stdout, and stderr to /dev/null when it puts itself into the
4106 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4107 log by default. The user can override this with suitable Perl magic.
4109 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4110 the log message when discard triggers, was not working for the DATA ACL or
4111 for the non-SMTP ACL.
4113 78. Error message wording change in sieve.c.
4115 79. If smtp_accept_max_per_host was set, the number of connections could be
4116 restricted to fewer than expected, because the daemon was trying to set up
4117 a new connection before checking whether the processes handling previous
4118 connections had finished. The check for completed processes is now done
4119 earlier. On busy systems, this bug wouldn't be noticed because something
4120 else would have woken the daemon, and it would have reaped the completed
4123 80. If a message was submitted locally by a user whose login name contained one
4124 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4125 It caused a spool format error. I have fixed the spool reading code. A
4126 related problem was that the "from" clause in the Received: line became
4127 illegal because of the space(s). It is now covered by ${quote_local_part.
4129 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4132 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4134 83. Patch from Sieve maintainer to fix unterminated string problem in
4135 "vacation" handling.
4137 84. Some minor changes to the Linux configuration files to help with other
4138 OS variants using glibc.
4140 85. One more patch for Sieve to update vacation handling to latest spec.
4143 ----------------------------------------------------
4144 See the note above about the 4.44 and 4.50 releases.
4145 ----------------------------------------------------
4151 1. Change 4.43/35 introduced a bug that caused file counts to be
4152 incorrectly computed when quota_filecount was set in an appendfile
4155 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4156 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4159 3. Give more explanation in the error message when the command for a transport
4160 filter fails to execute.
4162 4. There are several places where Exim runs a non-Exim command in a
4163 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4164 was being done only for the command run by the queryprogram router. It is
4165 now done for all such subprocesses. The other cases are: ${run, transport
4166 filters, and the commands run by the lmtp and pipe transports.
4168 5. Some older OS have a limit of 256 on the maximum number of file
4169 descriptors. Exim was using setrlimit() to set 1000 as a large value
4170 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4171 systems. I've change it so that if it can't get 1000, it tries for 256.
4173 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4174 was an oversight, and furthermore, ever since the addition of extra
4175 controls (e.g. 4.43/32), the checks on when to allow different forms of
4176 "control" were broken. There should now be diagnostics for all cases when a
4177 control that does not make sense is encountered.
4179 7. $recipients is now available in the predata ACL (oversight).
4181 8. Tidy the search cache before the fork to do a delivery from a message
4182 received from the command line. Otherwise the child will trigger a lookup
4183 failure and thereby defer the delivery if it tries to use (for example) a
4184 cached ldap connection that the parent has called unbind on.
4186 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4187 of $address_data from the recipient verification was clobbered by the
4188 sender verification.
4190 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4191 was its contents. (It was OK if the option was not defined at all.)
4193 11. A "Completed" log line is now written for messages that are removed from
4194 the spool by the -Mrm option.
4196 12. $host_address is now set to the target address during the checking of
4197 ignore_target_hosts.
4199 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4200 being passed; this would have caused $sender_host_name to have been used if
4201 matching the list had actually called for a host name (not very likely,
4202 since this list is usually IP addresses). A host name is now passed as
4205 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4206 code that specifies a non-restarting handler (typically sigaction() in
4207 modern systems) in an attempt to fix a rare and obscure crash bug.
4209 15. Narrowed the window for a race in the daemon that could cause it to ignore
4210 SIGCHLD signals. This is not a major problem, because they are used only to
4211 wake it up if nothing else does.
4213 16. A malformed maildirsize file could cause Exim to calculate negative values
4214 for the mailbox size or file count. Odd effects could occur as a result.
4215 The maildirsize information is now recalculated if the size or filecount
4218 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4219 support for a long time. Removed HAVE_SYS_VFS_H.
4221 18. Updated exipick to current release from John Jetmore.
4223 19. Allow an empty sender to be matched against a lookup in an address list.
4224 Previously the only cases considered were a regular expression, or an
4227 20. Exim went into a mad DNS lookup loop when doing a callout where the
4228 host was specified on the transport, if the DNS lookup yielded more than
4231 21. The RFC2047 encoding function was originally intended for short strings
4232 such as real names; it was not keeping to the 75-character limit for
4233 encoded words that the RFC imposes. It now respects the limit, and
4234 generates multiple encoded words if necessary. To be on the safe side, I
4235 have increased the buffer size for the ${rfc2047: expansion operator from
4238 22. Failure to deliver a bounce message always caused it to be frozen, even if
4239 there was an errors_to setting on the router. The errors_to setting is now
4242 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4243 canonical form (fully expanded) before being placed in
4244 $sender_host_address.
4246 24. Updated eximstats to version 1.33
4248 25. Include certificate and key file names in error message when GnuTLS fails
4249 to set them up, because the GnuTLS error message doesn't include the name
4250 of the failing file when there is a problem reading it.
4252 26. Expand error message when OpenSSL has problems setting up cert/key files.
4255 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4256 (this can affect the format of dates).
4258 28. exim_tidydb, when checking for the continued existence of a message for
4259 which it has found a message-specific retry record, was not finding
4260 messages that were in split spool directories. Consequently, it was
4261 deleting retry records that should have stayed in existence.
4263 29. eximstats updated to version 1.35
4264 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4265 1.35 - bugfix such that pie charts by volume are generated correctly
4267 30. The SPA authentication driver was not abandoning authentication and moving
4268 on to the next authenticator when an expansion was forced to fail,
4269 contradicting the general specification for all authenticators. Instead it
4270 was generating a temporary error. It now behaves as specified.
4272 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4273 (the order specifies the preference for clients). The order is now AES256,
4274 AES128, 3DES, ARCFOUR128.
4276 31. Small patch to Sieve code - explicitly set From: when generating an
4279 32. Exim crashed if a remote delivery caused a very long error message to be
4280 recorded - for instance if somebody sent an entire SpamAssassin report back
4281 as a large number of 550 error lines. This bug was coincidentally fixed by
4282 increasing the size of one of Exim's internal buffers (big_buffer) that
4283 happened as part of the Exiscan merge. However, to be on the safe side, I
4284 have made the code more robust (and fixed the comments that describe what
4287 33. Some experimental protocols are using DNS PTR records for new purposes. The
4288 keys for these records are domain names, not reversed IP addresses. The
4289 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4290 leaves it alone. Component reversal etc. now happens only for IP addresses.
4293 34. The host_aton() function is supposed to be passed a string that is known
4294 to be a valid IP address. However, in the case of IPv6 addresses, it was
4295 not checking this. This is a hostage to fortune. Exim now panics and dies
4296 if the condition is not met. A case was found where this could be provoked
4297 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4298 components; fortuitously, this particular loophole had already been fixed
4299 by change 4.50/55 or 4.44/33 above.
4301 If there are any other similar loopholes, the new check in host_aton()
4302 itself should stop them being exploited. The report I received stated that
4303 data on the command line could provoke the exploit when Exim was running as
4304 exim, but did not say which command line option was involved. All I could
4305 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4306 running as the user.
4309 35. There was a buffer overflow vulnerability in the SPA authentication code
4310 (which came originally from the Samba project). I have added a test to the
4311 spa_base64_to_bits() function which I hope fixes it.
4314 36. The daemon start-up calls getloadavg() while still root for those OS that
4315 need the first call to be done as root, but it missed one case: when
4316 deliver_queue_load_max is set with deliver_drop_privilege. This is
4317 necessary for the benefit of the queue runner, because there is no re-exec
4318 when deliver_drop_privilege is set.
4320 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4321 was in use, was not putting the data itself into the right store pool;
4322 consequently, it could be overwritten for a subsequent message in the same
4323 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4326 38. Sometimes the final signoff response after QUIT could fail to get
4327 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4328 < 0 before doing a fflush(). This bug looks as though it goes back to the
4329 introduction of TLS in release 3.20, but "sometimes" must have been rare
4330 because the tests only now provoked it.
4336 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4337 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4338 Now it returns an integer. A number of calls were still expecting a T/F
4339 return. Fortuitously, in all cases, the tests worked in OK situations,
4340 which is the norm. However, things would have gone wrong on any write
4341 failures on the smtp file descriptor. This function is used when sending
4342 messages over SMTP and also when doing verify callouts.
4344 2. When Exim is called to do synchronous delivery of a locally submitted
4345 message (the -odf or -odi options), it no longer closes stderr before doing
4348 3. Implemented the mua_wrapper option.
4350 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4352 5. Implemented the functions header_remove(), header_testname(),
4353 header_add_at_position(), and receive_remove_recipient(), and exported them
4356 6. If an ACL "warn" statement specified the addition of headers, Exim already
4357 inserted X-ACL-Warn: at the start if there was no header name. However, it
4358 was not making this test for the second and subsequent header lines if
4359 there were newlines in the string. This meant that an invalid header could
4360 be inserted if Exim was badly configured.
4362 7. Allow an ACL "warn" statement to add header lines at the start or after all
4363 the Received: headers, as well as at the end.
4365 8. Added the rcpt_4xx retry error code.
4367 9. Added postmaster_mailfrom=xxx to callout verification option.
4369 10. Added mailfrom=xxxx to the callout verification option, for verify=
4372 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4373 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4375 12. Inserted some casts to stop certain compilers warning when using pointer
4376 differences as field lengths or precisions in printf-type calls (mostly
4377 affecting debugging statements).
4379 13. Added optional readline() support for -be (dynamically loaded).
4381 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4382 same clock tick as a message's arrival, so that its received time was the
4383 same as the "first fail" time on the retry record, and that message
4384 remained on the queue past the ultimate address timeout, every queue runner
4385 would try a delivery (because it was past the ultimate address timeout) but
4386 after another failure, the ultimate address timeout, which should have then
4387 bounced the address, did not kick in. This was a "< instead of <=" error;
4388 in most cases the first failure would have been in the next clock tick
4389 after the received time, and all would be well.
4391 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4392 being recognized when the domain list was tested by the match_domain
4393 condition in an expansion string.
4395 16. Added the ${str2b64: operator.
4397 17. Exim was always calling setrlimit() to set a large limit for the number of
4398 processes, without checking whether the existing limit was already
4399 adequate. (It did check for the limit on file descriptors.) Furthermore,
4400 errors from getrlimit() and setrlimit() were being ignored. Now they are
4401 logged to the main and panic logs, but Exim does carry on, to try to do its
4402 job under whatever limits there are.
4404 18. Imported PCRE 5.0.
4406 19. Trivial typo in log message " temporarily refused connection" (the leading
4409 20. If the log selector return_path_on_delivery was set and an address was
4410 redirected to /dev/null, the delivery process crashed because it assumed
4411 that a return path would always be set for a "successful" delivery. In this
4412 case, the whole delivery is bypassed as an optimization, and therefore no
4415 21. Internal re-arrangement: the function for sending a challenge and reading
4416 a response while authentication was assuming a zero-terminated challenge
4417 string. It's now changed to take a pointer and a length, to allow for
4418 binary data in such strings.
4420 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4422 23. Exim was not respecting finduser_retries when seeking the login of the
4423 uid under which it was called; it was always trying 10 times. (The default
4424 setting of finduser_retries is zero.) Also, it was sleeping after the final
4425 failure, which is pointless.
4427 24. Implemented tls_on_connect_ports.
4429 25. Implemented acl_smtp_predata.
4431 26. If the domain in control=submission is set empty, Exim assumes that the
4432 authenticated id is a complete email address when it generates From: or
4433 Sender: header lines.
4435 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4436 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4437 chown and chgrp in /bin and hostname in /usr/bin.
4439 28. Exim was keeping the "process log" file open after each use, just as it
4440 does for the main log. This opens the possibility of it remaining open for
4441 long periods when the USR1 signal hits a daemon. Occasional processlog
4442 errors were reported, that could have been caused by this. Anyway, it seems
4443 much more sensible not to leave this file open at all, so that is what now
4446 29. The long-running daemon process does not normally write to the log once it
4447 has entered its main loop, and it closes the log before doing so. This is
4448 so that log files can straightforwardly be renamed and moved. However,
4449 there are a couple of unusual error situations where the daemon does write
4450 log entries, and I had neglected to close the log afterwards.
4452 30. The text of an SMTP error response that was received during a remote
4453 delivery was being truncated at 512 bytes. This is too short for some of
4454 the long messages that one sometimes sees. I've increased the limit to
4457 31. It is now possible to make retry rules that apply only when a message has a
4458 specific sender, in particular, an empty sender.
4460 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4461 it possible to be selective about when SMTP synchronization is enforced.
4463 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4465 32. Implemented hosts_connection_nolog.
4467 33. Added an ACL for QUIT.
4469 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4472 35. Added mailbox_size and mailbox_filecount to appendfile.
4474 36. Added control = no_multiline_responses to ACLs.
4476 37. There was a bug in the logic of the code that waits for the clock to tick
4477 in the case where the clock went backwards by a substantial amount such
4478 that the microsecond fraction of "now" was more than the microsecond
4479 fraction of "then" (but the whole seconds number was less).
4481 38. Added support for the libradius Radius client library this is found on
4482 FreeBSD (previously only the radiusclient library was supported).
4488 1. When certain lookups returned multiple values in the form name=value, the
4489 quoting of the values was not always being done properly. Specifically:
4490 (a) If the value started with a double quote, but contained no whitespace,
4492 (b) If the value contained whitespace other than a space character (i.e.
4493 tabs or newlines or carriage returns) it was not quoted.
4494 This fix has been applied to the mysql and pgsql lookups by writing a
4495 separate quoting function and calling it from the lookup code. The fix
4496 should probably also be applied to nisplus, ibase and oracle lookups, but
4497 since I cannot test any of those, I have not disturbed their existing code.
4499 2. A hit in the callout cache for a specific address caused a log line with no
4500 reason for rejecting RCPT. Now it says "Previous (cached) callout
4501 verification failure".
4503 3. There was an off-by-one bug in the queryprogram router. An over-long
4504 return line was truncated at 256 instead of 255 characters, thereby
4505 overflowing its buffer with the terminating zero. As well as fixing this, I
4506 have increased the buffer size to 1024 (and made a note to document this).
4508 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4509 when Exim is waiting for an SMTP response from a remote server, Exim
4510 restarts its select() call on the socket, thereby resetting its timeout.
4511 This is not a problem when such interrupts are rare. Somebody set up a cron
4512 job to run exiwhat every 2 minutes, which is less than the normal select()
4513 timeout (5 or 10 minutes). This meant that the select() timeout never
4514 kicked in because it was always reset. I have fixed this by comparing the
4515 time when an interrupt arrives with the time at the start of the first call
4516 to select(). If more time than the timeout has elapsed, the interrupt is
4517 treated as a timeout.
4519 5. Some internal re-factoring in preparation for the addition of Sieve
4520 extensions (by MH). In particular, the "personal" test is moved to a
4521 separate function, and given an option for scanning Cc: and Bcc: (which is
4522 not set for Exim filters).
4524 6. When Exim created an email address using the login of the caller as the
4525 local part (e.g. when creating a From: or Sender: header line), it was not
4526 quoting the local part when it contained special characters such as @.
4528 7. Installed new OpenBSD configuration files.
4530 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4531 try to make them clearer.
4533 9. Callout options, other than the timeout value, were being ignored when
4534 verifying sender addresses in header lines. For example, when using
4536 verify = header_sender/callout=no_cache
4538 the cache was (incorrectly) being used.
4540 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4541 only the Cygwin environment.
4543 11. When return_path_on_delivery was set as a log selector, if different remote
4544 addresses in the same message used different return paths and parallel
4545 remote delivery occurred, the wrong values would sometimes be logged.
4546 (Whenever a remote delivery process finished, the return path value from
4547 the most recently started remote delivery process was logged.)
4549 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4550 lines when AUTH and/or TLS are in use. This is the "received protocol"
4551 field. Exim used to use "asmtp" for authenticated SMTP, without any
4552 indication (in the protocol name) for TLS use. Now it follows the RFC and
4553 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4554 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4555 names appear in log lines as well as in Received: header lines.
4557 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4558 extensions, and comparison tests, and to fix some bugs.
4560 14. Changes to the "personal" filter test:
4562 (1) The test was buggy in that it was just doing the equivalent of
4563 "contains" tests on header lines. For example, if a user's address was
4564 anne@some.where, the "personal" test would incorrectly be true for
4566 To: susanne@some.where
4568 This test is now done by extracting each address from the header in turn,
4569 and checking the entire address. Other tests that are part of "personal"
4570 are now done using regular expressions (for example, to check local parts
4571 of addresses in From: header lines).
4573 (2) The list of non-personal local parts in From: addresses has been
4574 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4575 taken from the Sieve specification recommendations.
4577 (3) If the message contains any header line starting with "List-" it is
4578 treated as non-personal.
4580 (4) The test for "circular" in the Subject: header line has been removed
4581 because it now seems ill-conceived.
4583 15. Minor typos in src/EDITME comments corrected.
4585 16. Installed latest exipick from John Jetmore.
4587 17. If headers_add on a router specified a text string that was too long for
4588 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4589 of string_sprintf() is now avoided.
4591 18. $message_body_size was not set (it was always zero) when running the DATA
4592 ACL and the local_scan() function.
4594 19. For the "mail" command in an Exim filter, no default was being set for
4595 the once_repeat time, causing a random time value to be used if "once" was
4596 specified. (If the value happened to be <= 0, no repeat happened.) The
4597 default is now 0s, meaning "never repeat". The "vacation" command was OK
4598 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4599 (I found it when inspecting the code).
4601 20. There is now an overall timeout for performing a callout verification. It
4602 defaults to 4 times the callout timeout, which applies to individual SMTP
4603 commands during the callout. The overall timeout applies when there is more
4604 than one host that can be tried. The timeout is checked before trying the
4605 next host. This prevents very long delays if there are a large number of
4606 hosts and all are timing out (e.g. when the network connections are timing
4607 out). The value of the overall timeout can be changed by specifying an
4608 additional sub-option for "callout", called "maxwait". For example:
4610 verify = sender/callout=5s,maxwait=20s
4612 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4613 to the end before writing, but this should make it even safer).
4615 22. Exim was forgetting that it had advertised PIPELINING for the second and
4616 subsequent messages on an SMTP connection. It was also not resetting its
4617 memory on STARTTLS and an internal HELO.
4619 23. When Exim logs an SMTP synchronization error within a session, it now
4620 records whether PIPELINING has been advertised or not.
4622 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4623 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4624 rather than long int.
4626 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4628 26. Added the never_mail option to autoreply.
4634 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4635 crash if the getsockname() call failed; this can happen if a connection is
4636 closed very soon after it is established. The problem was simply in the
4637 order in which certain operations were done, causing Exim to try to write
4638 to the SMTP stream before it had set up the file descriptor. The bug has
4639 been fixed by making things happen in the correct order.
4645 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4646 before the connection was closed, thus losing the rejection response.
4648 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4649 some early Solaris releases, but causes trouble in current releases where
4650 socklen_t is defined.
4652 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4655 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4658 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4659 that releasing the top of it at the end releases what was used for sub-
4660 expansions (unless the block got too big). However, discard this block if
4661 the first thing is a variable or header, so that we can use its block when
4662 it is dynamic (useful for very large $message_headers, for example).
4664 6. Lookups now cache *every* query, not just the most recent. A new, separate
4665 store pool is used for this. It can be recovered when all lookup caches are
4666 flushed. Lookups now release memory at the end of their result strings.
4667 This has involved some general refactoring of the lookup sources.
4669 7. Some code has been added to the store_xxx() functions to reduce the amount
4670 of flapping under certain conditions.
4672 8. log_incoming_interface used to affect only the <= reception log lines. Now
4673 it causes the local interface and port to be added to several more SMTP log
4674 lines, for example "SMTP connection from", and rejection lines.
4676 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4678 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4680 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4681 could be overwritten at the end of the current message (or the start of a
4682 new message if it was set in a HELO ACL). The value is now preserved for
4683 the duration of the SMTP connection.
4685 12. If a transport had a headers_rewrite setting, and a matching header line
4686 contained an unqualified address, that address was qualified, even if it
4687 did not match any rewriting rules. The underlying bug was that the values
4688 of the flags that permit the existence of unqualified sender and recipient
4689 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4690 non-local messages, and by -bnq for local messages) were not being
4691 preserved with the message after it was received.
4693 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4694 "next input=" as part of the text comprising the host identity instead of
4695 the correct text. The code was using the same buffer for two different
4696 strings. However, depending on which order the printing function evaluated
4697 its arguments, the bug did not always show up. Under Linux, for example, my
4698 test suite worked just fine.
4700 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4701 doesn't work with some older versions of Perl. It has been changed to "my",
4702 which in any case is probably the better facility to use.
4704 15. A really picky compiler found some instances of statements for creating
4705 error messages that either had too many or two few arguments for the format
4708 16. The size of the buffer for calls to the DNS resolver has been increased
4709 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4710 for addresses that have a lot of PTR records. This alleviates a problem; it
4711 does not fully solve it.
4713 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4714 buffer now truncates the list and logs the incident, which is the same
4715 action as happens when Exim is looking up a host name and its aliases.
4716 Previously in this situation something unpredictable would happen;
4717 sometimes it was "internal error: store_reset failed".
4719 18. If a server dropped the connection unexpectedly when an Exim client was
4720 using GnuTLS and trying to read a response, the client delivery process
4721 crashed while trying to generate an error log message.
4723 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4724 string, for example:
4726 warn message = H1: something\nH2: something
4728 the text was added as a single header line from Exim's point of view
4729 though it ended up OK in the delivered message. However, searching for the
4730 second and subsequent header lines using $h_h2: did not work. This has been
4731 fixed. Similarly, if a system filter added multiple headers in this way,
4732 the routers could not see them.
4734 20. Expanded the error message when iplsearch is called with an invalid key to
4735 suggest using net-iplsearch in a host list.
4737 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4738 ACLs are no longer actually imposed (and a message to that effect is
4741 22. If a "gecos" field in a passwd entry contained escaped characters, in
4742 particular, if it contained a \" sequence, Exim got it wrong when building
4743 a From: or a Sender: header from that name. A second bug also caused
4744 incorrect handling when an unquoted " was present following a character
4745 that needed quoting.
4747 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4748 was not being matched caselessly.
4750 24. Arranged for all hyphens in the exim.8 source to be escaped with
4753 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4754 properly. Recipient callout cache records were still being keyed to include
4755 the sender, even when use_sender was set false. This led to far more
4756 callouts that were necessary. The sender is no longer included in the key
4757 when use_sender is false.
4759 26. Added "control = submission" modifier to ACLs.
4761 27. Added the ${base62d: operator to decode base 62 numbers.
4763 28. dnsdb lookups can now access SRV records.
4765 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4766 the configuration file.
4768 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4769 (-v) mode. This makes the output for a verbose queue run more intelligible.
4771 31. Added a use_postmaster feature to recipient callouts.
4773 32. Added the $body_zerocount variable, containing the number of binary zero
4774 bytes in the message body.
4776 33. The time of last modification of the "new" subdirectory is now used as the
4777 "mailbox time last read" when there is a quota error for a maildir
4780 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4782 35. Added +ignore_unknown as a special item in host lists.
4784 36. Code for decoding IPv6 addresses in host lists is now included, even if
4785 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4786 address was recognized as an IP address, but was then not correctly decoded
4787 into binary, causing unexpected and incorrect effects when compared with
4794 1. Very minor rewording of debugging text in manualroute to say "list of
4795 hosts" instead of "hostlist".
4797 2. If verify=header_syntax was set, and a header line with an unqualified
4798 address (no domain) and a large number of spaces between the end of the
4799 name and the colon was received, the reception process suffered a buffer
4800 overflow, and (when I tested it) crashed. This was caused by some obsolete
4801 code that should have been removed. The fix is to remove it!
4803 3. When running in the test harness, delay a bit after writing a bounce
4804 message to get a bit more predictability in the log output.
4806 4. Added a call to search_tidyup() just before forking a reception process. In
4807 theory, someone could use a lookup in the expansion of smtp_accept_max_
4808 per_host which, without the tidyup, could leave open a database connection.
4810 5. Added the variables $recipient_data and $sender_data which get set from a
4811 lookup success in an ACL "recipients" or "senders" condition, or a router
4812 "senders" option, similar to $domain_data and $local_part_data.
4814 6. Moved the writing of debug_print from before to after the "senders" test
4817 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4818 problems for message scanning, either using a data ACL, or using
4819 local_scan() because the Received: header was not generated till after they
4820 were called (in order to set the time as the time of reception completion).
4821 I have revised the way this works. The header is now generated after the
4822 body is received, but before the ACL or local_scan() are called. After they
4823 are run, the timestamp in the header is updated.
4829 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4830 before starting a queue runner without re-exec. This happened only when
4831 deliver_drop_privilege was set or when the Exim user was set to root. The
4832 effect of the bug was that timeouts during subsequent deliveries caused
4833 crashes instead of being properly handled. The handler is now left at its
4834 default (and expected) setting.
4836 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4837 message, again when deliver_drop_privilege is set or Exim is run as root.
4838 The bug described in (1) was not present in this case, but the tidying up
4839 of the other signals was missing. I have made the two cases consistent.
4841 3. The ignore_target_hosts setting on a manualroute router was being ignored
4842 for hosts that were looked up using the /MX notation.
4844 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4847 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4848 operated on the sender address. After changing the $sender_address to <>
4849 for the sender address verify, Exim was re-instated it as the original
4850 (before rewriting) address, but remembering that it had rewritten it, so it
4851 wasn't rewriting it again. This bug also had the effect of breaking the
4852 sender address verification caching when the sender address was rewritten.
4854 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4855 This has been changed so that if the ip literal address matches
4856 ignore_target_hosts, the router declines.
4858 7. Added expansion conditions match_domain, match_address, and match_local_
4859 part (NOT match_host).
4861 8. The placeholder for the Received: header didn't have a length field set.
4863 9. Added code to Exim itself and to exim_lock to test for a specific race
4864 condition that could lead to file corruption when using MBX delivery. The
4865 issue is with the lockfile that is created in /tmp. If this file is removed
4866 after a process has opened it but before that process has acquired a lock,
4867 there is the potential for a second process to recreate the file and also
4868 acquire a lock. This could lead to two Exim processes writing to the file
4869 at the same time. The added code performs the same test as UW imapd; it
4870 checks after acquiring the lock that its file descriptor still refers to
4871 the same named file.
4873 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4874 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4877 11. Added the smtp_active_hostname option. If used, this will typically be made
4878 to depend on the incoming interface address. Because $interface_address is
4879 not set up until the daemon has forked a reception process, error responses
4880 that can happen earlier (such as "too many connections") no longer contain
4883 12. If an expansion in a condition on a "warn" statement fails because a lookup
4884 defers, the "warn" statement is abandoned, and the next ACL statement is
4885 processed. Previously this caused the whole ACL to be aborted.
4887 13. Added the iplsearch lookup type.
4889 14. Added ident_timeout as a log selector.
4891 15. Added tls_certificate_verified as a log selector.
4893 16. Added a global option tls_require_ciphers (compare the smtp transport
4894 option of the same name). This controls incoming TLS connections.
4896 17. I finally figured out how to make tls_require_ciphers do a similar thing
4897 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4898 before starting the TLS session.
4900 18. Tabs are now shown as \t in -bP output.
4902 19. If the log selector return_path_on_delivery was set, Exim crashed when
4903 bouncing a message because it had too many Received: header lines.
4905 20. If two routers both had headers_remove settings, and the first one included
4906 a superfluous trailing colon, the final name in the first list and the
4907 first name in the second list were incorrectly joined into one item (with a
4908 colon in the middle).
4914 1. Added -C and -D options to the exinext utility, mainly to make it easier
4915 to include in the automated testing, but these could be helpful when
4916 multiple configurations are in use.
4918 2. The exinext utility was not formatting the output nicely when there was
4919 an alternate port involved in the retry record key, nor when there was a
4920 message id as well (for retries that were specific to a specific message
4921 and a specific host). It was also confused by IPv6 addresses, because of
4922 the additional colons they contain. I have fixed the IPv4 problem, and
4923 patched it up to do a reasonable job for IPv6.
4925 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4926 delivery, the log line now contains "pipelined" if PIPELINING was used.
4928 4. An SMTP transport process used to panic and die if the bind() call to set
4929 an explicit outgoing interface failed. This has been changed; it is now
4930 treated in the same way as a connect() failure.
4932 5. A reference to $sender_host_name in the part of a conditional expansion
4933 that was being skipped was still causing a DNS lookup. This no longer
4936 6. The def: expansion condition was not recognizing references to header lines
4937 that used bh_ and bheader_.
4939 7. Added the _cache feature to named lists.
4941 8. The code for checking quota_filecount in the appendfile transport was
4942 allowing one more file than it should have been.
4944 9. For compatibility with Sendmail, the command line option
4952 and sets the incoming protocol and host name (for trusted callers). The
4953 host name and its colon can be omitted when only the protocol is to be set.
4954 Note the Exim already has two private options, -pd and -ps, that refer to
4955 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4956 "s", but I don't think that's a major issue.
4958 10. A number of refactoring changes to the code, none of which should affect
4961 (a) The number of logging options was getting close to filling up the
4962 32-bit word that was used as a bit map. I have split them into two classes:
4963 those that are passed in the argument to log_write(), and those that are
4964 only ever tested independently outside of that function. These are now in
4965 separate 32-bit words, so there is plenty of room for expansion again.
4966 There is no change in the user interface or the logging behaviour.
4968 (b) When building, for example, log lines, the code previously used a
4969 macro that called string_cat() twice, in order to add two strings. This is
4970 not really sufficiently general. Furthermore, there was one instance where
4971 it was actually wrong because one of the argument was used twice, and in
4972 one call a function was used. (As it happened, calling the function twice
4973 did not affect the overall behaviour.) The macro has been replaced by a
4974 function that can join an arbitrary number of extra strings onto a growing
4977 (c) The code for expansion conditions now uses a table and a binary chop
4978 instead of a serial search (which was left over from when there were very
4979 few conditions). Also, it now recognizes conditions like "pam" even when
4980 the relevant support is not compiled in: a suitably worded error message is
4981 given if an attempt is made to use such a condition.
4983 11. Added ${time_interval:xxxxx}.
4985 12. A bug was causing one of the ddress fields not to be passed back correctly
4986 from remote delivery subprocesses. The field in question was not being
4987 subsequently used, so this caused to problems in practice.
4989 13. Added new log selectors queue_time and deliver_time.
4991 14. Might have fixed a bug in maildirsizefile handling that threw up
4992 "unexpected character" debug warnings, and recalculated the data
4993 unnecessarily. In any case, I expanded the warning message to give more
4996 15. Added the message "Restricted characters in address" to the statements in
4997 the default ACL that block characters like @ and % in local parts.
4999 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5000 Three changes have been made:
5002 (a) There was a serious bug; a negative response to MAIL caused the whole
5003 recipient domain to be cached as invalid, thereby blocking all messages
5004 to all local parts at the same domain, from all senders. This bug has
5005 been fixed. The domain is no longer cached after a negative response to
5006 MAIL if the sender used is not empty.
5008 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5011 (c) A new callout option, "use_sender" has been added for people who want
5012 the modified behaviour.
5018 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5021 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5022 indeed breaks things for older releases.
5024 3. Added additional logging to the case where there is a problem reading data
5025 from a filter that is running in a subprocess using a pipe, in order to
5026 try to track down a specific problem.
5028 4. Testing facility fudge: when running in the test harness and attempting
5029 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5030 getting "No route to host". Convert this to a timeout.
5032 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5035 6. Some OS don't have socklen_t but use size_t instead. This affects the
5036 fifth argument of getsockopt() amongst other things. This is now
5037 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5038 can be set for individual OS. I have set it for SunOS5, OSF1, and
5039 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5040 some earlier ones do not.
5042 7. Change 4.30/15 was not doing the test caselessly.
5044 8. The standard form for an IPv6 address literal was being rejected by address
5045 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5046 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5047 this, as well as the form without the "IPv6" on the front (but only when
5048 address literals are enabled, of course).
5050 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5052 10. Exim crashed if a message with an empty sender address specified by -f
5053 encountered a router with an errors_to setting. This could be provoked only
5054 by a command such as
5058 where an empty string was supplied; "<>" did not hit this bug.
5060 11. Installed PCRE release 4.5.
5062 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5063 remained set. It is now erased.
5065 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5066 times from message ids (which are base 36 rather than the normal 62).
5068 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5069 were being counted as actual protocol errors, and logged if the log
5070 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5071 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5072 and DATA following a set of rejected RCPTs do not count as protocol errors.
5073 In other words, Exim assumes they were pipelined, though this may not
5074 actually be the case. Of course, in all cases the client gets an
5075 appropriate error code.
5077 15. If a lookup fails in an ACL condition, a message about the failure may
5078 be available; it is used if testing the ACL cannot continue, because most
5079 such messages specify what the cause of the deferral is. However, some
5080 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5081 that caused an old message to be retained and used if a later statement
5082 caused a defer, replacing the real cause of the deferral.
5084 16. If an IP address had so many PTR records that the DNS lookup buffer
5085 was not large enough to hold them, Exim could crash while trying to process
5086 the truncated data. It now detects and logs this case.
5088 17. Further to 4.21/58, another change has been made: if (and only if) the
5089 first line of a message (the first header line) ends with CRLF, a bare LF
5090 in a subsequent header line has a space inserted after it, so as not to
5091 terminate the header.
5093 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5094 unnecessarily, used atoi() instead of strtol(), and didn't check the
5095 termination when getting file sizes from file names by regex.
5097 19. Completely re-implemented the support for maildirsize files, in the light
5098 of a number of problems with the previous contributed implementation
5099 (4.30/29). In particular:
5101 . If the quota is zero, the maildirsize file is maintained, but no quota is
5104 . If the maildir directory does not exist, it is created before any attempt
5105 to write a maildirsize file.
5107 . The quota value in the file is just a cache; if the quota is changed in
5108 the transport, the new value overrides.
5110 . A regular expression is available for excluding directories from the
5113 20. The autoreply transport checks the characters in options that define the
5114 message's headers; it allows continued headers, but it was checking with
5115 isspace() after an embedded newline instead of explicitly looking for a
5118 21. If all the "regular" hosts to which an address was routed had passed their
5119 expiry times, and had not reached their retry times, the address was
5120 bounced, even if fallback hosts were defined. Now Exim should go on to try
5123 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5124 equivalent code in the SMTP transport. Some hosts send humungous responses
5125 to HELO/EHLO, more than 1024 it seems.
5127 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5128 gives compiler warnings in some environments. I've now done it "properly",
5131 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5132 (because of problems with the built-in one) was declared to return uschar *
5133 instead of char *, causing compiler failure.
5135 25. Fixed a file descriptor leak when processing alias/forward files.
5137 26. Fixed a minor format string issue in dbfn.c.
5139 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5141 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5142 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5145 29. When a sender address is verified, it is cached, to save repeating the test
5146 when there is more than one recipient in a message. However, when the
5147 verification involves a callout, it is possible for different callout
5148 options to be set for different recipients. It is too complicated to keep
5149 track of this in the cache, so now Exim always runs a verification when a
5150 callout is required, relying on the callout cache for the optimization.
5151 The overhead is duplication of the address routing, but this should not be
5154 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5155 to be verified with callout=postmaster, and the main callout worked but the
5156 postmaster check failed, the verification correctly failed. However, if a
5157 subsequent RCPT command asked for sender verification *without* the
5158 postmaster check, incorrect caching caused this verification also to fail,
5161 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5162 it was not caching the DNS options (qualify_single, search_parents) that
5163 were used when the lookup failed. A subsequent lookup with different
5164 options therefore always gave the same answer, though there were cases
5165 where it should not have. (Example: a "domains = !$mx_any" option on a
5166 dnslookup router: the "domains" option is always processed without any
5167 widening, but the router might have qualify_single set.) Now Exim uses the
5168 cached value only when the same options are set.
5170 32. Added John Jetmore's "exipick" utility to the distribution.
5172 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5173 than a timeout (e.g. a certificate is required, and is not provided), an
5174 Exim server now closes the connection immediately. Previously it waited for
5175 the client to close - but if the client is SSL, it seems that they each
5176 wait for each other, leading to a delay before one of them times out.
5178 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5179 maintained 0.8.x compatibility because I don't think many are using it, and
5180 it is clearly obsolete.
5182 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5185 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5186 client certificate was expired. A simple patch fixes this, though I don't
5187 understand the full logic of why the verify callback is called multiple
5190 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5191 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5192 which causes problems with some clients (such as the Certicom SSL Plus
5193 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5194 disables the coutermeasure allowing Eudora to connect."
5196 38. Exim was not checking that a write() to a log file succeeded. This could
5197 lead to Bad Things if a log got too big, in particular if it hit a file
5198 size limit. Exim now panics and dies if it cannot write to a log file, just
5199 as it does if it cannot open a log file.
5201 39. Modified OS/Makefile-Linux so that it now contains
5203 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5205 The two -D definitions ensure that Exim is compiled with large file
5206 support, which makes it possible to handle log files that are bigger than
5209 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5210 instance) a domain was checked against a named list that involved a lookup,
5211 causing $domain_data to be set, then another domain was checked against the
5212 same list, then the first domain was re-checked, the value of $domain_data
5213 after the final check could be wrong. In particular, if the second check
5214 failed, it could be set empty. This bug probably also applied to
5217 41. The strip_trailing_dot option was not being applied to the address given
5218 with the -f command-line option.
5220 42. The code for reading a message's header from the spool was incrementing
5221 $received_count, but never initializing it. This meant that the value was
5222 incorrect (doubled) while delivering a message in the same process in which
5223 it was received. In the most common configuration of Exim, this never
5224 happens - a fresh exec is done - but it can happen when
5225 deliver_drop_privilege is set.
5227 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5228 it now includes up to 150 characters of the unexpected data in the log
5231 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5232 and building data strings. The size of both of these buffers was 10 000
5233 bytes - far larger than anybody would *ever* want, thought I. Needless to
5234 say, somebody hit the limit. I have increased the maximum line length to
5235 20 000 and the maximum data length of concatenated lines to 100 000. I have
5236 also fixed two bugs, because there was no checking on these buffers. Tsk,
5237 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5238 buffer is too small.
5240 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5241 lsearch lookups. Now it does.
5243 46. When parsing a route_list item in a manualroute router, a fixed-length
5244 buffer was used for the list of hosts. I made this 1024 bytes long,
5245 thinking that nobody would ever have a list of hosts that long. Wrong.
5246 Somebody had a whole pile of complicated expansion conditions, and the
5247 string was silently truncated, leading to an expansion error. It turns out
5248 that it is easier to change to an unlimited length (owing to other changes
5249 that have happened since this code was originally written) than to build
5250 structure for giving a limitation error. The length of the item that
5251 expands into the list of hosts is now unlimited.
5253 47. The lsearch lookup could not handle data where the length of text line was
5254 more than 4095 characters. Such lines were truncated, leading to shortened
5255 data being returned. It should now handle lines of any length.
5257 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5258 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5261 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5262 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5263 no longer happens in comment lines. A list of replacements is now placed
5264 at the head of all of the source files, except those whose only change is
5265 to replace PERL_COMMAND in the very first #! line.
5267 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5268 messages on the queue, with a bottom-up merge sort, using code contributed
5269 by Michael Haardt. This should make operations like -bp somewhat faster on
5270 large queues. It won't affect queue runners, except when queue_run_in_order
5273 51. Installed eximstats 1.31 in the distribution.
5275 52. Added support for SRV lookups to the dnslookup router.
5277 53. If an ACL referred to $message_body or $message_body_end, the value was not
5278 reset for any messages that followed in the same SMTP session.
5280 54. The store-handling optimization for building very long strings was not
5281 differentiating between the different store pools. I don't think this
5282 actually made any difference in practice, but I've tidied it.
5284 55. While running the routers to verify a sender address, $sender_address
5285 was still set to the sender address. This is wrong, because when routing to
5286 send a bounce to the sender, it would be empty. Therefore, I have changed
5287 it so that, while verifying a sender address, $sender_address is set to <>.
5288 (There is no change to what happens when verifying a recipient address.)
5290 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5291 target A or AAAA records (if not already returned) without resetting the
5292 qualify_single or search_parents options of the DNS resolver. These are
5293 inappropriate in this case because the targets of MX and SRV records must
5294 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5295 target that, when qualified, matched something in the local domain. These
5296 two options are now turned off when doing these lookups.
5298 57. It seems that at least some releases of Reiserfs (which does not have the
5299 concept of a fixed number of inodes) returns zero and not -1 for the
5300 number of available inodes. This interacted badly with check_spool_inodes,
5301 which assumed that -1 was the "no such thing" setting. What I have done is
5302 to check that the total number of inodes is greater than zero before doing
5303 the test of how many are available.
5305 58. When a "warn" ACL statement has a log_message modifier, the message is
5306 remembered, and not repeated. This is to avoid a lot of repetition when a
5307 message has many recipients that cause the same warning to be written.
5308 Howewer, Exim was preserving the list of already written lines for an
5309 entire SMTP session, which doesn't seem right. The memory is now reset if a
5310 new message is started.
5312 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5313 debugging output unless log_rewrite was also set.
5315 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5316 of (int)(handle) when we know that handle contains (void *)(-1).
5318 61. The Exim daemon panic-logs an error return when it closes the incoming
5319 connection. However "connection reset by peer" seems to be common, and
5320 isn't really an error worthy of noting specially, so that particular error
5323 62. When Exim is trying to find all the local interfaces, it used to panic and
5324 die if the ioctl to get the interface flags failed. However, it seems that
5325 on at least one OS (Solaris 9) it is possible to have an interface that is
5326 included in the list of interfaces, but for which you get a failure error
5327 for this call. This happens when the interface is not "plumbed" into a
5328 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5329 failure of the "get flags" call assumes that the interface is down.
5331 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5332 makes life easier for people who are doing arithmetic on fields extracted
5333 from dates, where you often get leading zeros that should not be
5334 interpreted as octal.
5336 64. Added qualify_domain to the redirect router, to override the global
5339 65. If a pathologically long header line contained very many addresses (the
5340 report of this problem mentioned 10 000) and each of them was rewritten,
5341 Exim could use up a very large amount of memory. (It kept on making new
5342 copies of the header line as it rewrote, and never released the old ones.)
5343 At the expense of a bit more processing, the header rewriting function has
5344 been changed so that it no longer eats memory in this way.
5346 66. The generation of the Received: header has been moved from the time that a
5347 message starts to be received, to the time that it finishes. The timestamp
5348 in the Received: header should now be very close to that of the <= log
5349 line. There are two side-effects of this change:
5351 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5352 logged header lines no longer include the local Received: line, because
5353 it has not yet been created. The same applies to a copy of the message
5354 that is returned to a non-SMTP sender when a message is rejected.
5356 (b) When a filter file is tested using -bf, no additional Received: header
5357 is added to the test message. After some thought, I decided that this
5360 This change does not affect the value of $received_for. It is still set
5361 after address rewriting, but before local_scan() is called.
5363 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5365 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5366 gave an unhelpful panic error message, and a defer error. I have managed to
5367 change this behaviour so that it now rejects any supplied certificate,
5368 which seems right, as the list of acceptable certificates is empty.
5370 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5371 gave an unhelpful defer error. I have not managed to make this reject any
5372 supplied certificates, but the error message it gives is "no certificate
5373 supplied", which is not helpful.
5375 70. exigrep's output now also includes lines that are not associated with any
5376 message, but which match the given pattern. Implemented by a patch from
5377 Martin Sluka, which also tidied up the Perl a bit.
5379 71. Recipient callout verification, like sender verification, was using <> in
5380 the MAIL FROM command. This isn't really the right thing, since the actual
5381 sender may affect whether the remote host accepts the recipient or not. I
5382 have changed it to use the actual sender in the callout; this means that
5383 the cache record is now keyed on a recipient/sender pair, not just the
5384 recipient address. There doesn't seem to be a real danger of callout loops,
5385 since a callout by the remote host to check the sender would use <>.
5386 [SEE ABOVE: changed after hitting problems.]
5388 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5389 temporary errors. However, in the case of such a code being given after
5390 the end of a data transmission (i.e. after ".") Exim was failing to write
5391 a retry record for the message. (Yes, there was some broken host that was
5392 actually sending 8xx at this point.)
5394 73. An unknown lookup type in a host list could cause Exim to panic-die when
5395 the list was checked. (An example that provoked this was putting <; in the
5396 middle of a list instead of at the start.) If this happened during a DATA
5397 ACL check, a -D file could be left lying around. This kind of configuration
5398 error no longer causes Exim to die; instead it causes a defer errror. The
5399 incident is still logged to the main and panic logs.
5401 74. Buglet left over from Exim 3 conversion. The message "too many messages
5402 in one connection" was written to the rejectlog but not the mainlog, except
5403 when address rewriting (yes!) was being logged.
5405 75. Added write_rejectlog option.
5407 76. When a system filter was run not as root (that is, when system_filter_user
5408 was set), the values of the $n variables were not being returned to the
5409 main process; thus, they were not subsequently available in the $sn
5412 77. Added +return_path_on_delivery log selector.
5414 78. A connection timeout was being treated differently from recipients deferred
5415 when testing hosts_max_try with a message that was older than the host's
5416 retry timeout. (The host should not be counted, thus allowing all hosts to
5417 be tried at least once before bouncing.) This may have been the cause of an
5418 occasionally reported bug whereby a message would remain on the queue
5419 longer than the retry timeout, but would be bounced if a delivery was
5420 forced. I say "may" because I never totally pinned down the problem;
5421 setting up timeout/retry tests is difficult. See also the next item.
5423 79. The ultimate address timeout was not being applied to errors that involved
5424 a combination of host plus message (for example, a timeout on a MAIL
5425 command). When an address resolved to a number of possible hosts, and they
5426 were not all tried for each delivery (e.g. because of hosts_max_try), a
5427 message could remain on the queue longer than the retry timeout.
5429 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5432 81. Fixed an obscure SMTP outgoing bug which required at least the following
5433 conditions: (a) there was another message waiting for the same server;
5434 (b) the server returned 5xx to all RCPT commands in the first message so
5435 that the message was not completed; (c) the server dropped the connection
5436 or gave a negative response to the RSET that Exim sends to abort the
5437 transaction. The observed case was a dropped connection after DATA that had
5438 been sent in pipelining mode. That is, the server had advertised PIPELINING
5439 but was not implementing it correctly. The effect of the bug was incorrect
5440 behaviour, such as trying another host, and this could lead to a crash.
5446 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5447 and daemon.c were passed as pointers to ints; they should have been
5448 pointers to socklen_t variables (which are typically unsigned ints).
5450 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5453 3. Fixed a really odd bug that affected only the testing scheme; patching a
5454 certain fixed string in the binary changed the value of another string that
5455 happened to be identical to the end of the original first string.
5457 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5458 name", it returns that address as the IP address. On some operating
5459 systems (e.g. Solaris), it also passes back the IP address string as the
5460 "host name". However, on others (e.g. Linux), it passes back an empty
5461 string. Exim wasn't checking for this, and was changing the host name to an
5462 empty string, assuming it had been canonicized.
5464 5. Although rare, it is permitted to have more than one PTR record for a given
5465 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5466 all the names associated with an address, because they do in Solaris.
5467 However, it seems that they do not in Linux for data that comes from the
5468 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5469 I found this out when I moved to a new Linux workstation and tried to run
5470 the Exim test suite.
5472 To get round this problem I have changed the code so that it now does its
5473 own call to the DNS to look up PTR records when searching for a host name.
5474 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5475 addresses that are only in /etc/hosts are still found.
5477 This behaviour is, however, controlled by an option called host_lookup_
5478 order, which defaults to "bydns:byaddr". If people want to use the other
5479 order, or indeed, just use one or the other means of lookup, they can
5480 specify it in this variable.
5482 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5483 some operating systems, this comes back from gethostbyaddr() as an empty
5484 string, and this is what Exim used to test for. However, it seems that in
5485 other systems, "." is yielded. Exim now tests for this case too.
5487 7. The values of check_spool_space and check_log_space are now held internally
5488 as a number of kilobytes instead of an absolute number of bytes. If a
5489 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5490 kilobyte. This means that much larger values can be stored.
5492 8. Exim monitor: an attempt to get the action menu when not actually pointing
5493 at a message produces an empty menu entitled "No message selected". This
5494 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5495 no entries in it ("Shell widget menu has zero width and/or height"). So I
5496 have added a single, blank menu entry in this case.
5498 9. Added ${quote_local_part.
5500 10. MIME decoding is now applied to the contents of Subject: header lines when
5503 11. Now that a reference to $sender_host_address automatically causes a reverse
5504 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5505 host lookup before query-style lookups in lists that might use this
5506 variable. This has therefore been abolished, and the "net-" prefix is no
5507 longer necessary for query-style lookups.
5509 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5510 have been a typo for LFLAGS, so it has been changed.
5512 13. The install script calls Exim with "-C /dev/null" in order to find the
5513 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5514 to be output. Howeve, since Exim outputs its version number before the
5515 error, it didn't break the script. It just looked ugly. I fixed this by
5516 always allowing "-C /dev/null" if the caller is root.
5518 14. Ignore overlarge ACL variable number when reading spool file - insurance
5519 against a later release with more variables having written the file.
5521 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5522 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5523 this, as well as the form without the "IPv6" on the front.
5525 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5526 OS/Makefile-Darwin file.
5528 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5529 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5531 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5532 variable. However, some OS use ints for inodes. I've added cast to long int
5533 to get rid of the compiler warning.
5535 19. I had forgotten to lock out "/../" in configuration file names when
5536 ALT_CONFIG_PREFIX was set.
5538 20. Routers used for verification do not need to specify transports. However,
5539 if such a router generated a host list, and callout was configured, Exim
5540 crashed, because it could not find a port number from the (non-existent)
5541 transport. It now assumes port 25 in this circumstance.
5543 21. Added the -t option to exigrep.
5545 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5546 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5547 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5548 them is set, LOOKUP_LSEARCH is forced.
5550 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5552 24. Added sender and host information to the "rejected by local_scan()" log
5553 line; previously there was no indication of these.
5555 25. Added .include_if_exists.
5557 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5558 Linux. It turns out that not all file systems support this. Apparently some
5559 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5560 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5561 not supported on the file descriptor, is now ignored when Exim is trying to
5562 sync a directory. This applies only to Linux.
5564 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5566 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5567 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5570 29. Added support for maildirsize files from supplied patch (modified a bit).
5572 30. The use of :fail: followed by an empty string could lead Exim to respond to
5573 sender verification failures with (e.g.):
5575 550 Verification failed for <xxx>
5576 550 Sender verify failed
5578 where the first response line was missing the '-' that indicates it is not
5579 the final line of the response.
5581 31. The loop for finding the name of the user that called Exim had a hardwired
5582 limit of 10; it now uses the value of finduser_retries, which is used for
5583 all other user lookups.
5585 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5588 33. Exim was neglecting to zero errno before one call of strtol() when
5589 expanding a string and expecting an integer value. On some systems this
5590 resulted in spurious "integer overflow" errors. Also, it was casting the
5591 result into an int without checking.
5593 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5594 did not work. The code looks as if it has *never* worked, though it appears
5595 to have been documented since at least releast 1.62. I have made it work.
5597 35. The "timeout_DNS" error in retry rules, also documented since at least
5598 1.62, also never worked. As it isn't clear exactly what this means, and
5599 clearly it isn't a major issue, I have abolished the feature by treating it
5600 as "timeout", and writing a warning to the main and panic logs.
5602 36. The display of retry rules for -brt wasn't always showing the error code
5605 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5606 timeout_connect_A, timeout_connect_MX.
5608 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5609 to the empty sender.
5611 39. The daemon was not analysing the content of -oX till after it had closed
5612 stderr and disconnected from the controlling terminal. This meant that any
5613 syntax errors were only noted on the panic log, and the return code from
5614 the command was 0. By re-arranging the code a little, I've made the
5615 decoding happen first, so such errors now appear on stderr, and the return
5616 code is 1. However, the actual setting up of the sockets still happens in
5617 the disconnected process, so errors there are still only recorded on the
5620 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5621 connections (as happens on some IP stacks) was logged at start up time as
5622 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5623 it from "IPv6 and IPv4", which means that two separate sockets are being
5626 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5627 says whether AF_INET or AF_INET6 was passed as an argument.
5629 42. Exiwhat output was messed up when time zones were included in log
5632 43. Exiwhat now gives more information about the daemon's listening ports,
5633 and whether -tls-on-connect was used.
5635 44. The "port" option of the smtp transport is now expanded.
5637 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5638 silently ignored. Now an error message is written to the main and panic
5641 46. There's a new ACL modifier called "logwrite" which writes to a log file
5642 as soon as it is encountered.
5644 47. Added $local_user_uid and $local_user_gid at routing time.
5646 48. Exim crashed when trying to verify a sender address that was being
5649 49. Exim was recognizing only a space character after ".include". It now also
5650 recognizes a tab character.
5652 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5653 extracting the relevant information from the specification. The man page no
5654 longer contains scrambled data for the -d option, and I've added a section
5655 at the front about calling Exim under different names.
5657 51. Added "extra_headers" argument to the "mail" command in filter files.
5659 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5662 53. Installed eximstats 1.29.
5664 54. Added transport_filter_timeout as a generic transport option.
5666 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5667 Cc: header lines. This was required by RFC 822, but it not required by RFC
5670 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5671 incoming messages that did not have them. Now it does so only if the
5672 message originates locally, that is, if there is no associated remote host
5673 address. When Resent- header lines are present, this applies to the Resent-
5674 lines rather than the non-Resent- lines.
5676 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5677 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5679 58. Messages for configuration errors now include the name of the main
5680 configuration file - useful now that there may be more than one file in a
5681 list (.included file names were always shown).
5683 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5684 for those rare installations that do not start the daemon as root or run it
5685 setuid root. I've cut out the call to initgroups() if the daemon is not
5688 60. The Exim user and group can now be bound into the binary as text strings
5689 that are looked up at the start of Exim's processing.
5691 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5693 62. Added $mailstore_basename variable.
5695 63. Installed patch to sieve.c from Michael Haardt.
5697 64. When Exim failed to open the panic log after failing to open the main log,
5698 the original message it was trying to log was written to stderr and debug
5699 output, but if they were not available (the usual case in production), it
5700 was lost. Now it is written to syslog before the two lines that record the
5701 failures to open the logs.
5703 65. Users' Exim filters run in subprocesses under the user's uid. It is
5704 possible for a "deliver" command or an alias in a "personal" command to
5705 provoke an address rewrite. If logging of address rewriting is configured,
5706 this fails because the process is not running as root or exim. There may be
5707 a better way of dealing with this, but for the moment (because 4.30 needs
5708 to be released), I have disabled address rewrite logging when running a
5709 filter in a non-root, non-exim process.
5715 1. The buildconfig auxiliary program wasn't quoting the value set for
5716 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5717 not defined. This bug was masked in 4.22 by the effect that was fixed in
5720 2. Some messages that were rejected after a message id was allocated were
5721 shown as "incomplete" by exigrep. It no longer does this for messages that
5722 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5724 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5725 have allow_domain_literals set, the ID did not get logged in the <= line.
5726 Domain literals are now always recognized in Message-ID: header lines.
5728 4. The first argument for a ${extract expansion item is the key name or field
5729 number. Leading and trailing spaces in this item were not being ignored,
5730 causing some misleading effects.
5732 5. When deliver_drop_privilege was set, single queue runner processes started
5733 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5734 same command in the process it spins off) were not dropping privilege.
5736 6. When the daemon running as "exim" started a queue runner, it always
5737 re-executed Exim in the spun-off process. This is a waste of effort when
5738 deliver_drop_privilege is set. The new process now just calls the
5739 queue-runner function directly.
5745 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5748 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5749 ignored. Though the use of -oP was forcing the writing of a pid file, it
5750 was always written to the default place.
5752 3. If the message "no IP address found for host xxxx" is generated during
5753 incoming verification, it is now followed by identification of the incoming
5754 connection (so you can more easily find what provoked it).
5756 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5758 5. Added some features to "harden" Exim a bit more against certain attacks:
5760 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5761 be put in Local/Makefile. This is like the never_users runtime option,
5762 but it cannot be overridden. The default setting is "root".
5764 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5765 prefix string with which any file named in a -C command line option
5768 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5769 is retained for -C and -D only if the caller of Exim is root. Without
5770 it, the exim user may also use -C and -D and retain privilege.
5772 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5773 command line option is disabled.
5775 6. Macro names set by the -D option must start with an upper case letter, just
5776 like macro names defined in the configuration file.
5778 7. Added "dereference=" facility to LDAP.
5780 8. Two instances of the typo "uknown" in the source files are fixed.
5782 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5783 the Configure-Makefile script screwed up while processing it.
5785 10. Incorporated PCRE 4.4.
5787 11. The SMTP synchronization check was not operating right at the start of an
5788 SMTP session. For example, it could not catch a HELO sent before the client
5789 waited for the greeting. There is now a check for outstanding input at the
5790 point when the greeting is written. Because of the duplex, asynchronous
5791 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5792 way, but not yet received, when the check is performed.
5794 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5795 on TCP/IP sockets, because this apparently causes some broken clients to
5798 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5799 unchanged) from the Cygwin maintainer.
5801 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5802 was supported instead of testing for mbx. Effectively a typo.
5804 15. The spa authenticator server code was not checking that the input it
5805 received was valid base64.
5807 16. The debug output line for the "set" modifier in ACLs was not showing the
5808 name of the variable that was being set.
5810 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5812 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5813 lookup on its own. Something else was needed to trigger the lookup. For
5814 example, a match in host_lookup or the need for a host name in a host list.
5815 Now, if $sender_host_name is referenced and the host name has not yet been
5816 looked up, a lookup is performed. If the lookup fails, the variable remains
5817 empty, and $host_lookup_failed is set to "1".
5819 19. Added "eqi" as a case-independent comparison operator.
5821 20. The saslauthd authentication condition could segfault if neither service
5822 nor realm was specified.
5824 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5825 error message that was logged was misleading, and incoming SMTP
5826 connections were dropped. The message is now more accurate, and temporary
5827 errors are given to SMTP connections.
5829 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5830 (except RSET) with a 421 error, until QUIT is received. However, it was
5831 failing to send a response to QUIT.
5833 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5834 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5835 ACL is now run after the helo_try_verify_hosts code.
5837 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5838 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5839 case-independent, so other case variants are also recognized). Apparently
5840 some systems use these upper case variants.
5842 25. If more than two messages were waiting for the same host, and a transport
5843 filter was specified for the transport, Exim sent two messages over the
5844 same TCP/IP connection, and then failed with "socket operation on non-
5845 socket" when it tried to send the third.
5847 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5849 27. The extern definition of crypt16() in expand.c was not being excluded when
5850 the OS had its own crypt16() function.
5852 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5853 as a preferred synonym for return_size_limit, both as an option and as an
5856 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5858 30. Changed the default configuration ACL to relax the local part checking rule
5859 for addresses that are not in any local domains. For these addresses,
5860 slashes and pipe symbols are allowed within local parts, but the sequence
5861 /../ is explicitly forbidden.
5863 31. SPA server authentication was not clearing the challenge buffer before
5866 32. log_message in a "warn" ACL statement was writing to the reject log as
5867 well as to the main log, which contradicts the documentation and doesn't
5868 seem right (because no rejection is happening). So I have stopped it.
5870 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5871 However, I am unable to do any testing of this.
5873 34. Fixed an infelicity in the appendfile transport. When checking directories
5874 for a mailbox, to see if any needed to be created, it was accidentally
5875 using path names with one or more superfluous leading slashes; tracing
5876 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5878 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5879 contents are added to the log line that is written for every discarded
5880 recipient. (Previously a log_message setting was ignored.)
5882 36. The ${quote: operator now quotes the string if it is empty.
5884 37. The install script runs exim in order to find its version number. If for
5885 some reason other than non-existence or emptiness, which it checks, it
5886 could not run './exim', it was installing it with an empty version number,
5887 i.e. as "exim-". This error state is now caught, and the installation is
5890 38. An argument was missing from the function that creates an error message
5891 when Exim fails to connect to the socket for saslauthd authentication.
5892 This could cause Exim to crash, or give a corrupted message.
5894 39. Added isip, isip4, and isip6 to ${if conditions.
5896 40. The ACL variables $acl_xx are now saved with the message, and can be
5897 accessed later in routers, transports, and filters.
5899 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5900 strings in the file are not string-expanded.
5902 42. If a MAIL command specified a SIZE value that was too large to fit into an
5903 int variable, the check against message_size_limit failed. Such values are
5904 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5905 day this will have to be increased, but I don't think I want to be around
5906 when emails are that large.
5913 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5914 iconv() is not standard in FreeBSD.
5916 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5917 IPv6 enabled. The observed symptom was a segmentation fault on return from
5918 the function os_common_find_running_interfaces() in src/os.c.
5920 3. In the check_special_case() function in daemon.c I had used "errno" as an
5921 argument name, which causes warnings on some systems. This was basically a
5922 typo, since it was named "eno" in the comments!
5924 4. The code that waits for the clock to tick (at a resolution of some fraction
5925 of a second) so as to ensure message-id uniqueness was always waiting for
5926 at least one whole tick, when it could have waited for less. [This is
5927 almost certainly not relevant at current processor speeds, where it is
5928 unlikely to ever wait at all. But we try to future-proof.]
5930 5. The function that sleeps for a time interval that includes fractions of a
5931 second contained a race. It did not block SIGALRM between setting the
5932 timer, and suspending (a couple of lines later). If the interval was short
5933 and the sigsuspend() was delayed until after it had expired, the suspension
5934 never ended. On busy systems this could lead to processes getting stuck for
5937 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5938 process, before it forks any delivery processes. The open lookup caching
5939 mechanism meant that the open file or database connection was passed into
5940 the delivery process. The problem was that delivery processes always tidy
5941 up cached lookup data. This could cause a problem for the next delivery
5942 process started by the queue runner, because the external queue runner
5943 process does not know about the closure. So the next delivery process
5944 still has data in the lookup cache. In the case of a file lookup, there was
5945 no problem because closing a file descriptor in a subprocess doesn't affect
5946 the parent. However, if the lookup was caching a connection to a database,
5947 the connection was closed, and the second delivery process was likely to
5948 see errors such as "PGSQL: query failed: server closed the connection
5949 unexpectedly". The problem has been fixed by closing all cached lookups
5950 in a queue runner before running a delivery process.
5952 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5953 seem to have the "const" qualifier which it has on other OS. I've
5956 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5957 *of the same type* (client or server) with the same public name that an
5958 error should be diagnosed.
5960 9. When Exim looked up a host name for an IP address, but failed to find the
5961 original IP address when looking up the host name (a safety check), it
5962 output the message "<ip address> does not match any IP for NULL", which was
5963 confusing, to say the least. The bug was that the host name should have
5964 appeared instead of "NULL".
5966 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5967 user that is built into the binary, and the -C or -D options is used, root
5968 privilege is dropped before the configuration file is read. In addition,
5969 logging is switched to stderr instead of the normal log files. If the
5970 configuration then re-defines the Exim user, the unprivileged environment
5971 is probably not what is expected, so Exim logs a panic warning message (but
5974 However, if deliver_drop_privilege is set, the unprivileged state may well
5975 be exactly what is intended, so the warning has been cut out in that case,
5976 and Exim is allowed to try to write to its normal log files.
5982 1. smtp_return_error_details was not giving details for temporary sender
5983 or receiver verification errors.
5985 2. Diagnose a configuration error if two authenticators have the same public
5988 3. Exim used not to create the message log file for a message until the first
5989 delivery attempt. This could be confusing when incoming messages were held
5990 for policy or load reasons. The message log file is now created at the time
5991 the message is received, and an initial "Received" line is written to it.
5993 4. The automatically generated man page for command line options had a minor
5994 bug that caused no ill effects; however, a more serious problem was that
5995 the procedure for building the man page automatically didn't always
5996 operate. Consequently, release 4.20 contains an out-of-date version. This
5997 shouldn't happen again.
5999 5. When building Exim with embedded Perl support, the script that builds the
6000 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6001 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6003 6. The freeze_tell option was not being used for messages that were frozen on
6004 arrival, either by an ACL or by local_scan().
6006 7. Added the smtp_incomplete_transaction log selector.
6008 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6009 was accepting AUTH without a new EHLO.
6011 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6012 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6015 10. Logging of TCP/IP connections (when configured) now happens in the main
6016 daemon process instead of the child process, so that the TCP/IP connection
6017 count is more accurate (but it can never be perfect).
6019 11. The use of "drop" in a nested ACL was not being handled correctly in the
6020 outer ACL. Now, if condition failure induced by the nested "drop" causes
6021 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6022 or "require"), the connection is dropped.
6024 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6025 that yield "discard" can now be used with an "accept" or a "discard" verb,
6026 but an error is generated for any others (because I can't see a useful way
6027 to define what should happen).
6029 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6030 are now processed in the same way as lines in the Exim configuration file.
6031 In particular, continuation lines are supported.
6033 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6035 15. Added -ti meaning -t -i.
6037 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6038 domains, and warn by logging if others are found.
6040 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6041 returned from GIFCONF when Exim is trying to find the list of interfaces on
6042 a host. The code in os.c has been modified to copy each ifreq to an aligned
6043 structure in all cases.
6045 Also, in some cases, the returned ifreq's were being copied to a 'struct
6046 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6047 means the last couple of bytes of an IPv6 address could be chopped if the
6048 ifreq contained only a normal sockaddr (14 bytes storage).
6050 18. Named domain lists were not supported in the hosts_treat_as_local option.
6051 An entry such as +xxxx was not recognized, and was treated as a literal
6054 19. Ensure that header lines added by a DATA ACL are included in the reject log
6055 if the ACL subsequently rejects the message.
6057 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6058 MD5 (which is deprecated).
6060 21. When testing a filter file using -bf, Exim was writing a message when it
6061 took the sender from a "From " line in the message, but it was not doing so
6062 when it took $return_path from a Return-Path: header line. It now does.
6064 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6065 with a valid header line field name (a series of printing characters
6066 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6068 23. Changed "disc" in the source to "disk" to conform to the documentation and
6069 the book and for uniformity.
6071 24. Ignore Sendmail's -Ooption=value command line item.
6073 25. When execve() failed while trying to run a command in a pipe transport,
6074 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6075 could be confused with a return value of 69 from the command itself. This
6076 has been changed to 127, the value the shell returns if it is asked to run
6077 a non-existent command. The wording for the related log line suggests a
6078 non-existent command as the problem.
6080 26. If received_header_text expands to an empty string, do not add a Received:
6081 header line to the message. (Well, it adds a token one on the spool, but
6082 marks it "old" so that it doesn't get used or transmitted.)
6084 27. Installed eximstats 1.28 (addition of -nt option).
6086 28. There was no check for failure on the call to getsockname() in the daemon
6087 code. This can fail if there is a shortage of resources on the system, with
6088 ENOMEM, for example. A temporary error is now given on failure.
6090 29. Contrary to the C standard, it seems that in some environments, the
6091 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6092 program. Exim now does this explicitly; it affects the formatting of
6093 timestamps using strftime().
6095 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6096 complaints. I've now initialized all the variables, to avoid this.
6098 32. Header lines added by a system filter were not being "seen" during
6099 transport-time rewrites.
6101 33. The info_callback() function passed to OpenSSL is set up with type void
6102 (*)(SSL *, int, int), as described somewhere. However, when calling the
6103 function (actually a macro) that sets it up, the type void(*)() is
6104 expected. I've put in a cast to prevent warnings from picky compilers.
6106 34. If a DNS black list lookup found a CNAME record, but there were no A
6107 records associated with the domain it pointed at, Exim crashed.
6109 35. If a DNS black list lookup returned more than one A record, Exim ignored
6110 all but the first. It now scans all returned addresses if a particular IP
6111 value is being sought. In this situation, the contents of the
6112 $dnslist_value variable are a list of all the addresses, separated by a
6115 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6116 to accept a host name and all its aliases if the forward lookup for any of
6117 them yielded the IP address of the incoming connection. Now it accepts only
6118 those names whose forward lookup yields the correct IP address. Any other
6119 names are discarded. This closes a loophole whereby a rogue DNS
6120 administrator could create reverse DNS records to break through a
6121 wildcarded host restriction in an ACL.
6123 37. If a user filter or a system filter that ran in a subprocess used any of
6124 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6125 the wrong values were passed to the pipe command ($thisaddress had the
6126 value of $0, $0 had the value of $1, etc). This bug was introduced by
6127 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6129 38. Improved the line breaking for long SMTP error messages from ACLs.
6130 Previously, if there was no break point between 40 and 75 characters, Exim
6131 left the rest of the message alone. Two changes have been made: (a) I've
6132 reduced the minimum length to 35 characters; (b) if it can't find a break
6133 point between 35 and 75 characters, it looks ahead and uses the first one
6134 that it finds. This may give the occasional overlong line, but at least the
6135 remaining text gets split now.
6137 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6138 file descriptors might be low, and that setting 1000 would always raise it.
6139 It turns out that in some environments, the limit is already over 1000 and
6140 that lowering it causes trouble. So now Exim takes care not to decrease it.
6142 40. When delivering a message, the value of $return_path is set to $sender_
6143 address at the start of routing (routers may change the value). By an
6144 oversight, this default was not being set up when an address was tested by
6145 -bt or -bv, which affected the outcome if any router or filter referred to
6148 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6149 log only when "message" or "log_message" are set. However, if one of the
6150 conditions was an address verification, or a call to a nested ACL, the
6151 messages generated by the underlying test were being passed through. This
6152 no longer happens. The underlying message is available in $acl_verify_
6153 message for both "message" and "log_message" expansions, so it can be
6154 passed through if needed.
6156 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6157 new expansion $bh_ to give the encoded byte string without charset
6158 translation. Translation happens only if iconv() is available; HAVE_ICONV
6159 indicates this at build time. HEADERS_CHARSET gives the charset to
6160 translate to; headers_charset can change it in the configuration, and
6161 "headers charset" can change it in an individual filter file.
6163 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6164 that creates RFC 2047 encoded "words" labels them as that charset instead
6165 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6166 expansion operator; (ii) when Exim creates a From: line for a local
6167 message; (iii) when a header line is rewritten to include a "phrase" part.
6169 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6170 buggy, causing it to skip the first lines of messages whose message ID
6171 ended in 'D'. This would not have bitten before Exim release 4.14, because
6172 message IDs were unlikely to end in 'D' before then. The effect was to have
6173 incorrect size information for certain domains.
6175 45. #include "config.h" was missing at the start of the crypt16.c module. This
6176 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6179 46. If there was a timeout during a "random" callout check, Exim treated it as
6180 a failure of the random address, and carried on sending RSET and the real
6181 address. If the delay was just some slowness somewhere, the response to the
6182 original RCPT would be taken as a response to RSET and so on, causing
6183 mayhem of various kinds.
6185 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6186 when I implemented it. It didn't allow for the fact that some option values
6187 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6188 the right test for positive values.
6190 48. Domain names in DNS records are case-independent. Exim always looks them up
6191 in lower case. Some resolvers return domain names in exactly the case they
6192 appear in the zone file, that is, they may contain uppercase letters. Not
6193 all resolvers do this - some return always lower case. Exim was treating a
6194 change of case by a resolver as a change of domain, similar to a widening
6195 of a domain abbreviation. This triggered its re-routing code and so it was
6196 trying to route what was effectively the same domain again. This normally
6197 caused routing to fail (because the router wouldn't handle the domain
6198 twice). Now Exim checks for this case specially, and just changes the
6199 casing of the domain that it ultimately uses when it transmits the message
6202 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6205 50. If a filter generated a file delivery with a non-absolute name (possible if
6206 no home directory exists for the router), the forbid_file option was not
6209 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6210 the existing equality matching.
6212 52. Exim was using ints instead of ino_t variables in some places where it was
6213 dealing with inode numbers.
6215 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6216 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6217 called TMPDIR, and if it finds it is different, it changes its value.
6219 54. The smtp_printf() function is now made available to local_scan() so
6220 additional output lines can be written before returning. There is also an
6221 smtp_fflush() function to enable the detection of a dropped connection.
6222 The variables smtp_input and smtp_batched_input are exported to
6225 55. Changed the default runtime configuration: the message "Unknown user"
6226 has been removed from the ACL, and instead placed on the localuser router,
6227 using the cannot_route_message feature. This means that any verification
6228 failures that generate their own messages won't get overridden. Similarly,
6229 the "Unrouteable address" message that was in the ACL for unverifiable
6230 relay addresses has also been removed.
6232 56. Added hosts_avoid_esmtp to the smtp transport.
6234 57. The exicyclog script was not checking for the esoteric option
6235 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6236 will work only if exicyclog is run under the appropriate euid.
6238 58. Following a discussion on the list, the rules by which Exim recognises line
6239 endings on incoming messages have been changed. The -dropcr and drop_cr
6240 options are now no-ops, retained only for backwards compatibility. The
6241 following line terminators are recognized: LF CRLF CR. However, special
6242 processing applies to CR:
6244 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6245 nor a local message in the state where . is a terminator.
6247 (ii) If a bare CR is encountered in a header line, an extra space is added
6248 after the line terminator so as not to end the header. The reasoning
6249 behind this is that bare CRs in header lines are most likely either
6250 to be mistakes, or people trying to play silly games.
6252 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6253 was being incorrectly given as 18 bytes larger than it should have been.
6254 This is a VOB (very old bug).
6256 60. This may never have affected anything current, but just in case it has:
6257 When the local host is found other than at the start of a list of hosts,
6258 the local host, those with the same MX, and any that follow, are discarded.
6259 When the list in question was part of a longer list of hosts, the following
6260 hosts (not currently being processed) were also being discarded. This no
6261 longer happens. I'm not sure if this situation could ever has previously
6264 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6267 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6268 header. This is something that is recommended in a new Internet Draft, and
6269 is something that is documented as being done by Sendmail. There are two
6270 possible values. For messages generated by the autoreply transport, Exim
6273 Auto-Submitted: auto-replied
6275 whereas for all other generated messages (e.g. bounces) it adds
6277 Auto-Submitted: auto-generated
6279 63. The "personal" condition in filters now includes a test for the
6280 Auto-Submitted: header. If it contains the string "auto-" the message it
6281 not considered personal.
6283 64. Added rcpt_include_affixes as a generic transport option.
6285 65. Added queue_only_override (default true).
6287 66. Added the syslog_duplication option.
6289 67. If what should have been the first header line of a message consisted of
6290 a space followed by a colon, Exim was mis-interpreting it as a header line.
6291 It isn't of course - it is syntactically invalid and should therefore be
6292 treated as the start of the message body. The misbehaviour could have
6293 caused a number of strange effects, including loss of data in subsequent
6294 header lines, and spool format errors.
6296 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6297 client host had authenticated. This control can now be exercised by an ACL
6298 for more flexibility.
6300 69. By default, callouts do not happen when testing with -bh. There is now a
6301 variant, -bhc, which does actually run the callout code, including
6302 consulting and updating the callout cache.
6304 70. Added support for saslauthd authentication, courtesy of Alexander
6307 71. If statvfs() failed on the spool or log directories while checking their
6308 size for availability, Exim confusingly gave the error "space shortage".
6309 Furthermore, in debugging mode it crashed with a floating point exception.
6310 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6311 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6312 serious problem, Exim now writes to the main and panic logs when this
6313 happens, with details of the failure. It then refuses to accept the
6314 incoming message, giving the message "spool directory problem" or "log
6315 directory problem" with a 421 code for SMTP messages.
6317 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6318 0, 1, and 2 exist, because some OS complain for execs without them (see
6319 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6320 descriptors. However, the code omitted to check that the open succeeded,
6321 causing mysterious errors if for some reason the permissions on /dev/null
6322 got screwed. Now Exim writes a message to the main and panic logs, and
6323 bombs out if it can't open /dev/null.
6325 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6326 interact so that it is all more flexible. It is supposed to remain
6327 backwards compatible. Also added extra_local_interfaces.
6329 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6330 to bomb out with an assertion failure - to the client this appears as a
6331 connection drop. This problem occurs in the part of the code that was taken
6332 from the Samba project. Fortunately, the assertion is in a very simple
6333 function, so I have fixed this by reproducing the function inline in the
6334 one place where it is called, and arranging for authentication to fail
6335 instead of killing the process with assert().
6337 75. The SPA client code was not working when the server requested OEM rather
6338 than Unicode encoding.
6340 76. Added code to make require_files with a specific uid setting more usable in
6341 the case where statting the file as root fails - usually a non-root-mounted
6342 NFS file system. When this happens and the failure is EACCES, Exim now
6343 forks a subprocess and does the per-uid checking as the relevant uid.
6345 77. Added process_log_path.
6347 78. If log_file_path was not explicitly set, a setting of check_log_space or
6348 check_log_inodes was ignored.
6350 79. If a space check for the spool or log partitions fails, the incident is now
6351 logged. Of course, in the latter case the data may get lost...
6353 80. Added the %p formatting code to string_format() so that it can be used to
6354 print addresses in debug_print(). Adjusted all the address printing in the
6355 debugging in store.c to use %p rather than %d.
6357 81. There was a concern that a line of code in smtp_in.c could overflow a
6358 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6359 initially expressed, the concern was not well-founded, because trailing
6360 spaces are removed early. However, if the trailing spaces were followed by
6361 a NULL, they did not get removed, so the overflow was possible. Two fixes
6364 (a) I re-wrote the offending code in a cleaner fashion.
6365 (b) If an incoming SMTP command contains a NULL character, it is rejected
6368 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6369 runs initgroups(), so that if the Exim user is in any additional groups,
6370 they will be used during message reception.
6376 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob