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.
37 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
40 JH/02 The smtp transport option "multi_domain" is now expanded.
42 JH/03 The smtp transport now requests PRDR by default, if the server offers
45 JH/04 Certificate name checking on server certificates, when exim is a client,
46 is now done by default. The transport option tls_verify_cert_hostnames
47 can be used to disable this per-host. The build option
48 EXPERIMENTAL_CERTNAMES is withdrawn.
50 JH/05 The value of the tls_verify_certificates smtp transport and main options
51 default to the word "system" to access the system default CA bundle.
52 For GnuTLS, only version 3.0.20 or later.
54 JH/06 Verification of the server certificate for a TLS connection is now tried
55 (but not required) by default. The verification status is now logged by
56 default, for both outbound TLS and client-certificate supplying inbound
59 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
62 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
63 Status Notification (bounce) messages are now MIME format per RFC 3464.
64 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
65 under the control of the dsn_advertise_hosts option, and routers may
66 have a dsn_lasthop option.
68 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
69 default, modifiable by a malware= option. The list separator for
70 the options can now be changed in the usual way. Bug 68.
72 JH/10 The smtp_receive_timeout main option is now expanded before use.
74 JH/11 The incoming_interface log option now also enables logging of the
75 local interface on delivery outgoing connections.
77 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
78 if the interface and destination host and port all match.
80 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
83 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
84 Patch from Andrew Lewis.
86 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
87 now supports optional time-restrictions, weighting, and priority
88 modifiers per server. Patch originally by <rommer@active.by>.
90 JH/16 The spamd_address main option now supports a mixed list of local
91 and remote servers. Remote servers can be IPv6 addresses, and
94 JH/17 Bug 68: The spamd_address main option now supports an optional
95 timeout value per server.
97 JH/18 Bug 1581: Router and transport options headers_add/remove can
98 now have the list separator specified.
100 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
103 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
106 JH/21 Support for the A6 type of dns record is withdrawn.
108 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
109 rather than the verbs used.
111 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
112 from 255 to 1024 chars.
114 JH/24 Verification callouts now attempt to use TLS by default.
116 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
117 are generic router options now. The defaults didn't change.
119 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
120 Original patch from Alexander Shikoff, worked over by JH.
122 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
123 files if ZCAT_COMMAND is not executable.
125 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
127 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
129 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
130 Normally benign, it bites when the pair was led to by a CNAME;
131 modern usage is to not canoicalize the domain to a CNAME target
132 (and we were inconsistent anyway for A-only vs AAAA+A).
134 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
136 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
137 when evaluating $sender_host_dnssec.
139 JH/31 Check the HELO verification lookup for DNSSEC, adding new
140 $sender_helo_dnssec variable.
142 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
144 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
146 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
148 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
149 documented as working, but never had. Support all but $spam_report.
151 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
152 added for tls authenticator.
157 TL/01 When running the test suite, the README says that variables such as
158 no_msglog_check are global and can be placed anywhere in a specific
159 test's script, however it was observed that placement needed to be near
160 the beginning for it to behave that way. Changed the runtest perl
161 script to read through the entire script once to detect and set these
162 variables, reset to the beginning of the script, and then run through
163 the script parsing/test process like normal.
165 TL/02 The BSD's have an arc4random API. One of the functions to induce
166 adding randomness was arc4random_stir(), but it has been removed in
167 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
168 function when detected.
170 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
171 cause callback expansion.
173 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
174 syntax errors in an expansion can be treated as a string instead of
175 logging or causing an error, due to the internal use of bool_lax
176 instead of bool when processing it.
178 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
179 server certificates when making smtp deliveries.
181 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
183 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
185 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
187 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
188 Merged patch from Sebastian Wiedenroth.
190 JH/05 Fix results-pipe from transport process. Several recipients, combined
191 with certificate use, exposed issues where response data items split
192 over buffer boundaries were not parsed properly. This eventually
193 resulted in duplicates being sent. This issue only became common enough
194 to notice due to the introduction of conection certificate information,
195 the item size being so much larger. Found and fixed by Wolfgang Breyha.
197 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
198 size buffer was used, resulting in syntax errors when an expansion
201 JH/07 Add support for directories of certificates when compiled with a GnuTLS
202 version 3.3.6 or later.
204 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
205 is EXPERIMENTAL_EVENT, the main-configuration and transport options
206 both become "event_action", the variables become $event_name, $event_data
207 and $event_defer_errno. There is a new variable $verify_mode, usable in
208 routers, transports and related events. The tls:cert event is now also
209 raised for inbound connections, if the main configuration event_action
212 TL/06 In test suite, disable OCSP for old versions of openssl which contained
213 early OCSP support, but no stapling (appears to be less than 1.0.0).
215 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
216 server certificate names available under the smtp transport option
217 "tls_verify_cert_hostname" now do not permit multi-component wildcard
220 JH/10 Time-related extraction expansions from certificates now use the main
221 option "timezone" setting for output formatting, and are consistent
222 between OpenSSL and GnuTLS compilations. Bug 1541.
224 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
225 encoded parameter in the incoming message. Bug 1558.
227 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
228 include certificate info, eximon was claiming there were spoolfile
231 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
233 JH/14 Log delivery-related information more consistently, using the sequence
234 "H=<name> [<ip>]" wherever possible.
236 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
237 are problematic for Debian distribution, omit them from the release
240 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
242 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
245 JH/17 Fix a null-indirection in certextract expansions when a nondefault
246 output list separator was used.
251 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
252 checkers that were complaining about end of non-void function with no
255 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
256 This was a regression intruduced in 4.83 by another bugfix.
258 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
260 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
261 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
267 TF/01 Correctly close the server side of TLS when forking for delivery.
269 When a message was received over SMTP with TLS, Exim failed to clear up
270 the incoming connection properly after forking off the child process to
271 deliver the message. In some situations the subsequent outgoing
272 delivery connection happened to have the same fd number as the incoming
273 connection previously had. Exim would try to use TLS and fail, logging
274 a "Bad file descriptor" error.
276 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
277 utilities have not been installed.
279 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
280 temporary space as the ACL may create new global variables.
282 TL/01 LDAP support uses per connection or global context settings, depending
283 upon the detected version of the libraries at build time.
285 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
286 to extract and use the src ip:port in logging and expansions as if it
287 were a direct connection from the outside internet. PPv2 support was
288 updated based on HAProxy spec change in May 2014.
290 JH/02 Add ${listextract {number}{list}{success}{fail}}.
292 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
293 Properly escape header and check for NULL return.
295 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
298 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
300 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
301 characters in header names, implemented as a verify condition.
302 Contributed by Michael Fischer v. Mollard.
304 TL/05 Rename SPF condition results err_perm and err_temp to standardized
305 results permerror and temperror. Previous values are deprecated but
306 still accepted. In a future release, err_perm and err_temp will be
307 completely removed, which will be a backward incompatibility if the
308 ACL tests for either of these two old results. Patch contributed by
309 user bes-internal on the mailing list.
311 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
313 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
314 selectors, in both main and reject logs.
316 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
319 JH/07 Add malware type "sock" for talking to simple daemon.
321 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
323 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
324 routers/transports under cutthrough routing.
326 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
327 numbers. Touch up "bool" conditional to keep the same definition.
329 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
331 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
333 JH/12 Expand items in router/transport headers_add or headers_remove lists
334 individually rather than the list as a whole. Bug 1452.
336 Required for reasonable handling of multiple headers_ options when
337 they may be empty; requires that headers_remove items with embedded
338 colons must have them doubled (or the list-separator changed).
340 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
341 view the policy declared in the DMARC record. Currently, $dmarc_status
342 is a combined value of both the record presence and the result of the
345 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
347 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
348 dnslookup router and the smtp transport (applying to the forward
351 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
352 of ldap servers used for a specific lookup. Patch provided by Heiko
355 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
356 New variable $lookup_dnssec_authenticated for observability.
358 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
359 Patch submitted by Lars Timman.
361 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
363 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
364 Requires trusted mode and valid format message id, aborts otherwise.
365 Patch contributed by Heiko Schlichting.
367 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
368 certextract with support for various fields. Bug 1358.
370 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
371 is requested by default, modifiable by smtp transport option
374 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
375 operate on certificate variables to give certificate fingerprints
376 Also new ${sha256:cert_variable}.
378 JH/23 The PRDR feature is moved from being Experimental into the mainline.
380 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
381 Christian Aistleitner.
383 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
385 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
386 file. Patch from Wolfgang Breyha.
388 JH/25 Expand the coverage of the delivery $host and $host_address to
389 client authenticators run in verify callout. Bug 1476.
391 JH/26 Port service names are now accepted for tls_on_connect_ports, to
392 align with daemon_smtp_ports. Bug 72.
394 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
395 support and error reporting did not work properly.
397 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
398 and is readable. Patch from Andrew Colin Kissa.
400 TL/14 Enhance documentation of ${run expansion and how it parses the
401 commandline after expansion, particularly in the case when an
402 unquoted variable expansion results in an empty value.
404 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
406 PP/02 Fix internal collision of T_APL on systems which support RFC3123
407 by renaming away from it. Addresses GH issue 15, reported by
410 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
412 TL/15 SECURITY: prevent double expansion in math comparison functions
413 (can expand unsanitized data). Not remotely exploitable.
420 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
422 PP/02 Make -n do something, by making it not do something.
423 When combined with -bP, the name of an option is not output.
425 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
428 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
429 $sender_host_name and config options to manage this, and basic check
432 PP/05 DSCP support for outbound connections and control modifier for inbound.
434 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
435 (Only plugin which currently uses this is kerberos4, which nobody should
436 be using, but we should make it available and other future plugins might
437 conceivably use it, even though it would break NAT; stuff *should* be
438 using channel bindings instead).
440 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
441 name; added for Sendmail compatibility; requires admin caller.
442 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
443 just ignore it); requires trusted caller.
444 Also parse but ignore: -Ac -Am -X<logfile>
447 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
449 TL/02 Add +smtp_confirmation as a default logging option.
451 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
452 Patch by Magnus Holmgren from 2007-02-20.
454 TL/04 Bugzilla 1281 - Spec typo.
455 Bugzilla 1283 - Spec typo.
456 Bugzilla 1290 - Spec grammar fixes.
458 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
460 TL/06 Add Experimental DMARC support using libopendmarc libraries.
462 TL/07 Fix an out of order global option causing a segfault. Reported to dev
463 mailing list by by Dmitry Isaikin.
465 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
467 JH/02 Support "G" suffix to numbers in ${if comparisons.
469 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
471 NM/01 Bugzilla 1197 - Spec typo
472 Bugzilla 1196 - Spec examples corrections
474 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
476 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
477 gnutls_enable_pkcs11, but renamed to more accurately indicate its
480 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
481 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
483 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
484 "acl {{name}{arg}...}", and optional args on acl condition
487 JH/05 Permit multiple router/transport headers_add/remove lines.
489 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
491 JH/07 Avoid using a waiting database for a single-message-only transport.
492 Performance patch from Paul Fisher. Bugzilla 1262.
494 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
497 JH/09 Add $headers_added variable, with content from use of ACL modifier
498 add_header (but not yet added to the message). Bugzilla 199.
500 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
501 Pulled from Bugzilla 817 by Wolfgang Breyha.
503 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
505 (nb: this is the same fix as in Exim 4.80.1)
507 JH/11 Add A= logging on delivery lines, and a client_set_id option on
510 JH/12 Add optional authenticated_sender logging to A= and a log_selector
513 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
515 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
516 advertise SMTP AUTH mechanism to us, instead of a generic
517 protocol violation error. Also, make Exim more robust to bad
518 data from the Dovecot auth socket.
520 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
522 When a queue runner is handling a message, Exim first routes the
523 recipient addresses, during which it prunes them based on the retry
524 hints database. After that it attempts to deliver the message to
525 any remaining recipients. It then updates the hints database using
528 So if a recipient address works intermittently, it can get repeatedly
529 deferred at routing time. The retry hints record remains fresh so the
530 address never reaches the final cutoff time.
532 This is a fairly common occurrence when a user is bumping up against
533 their storage quota. Exim had some logic in its local delivery code
534 to deal with this. However it did not apply to per-recipient defers
535 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
537 This change adds a proper retry rule check during routing so that the
538 final cutoff time is checked against the message's age. We only do
539 this check if there is an address retry record and there is not a
540 domain retry record; this implies that previous attempts to handle
541 the address had the retry_use_local_parts option turned on. We use
542 this as an approximation for the destination being like a local
543 delivery, as in LMTP.
545 I suspect this new check makes the old local delivery cutoff check
546 redundant, but I have not verified this so I left the code in place.
548 TF/02 Correct gecos expansion when From: is a prefix of the username.
550 Test 0254 submits a message to Exim with the header
554 When I ran the test suite under the user fanf2, Exim expanded
555 the header to contain my full name, whereas it should have added
556 a Resent-Sender: header. It erroneously treats any prefix of the
557 username as equal to the username.
559 This change corrects that bug.
561 GF/01 DCC debug and logging tidyup
562 Error conditions log to paniclog rather than rejectlog.
563 Debug lines prefixed by "DCC: " to remove any ambiguity.
565 TF/03 Avoid unnecessary rebuilds of lookup-related code.
567 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
568 Bug spotted by Jeremy Harris; was flawed since initial commit.
569 Would have resulted in OCSP responses post-SNI triggering an Exim
570 NULL dereference and crash.
572 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
574 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
575 Bug detection, analysis and fix by Samuel Thibault.
576 Bugzilla 1331, Debian bug #698092.
578 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
580 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
581 Server implementation by Todd Lyons, client by JH.
582 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
583 config variable "prdr_enable" controls whether the server
584 advertises the facility. If the client requests PRDR a new
585 acl_data_smtp_prdr ACL is called once for each recipient, after
586 the body content is received and before the acl_smtp_data ACL.
587 The client is controlled by bolth of: a hosts_try_prdr option
588 on the smtp transport, and the server advertisement.
589 Default client logging of deliveries and rejections involving
590 PRDR are flagged with the string "PRDR".
592 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
593 fclose(). Diagnosis by Todd Lyons.
595 PP/17 Update configure.default to handle IPv6 localhost better.
596 Patch by Alain Williams (plus minor tweaks).
599 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
600 This is now consistent with GnuTLS, and is now documented: the
601 previous undocumented portable approach to treating the option as
602 unset was to force an expansion failure. That still works, and
603 an empty string is now equivalent.
605 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
606 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
607 not performing validation itself.
609 PP/20 Added force_command boolean option to pipe transport.
610 Patch from Nick Koston, of cPanel Inc.
612 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
615 TF/04 Added udpsend ACL modifer and hexquote expansion operator
617 PP/21 Fix eximon continuous updating with timestamped log-files.
618 Broken in a format-string cleanup in 4.80, missed when I repaired the
619 other false fix of the same issue.
620 Report and fix from Heiko Schlichting.
623 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
624 Report from Prashanth Katuri.
626 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
627 It's SecureTransport, so affects any MacOS clients which use the
628 system-integrated TLS libraries, including email clients.
630 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
631 using a MIME ACL for non-SMTP local injection.
632 Report and assistance in diagnosis by Warren Baker.
634 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
636 JH/16 Fix comparisons for 64b. Bugzilla 1385.
638 TL/09 Add expansion variable $authenticated_fail_id to keep track of
639 last id that failed so it may be referenced in subsequent ACL's.
641 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
644 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
645 ldap library initialization, allowing self-signed CA's to be
646 used. Also properly sets require_cert option later in code by
647 using NULL (global ldap config) instead of ldap handle (per
648 session). Bug diagnosis and testing by alxgomz.
650 TL/12 Enhanced documentation in the ratelimit.pl script provided in
651 the src/util/ subdirectory.
653 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
654 renamed to Transport Post Delivery Action by Jeremy Harris, as
657 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
658 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
659 redis_servers = needs to be configured which will be used by the redis
660 lookup. Patch from Warren Baker, of The Packet Hub.
662 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
664 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
665 hostname or reverse DNS when processing a host list. Used suggestions
666 from multiple comments on this bug.
668 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
670 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
671 Missed a few lines, added it to make the runtest require no keyboard
674 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
675 contains upper case chars. Make router use caseful_local_part.
677 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
678 support when GnuTLS has been built with p11-kit.
684 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
686 This, or similar/improved, will also be change PP/11 of 4.82.
692 PP/01 Handle short writes when writing local log-files.
693 In practice, only affects FreeBSD (8 onwards).
694 Bugzilla 1053, with thanks to Dmitry Isaikin.
696 NM/01 Bugzilla 949 - Documentation tweak
698 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
701 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
703 PP/02 Implemented gsasl authenticator.
705 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
707 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
708 `pkg-config foo` for cflags/libs.
710 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
711 with rest of GSASL and with heimdal_gssapi.
713 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
714 `pkg-config foo` for cflags/libs for the TLS implementation.
716 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
717 properties get this fed in as external SSF. A number of robustness
718 and debugging improvements to the cyrus_sasl authenticator.
720 PP/08 cyrus_sasl server now expands the server_realm option.
722 PP/09 Bugzilla 1214 - Log authentication information in reject log.
723 Patch by Jeremy Harris.
725 PP/10 Added dbmjz lookup type.
727 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
729 PP/12 MAIL args handles TAB as well as SP, for better interop with
730 non-compliant senders.
731 Analysis and variant patch by Todd Lyons.
733 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
734 Bug report from Lars Müller <lars@samba.org> (via SUSE),
735 Patch from Dirk Mueller <dmueller@suse.com>
737 PP/13 tls_peerdn now print-escaped for spool files.
738 Observed some $tls_peerdn in wild which contained \n, which resulted
739 in spool file corruption.
741 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
742 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
743 or write after TLS renegotiation, which otherwise led to messages
746 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
747 as a tracking header (ie: a signed header comes before the signature).
748 Patch from Wolfgang Breyha.
750 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
751 comma-sep list; embedded commas doubled.
753 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
755 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
757 Report and patch from Dmitry Banschikov.
759 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
760 Removed SSL_clear() after SSL_new() which led to protocol negotiation
761 failures. We appear to now support TLS1.1+ with Exim.
763 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
764 lets Exim select keys and certificates based upon TLS SNI from client.
765 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
766 before an outbound SMTP session. New log_selector, +tls_sni.
768 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
769 NULL dereference. Report and patch from Alun Jones.
771 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
772 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
773 Not seeing resolver debug output on NetBSD, but suspect this is a
774 resolver implementation change.
776 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
777 Left warnings. Added "eximon gdb" invocation mode.
779 PP/21 Defaulting "accept_8bitmime" to true, not false.
781 PP/22 Added -bw for inetd wait mode support.
783 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
784 locate the relevant includes and libraries. Made this the default.
786 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
787 Bugzilla 1246, report and most of solution from Tomasz Kusy.
789 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
790 This may cause build issues on older platforms.
792 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
793 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
794 gnutls_require_mac & gnutls_require_protocols (no longer supported).
795 Added SNI support via GnuTLS too.
796 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
798 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
800 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
801 Applied second patch from Janne, implementing suggestion to default
802 multiple-strings-in-record handling to match SPF spec.
804 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
806 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
807 read-only, out of scope).
808 Patch from Wolfgang Breyha, report from Stuart Northfield.
810 PP/29 Fix three issues highlighted by clang analyser static analysis.
811 Only crash-plausible issue would require the Cambridge-specific
812 iplookup router and a misconfiguration.
813 Report from Marcin Mirosław.
815 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
817 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
818 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
819 As part of this, removing so much warning spew let me fix some minor
820 real issues in debug logging.
822 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
823 assignment on my part. Fixed.
825 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
826 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
827 Janne Snabb (who went above and beyond: thank you).
829 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
830 string otherwise requires a connection and a bunch more work and it's
831 relatively easy to get wrong. Should also expose TLS library linkage
834 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
835 64-bit ${eval} (JH/03).
837 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
838 GNU libc to support some of the 64-bit stuff, should not lead to
839 conflicts. Defined before os.h is pulled in, so if a given platform
840 needs to override this, it can.
842 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
843 protection layer was required, which is not implemented.
844 Bugzilla 1254, patch from Wolfgang Breyha.
846 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
847 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
848 tls_dhparam take prime identifiers. Also unbreak combination of
849 OpenSSL+DH_params+TLSSNI.
851 PP/39 Disable SSLv2 by default in OpenSSL support.
857 PP/01 Solaris build fix for Oracle's LDAP libraries.
858 Bugzilla 1109, patch from Stephen Usher.
860 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
862 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
865 TF/02 Fix a couple more cases where we did not log the error message
866 when unlink() failed. See also change 4.74-TF/03.
868 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
869 lock up or crash if it happened to be inside a call to libc when it
870 got a SIGUSR1 from exiwhat.
872 The SIGUSR1 handler appends the current process status to the process
873 log which is later printed by exiwhat. It used to use the general
874 purpose logging code to do this, but several functions it calls are
875 not safe for signals.
877 The new output code in the SIGUSR1 handler is specific to the process
878 log, and simple enough that it's easy to inspect for signal safety.
879 Removing some special cases also simplifies the general logging code.
880 Removing the spurious timestamps from the process log simplifies
883 TF/04 Improved ratelimit ACL condition.
885 The /noupdate option has been deprecated in favour of /readonly which
886 has clearer semantics. The /leaky, /strict, and /readonly update modes
887 are mutually exclusive. The update mode is no longer included in the
888 database key; it just determines when the database is updated. (This
889 means that when you upgrde Exim will forget old rate measurements.)
891 Exim now checks that the per_* options are used with an update mode that
892 makes sense for the current ACL. For example, when Exim is processing a
893 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
894 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
895 must specify per_mail/readonly. If you omit the update mode it defaults to
896 /leaky where that makes sense (as before) or /readonly where required.
898 The /noupdate option is now undocumented but still supported for
899 backwards compatibility. It is equivalent to /readonly except that in
900 ACLs where /readonly is required you may specify /leaky/noupdate or
901 /strict/noupdate which are treated the same as /readonly.
903 A useful new feature is the /count= option. This is a generalization
904 of the per_byte option, so that you can measure the throughput of other
905 aggregate values. For example, the per_byte option is now equivalent
906 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
908 The per_rcpt option has been generalized using the /count= mechanism
909 (though it's more complicated than the per_byte equivalence). When it is
910 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
911 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
912 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
913 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
914 Note that using per_rcpt with a non-readonly update mode in more than
915 one ACL will cause the recipients to be double-counted. (The per_mail
916 and per_byte options don't have this problem.)
918 The handling of very low rates has changed slightly. If the computed rate
919 is less than the event's count (usually one) then this event is the first
920 after a long gap. In this case the rate is set to the same as this event's
921 count, so that the first message of a spam run is counted properly.
923 The major new feature is a mechanism for counting the rate of unique
924 events. The new per_addr option counts the number of different
925 recipients that someone has sent messages to in the last time period. It
926 behaves like per_rcpt if all the recipient addresses are different, but
927 duplicate recipient addresses do not increase the measured rate. Like
928 the /count= option this is a general mechanism, so the per_addr option
929 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
930 example, measure the rate that a client uses different sender addresses
931 with the options per_mail/unique=$sender_address. There are further
932 details in the main documentation.
934 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
936 TF/06 Removed a few PCRE remnants.
938 TF/07 Automatically extract Exim's version number from tags in the git
939 repository when doing development or release builds.
941 PP/02 Raise smtp_cmd_buffer_size to 16kB.
942 Bugzilla 879. Patch from Paul Fisher.
944 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
945 Heavily based on revision 40f9a89a from Simon Arlott's tree.
948 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
950 PP/05 Variable $av_failed, true if the AV scanner deferred.
951 Bugzilla 1078. Patch from John Horne.
953 PP/06 Stop make process more reliably on build failure.
954 Bugzilla 1087. Patch from Heiko Schlittermann.
956 PP/07 Make maildir_use_size_file an _expandable_ boolean.
957 Bugzilla 1089. Patch from Heiko Schlittermann.
959 PP/08 Handle ${run} returning more data than OS pipe buffer size.
960 Bugzilla 1131. Patch from Holger Weiß.
962 PP/09 Handle IPv6 addresses with SPF.
963 Bugzilla 860. Patch from Wolfgang Breyha.
965 PP/10 GnuTLS: support TLS 1.2 & 1.1.
967 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
970 PP/11 match_* no longer expand right-hand-side by default.
971 New compile-time build option, EXPAND_LISTMATCH_RHS.
972 New expansion conditions, "inlist", "inlisti".
974 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
976 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
978 PP/14 fix log_write() format string regression from TF/03.
979 Bugzilla 1152. Patch from Dmitry Isaikin.
985 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
987 PP/02 Harmonised TLS library version reporting; only show if debugging.
988 Layout now matches that introduced for other libraries in 4.74 PP/03.
990 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
992 PP/04 New "dns_use_edns0" global option.
994 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
997 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
998 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1000 TK/01 Updated PolarSSL code to 0.14.2.
1001 Bugzilla 1097. Patch from Andreas Metzler.
1003 PP/07 Catch divide-by-zero in ${eval:...}.
1004 Fixes bugzilla 1102.
1006 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1009 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1010 format-string attack -- SECURITY: remote arbitrary code execution.
1012 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1013 time unintentionally subject to list matching rules, letting the header
1014 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1015 arbitrary string expansion). This allowed for information disclosure.
1017 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1018 INT_MIN/-1 -- value coerced to INT_MAX.
1024 NM/01 Workround for PCRE version dependency in version reporting
1027 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1028 This fixes portability to compilers other than gcc, notably
1029 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1031 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1032 makefiles for portability to HP-UX and POSIX correctness.
1034 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1035 Also via indented variable definition in the Makefile.
1036 (Debugging by Oliver Heesakkers).
1038 PP/02 Restore caching of spamd results with expanded spamd_address.
1039 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1041 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1042 Improves build reliability. Fix from: Frank Elsner
1044 NM/02 Fix wide character breakage in the rfc2047 coding
1045 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1047 NM/03 Allow underscore in dnslist lookups
1048 Fixes bug 1026. Patch from Graeme Fowler
1050 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1051 Code patches from Adam Ciarcinski of NetBSD.
1053 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1056 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1057 is logged, to avoid truncation. Patch from John Horne.
1059 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1060 Patch from Jakob Hirsch.
1062 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1063 SQL string expansion failure details.
1064 Patch from Andrey Oktyabrski.
1066 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1067 Patch from Simon Arlott.
1069 PP/09 New lookups functionality failed to compile on old gcc which rejects
1070 extern declarations in function scope.
1071 Patch from Oliver Fleischmann
1073 PP/10 Use sig_atomic_t for flags set from signal handlers.
1074 Check getgroups() return and improve debugging.
1075 Fixed developed for diagnosis in bug 927 (which turned out to be
1078 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1079 Patch from Mark Zealey.
1081 PP/12 Bugzilla 1056: Improved spamd server selection.
1082 Patch from Mark Zealey.
1084 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1085 Based on patch from Heiko Schlittermann.
1087 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1088 Patch from Uwe Doering, sign-off by Michael Haardt.
1090 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1091 variable declaration deep within a block. Bug and patch from
1094 PP/15 lookups-Makefile IRIX compatibilty coercion.
1096 PP/16 Make DISABLE_DKIM build knob functional.
1098 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1099 Patch by Simon Arlott
1101 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1102 variable argument macros. Our copy now differs from upstream.
1108 TF/01 Failure to get a lock on a hints database can have serious
1109 consequences so log it to the panic log.
1111 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1112 controlled using the smtp_confirmation log selector.
1114 TF/03 Include the error message when we fail to unlink a spool file.
1116 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1117 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1118 for maintaining out-of-tree patches for some time.
1120 PP/01 Bugzilla 139: Documentation and portability issues.
1121 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1122 Handle per-OS dynamic-module compilation flags.
1124 PP/02 Let /dev/null have normal permissions.
1125 The 4.73 fixes were a little too stringent and complained about the
1126 permissions on /dev/null. Exempt it from some checks.
1127 Reported by Andreas M. Kirchwitz.
1129 PP/03 Report version information for many libraries, including
1130 Exim version information for dynamically loaded libraries. Created
1131 version.h, now support a version extension string for distributors
1132 who patch heavily. Dynamic module ABI change.
1134 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1135 privilege escalation vulnerability whereby the Exim run-time user
1136 can cause root to append content of the attacker's choosing to
1139 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1142 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1143 If dropping privileges for untrusted macros, we disabled normal logging
1144 on the basis that it would fail; for the Exim run-time user, this is not
1145 the case, and it resulted in successful deliveries going unlogged.
1146 Fixed. Reported by Andreas Metzler.
1152 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1153 only prepend for the Resent-* case. Fixes regression introduced in
1154 Exim 4.70 by NM/22 for Bugzilla 607.
1156 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1157 increasing numbers of administrators be bitten by this.
1159 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1161 PP/03 Bugzilla 994: added openssl_options main configuration option.
1163 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1165 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1167 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1169 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1170 without a peer certificate, leading to a segfault because of an
1171 assumption that peers always have certificates. Be a little more
1172 paranoid. Problem reported by Martin Tscholak.
1174 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1175 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1176 NB: ClamAV planning to remove STREAM in "middle of 2010".
1177 CL also introduces -bmalware, various -d+acl logging additions and
1178 more caution in buffer sizes.
1180 PP/09 Implemented reverse_ip expansion operator.
1182 PP/10 Bugzilla 937: provide a "debug" ACL control.
1184 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1186 PP/12 Bugzilla 973: Implement --version.
1188 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1190 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1192 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1194 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1195 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1196 ignore trailing whitespace.
1198 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1200 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1203 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1204 Notification from Dr Andrew Aitchison.
1206 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1207 ExtendedDetectionInfo response format.
1208 Notification from John Horne.
1210 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1213 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1214 XSL and documented dependency on system catalogs, with examples of how
1217 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1220 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1221 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1222 configuration file which is writeable by the Exim user or group.
1224 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1225 of configuration files to cover files specified with the -C option if
1226 they are going to be used with root privileges, not just the default
1229 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1230 option (effectively making it always true).
1232 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1233 files to be used while preserving root privileges.
1235 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1236 that rogue child processes cannot use them.
1238 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1239 run-time user, instead of root.
1241 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1242 Exim run-time user without dropping privileges.
1244 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1245 result string, instead of calling string_vformat() twice with the same
1248 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1249 for other users. Others should always drop root privileges if they use
1250 -C on the command line, even for a whitelisted configure file.
1252 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1254 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1260 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1261 $data_path, and $header_path variables; fixed documentation bugs and
1264 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1265 exipick to access non-standard spools, including the "frozen" queue
1268 NM/01 Bugzilla 965: Support mysql stored procedures.
1269 Patch from Alain Williams
1271 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1273 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1274 Patch from Andreas Metzler
1276 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1277 Patch from Kirill Miazine
1279 NM/05 Bugzilla 671: Added umask to procmail example.
1281 JJ/03 installed exipick 20100323.0, fixing doc bug
1283 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1284 directory. Notification and patch from Dan Rosenberg.
1286 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1288 TK/02 Improve log output when DKIM signing operation fails.
1290 MH/01 Treat the transport option dkim_domain as a colon separated
1291 list, not as a single string, and sign the message with each element,
1292 omitting multiple occurences of the same signer.
1294 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1295 Bugzilla 985, 986. Patch by Simon Arlott
1297 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1298 Patch by Simon Arlott
1300 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1301 MBX locking. Notification from Dan Rosenberg.
1307 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1309 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1311 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1313 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1315 NM/04 Bugzilla 915: Flag broken perl installation during build.
1321 TK/01 Added patch by Johannes Berg that expands the main option
1322 "spamd_address" if it starts with a dollar sign.
1324 TK/02 Write list of recipients to X-Envelope-Sender header when building
1325 the mbox-format spool file for content scanning (suggested by Jakob
1328 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1329 (http://www.dcc-servers.net/) support via dccifd. Activated by
1330 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1332 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1333 by Mark Daniel Reidel <mr@df.eu>.
1335 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1336 When building exim an external PCRE library is now needed -
1337 PCRE is a system library on the majority of modern systems.
1338 See entry on PCRE_LIBS in EDITME file.
1340 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1341 conversation. Added nologin parameter to request.
1342 Patch contributed by Kirill Miazine.
1344 TF/01 Do not log submission mode rewrites if they do not change the address.
1346 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1348 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1349 log files in place. Contributed by Roberto Lima.
1351 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1353 TF/03 Bugzilla 615: When checking the local_parts router precondition
1354 after a local_part_suffix or local_part_prefix option, Exim now
1355 does not use the address's named list lookup cache, since this
1356 contains cached lookups for the whole local part.
1358 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1359 Robert Millan. Documentation is in experimental-spec.txt.
1361 TF/04 Bugzilla 668: Fix parallel build (make -j).
1363 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1365 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1366 Patch provided by Jan Srzednicki.
1368 TF/05 Leading white space used to be stripped from $spam_report which
1369 wrecked the formatting. Now it is preserved.
1371 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1372 that they are available at delivery time.
1374 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1376 TF/08 TLS error reporting now respects the incoming_interface and
1377 incoming_port log selectors.
1379 TF/09 Produce a more useful error message if an SMTP transport's hosts
1380 setting expands to an empty string.
1382 NM/06 Bugzilla 744: EXPN did not work under TLS.
1383 Patch provided by Phil Pennock.
1385 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1386 Patch provided by Richard Godbee.
1388 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1389 acl_smtp_notquit, added index entry.
1391 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1392 Patch provided by Eugene Bujak.
1394 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1395 accept(). Patch provided by Maxim Dounin.
1397 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1398 Patch provided by Phil Pennock.
1400 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1402 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1403 Patch provided by Brad "anomie" Jorsch.
1405 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1406 Patch provided by Dean Brooks.
1408 TK/05 Add native DKIM support (does not depend on external libraries).
1410 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1411 Patch provided by Graeme Fowler.
1413 NM/16 Bugzilla 851: Documentation example syntax fix.
1415 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1417 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1420 NM/19 Bugzilla 745: TLS version reporting.
1421 Patch provided by Phil Pennock.
1423 NM/20 Bugzilla 167: bool: condition support.
1424 Patch provided by Phil Pennock.
1426 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1427 clients. Patch provided by Phil Pennock.
1429 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1430 Patch provided by Brad "anomie" Jorsch.
1432 NM/23 Bugzilla 687: Fix misparses in eximstats.
1433 Patch provided by Heiko Schlittermann.
1435 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1436 Patch provided by Heiko Schlittermann.
1438 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1439 plus update to original patch.
1441 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1443 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1444 Patch provided by David Brownlee.
1446 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1448 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1450 NM/30 Bugzilla 888: TLS documentation bugfixes.
1452 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1454 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1455 Unlike the original bugzilla I have changed all shell scripts in src tree.
1457 NM/33 Bugzilla 898: Transport filter timeout fix.
1458 Patch by Todd Rinaldo.
1460 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1461 Patch by Serge Demonchaux.
1463 NM/35 Bugzilla 39: Base64 decode bug fixes.
1464 Patch by Jakob Hirsch.
1466 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1468 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1470 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1472 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1478 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1479 ALT-N's libdkim that I have put here:
1480 http://duncanthrax.net/exim-experimental/
1482 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1483 were called 'true' and it seems that C99 defines that as a reserved
1484 keyword to be used with 'bool' variable types. That means you could
1485 not include C99-style headers which use bools without triggering
1486 build errors in sieve.c.
1488 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1489 as mailq or other aliases. Changed the --help handling significantly
1490 to do whats expected. exim_usage() emits usage/help information.
1492 SC/01 Added the -bylocaldomain option to eximstats.
1494 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1496 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1498 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1504 PH/01 Another patch from the Sieve maintainer.
1506 PH/02 When an IPv6 address is converted to a string for single-key lookup
1507 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1508 dots are used instead of colons so that keys in lsearch files need not
1509 contain colons. This was done some time before quoting was made available
1510 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1511 (notated using the quote facility) so as to distinguish them from IPv4
1512 keys. This meant that lookups for IP addresses in host lists did not work
1513 for iplsearch lookups.
1515 This has been fixed by arranging for IPv6 addresses to be expressed with
1516 colons if the lookup type is iplsearch. This is not incompatible, because
1517 previously such lookups could never work.
1519 The situation is now rather anomolous, since one *can* have colons in
1520 ordinary lsearch keys. However, making the change in all cases is
1521 incompatible and would probably break a number of configurations.
1523 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1526 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1527 conversion specification without a maximum field width, thereby enabling
1528 a rogue spamd server to cause a buffer overflow. While nobody in their
1529 right mind would setup Exim to query an untrusted spamd server, an
1530 attacker that gains access to a server running spamd could potentially
1531 exploit this vulnerability to run arbitrary code as the Exim user.
1533 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1534 $primary_hostname instead of what libspf2 thinks the hosts name is.
1536 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1537 a directory entry by the name of the lookup key. Previously, if a
1538 symlink pointed to a non-existing file or a file in a directory that
1539 Exim lacked permissions to read, a lookup for a key matching that
1540 symlink would fail. Now it is enough that a matching directory entry
1541 exists, symlink or not. (Bugzilla 503.)
1543 PH/03 The body_linecount and body_zerocount variables are now exported in the
1546 PH/04 Added the $dnslist_matched variable.
1548 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1549 This means they are set thereafter only if the connection becomes
1552 PH/06 Added the client_condition to authenticators so that some can be skipped
1553 by clients under certain conditions.
1555 PH/07 The error message for a badly-placed control=no_multiline_responses left
1556 "_responses" off the end of the name.
1558 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1560 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1561 (without spaces) instead of just copying the configuration text.
1563 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1565 PH/11 Added $max_received_linelength.
1567 PH/12 Added +ignore_defer and +include_defer to host lists.
1569 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1570 way in which PCRE > 7.0 is built.
1572 PH/14 Implemented queue_only_load_latch.
1574 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1575 MAIL command. The effect was to mangle the value on 64-bit systems.
1577 PH/16 Another patch from the Sieve maintainer.
1579 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1581 PH/18 If a system quota error occurred while trying to create the file for
1582 a maildir delivery, the message "Mailbox is full" was not appended to the
1583 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1584 only to a quota excession during the actual writing of the file.
1586 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1587 characters?) which causes problems in log lines. The DN values are now
1588 passed through string_printing() before being added to log lines.
1590 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1591 and InterBase are left for another time.)
1593 PH/21 Added message_body_newlines option.
1595 PH/22 Guard against possible overflow in moan_check_errorcopy().
1597 PH/23 POSIX allows open() to be a macro; guard against that.
1599 PH/24 If the recipient of an error message contained an @ in the local part
1600 (suitably quoted, of course), incorrect values were put in $domain and
1601 $local_part during the evaluation of errors_copy.
1607 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1608 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1611 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1612 issue a MAIL command.
1614 PH/02 In an ACL statement such as
1616 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1618 if a client was not listed at all, or was listed with a value other than
1619 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1620 the condition was not true (as it should be), so access was not denied.
1621 The bug was that the ! inversion was incorrectly passed on to the second
1622 item. This has been fixed.
1624 PH/03 Added additional dnslists conditions == and =& which are different from
1625 = and & when the dns lookup returns more than one IP address.
1627 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1628 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1630 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1631 FSYNC, which compiles an option called disable_fsync that allows for
1632 bypassing fsync(). The documentation is heavily laced with warnings.
1634 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1636 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1637 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1638 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1639 including adding "make clean"; (3) Added -fPIC when compiling the test
1640 dynamically loaded module, to get rid of a warning.
1642 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1643 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1644 = 0s. The bug is otherwise harmless.
1646 PH/07 There was a bug in the dovecot authenticator such that the value of
1647 $auth1 could be overwritten, and so not correctly preserved, after a
1648 successful authentication. This usually meant that the value preserved by
1649 the server_setid option was incorrect.
1651 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1653 PH/09 Installed PCRE release 7.0.
1655 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1656 run for batched SMTP input. It is now run at the start of every message
1657 in the batch. While fixing this I discovered that the process information
1658 (output by running exiwhat) was not always getting set for -bs and -bS
1659 input. This is fixed, and it now also says "batched" for BSMTP.
1661 PH/11 Added control=no_pipelining.
1663 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1664 patch, slightly modified), and move the expansion of helo_data till after
1665 the connection is made in the smtp transport (so it can use these
1668 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1670 PH/14 Added log_selector = +pid.
1672 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1674 PH/16 Add ${if forany and ${if forall.
1676 PH/17 Added dsn_from option to vary the From: line in DSNs.
1678 PH/18 Flush SMTP output before performing a callout, unless control =
1679 no_callout_flush is set.
1681 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1682 was true (the default) a successful delivery failed to delete the retry
1683 item, thus causing premature timeout of the address. The bug is now
1686 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1688 PH/21 Long custom messages for fakedefer and fakereject are now split up
1689 into multiline reponses in the same way that messages for "deny" and
1690 other ACL rejections are.
1692 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1693 with slight modification.
1695 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1696 draft, changing the syntax and factoring some duplicate code".
1698 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1699 for deliveries of the second and subsequent messages over the same SMTP
1702 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1703 ${reduce, with only minor "tidies".
1705 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1707 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1708 expansion side effects.
1710 PH/27 When a message times out after an over-quota error from an Exim-imposed
1711 quota, the bounce message says "mailbox is full". This message was not
1712 being given when it was a system quota that was exceeded. It now should
1715 MH/03 Made $recipients available in local_scan(). local_scan() already has
1716 better access to the recipient list through recipients_list[], but
1717 $recipients can be useful in postmaster-provided expansion strings.
1719 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1720 in the case of a MAIL command with additional options following the
1721 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1722 were accidentally chopped off.
1724 PH/29 SMTP synchronization checks are implemented when a command is read -
1725 there is a check that no more input is waiting when there shouldn't be
1726 any. However, for some commands, a delay in an ACL can mean that it is
1727 some time before the response is written. In this time, more input might
1728 arrive, invalidly. So now there are extra checks after an ACL has run for
1729 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1730 pipelining has not been advertised.
1732 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1734 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1735 returned in the SMTP response when smtp_return_error_details was false.
1736 This has been fixed.
1738 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1739 instead of the C I/O that was originally supplied, because problems were
1740 reported on Solaris.
1742 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1743 Exim which did not show up earlier: it was assuming that a call to
1744 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1745 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1746 was a macro that became an assignment, so it seemed to work. This has
1747 changed to a proper function call with a void return, hence the compile
1748 error. Exim's code has been fixed.
1750 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1753 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1755 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1757 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1758 because it turns out to be too big or there is a timeout) there is no
1759 "Completed" line in the log. When some messages of this type were
1760 selected by exigrep, they were listed as "not completed". Others were
1761 picked up by some special patterns. I have improved the selection
1762 criteria to be more general.
1764 PH/38 The host_find_failed option in the manualroute router can now be set
1765 to "ignore", to completely ignore a host whose IP address cannot be
1766 found. If all hosts are ignored, the behaviour is controlled by the new
1767 host_all_ignored option.
1769 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1770 homing or because of multiple MX records with /mx) generated more than
1771 one IP address, and the following item turned out to be the local host,
1772 all the secondary addresses of the first item were incorrectly removed
1773 from the list, along with the local host and any following hosts (which
1774 is what is supposed to happen).
1776 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1777 whoever called Exim into the -H file. In the case of the daemon it was
1778 behaving confusingly. When first started, it used values for whoever
1779 started the daemon, but after a SIGHUP it used the Exim user (because it
1780 calls itself on a restart). I have changed the code so that it now always
1783 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1784 message are rejected with the same error (e.g. no authentication or bad
1785 sender address), and a DATA command is nevertheless sent (as can happen
1786 with PIPELINING or a stupid MUA), the error message that was given to the
1787 RCPT commands is included in the rejection of the DATA command. This is
1788 intended to be helpful for MUAs that show only the final error to their
1791 PH/42 Another patch from the Sieve maintainer.
1793 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1794 Eximstats - Fixed some broken HTML links and added missing column headers
1796 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1797 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1799 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1801 PH/43 Yet another patch from the Sieve maintainer.
1803 PH/44 I found a way to check for a TCP/IP connection going away before sending
1804 the response to the final '.' that terminates a message, but only in the
1805 case where the client has not sent further data following the '.'
1806 (unfortunately, this is allowed). However, in many cases there won't be
1807 any further data because there won't be any more messages to send. A call
1808 to select() can be used: if it shows that the input is "ready", there is
1809 either input waiting, or the socket has been closed. An attempt to read
1810 the next input character can distinguish the two cases. Previously, Exim
1811 would have sent an OK response which the client would never have see.
1812 This could lead to message repetition. This fix should cure that, at
1813 least in a lot of common cases.
1815 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1816 advertised in response to EHLO.
1822 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1823 fixed by 4.65/MH/01 (is this a record?) are fixed:
1825 (i) An empty string was always treated as zero by the numeric comparison
1826 operators. This behaviour has been restored.
1828 (ii) It is documented that the numeric comparison operators always treat
1829 their arguments as decimal numbers. This was broken in that numbers
1830 starting with 0 were being interpreted as octal.
1832 While fixing these problems I realized that there was another issue that
1833 hadn't been noticed. Values of message_size_limit (both the global option
1834 and the transport option) were treated as octal if they started with 0.
1835 The documentation was vague. These values are now always treated as
1836 decimal, and I will make that clear in the documentation.
1842 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1843 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1846 MH/01 Don't check that the operands of numeric comparison operators are
1847 integers when their expansion is in "skipping" mode (fixes bug
1848 introduced by 4.64-PH/07).
1850 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1851 child addresses, Exim now panics and dies. Previously, because the count
1852 is held in a short int, deliveries were likely to be lost. As such a
1853 large number of recipients for a single message is ridiculous
1854 (performance will be very, very poor), I have chosen to impose a limit
1855 rather than extend the field.
1861 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1862 leftover -K file (the existence of which was triggered by #402).
1863 While we were at it, introduced process PID as part of the -K
1864 filename. This should rule out race conditions when creating
1867 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1868 processing considerably. Previous code took too long for large mails,
1869 triggering a timeout which in turn triggers #401.
1871 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1872 in the DK code in transports.c. sendfile() is not really portable,
1873 hence the _LINUX specificness.
1875 TF/01 In the add_headers option to the mail command in an Exim filter,
1876 there was a bug that Exim would claim a syntax error in any
1877 header after the first one which had an odd number of characters
1880 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1881 callout verification, Exim cached a "reject" for the entire domain. This
1882 is correct for most verifications, but it is not correct for a recipient
1883 verification with use_sender or use_postmaster set, because in that case
1884 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1885 case of MAIL FROM:<> rejection from other early rejections (e.g.
1886 rejection of HELO). When verifying a recipient using a non-null MAIL
1887 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1888 Whatever the result of the callout, the value of the domain cache is
1889 left unchanged (for any other kind of callout, getting as far as trying
1890 RCPT means that the domain itself is ok).
1892 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1895 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1896 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1897 session, the daemon ignores ECONNECT errors and logs others; it now
1898 ignores EPIPE as well.
1900 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1901 (quoted-printable decoding).
1903 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1904 later the small subsequent patch to fix an introduced bug.
1906 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1908 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1910 PH/08 An error is now given if message_size_limit is specified negative.
1912 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1913 to be given (somewhat) arbitrary names.
1915 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1918 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1919 miscellaneous code fixes
1921 PH/10 Added the log_reject_target ACL modifier to specify where to log
1924 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1925 hostname. This is wrong, because it relates to the incoming message (and
1926 probably the interface on which it is arriving) and not to the outgoing
1927 callout (which could be using a different interface). This has been
1928 changed to use the value of the helo_data option from the smtp transport
1929 instead - this is what is used when a message is actually being sent. If
1930 there is no remote transport (possible with a router that sets up host
1931 addresses), $smtp_active_hostname is used.
1933 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1934 tweaks were necessary in order to get it to work (see also 21 below):
1935 (a) The code assumed that strncpy() returns a negative number on buffer
1936 overflow, which isn't the case. Replaced with Exim's string_format()
1938 (b) There were several signed/unsigned issues. I just did the minimum
1939 hacking in of casts. There is scope for a larger refactoring.
1940 (c) The code used strcasecmp() which is not a standard C function.
1941 Replaced with Exim's strcmpic() function.
1942 (d) The code set only $1; it now sets $auth1 as well.
1943 (e) A simple test gave the error "authentication client didn't specify
1944 service in request". It would seem that Dovecot has changed its
1945 interface. Fortunately there's a specification; I followed it and
1946 changed what the client sends and it appears to be working now.
1948 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1951 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1952 address A is aliased to B and C, where B exists and C does not. Without
1953 -v the output is "A verified" because verification stops after a
1954 successful redirection if more than one address is generated. However,
1955 with -v the child addresses are also verified. Exim was outputting "A
1956 failed to verify" and then showing the successful verification for C,
1957 with its parentage. It now outputs "B failed to verify", showing B's
1958 parentage before showing the successful verification of C.
1960 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1961 look up a TXT record in a specific list after matching in a combined
1964 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1965 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1966 they consult the DNS. I had assumed they would set it the way they
1967 wanted; and indeed my experiments on Linux seem to show that in some
1968 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1969 To be on the safe side, however, I have now made the interface to
1970 host_find_byname() similar to host_find_bydns(), with an argument
1971 containing the DNS resolver options. The host_find_byname() function now
1972 sets these options at its start, just as host_find_bydns() does. The smtp
1973 transport options dns_qualify_single and dns_search_parents are passed to
1974 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1975 of host_find_byname() use the default settings of RES_DEFNAMES
1976 (qualify_single) but not RES_DNSRCH (search_parents).
1978 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1979 spool_read_header() do less string testing, by means of a preliminary
1980 switch on the second character of optional "-foo" lines. (This is
1981 overdue, caused by the large number of possibilities that now exist.
1982 Originally there were few.) While I was there, I also converted the
1983 str(n)cmp tests so they don't re-test the leading "-" and the first
1984 character, in the hope this might squeeze out yet more improvement.
1986 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1987 flag allowing group syntax was set by the header_syntax check but not
1988 turned off, possible causing trouble later; (2) The flag was not being
1989 set at all for the header_verify test, causing "group"-style headers to
1990 be rejected. I have now set it in this case, and also caused header_
1991 verify to ignore an empty address taken from a group. While doing this, I
1992 came across some other cases where the code for allowing group syntax
1993 while scanning a header line wasn't quite right (mostly, not resetting
1994 the flag correctly in the right place). These bugs could have caused
1995 trouble for malformed header lines. I hope it is now all correct.
1997 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1998 with the "reply" argument non-NULL. The code, however (which originally
1999 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2000 but it didn't always do it. This confused somebody who was copying the
2001 code for some other use. I have removed all the tests.
2003 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2004 feature that was used to support insecure browsers during the U.S. crypto
2005 embargo. It requires special client support, and Exim is probably the
2006 only MTA that supported it -- and would never use it because real RSA is
2007 always available. This code has been removed, because it had the bad
2008 effect of slowing Exim down by computing (never used) parameters for the
2009 RSA_EXPORT functionality.
2011 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2012 authenticator to fail if there's a tab character in the incoming data
2013 (there should never be unless someone is messing about, as it's supposed
2014 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2015 if the connection is using TLS or if the remote IP is the same as the
2016 local IP, and the "valid-client-cert option" if a client certificate has
2019 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2020 authenticators. This can be used for authorization after authentication
2021 succeeds. (In the case of plaintext, it servers for both authentication
2024 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2025 if any retry times were supplied.
2027 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2028 connection, where there is no client IP address to check. In this
2029 situation, the verify now always succeeds.
2031 PH/25 Applied John Jetmore's -Mset patch.
2033 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2035 PH/27 In a string expansion for a processed (not raw) header when multiple
2036 headers of the same name were present, leading whitespace was being
2037 removed from all of them, but trailing whitespace was being removed only
2038 from the last one. Now trailing whitespace is removed from each header
2039 before concatenation. Completely empty headers in a concatenation (as
2040 before) are ignored.
2042 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2043 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2045 PH/29 [Removed. This was a change that I later backed out, and forgot to
2046 correct the ChangeLog entry (that I had efficiently created) before
2047 committing the later change.]
2049 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2050 address errors (4xx response to RCPT) over the same connection as other
2051 messages routed to the same hosts. Such deliveries are always "forced",
2052 so retry times are not inspected. This resulted in far too many retries
2053 for the affected addresses. The effect occurred only when there were more
2054 hosts than the hosts_max_try setting in the smtp transport when it had
2055 the 4xx errors. Those hosts that it had tried were not added to the list
2056 of hosts for which the message was waiting, so if all were tried, there
2057 was no problem. Two fixes have been applied:
2059 (i) If there are any address or message errors in an SMTP delivery, none
2060 of the hosts (tried or untried) are now added to the list of hosts
2061 for which the message is waiting, so the message should not be a
2062 candidate for sending over the same connection that was used for a
2063 successful delivery of some other message. This seems entirely
2064 reasonable: after all the message is NOT "waiting for some host".
2065 This is so "obvious" that I'm not sure why it wasn't done
2066 previously. Hope I haven't missed anything, but it can't do any
2067 harm, as the worst effect is to miss an optimization.
2069 (ii) If, despite (i), such a delivery is accidentally attempted, the
2070 routing retry time is respected, so at least it doesn't keep
2071 hammering the server.
2073 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2074 in ${readsocket because some servers need this prod.
2076 PH/32 Added some extra debug output when updating a wait-xxx database.
2078 PH/33 The hint "could be header name not terminated by colon", which has been
2079 given for certain expansion errors for a long time, was not being given
2080 for the ${if def:h_colon_omitted{... case.
2082 PH/34 The spec says: "With one important exception, whenever a domain list is
2083 being scanned, $domain contains the subject domain." There was at least
2084 one case where this was not true.
2086 PH/35 The error "getsockname() failed: connection reset by peer" was being
2087 written to the panic log as well as the main log, but it isn't really
2088 panic-worthy as it just means the connection died rather early on. I have
2089 removed the panic log writing for the ECONNRESET error when getsockname()
2092 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2093 runs only) independently of the message's sender address. This meant
2094 that, if the 4xx error was in fact related to the sender, a different
2095 message to the same recipient with a different sender could confuse
2096 things. In particualar, this can happen when sending to a greylisting
2097 server, but other circumstances could also provoke similar problems.
2098 I have changed the default so that the retry time for these errors is now
2099 based a combination of the sender and recipient addresses. This change
2100 can be overridden by setting address_retry_include_sender=false in the
2103 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2104 remote server are returned as part of bounce messages. This was not
2105 happening for LMTP over a pipe (the lmtp transport), but now it is the
2106 same for both kinds of LMTP.
2108 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2109 in header lines that were in fact CNAMEs. This is no longer the case.
2111 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2112 and queue runs started by the daemon processed all messages. This has
2113 been fixed so that -R and -S can now usefully be given with -q<time>.
2115 PH/40 Import PCRE release 6.7 (fixes some bugs).
2117 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2119 PH/42 Give an error if -q is specified more than once.
2121 PH/43 Renamed the variables $interface_address and $interface_port as
2122 $received_ip_address and $received_port, to make it clear that these
2123 values apply to message reception, and not to the outgoing interface when
2124 a message is delivered. (The old names remain recognized, of course.)
2126 PH/44 There was no timeout on the connect() call when using a Unix domain
2127 socket in the ${readsocket expansion. There now is.
2129 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2130 be meaningful with "accept".
2132 SC/01 Eximstats V1.43
2133 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2135 SC/02 Eximstats V1.44
2136 Use a glob alias rather than an array ref in the generated
2137 parser. This improves both readability and performance.
2139 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2140 Collect SpamAssassin and rejection statistics.
2141 Don't display local sender or destination tables unless
2142 there is data to show.
2143 Added average volumes into the top table text output.
2145 SC/04 Eximstats V1.46
2146 Collect data on the number of addresses (recipients)
2147 as well as the number of messages.
2149 SC/05 Eximstats V1.47
2150 Added 'Message too big' to the list of mail rejection
2151 reasons (thanks to Marco Gaiarin).
2153 SC/06 Eximstats V1.48
2154 Mainlog lines which have GMT offsets and are too short to
2155 have a flag are now skipped.
2157 SC/07 Eximstats V1.49 (Alain Williams)
2158 Added the -emptyok flag.
2160 SC/08 Eximstats V1.50
2161 Fixes for obtaining the IP address from reject messages.
2163 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2164 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2165 whitesspace changes from 4.64-PH/27
2167 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2170 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2171 are found, allow negative numbers in numeric criteria)
2173 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2175 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2176 to match changes made in 4.64-PH/43
2178 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2180 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2181 "domain" is actually an IP address, and force a failure. This locks out
2182 those revolvers/nameservers that support "A-for-A" lookups, in
2183 contravention of the specifications.
2185 PH/48 When a host name was looked up from an IP address, and the subsequent
2186 forward lookup of the name timed out, the host name was left in
2187 $sender_host_name, contrary to the specification.
2189 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2190 restricted to single-key lookups, Exim was not diagnosing an error if
2191 * or *@ was used with a query-style lookup.
2193 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2195 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2196 long ago, but noone interested enough thought of it. Let's just say that
2197 the "1.1" means that there are some new functions that weren't there at
2198 some point in the past.
2200 PH/51 Error processing for expansion failure of helo_data from an smtp
2201 transport during callout processing was broken.
2203 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2204 tested/used via the -bh/-bhc/-bs options.
2206 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2207 bug, fixed in subsequent PCRE releases).
2209 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2210 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2212 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2218 SC/01 Use a glob alias rather than an array ref in eximstats generated
2219 parser. This improves both readability and performance.
2221 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2222 Don't display local sender or destination tables in eximstats unless
2223 there is data to show.
2224 Added average volumes into the eximstats top table text output.
2226 SC/03 Collect data on the number of addresses (recipients) as well
2227 as the number of messages in eximstats.
2229 TF/01 Correct an error in the documentation for the redirect router. Exim
2230 does (usually) call initgroups() when daemonizing.
2232 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2233 with consistent privilege compared to when running as a daemon.
2235 TF/03 Note in the spec that $authenticated_id is not set for local
2236 submissions from trusted users.
2238 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2239 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2241 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2242 by adding some example configuration directives to the default
2243 configuration file. A little bit of work is required to uncomment the
2244 directives and define how usernames and passwords are checked, but
2245 there is now a framework to start from.
2247 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2248 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2249 without this. I don't know how relevant this is to other LDAP libraries.
2251 PH/02 Add the verb name to the "unknown ACL verb" error.
2253 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2255 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2257 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2258 directory not expanded when it should be if an expanded home directory
2259 was set for the address (which is overridden by the transport).
2261 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2264 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2265 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2266 because it is too late at that time, and has no effect.
2268 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2269 security issue with \' (bugzilla #107). I could not use the
2270 PQescapeStringConn() function, because it needs a PGconn value as one of
2273 PH/08 When testing addresses using -bt, indicate those final addresses that
2274 are duplicates that would not cause an additional delivery. At least one
2275 person was confused, thinking that -bt output corresponded to deliveries.
2276 (Suppressing duplicates isn't a good idea as you lose the information
2277 about possibly different redirections that led to the duplicates.)
2279 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2280 systems where poll() doesn't work, in particular OS X.
2282 PH/10 Added more information to debugging output for retry time not reached.
2284 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2285 operations in malware.c.
2287 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2290 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2291 syslog_duplication set false, log lines that would normally be written
2292 both the the main log and to the reject log were not written to syslog at
2295 PH/14 In the default configuration, change the use of "message" in ACL warn
2296 statements to "add_header".
2298 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2299 not followed by a command (e.g. "seen endif").
2301 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2302 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2305 PH/17 Added extra conditions to the default value of delay_warning_condition
2309 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2310 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2311 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2314 The Auto-Submitted: and various List- headers are standardised, whereas I
2315 don't think Precedence: ever was.
2317 PH/18 Refactored debugging code in route_finduser() to show more information,
2318 in particular, the error code if getpwnam() issues one.
2320 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2321 This is apparently needed in addition to the PH/07 change above to avoid
2322 any possible encoding problems.
2324 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2325 but not after initializing Perl.
2327 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2328 output them only if debugging. By default they are written stderr,
2329 apparently, which is not desirable.
2331 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2334 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2337 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2339 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2340 authenticated or an ident call has been made. Suppress the default
2341 values for $authenticated_id and $authenticated_sender (but permit -oMai
2342 and -oMas) when testing with -bh.
2344 PH/24 Re-jigged the order of the tests in the default configuration so that the
2345 tests for valid domains and recipients precede the DNS black list and CSA
2346 tests, on the grounds that those ones are more expensive.
2348 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2349 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2350 This bug exists in every version of Exim that I still have, right back to
2353 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2354 However, an attempt to turn on case-sensitivity in a regex key by
2355 including (?-i) didn't work because the subject string was already
2356 lowercased, and the effects were non-intuitive. It turns out that a
2357 one-line patch can be used to allow (?-i) to work as expected.
2363 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2364 other effects) broke the use of negated acl sub-conditions.
2366 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2369 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2370 "Deny" causes Exim to reject the incoming connection with a 554 error.
2371 Unfortunately, if there is a major crisis, such as a disk failure,
2372 tcp-wrappers gives "deny", whereas what one would like would be some
2373 kind of temporary error. A kludge has been added to help with this.
2374 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2375 554 error is used if errno is still zero or contains ENOENT (which occurs
2376 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2379 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2381 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2382 errors. Otherwise a message that provokes a temporary error (when other
2383 messages do not) can cause a whole host to time out.
2385 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2386 addresses were routed directly to files or pipes from a redirect router.
2387 File deliveries just didn't batch; pipe deliveries might have suffered
2390 PH/06 A failure to get a lock for a hints database would erroneously always say
2391 "Failed to get write lock", even when it was really a read lock.
2393 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2394 of 0600. This has been changed to use the value of the lockfile_mode
2395 option (which defaults to 0600).
2397 PH/08 Applied small patch from the Sieve maintainer.
2399 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2400 folder from quota calculations, a direct delivery into this folder messed
2401 up the contents of the maildirsize file. This was because the regex was
2402 used only to exclude .Trash (or whatever) when the size of the mailbox
2403 was calculated. There was no check that a delivery was happening into an
2404 excluded directory. This bug has been fixed by ignoring all quota
2405 processing for deliveries into excluded directories.
2407 PH/10 Added the maildirfolder_create_regex option to appendfile.
2413 PH/01 The code for finding all the local interface addresses on a FreeBSD
2414 system running IPv6 was broken. This may well have applied to all BSD
2415 systems, as well as to others that have similar system calls. The broken
2416 code found IPv4 interfaces correctly, but gave incorrect values for the
2417 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2418 that it would not match correctly against @[] and not recognize the IPv6
2421 PH/02 The ipliteral router was not recognizing addresses of the form user@
2422 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2424 PH/03 Added disable_ipv6.
2426 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2427 decoded form, because it is most often used to construct To: headers
2428 lines in autoreplies, and the decoded form may well be syntactically
2429 invalid. However, $reply_address has leading white space removed, and all
2430 newlines turned into spaces so that the autoreply transport does not
2433 PH/05 If group was specified without a user on a router, and no group or user
2434 was specified on a transport, the group from the router was ignored.
2436 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2437 for visible compile-time settings that can be used to change these
2438 numbers, for those that want even more. Backwards compatibility with old
2439 spool files has been maintained. However, going back to a previous Exim
2440 release will lost any variables that are in spool files.
2442 PH/07 Two small changes when running in the test harness: increase delay when
2443 passing a TCP/IP connection to a new process, in case the original
2444 process has to generate a bounce, and remove special handling of
2445 127.0.0.2 (sic), which is no longer necessary.
2447 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2448 be the same on different OS.
2450 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2453 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2454 whether --show-vars was specified or not
2456 JJ/02 exipick: Added support for new ACL variable spool format introduced
2459 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2460 syntactically invalid From: or Reply-to: line, and a filter used this to
2461 generate an autoreply, and therefore failed to obtain an address for the
2462 autoreply, Exim could try to deliver to a non-existent relative file
2463 name, causing unrelated and misleading errors. What now happens is that
2464 it logs this as a hard delivery error, but does not attempt to create a
2467 PH/11 The exinext utility has a -C option for testing purposes, but although
2468 the given file was scanned by exinext itself; it wasn't being passed on
2469 when Exim was called.
2471 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2472 an end-of-file indication when reading a command response.
2474 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2475 compiled. In many other places in Exim, IPv6 addresses are always
2476 recognized, so I have changed this. It also means that IPv4 domain
2477 literals of the form [IPV4:n.n.n.n] are now always recognized.
2479 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2480 used if the router is not running as root, for example, when verifying at
2481 ACL time, or when using -bh. The debugging output from this situation was
2482 non-existent - all you got was a failure to exec. I have made two
2485 (a) Failures to set uid/gid, the current directory, or a process leader
2486 in a subprocess such as that created by queryprogram now generate
2487 suitable debugging ouput when -d is set.
2489 (b) The queryprogram router detects when it is not running as root,
2490 outputs suitable debugging information if -d is set, and then runs
2491 the subprocess without attempting to change uid/gid.
2493 PH/15 Minor change to Makefile for building test_host (undocumented testing
2496 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2497 additional section of a DNS packet that returns MX or SRV records.
2498 Instead, it always explicitly searches for A/AAAA records. This avoids
2499 major problems that occur when a DNS server includes only records of one
2500 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2501 fixed another bug: if SRV records were looked up and the corresponding
2502 address records were *not* found in the additional section, the port
2503 values from the SRV records were lost.
2505 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2506 using the correct key (the original address) when searching the retry
2507 rules in order to find which one to use for generating the retry hint.
2509 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2510 adding the default one. Similarly, if it contains a Reply-To: header, the
2511 errors_reply_to option, if set, is not used.
2513 PH/19 When calculating a retry time, Exim used to measure the "time since
2514 failure" by looking at the "first failed" field in the retry record. Now
2515 it does not use this if it is later than than the arrival time of the
2516 message. Instead it uses the arrival time. This makes for better
2517 behaviour in cases where some deliveries succeed, thus re-setting the
2518 "first failed" field. An example is a quota failure for a huge message
2519 when small messages continue to be delivered. Without this change, the
2520 "time since failure" will always be short, possible causing more frequent
2521 delivery attempts for the huge message than are intended.
2522 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2524 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2525 $1, $2, $3) because the numerical variables can be reset during some
2526 expansion items (e.g. "match"), thereby losing the authentication data.
2528 PH/21 Make -bV show the size of off_t variables so that the test suite can
2529 decide whether to run tests for quotas > 2G.
2531 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2532 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2533 filecount value is greater than 2G or if a quota value is greater than 2G
2534 on a system where the size of off_t is not greater than 4, a panic error
2537 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2538 never match. The debug and -bh output now contains an explicit error
2539 message indicating a malformed IPv4 address or mask.
2541 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2542 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2543 PH/23 above applies.
2545 PH/25 Do not write to syslog when running in the test harness. The only
2546 occasion when this arises is a failure to open the main or panic logs
2547 (for which there is an explicit test).
2549 PH/26 Added the /no_tell option to "control=freeze".
2551 PH/27 If a host name lookup failed very early in a connection, for example, if
2552 the IP address matched host_lookup and the reverse lookup yielded a name
2553 that did not have a forward lookup, an error message of the form "no IP
2554 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2555 could be logged. Now it outputs the IP address instead of "NULL".
2557 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2558 allows the sender and the authenticated sender to be set when
2559 submitting a message from within Exim. Since child_open_exim() is
2560 documented for local_scan(), the new function should be too.
2562 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2563 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2564 results in an empty string is now treated as unset.
2566 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2568 PH/31 Added sender_verify_fail logging option.
2570 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2571 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2572 tidied the source and removed it altogether.
2574 PH/33 When a queue run was abandoned because the load average was too high, a
2575 log line was always written; now it is written only if the queue_run log
2576 selector is set. In addition, the log line for abandonment now contains
2577 information about the queue run such as the pid. This is always present
2578 in "start" and "stop" lines but was omitted from the "abandon" line.
2580 PH/34 Omit spaces between a header name and the colon in the error message that
2581 is given when verify = headers_syntax fails (if there are lots of them,
2582 the message gets confusing).
2584 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2585 names, as there are now some PTR records that contain slashes. This check
2586 is only to protect against broken name servers that fall over on strange
2587 characters, so the fact that it applies to all lookups doesn't matter.
2589 PH/36 Now that the new test suite is complete, we can remove some of the
2590 special code in Exim that was needed for the old test suite. For example,
2591 sorting DNS records because real resolvers return them in an arbitrary
2592 order. The new test suite's fake resolver always returns records in the
2595 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2596 bounces) except when queue_only is set, to avoid logging races between
2597 the different processes.
2599 PH/38 Panic-die if .include specifies a non-absolute path.
2601 PH/39 A tweak to the "H" retry rule from its user.
2603 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2604 a label. They prevented compilation on older perls.
2606 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2607 a warning to be raised on newish perls.
2609 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2610 on queue. Changes to match documented behaviour of showing count of
2611 messages matching specified criteria.
2613 PH/40 Changed the default ident timeout from 30s to 5s.
2615 PH/41 Added support for the use of login_cap features, on those BSD systems
2616 that have them, for controlling the resources used by pipe deliveries.
2618 PH/42 The content-scanning code uses fopen() to create files in which to put
2619 message data. Previously it was not paying any attention to the mode of
2620 the files. Exim runs with umask(0) because the rest of the code creates
2621 files with open(), and sets the required mode explicitly. Thus, these
2622 files were ending up world-writeable. This was not a big issue, because,
2623 being within the spool directory, they were not world-accessible. I have
2624 created a function called modefopen, which takes an additional mode
2625 argument. It sets umask(777), creates the file, chmods it to the required
2626 mode, then resets the umask. All the relevant calls to fopen() in the
2627 content scanning code have been changed to use this function.
2629 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2630 to 24 hours. This avoids potential overflow problems when processing G
2631 and H retry rules. I suspect nobody ever tinkers with this value.
2633 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2635 PH/45 When the plaintext authenticator is running as a client, the server's
2636 challenges are checked to ensure they are valid base64 strings. By
2637 default, the authentication attempt is cancelled if an invalid string is
2638 received. Setting client_ignore_invalid_base64 true ignores these errors.
2639 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2640 they are received. Thus, the responses can be made to depend on the
2641 challenges. If an invalid string is ignored, an empty string is placed in
2644 PH/46 Messages that are created by the autoreply transport now contains a
2645 References: header, in accordance with RFCs 2822 and 3834.
2647 PH/47 Added authenticated_sender_force to the smtp transport.
2649 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2651 PH/49 Installed latest patch from the Sieve maintainer.
2653 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2654 was also set, the appendfile transport was unnecessarily scanning a
2655 directory of message files (e.g. for maildir delivery) to find the count
2656 of files (along with the size), even though it did not need this
2657 information. It now does the scan only if it needs to find either the
2658 size of the count of files.
2660 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2662 PH/52 Two bugs concerned with error handling when the smtp transport is
2665 (i) Exim was not creating retry information for temporary errors given
2666 for individual recipients after the DATA command when the smtp transport
2667 was used in LMTP mode. This meant that they could be retried too
2668 frequently, and not timed out correctly.
2670 (ii) Exim was setting the flag that allows error details to be returned
2671 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2672 recipients that were returned after the DATA command.
2674 PH/53 This is related to PH/52, but is more general: for any failing address,
2675 when detailed error information was permitted to be returned to the
2676 sender, but the error was temporary, then after the final timeout, only
2677 "retry timeout exceeded" was returned. Now it returns the full error as
2678 well as "retry timeout exceeded".
2680 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2681 do this, and (what is worse) MTAs that accept it.
2683 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2684 will now be deprecated.
2686 PH/56 New os.c-cygwin from the Cygwin maintainer.
2688 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2689 formats (previously only available in exim formats via -bpr, -bpru,
2690 and -bpra. Now also available in native and exiqgrep formats)
2692 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2693 with very large, slow to parse queues
2695 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2697 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2699 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2700 responses to authentication challenges, though it was showing the
2701 challenges; (ii) I've removed the CR characters from the debug output for
2704 PH/58 Allow for the insertion of a newline as well as a space when a string
2705 is turned into more than one encoded-word during RFC 2047 encoding. The
2706 Sieve code now uses this.
2708 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2709 data_4xx, lost_connection, tls_required.
2711 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2712 message was being sent as an SMTP response.
2714 PH/61 Add -l and -k options to exicyclog.
2716 PH/62 When verifying, if an address was redirected to one new address, so that
2717 verification continued, and the new address failed or deferred after
2718 having set something in $address_data, the value of $address_data was not
2719 passed back to the ACL. This was different to the case when no
2720 redirection occurred. The value is now passed back in both cases.
2722 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2723 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2724 use login_cap.h, so on its own it isn't the distinguishing feature. The
2725 new name refers directly to the setclassresources() function.
2727 PH/65 Added configuration files for NetBSD3.
2729 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2731 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2732 is preferred over IPv4.
2734 PH/68 The bounce_return_message and bounce_return_body options were not being
2735 honoured for bounces generated during the reception of non-SMTP messages.
2736 In particular, this applied to messages rejected by the ACL. This bug has
2737 been fixed. However, if bounce_return_message is true and bounce_return_
2738 body is false, the headers that are returned for a non-SMTP message
2739 include only those that have been read before the error was detected.
2740 (In the case of an ACL rejection, they have all been read.)
2742 PH/69 The HTML version of the specification is now built in a directory called
2743 spec_html instead of spec.html, because the latter looks like a path with
2744 a MIME-type, and this confuses some software.
2746 PH/70 Catch two compiler warnings in sieve.c.
2748 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2749 function verify_get_ident() calls ip_connect() to connect a socket, but
2750 if the "connect()" function timed out, ip_connect() used to close the
2751 socket. However, verify_get_ident() also closes the socket later, and in
2752 between Exim writes to the log, which may get opened at this point. When
2753 the socket was closed in ip_connect(), the log could get the same file
2754 descriptor number as the socket. This naturally causes chaos. The fix is
2755 not to close the socket in ip_connect(); the socket should be closed by
2756 the function that creates it. There was only one place in the code where
2757 this was missing, in the iplookup router, which I don't think anybody now
2758 uses, but I've fixed it anyway.
2760 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2761 well as to direct DNS lookups. Otherwise the handling of names in host
2762 lists is inconsistent and therefore confusing.
2768 PH/01 Two changes to the default runtime configuration:
2770 (1) Move the checks for relay_from_hosts and authenticated clients from
2771 after to before the (commented out) DNS black list checks.
2773 (2) Add control=submission to the relay_from_hosts and authenticated
2774 clients checks, on the grounds that messages accepted by these
2775 statements are most likely to be submissions.
2777 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2779 (1) Generate an error if the third argument for the ${prvs expansion is
2782 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2785 (3) Reset the variables that are obtained from the first argument of
2786 ${prvscheck and used in the second argument before leaving the code,
2787 because their memory is reclaimed, so using them afterwards may do
2790 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2791 one (it's much easier than Tom thought :-).
2793 (5) Because of (4), we can now allow for the use of $prvscheck_result
2794 inside the third argument.
2796 PH/03 For some reason, the default setting of PATH when running a command from
2797 a pipe transport was just "/usr/bin". I have changed it to
2800 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2801 anything to be listed in the output from -bV.
2803 PH/05 When a filter generated an autoreply, the entire To: header line was
2804 quoted in the delivery log line, like this:
2806 => >A.N.Other <ano@some.domain> <original@ddress> ...
2808 This has been changed so that it extracts the operative address. There
2809 may be more than one such address. If so, they are comma-separated, like
2812 => >ano@some.domain,ona@other.domain <original@ddress> ...
2814 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2815 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2816 not being looked up in the rDNS to get a host name, Exim was showing the
2817 IP address twice in Received: lines, even though the IP addresses were
2818 identical. For example:
2820 Received: from [1.2.3.4] (helo=[1.2.3.4])
2822 However, if the real host name was known, it was omitting the HELO data
2823 if it matched the actual IP address. This has been tidied up so that it
2824 doesn't show the same IP address twice.
2826 PH/07 When both +timestamp and +memory debugging was on, the value given by
2827 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2828 called by the debug printing, thereby overwriting the timestamp buffer.
2829 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2831 PH/08 When the original message was included in an autoreply transport, it
2832 always said "this is a copy of the message, including all the headers",
2833 even if body_only or headers_only was set. It now gives an appropriate
2836 PH/09 Applied a patch from the Sieve maintainer which:
2838 o fixes some comments
2839 o adds the (disabled) notify extension core
2840 o adds some debug output for the result of if/elsif tests
2841 o points to the current vacation draft in the documentation
2842 and documents the missing references header update
2846 o fixes a bug in processing the envelope test (when testing
2847 multiple envelope elements, the last element determinted the
2850 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2851 Electronic Mail") by including:
2853 Auto-submitted: auto-generated
2855 in the messages that it generates (bounce messages and others, such as
2856 warnings). In the case of bounce messages for non-SMTP mesages, there was
2857 also a typo: it was using "Auto_submitted" (underscore instead of
2858 hyphen). Since every message generated by Exim is necessarily in response
2859 to another message, thes have all been changed to:
2861 Auto-Submitted: auto-replied
2863 in accordance with these statements in the RFC:
2865 The auto-replied keyword:
2867 - SHOULD be used on messages sent in direct response to another
2868 message by an automatic process,
2870 - MUST NOT be used on manually-generated messages,
2872 - MAY be used on Delivery Status Notifications (DSNs) and Message
2873 Disposition Notifications (MDNs),
2875 - MUST NOT be used on messages generated by automatic or periodic
2876 processes, except for messages which are automatic responses to
2879 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2880 to the default Received: header definition.
2882 PH/12 Added log selector acl_warn_skipped (default on).
2884 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2885 variables because (a) they are in the wrong storage pool and (b) even if
2886 they were copied, it wouldn't work properly because of the caching.
2888 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2889 checking when decoding. Apparently there are clients that generate
2890 overlong encoded strings. Why am I not surprised?
2892 PH/15 If the first argument of "${if match_address" was not empty, but did not
2893 contain an "@" character, Exim crashed. Now it writes a panic log message
2894 and treats the condition as false.
2896 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2898 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2899 extension "envelope-auth". The code is finished and in agreement with
2900 other implementations, but there is no documentation so far and in fact,
2901 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2902 not changing the active code.
2904 Print executed "if" and "elsif" statements when debugging is used. This
2905 helps a great deal to understand what a filter does.
2907 Document more things not specified clearly in RFC3028. I had all this
2908 sorted out, when out of a sudden new issues came to my mind. Oops."
2910 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2913 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2914 canonical form (as documented). However, after a host name lookup from
2915 the IP address, check_host() was doing a simple string comparison with
2916 addresses acquired from the DNS when checking that the found name did
2917 have the original IP as one of its addresses. Since any found IPv6
2918 addresses are likely to be in abbreviated form, the comparison could
2919 fail. Luckily, there already exists a function for doing the comparison
2920 by converting both addresses to binary, so now that is used instead of
2921 the text comparison.
2923 PH/20 There was another similar case to PH/19, when a complete host name was
2924 given in a host list; looking up its IP address could give an abbreviated
2925 form, whereas the current host's name might or might not be abbreviated.
2926 The same fix has been applied.
2932 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2933 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2936 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2937 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2939 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2941 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2942 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2943 Also, for "Auto-Submitted", treat anything other than "no" as
2944 non-personal, in accordance with RFC 3834. (Previously it treated
2945 anything starting "auto-" as non-personal.)
2947 TF/01 The control=submission/name=... option had a problem with syntax
2948 errors if the name included a slash character. The /name= option
2949 now slurps the rest of the string, so it can include any characters
2950 but it must come last in the list of options (after /sender_retain
2953 PH/05 Some modifications to the interface to the fake nameserver for the new
2961 TK/01 Added the "success_on_redirect" address verification option. See
2962 NewStuff for rationale and an example.
2964 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2966 PH/02 Patch to exigrep to allow it to work on syslog lines.
2968 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2969 fread() to skip over the body file's header line, because in Cygwin the
2970 header line is locked and is inaccessible.
2972 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2973 co-exist for some time) to make it clear that it is the Exim ID that is
2974 referenced, not the Message-ID: header line.
2976 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2977 string_format() function, because snprintf() does not exist on all
2980 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2981 ${stat: expansion item.
2983 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2984 protocol synchronization error", to keep the pedants happy.
2986 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2987 well as for IRIX systems, when gcc is being used. See the host.c source
2990 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2992 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2995 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2996 transport and to the smtp transport in LMTP mode.
2998 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3000 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3001 run a filter in a subprocess. This could lead to confusion in subsequent
3002 lookups in the parent process. There should also be a search_tidyup() at
3003 the end of the subprocess.
3005 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3006 only if the host matched helo_try_verify_hosts, which caused the
3007 verification to occur when the EHLO/HELO command was issued. The ACL just
3008 tested the remembered result. Now, if a previous verification attempt has
3009 not happened, "verify = helo" does it there and then.
3011 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3013 TK/03 Fix log output including CR from clamd.
3015 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3016 exist provoked a memory error which could cause a segfault.
3018 PH/15 Installed PCRE 6.2
3020 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3022 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3023 of the problem. Specifically, suggested +O2 rather than +O1 for the
3026 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3028 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3029 record, the port was not correctly logged when the outgoing_port log
3030 selector was set (it logged the transort's default port).
3032 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3033 fallback_hosts, and "hosts" in the smtp transport.
3035 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3036 host errors such as "Connection refused".
3038 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3039 authentication with radiusclient 0.4.9:
3041 - Error returned from rc_read_config was caught wrongly
3042 - Username/password not passed on to radius server due to wrong length.
3044 The presumption is that some radiusclient API changes for 4.51/PH/17
3045 were not taken care of correctly. The code is still untested by me (my
3046 Linux distribution still has 0.3.2 of radiusclient), but it was
3047 contributed by a Radius user.
3049 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3050 expanding the "port" option of the smtp transport.
3052 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3053 while reading a MIME header. Thanks to Tom Hughes for a patch.
3055 PH/24 Include config.h inside local_scan.h so that configuration settings are
3058 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3059 that in an ACL for RCPT (for example), you can examine exactly what was
3062 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3063 commands, but it was not correctly comparing the address with the actual
3064 client host address. Thus, it would show the EHLO address in Received:
3065 header lines when this was not necessary.
3067 PH/27 Added the % operator to ${eval:}.
3069 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3070 it should be ignoring failures (because with -C, for example, it has lost
3071 privilege). It wasn't ignoring creation failures other than "already
3074 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3077 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3078 because an input line was too long, either on its own, or by virtue of
3079 too many continuations, the temporary file was not being removed, and the
3080 return code was incorrect.
3082 PH/31 Missing "BOOL" in function definition in filtertest.c.
3084 PH/32 Applied Sieve patches from the maintainer.
3086 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3088 PH/33 Added "verify = not_blind".
3090 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3091 Local/Makefile (with some defaults set). These are used in built scripts
3092 such as exicyclog, but they have never been used in the exim_install
3093 script (though there are many overriding facilities there). I have
3094 arranged that the exim_install script now takes note of these two
3097 PH/35 Installed configuration files for Dragonfly.
3099 PH/36 When a locally submitted message by a trusted user did not contain a
3100 From: header, and the sender address was obtained from -f or from an SMTP
3101 MAIL command, and the trusted user did not use -F to supply a sender
3102 name, $originator_name was incorrectly used when constructing a From:
3103 header. Furthermore, $originator_name was used for submission mode
3104 messages from external hosts without From: headers in a similar way,
3105 which is clearly wrong.
3107 PH/37 Added control=suppress_local_fixups.
3109 PH/38 When log_selector = +received_sender was set, and the addition of the
3110 sender made the log line's construction buffer exactly full, or one byte
3111 less than full, an overflow happened when the terminating "\n" was
3114 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3115 when the result of a list match is failure because a DNS lookup failed.
3117 PH/40 RM_COMMAND is now used in the building process.
3119 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3120 the "build-* directories that it finds.
3122 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3123 address in a domain literal was a prefix of an interface address.
3125 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3126 when verifying a sender address, unless rewrite_headers is false.
3128 PH/44 Wrote a long comment about why errors_to addresses are verified as
3129 recipients, not senders.
3131 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3132 the ratelimit ACL was added.
3134 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3136 PH/46 Added extra information about PostgreSQL errors to the error string.
3138 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3139 avoiding the need to install special zones in a real server. This is
3140 backwards compatible; if it can't find the fake resolver, it drops back.
3141 Thus, both old and new test suites can be run.
3143 TF/02 Added util/ratelimit.pl
3145 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3146 clock is set back in time.
3148 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3149 Candler <B.Candler@pobox.com>.
3151 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3152 OK for addresses that are the result of redirections.
3154 PH/48 A number of further additions for the benefit of the new test suite,
3155 including a fake gethostbyname() that interfaces to the fake DNS resolver
3158 TF/06 The fix for widen_domains has also been applied to qualify_single and
3159 search_parents which are the other dnslookup options that can cause
3162 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3165 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3167 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3168 service" if it encountered an MX record with an empty target hostname.
3169 The message is now "an MX or SRV record indicated no SMTP service".
3171 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3172 if the DNS of the sending site is misconfigured. This is quite a
3173 common situation. This change restores the behaviour of treating a
3174 helo verification defer as a failure.
3176 PH/51 If self=fail was set on a router, the bounce message did not include the
3177 actual error message.
3183 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3185 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3186 command itself ended in error, the underlying message about the transport
3187 filter timeout was being overwritten with the pipe command error. Now the
3188 underlying error message should be appended to the second error message.
3190 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3192 PH/02 Reduce the amount of output that "make" produces by default. Full output
3193 can still be requested.
3195 PH/03 The warning log line about a condition test deferring for a "warn" verb
3196 was being output only once per connection, rather than after each
3197 occurrence (because it was using the same function as for successful
3198 "warn" verbs). This seems wrong, so I have changed it.
3200 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3201 it should not have, which might have caused a crash in the right
3202 circumstances, but probably never did.
3204 PH/04 Installed a modified version of Tony Finch's patch to make submission
3205 mode fix the return path as well as the Sender: header line, and to
3206 add a /name= option so that you can make the user's friendly name appear
3209 TF/03 Added the control = fakedefer ACL modifier.
3211 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3212 Mark Lowes for thorough testing.
3214 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3216 TK/03 Merged latest SRS patch from Miles Wilton.
3218 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3219 with the definition in sysexits.h (which is #included earlier).
3220 Fortunately, Exim does not actually use EX_OK. The code used to try to
3221 preserve the sysexits.h value, by assumimg that macro definitions were
3222 scanned for macro replacements. I have been disabused of this notion,
3223 so now the code just undefines EX_OK before #including unistd.h.
3225 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3226 in the smtp transport. When a block could not be written in a single
3227 write() function, the timeout was being re-applied to each part-write.
3228 This seems wrong - if the receiver was accepting one byte at a time it
3229 would take for ever. The timeout is now adjusted when this happens. It
3230 doesn't have to be particularly precise.
3232 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3233 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3235 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3236 without a domain if the check to <postmaster@domain> fails.
3238 SC/01 Eximstats: added -xls and the ability to specify output files
3239 (patch written by Frank Heydlauf).
3241 SC/02 Eximstats: use FileHandles for outputing results.
3243 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3245 SC/04 Eximstats: fixed display of large numbers with -nvr option
3247 SC/05 Eximstats: fixed merging of reports with empty tables.
3249 SC/06 Eximstats: added the -include_original_destination flag
3251 SC/07 Eximstats: removed tabs and trailing whitespace.
3253 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3255 TK/06 MBOX spool code: Add real "From " MBOX separator line
3256 so the .eml file is really in mbox format (even though
3257 most programs do not really care). Patch from Alex Miller.
3259 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3260 The latter is generated from $received_to and is only set if the
3261 message has one envelope recipient. SA can use these headers,
3262 obviously out-of-the-box. Patch from Alex Miller.
3264 PH/08 The ${def test on a variable was returning false if the variable's
3265 value was "0", contrary to what the specification has always said!
3266 The result should be true unless the variable is empty.
3268 PH/09 The syntax error of a character other than { following "${if
3269 def:variable_name" (after optional whitespace) was not being diagnosed.
3270 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3271 accidental colon was present, for example, could give incorrect results.
3273 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3274 result is used (not including appendfile, where other changes are about
3277 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3278 This involved changing a lot of size variables from int to off_t. It
3279 should work with maildirs and everything.
3281 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3282 spamd dying while we are connected to it.
3284 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3287 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3288 function for BDB 4.3.
3290 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3292 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3293 the sources for PCRE have been re-arranged and more files are now
3296 PH/15 The code I had for printing potentially long long variables in PH/11
3297 above was not the best (it lost precision). The length of off_t variables
3298 is now inspected at build time, and an appropriate printing format (%ld
3299 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3300 to be "long long int" or "long int". This is needed for the internal
3301 formatting function string_vformat().
3303 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3304 the configuration file to be ":syslog", then the script "guesses" where
3305 the logs files are, rather than using the compiled in default. In our
3306 case the guess is not the same as the compiled default, so the script
3307 suddenly stopped working when I started to use syslog. The patch checks
3308 to see if log_file_path is "". If so, it attempts to read it from exim
3309 with no configuration file to get the compiled in version, before it
3310 falls back to the previous guessing code."
3312 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3313 implementing BATV in an Exim configuration. See NewStuff for the gory
3316 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3317 Makefile that are specific to HP-UX.
3319 PH/18 If the "use_postmaster" option was set for a recipient callout together
3320 with the "random" option, the postmaster address was used as the MAIL
3321 FROM address for the random test, but not for the subsequent recipient
3322 test. It is now used for both.
3324 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3325 patch removes a few documentation additions to RFC 3028, because the
3326 latest draft now contains them. It adds the new en;ascii-case comparator
3327 and a new error check for 8bit text in MIME parts. Comparator and
3328 require names are now matched exactly. I enabled the subaddress
3329 extension, but it is not well tested yet (read: it works for me)."
3331 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3332 rework some of the code of TK/09 above to avoid the hardwired use of
3333 "%lld" and "long long". Replaced the call to snprintf() with a call to
3336 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3337 records point to non-existent hosts", "retry timeout exceeded", and
3338 "retry time not reached for any host after a long failure period".
3340 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3341 experimental DomainKeys support:
3343 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3344 (2) On an error such as an illegally used "control", the wrong name for
3345 the control was given.
3347 These problems did NOT occur unless DomainKeys support was compiled.
3349 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3351 PH/24 Added ${if match_ip condition.
3353 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3354 databases so that it will be absolutely obvious if a crash occurs in the
3355 DB library. This is a regular occurrence (often caused by mis-matched
3358 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3359 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3360 picked up on a user's system that detects such things. There doesn't seem
3361 to be a gcc warning option for this - only an attribute that has to be
3362 put on the function's prototype. It seems that in Fedora Core 4 they have
3363 set this on a number of new functions. No doubt there will be more in due
3366 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3367 specify a transport. However, if an address that was verified by such a
3368 router was the subject of a callout, Exim crashed because it tried to
3369 read the rcpt_include_affixes from the non-existent transport. Now it
3370 just assumes that the setting of that option is false. This bug was
3371 introduced by 4.51/PH/31.
3373 PH/28 Changed -d+all to exclude +memory, because that information is very
3374 rarely of interest, but it makes the output a lot bigger. People tend to
3375 do -d+all out of habit.
3377 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3378 code in os-type was giving problems when libc.so lives in lib64, like on
3381 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3382 aren't the modern standard, and it seems that some systems' include files
3383 don't always have them. Exim was already checking for some of the newer
3384 ones like T_AAAA, and defining it itself. I've added checks for all the
3385 record types that Exim uses.
3387 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3388 not automatically generating a new one, as it is supposed to. This
3389 prevented TLS from working. If the file did exist, but contained invalid
3390 data, a new version was generated, as expected. It was only the case of a
3391 non-existent file that was broken.
3393 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3394 with a change in libdomainkeys > 0.64.
3396 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3397 from DNS. If the selector record carries the flag, it now has
3398 precedence over the domain-wide flag.
3400 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3402 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3403 the use of a port name that isn't defined in /etc/services) occurred, the
3404 message was deferred as in a normal delivery, and thus remained on the
3405 spool, instead of being failed because of the mua_wrapper setting. This
3406 is now fixed, and I tidied up some of the mua_wrapper messages at the
3409 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3410 the messages in a hash of arrays rather than using individual hashes.
3411 This is a bit cleaner and results in dramatic memory savings, albeit
3412 at a slight CPU cost.
3414 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3415 as requested by Marc Sherman.
3417 SC/10 Eximstats: added histograms for user specified patterns as requested
3420 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3422 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3423 fopen() in the content-scanning modules that did not already have it.
3429 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3430 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3432 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3434 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3436 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3437 to test Sieve filters that use "vacation".
3439 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3440 that changes the way the GnuTLS parameters are stored in the cache file.
3441 The new format can be generated externally. For backward compatibility,
3442 if the data in the cache doesn't make sense, Exim assumes it has read an
3443 old-format file, and it generates new data and writes a new file. This
3444 means that you can't go back to an older release without removing the
3447 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3448 work if there are any delivery delays because "one_time" forces the
3449 parent to be marked "delivered", so its unseen clone is never tried
3450 again. For this reason, Exim now forbids the simultaneous setting of
3453 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3454 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3455 entry if you want to know the details. The fix, however, neglected to
3456 consider the case when local delivery batching is involved. The test for
3457 "previously delivered" was not happening when checking to see if an
3458 address could be batched with a previous (undelivered) one; under
3459 certain circumstances this could lead to multiple deliveries to the same
3462 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3463 in its include files, and this causes problems building Exim.
3465 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3466 header_syntax) but Exim was just ignoring anything given after a slash.
3467 In particular, this caused confusion with an attempt to use "verify =
3468 reverse_host_lookup/defer_ok". An error is now given when options are
3469 supplied for verify items that do not have them. (Maybe reverse_host_
3470 lookup should have a defer_ok option, but that's a different point.)
3472 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3473 defined by RFC 821) to 2048, because there were problems with some AUTH
3474 commands, and RFC 1869 says the size should be increased for extended
3475 SMTP commands that take arguments.
3477 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3480 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3481 "unknown" error; now it says that the functionality isn't in the binary.
3483 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3484 an address' error message when a string expansion fails (syntax or
3485 whatever). Otherwise the password may appear in the log. Following change
3486 PH/42 below, there is no longer a chance of it appearing in a bounce
3489 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3491 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3492 of its addresses was ever tried. (Bugzilla bug #2.)
3494 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3495 the result incorrectly in the debug output. (It correctly added a newline
3496 to what was transported.)
3498 TF/01 Added $received_time.
3500 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3501 commented out examples of how to interface to a virus scanner and to
3502 SpamAssassin. Also added commented examples of av_scanner and
3503 spamd_address settings.
3505 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3506 and controls are allowed in which ACLs. There were a couple of minor
3507 errors. Some of the entries in the conditions table (which is a table of
3508 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3509 negation of where the condition IS allowed.
3511 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3513 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3514 header file does not have a version number, so I've had to invent a new
3515 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3516 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3517 radiusclient), but it was contributed by a Radius user.
3519 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3520 files or directories, for OpenSSL.
3522 PH/19 When an Exim process that is running as root has to create an Exim log
3523 file, it does so in a subprocess that runs as exim:exim so as to get the
3524 ownership right at creation (otherwise, other Exim processes might see
3525 the file with the wrong ownership). There was no test for failure of this
3526 fork() call, which would lead to the process getting stuck as it waited
3527 for a non-existent subprocess. Forks do occasionally fail when resources
3528 run out. I reviewed all the other calls to fork(); they all seem to check
3531 PH/20 When checking for unexpected SMTP input at connect time (before writing
3532 the banner), Exim was not dealing correctly with a non-positive return
3533 from the read() function. If the client had disconnected by this time,
3534 the result was a log entry for a synchronization error with an empty
3535 string after "input=" when read() returned zero. If read() returned -1
3536 (an event I could not check), uninitialized data bytes were printed.
3537 There were reports of junk text (parts of files, etc) appearing after
3540 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3542 PH/22 Added support for macro redefinition, and (re)definition in between
3543 driver and ACL definitions.
3545 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3546 forgetting to use the resulting value; it was using the unexpanded value.
3548 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3549 hadn't been configured. The fix is from Juergen Kreileder, who
3550 understands it better than I do:
3552 "Here's what I see happening with three configured cyrus_sasl
3553 authenticators configured (plain, login, cram-md5):
3555 On startup auth_cyrus_sasl_init() gets called for each of these.
3556 This means three calls to sasl_listmech() without a specified mech_list.
3557 => SASL tests which mechs of all available mechs actually work
3558 => three warnings about OTP not working
3559 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3561 With the patch, sasl_listmech() also gets called three times. But now
3562 SASL's mech_list option is set to the server_mech specified in the the
3563 authenticator. Or in other words, the answer from sasl_listmech()
3564 gets limited to just the mech you're testing for (which is different
3566 => the return list contains just 'plain' or 'login', 'cram-md5' or
3567 nothing depending on the value of ob->server_mech.
3569 I've just tested the patch: Authentication still works fine,
3570 unavailable mechs specified in the exim configuration are still
3571 caught, and the auth.log warnings about OTP are gone."
3573 PH/25 When debugging is enabled, the contents of the command line are added
3574 to the debugging output, even when log_selector=+arguments is not
3577 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3578 answer is "GNU", and only if the return is "GNU/something" is the answer
3581 PH/27 $acl_verify_message is now set immediately after the failure of a
3582 verification in an ACL, and so is available in subsequent modifiers. In
3583 particular, the message can be preserved by coding like this:
3585 warn !verify = sender
3586 set acl_m0 = $acl_verify_message
3588 Previously, $acl_verify_message was set only while expanding "message"
3589 and "log_message" when a very denied access.
3591 PH/28 Modified OS/os.c-Linux with
3593 -#ifndef OS_LOAD_AVERAGE
3594 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3596 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3597 nomenclature these days.)
3599 PH/29 Installed patch from the Sieve maintainer that adds the options
3600 sieve_useraddress and sieve_subaddress to the redirect router.
3602 PH/30 In these circumstances:
3603 . Two addresses routed to the same list of hosts;
3604 . First host does not offer TLS;
3605 . First host accepts first address;
3606 . First host gives temporary error to second address;
3607 . Second host offers TLS and a TLS session is established;
3608 . Second host accepts second address.
3609 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3610 and peerdn, if requested) that were in fact used only for the second
3613 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3614 attention to any local part prefix or suffix that was matched by the
3615 router that accepted the address. It now behaves in the same way as it
3616 does for delivery: the affixes are removed from the local part unless
3617 rcpt_include_affixes is set on the transport.
3619 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3620 timeout during the DATA phase of an incoming message.
3622 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3623 applied a patch sanctioned by the Sieve maintainer.
3625 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3626 the uid or gid is negative. A case of a negative gid caused this to be
3627 noticed. The fix allows for either to be negative.
3629 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3630 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3633 PH/36 The change PH/12 above was broken. Fixed it.
3635 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3636 the grounds that routing the same address twice would always produce the
3637 same answer. This might have been true once, but it is certainly no
3638 longer true now. Routing a child address may depend on the previous
3639 routing that produced that child. Some complicated redirection strategies
3640 went wrong when messages had multiple recipients, and made Exim's
3641 behaviour dependent on the order in which the addresses were given.
3643 I have moved the duplicate checking until after the routing is complete.
3644 Exim scans the addresses that are assigned to local and remote
3645 transports, and removes any duplicates. This means that more work will be
3646 done, as duplicates will always all be routed, but duplicates are
3647 presumably rare, so I don't expect this is of any significance.
3649 For deliveries to pipes, files, and autoreplies, the duplicate checking
3650 still happens during the routing process, since they are not going to be
3653 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3654 It corrects a timeout issue with spamd. This is Ian's comment: "The
3655 background is that sometimes spamd either never reads data from a
3656 connection it has accepted, or it never writes response data. The exiscan
3657 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3658 blindly assumes that writes won't block so it may never time out."
3660 PH/39 Allow G after quota size as well as K and M.
3662 PH/40 The value set for $authenticated_id in an authenticator may not contain
3663 binary zeroes or newlines because the value is written to log lines and
3664 to spool files. There was no check on this. Now the value is run through
3665 the string_printing() function so that such characters are converted to
3666 printable escape sequences.
3668 PH/41 $message_linecount is a new variable that contains the total number of
3669 lines in the message. Compare $body_linecount, which is the count for the
3672 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3673 bounce and delay warning messages, except in certain special cases, which
3676 (a) An SMTP error message from a remote host;
3677 (b) A message specified in a :fail: redirection;
3678 (c) A message specified in a "fail" command in a system filter;
3679 (d) A message specified in a FAIL return from the queryprogram router;
3680 (e) A message specified by the cannot_route_message router option.
3682 In these cases only, Exim does include the error details in bounce and
3683 warning messages. There are also a few cases where bland messages such
3684 as "unrouteable address" or "local delivery error" are given.
3686 PH/43 $value is now also set for the "else" part of a ${run expansion.
3688 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3689 being worked on, but at least Exim now implements the latest version to
3692 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3693 process to complete was treated as a delivery failure, a timeout while
3694 writing the message to the pipe was logged, but erroneously treated as a
3695 successful delivery. Such timeouts include transport filter timeouts. For
3696 consistency with the overall process timeout, these timeouts are now
3697 treated as errors, giving rise to delivery failures by default. However,
3698 there is now a new Boolean option for the pipe transport called
3699 timeout_defer, which, if set TRUE, converts the failures into defers for
3700 both kinds of timeout. A transport filter timeout is now identified in
3703 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3704 systems where "make" and "gmake" are different, calling "gmake" at top
3705 level broke things. I've arranged for the value of $(MAKE) to be passed
3706 from the Makefile to this script so that it can call the same version of
3710 A note about Exim versions 4.44 and 4.50
3711 ----------------------------------------
3713 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3714 changes of various kinds. As a consequence, a big documentation update was
3715 needed. This delayed the release for rather longer than seemed good, especially
3716 in the light of a couple of (minor) security issues. Therefore, the changes
3717 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3718 release. So 4.44 and 4.50 are in effect two different branches that both start
3721 I have left the 4.50 change log unchanged; it contains all the changes since
3722 4.43. The change log for 4.44 is below; many of its items are identical to
3723 those for 4.50. This seems to be the most sensible way to preserve the
3724 historical information.
3730 1. Minor wording change to the doc/README.SIEVE file.
3732 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3733 computation of the current number of files was incorrect.
3735 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3736 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3739 4. Give more explanation in the error message when the command for a transport
3740 filter fails to execute.
3742 5. There are several places where Exim runs a non-Exim command in a
3743 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3744 was being done only for the command run by the queryprogram router. It is
3745 now done for all such subprocesses. The other cases are: ${run, transport
3746 filters, and the commands run by the lmtp and pipe transports.
3748 6. Added CONFIGURE_GROUP build-time option.
3750 7. Some older OS have a limit of 256 on the maximum number of file
3751 descriptors. Exim was using setrlimit() to set 1000 as a large value
3752 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3753 systems. I've change it so that if it can't get 1000, it tries for 256.
3755 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3756 was an oversight, and furthermore, ever since the addition of extra
3757 controls (e.g. 4.43/32), the checks on when to allow different forms of
3758 "control" were broken. There should now be diagnostics for all cases when a
3759 control that does not make sense is encountered.
3761 9. Added the /retain_sender option to "control=submission".
3763 10. $recipients is now available in the predata ACL (oversight).
3765 11. Tidy the search cache before the fork to do a delivery from a message
3766 received from the command line. Otherwise the child will trigger a lookup
3767 failure and thereby defer the delivery if it tries to use (for example) a
3768 cached ldap connection that the parent has called unbind on.
3770 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3771 of $address_data from the recipient verification was clobbered by the
3772 sender verification.
3774 13. The value of address_data from a sender verification is now available in
3775 $sender_address_data in subsequent conditions in the ACL statement.
3777 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3779 15. Added a new option "connect=<time>" to callout options, to set a different
3782 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3783 was its contents. (It was OK if the option was not defined at all.)
3785 17. A "Completed" log line is now written for messages that are removed from
3786 the spool by the -Mrm option.
3788 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3789 information about exactly what failed.
3791 19. Added -dd to debug only the daemon process.
3793 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3794 handles timeouts, both on the server side and network timeouts. Renamed the
3795 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3797 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3798 It is now set to "smtps".
3800 22. $host_address is now set to the target address during the checking of
3801 ignore_target_hosts.
3803 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3804 being passed; this would have caused $sender_host_name to have been used if
3805 matching the list had actually called for a host name (not very likely,
3806 since this list is usually IP addresses). A host name is now passed as
3809 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3810 code that specifies a non-restarting handler (typically sigaction() in
3811 modern systems) in an attempt to fix a rare and obscure crash bug.
3813 25. Narrowed the window for a race in the daemon that could cause it to ignore
3814 SIGCHLD signals. This is not a major problem, because they are used only to
3815 wake it up if nothing else does.
3817 26. A malformed maildirsize file could cause Exim to calculate negative values
3818 for the mailbox size or file count. Odd effects could occur as a result.
3819 The maildirsize information is now recalculated if the size or filecount
3822 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3823 support for a long time. Removed HAVE_SYS_VFS_H.
3825 28. Installed the latest version of exipick from John Jetmore.
3827 29. In an address list, if the pattern was not a regular expression, an empty
3828 subject address (from a bounce message) matched only if the pattern was an
3829 empty string. Non-empty patterns were not even tested. This was the wrong
3830 because it is perfectly reasonable to use an empty address as part of a
3831 database query. An empty address is now tested by patterns that are
3832 lookups. However, all the other forms of pattern expect the subject to
3833 contain a local part and a domain, and therefore, for them, an empty
3834 address still always fails if the pattern is not itself empty.
3836 30. Exim went into a mad DNS loop when attempting to do a callout where the
3837 host was specified on an smtp transport, and looking it up yielded more
3838 than one IP address.
3840 31. Re-factored the code for checking spool and log partition space into a
3841 function that finds that data and another that does the check. The former
3842 is then used to implement four new variables: $spool_space, $log_space,
3843 $spool_inodes, and $log_inodes.
3845 32. The RFC2047 encoding function was originally intended for short strings
3846 such as real names; it was not keeping to the 75-character limit for
3847 encoded words that the RFC imposes. It now respects the limit, and
3848 generates multiple encoded words if necessary. To be on the safe side, I
3849 have increased the buffer size for the ${rfc2047: expansion operator from
3852 33. It is now permitted to omit both strings after an "if" condition; if the
3853 condition is true, the result is "true". As before, when the second string
3854 is omitted, a false condition yields an empty string. This makes it less
3855 cumbersome to write custom ACL and router conditions.
3857 34. Failure to deliver a bounce message always caused it to be frozen, even if
3858 there was an errors_to setting on the router. The errors_to setting is now
3861 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3862 canonical form (fully expanded) before being placed in
3863 $sender_host_address.
3865 36. The table in the code that translates DNS record types into text (T_A to
3866 "A" for instance) was missing entries for NS and CNAME. It is just possible
3867 that this could have caused confusion if both these types were looked up
3868 for the same domain, because the text type is used as part of Exim's
3869 per-process caching. But the chance of anyone hitting this buglet seems
3872 37. The dnsdb lookup has been extended in a number of ways.
3874 (1) There is a new type, "zns", which walks up the domain tree until it
3875 finds some nameserver records. It should be used with care.
3877 (2) There is a new type, "mxh", which is like "mx" except that it returns
3878 just the host names, not the priorities.
3880 (3) It is now possible to give a list of domains (or IP addresses) to be
3881 looked up. The behaviour when one of the lookups defers can be
3882 controlled by a keyword.
3884 (4) It is now possible to specify the separator character for use when
3885 multiple records are returned.
3887 38. The dnslists ACL condition has been extended: it is now possible to supply
3888 a list of IP addresses and/or domains to be looked up in a particular DNS
3891 39. Added log_selector=+queue_time_overall.
3893 40. When running the queue in the test harness, wait just a tad after forking a
3894 delivery process, to get repeatability of debugging output.
3896 41. Include certificate and key file names in error message when GnuTLS fails
3897 to set them up, because the GnuTLS error message doesn't include the name
3898 of the failing file when there is a problem reading it.
3900 42. Allow both -bf and -bF in the same test run.
3902 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3904 44. The "Exiscan patch" is now merged into the mainline Exim source.
3906 45. Sometimes the final signoff response after QUIT could fail to get
3907 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3908 < 0 before doing a fflush(). This bug looks as though it goes back to the
3909 introduction of TLS in release 3.20, but "sometimes" must have been rare
3910 because the tests only now provoked it.
3912 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3913 (this can affect the format of dates).
3915 47. exim_tidydb, when checking for the continued existence of a message for
3916 which it has found a message-specific retry record, was not finding
3917 messages that were in split spool directories. Consequently, it was
3918 deleting retry records that should have stayed in existence.
3920 48. Steve fixed some bugs in eximstats.
3922 49. The SPA authentication driver was not abandoning authentication and moving
3923 on to the next authenticator when an expansion was forced to fail,
3924 contradicting the general specification for all authenticators. Instead it
3925 was generating a temporary error. It now behaves as specified.
3927 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3928 (the order specifies the preference for clients). The order is now AES256,
3929 AES128, 3DES, ARCFOUR128.
3931 51. Small patch to Sieve code - explicitly set From: when generating an
3934 52. Exim crashed if a remote delivery caused a very long error message to be
3935 recorded - for instance if somebody sent an entire SpamAssassin report back
3936 as a large number of 550 error lines. This bug was coincidentally fixed by
3937 increasing the size of one of Exim's internal buffers (big_buffer) that
3938 happened as part of the Exiscan merge. However, to be on the safe side, I
3939 have made the code more robust (and fixed the comments that describe what
3942 53. Now that there can be additional text after "Completed" in log lines (if
3943 the queue_time_overall log selector is set), a one-byte patch to exigrep
3944 was needed to allow it to recognize "Completed" as not the last thing in
3947 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3948 patch that reportedly fixes this has been added. I am not expert enough to
3949 create a test for it. This is what the patch creator wrote:
3951 "I found a little strange behaviour of ldap code when working with
3952 Windows 2003 AD Domain, where users was placed in more than one
3953 Organization Units. When I tried to give exim partial DN, the exit code
3954 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3955 But simultaneously result of request was absolutely normal ldap result,
3956 so I produce this patch..."
3958 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3959 REFERENCE, so I have modified the code to exclude the patch when that macro
3962 55. Some experimental protocols are using DNS PTR records for new purposes. The
3963 keys for these records are domain names, not reversed IP addresses. The
3964 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3965 leaves it alone. Component reversal etc. now happens only for IP addresses.
3968 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3970 57. Double the size of the debug message buffer (to 2048) so that more of very
3971 long debug lines gets shown.
3973 58. The exicyclog utility now does better if the number of log files to keep
3974 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3976 59. Two changes related to the smtp_active_hostname option:
3978 (1) $smtp_active_hostname is now available as a variable.
3979 (2) The default for smtp_banner uses $smtp_active_hostname instead
3980 of $primary_hostname.
3982 60. The host_aton() function is supposed to be passed a string that is known
3983 to be a valid IP address. However, in the case of IPv6 addresses, it was
3984 not checking this. This is a hostage to fortune. Exim now panics and dies
3985 if the condition is not met. A case was found where this could be provoked
3986 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3987 components; fortuitously, this particular loophole had already been fixed
3988 by change 4.50/55 above.
3990 If there are any other similar loopholes, the new check in host_aton()
3991 itself should stop them being exploited. The report I received stated that
3992 data on the command line could provoke the exploit when Exim was running as
3993 exim, but did not say which command line option was involved. All I could
3994 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3995 running as the user.
3998 61. There was a buffer overflow vulnerability in the SPA authentication code
3999 (which came originally from the Samba project). I have added a test to the
4000 spa_base64_to_bits() function which I hope fixes it.
4003 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4004 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4006 63. The daemon start-up calls getloadavg() while still root for those OS that
4007 need the first call to be done as root, but it missed one case: when
4008 deliver_queue_load_max is set with deliver_drop_privilege. This is
4009 necessary for the benefit of the queue runner, because there is no re-exec
4010 when deliver_drop_privilege is set.
4012 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4013 This has been fixed.
4015 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4016 was in use, was not putting the data itself into the right store pool;
4017 consequently, it could be overwritten for a subsequent message in the same
4018 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4021 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4023 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4024 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4025 the function were treating the return as a boolean value, which happened to
4026 work because 0=false and not-0=true, but is not correct code.
4028 68. The host_aton() function was not handling scoped IPv6 addresses (those
4029 with, for example, "%eth0" on the end) correctly.
4031 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4032 negated items (that is, ~something) in unsigned ints. Some compilers
4033 apparently mutter when there is no cast.
4035 70. If an address verification called from an ACL failed, and did not produce a
4036 user-specific message (i.e. there was only a "system" message), nothing was
4037 put in $acl_verify_message. In this situation, it now puts the system
4040 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4041 session; change 4.31/43 added the unwanted input to the log line - except
4042 that it did not do this in the start of session case. It now does.
4044 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4045 This is wrong and can cause the other end to generate a synchronization
4046 error if it is another Exim or anything else that does the synchronization
4047 check. A QUIT command is no longer sent after a timeout.
4049 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4050 during host lookups.
4052 74. The defer_ok option of callout verification was not working if it was used
4053 when verifying addresses in header lines, that is, for this case:
4055 verify = header_sender/callout=defer_ok
4057 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4058 those file descriptors could be used for SMTP connections. If anything
4059 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4060 could be sent to the SMTP client, causing chaos. The daemon now opens
4061 stdin, stdout, and stderr to /dev/null when it puts itself into the
4064 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4065 log by default. The user can override this with suitable Perl magic.
4067 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4068 the log message when discard triggers, was not working for the DATA ACL or
4069 for the non-SMTP ACL.
4071 78. Error message wording change in sieve.c.
4073 79. If smtp_accept_max_per_host was set, the number of connections could be
4074 restricted to fewer than expected, because the daemon was trying to set up
4075 a new connection before checking whether the processes handling previous
4076 connections had finished. The check for completed processes is now done
4077 earlier. On busy systems, this bug wouldn't be noticed because something
4078 else would have woken the daemon, and it would have reaped the completed
4081 80. If a message was submitted locally by a user whose login name contained one
4082 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4083 It caused a spool format error. I have fixed the spool reading code. A
4084 related problem was that the "from" clause in the Received: line became
4085 illegal because of the space(s). It is now covered by ${quote_local_part.
4087 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4090 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4092 83. Patch from Sieve maintainer to fix unterminated string problem in
4093 "vacation" handling.
4095 84. Some minor changes to the Linux configuration files to help with other
4096 OS variants using glibc.
4098 85. One more patch for Sieve to update vacation handling to latest spec.
4101 ----------------------------------------------------
4102 See the note above about the 4.44 and 4.50 releases.
4103 ----------------------------------------------------
4109 1. Change 4.43/35 introduced a bug that caused file counts to be
4110 incorrectly computed when quota_filecount was set in an appendfile
4113 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4114 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4117 3. Give more explanation in the error message when the command for a transport
4118 filter fails to execute.
4120 4. There are several places where Exim runs a non-Exim command in a
4121 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4122 was being done only for the command run by the queryprogram router. It is
4123 now done for all such subprocesses. The other cases are: ${run, transport
4124 filters, and the commands run by the lmtp and pipe transports.
4126 5. Some older OS have a limit of 256 on the maximum number of file
4127 descriptors. Exim was using setrlimit() to set 1000 as a large value
4128 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4129 systems. I've change it so that if it can't get 1000, it tries for 256.
4131 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4132 was an oversight, and furthermore, ever since the addition of extra
4133 controls (e.g. 4.43/32), the checks on when to allow different forms of
4134 "control" were broken. There should now be diagnostics for all cases when a
4135 control that does not make sense is encountered.
4137 7. $recipients is now available in the predata ACL (oversight).
4139 8. Tidy the search cache before the fork to do a delivery from a message
4140 received from the command line. Otherwise the child will trigger a lookup
4141 failure and thereby defer the delivery if it tries to use (for example) a
4142 cached ldap connection that the parent has called unbind on.
4144 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4145 of $address_data from the recipient verification was clobbered by the
4146 sender verification.
4148 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4149 was its contents. (It was OK if the option was not defined at all.)
4151 11. A "Completed" log line is now written for messages that are removed from
4152 the spool by the -Mrm option.
4154 12. $host_address is now set to the target address during the checking of
4155 ignore_target_hosts.
4157 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4158 being passed; this would have caused $sender_host_name to have been used if
4159 matching the list had actually called for a host name (not very likely,
4160 since this list is usually IP addresses). A host name is now passed as
4163 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4164 code that specifies a non-restarting handler (typically sigaction() in
4165 modern systems) in an attempt to fix a rare and obscure crash bug.
4167 15. Narrowed the window for a race in the daemon that could cause it to ignore
4168 SIGCHLD signals. This is not a major problem, because they are used only to
4169 wake it up if nothing else does.
4171 16. A malformed maildirsize file could cause Exim to calculate negative values
4172 for the mailbox size or file count. Odd effects could occur as a result.
4173 The maildirsize information is now recalculated if the size or filecount
4176 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4177 support for a long time. Removed HAVE_SYS_VFS_H.
4179 18. Updated exipick to current release from John Jetmore.
4181 19. Allow an empty sender to be matched against a lookup in an address list.
4182 Previously the only cases considered were a regular expression, or an
4185 20. Exim went into a mad DNS lookup loop when doing a callout where the
4186 host was specified on the transport, if the DNS lookup yielded more than
4189 21. The RFC2047 encoding function was originally intended for short strings
4190 such as real names; it was not keeping to the 75-character limit for
4191 encoded words that the RFC imposes. It now respects the limit, and
4192 generates multiple encoded words if necessary. To be on the safe side, I
4193 have increased the buffer size for the ${rfc2047: expansion operator from
4196 22. Failure to deliver a bounce message always caused it to be frozen, even if
4197 there was an errors_to setting on the router. The errors_to setting is now
4200 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4201 canonical form (fully expanded) before being placed in
4202 $sender_host_address.
4204 24. Updated eximstats to version 1.33
4206 25. Include certificate and key file names in error message when GnuTLS fails
4207 to set them up, because the GnuTLS error message doesn't include the name
4208 of the failing file when there is a problem reading it.
4210 26. Expand error message when OpenSSL has problems setting up cert/key files.
4213 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4214 (this can affect the format of dates).
4216 28. exim_tidydb, when checking for the continued existence of a message for
4217 which it has found a message-specific retry record, was not finding
4218 messages that were in split spool directories. Consequently, it was
4219 deleting retry records that should have stayed in existence.
4221 29. eximstats updated to version 1.35
4222 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4223 1.35 - bugfix such that pie charts by volume are generated correctly
4225 30. The SPA authentication driver was not abandoning authentication and moving
4226 on to the next authenticator when an expansion was forced to fail,
4227 contradicting the general specification for all authenticators. Instead it
4228 was generating a temporary error. It now behaves as specified.
4230 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4231 (the order specifies the preference for clients). The order is now AES256,
4232 AES128, 3DES, ARCFOUR128.
4234 31. Small patch to Sieve code - explicitly set From: when generating an
4237 32. Exim crashed if a remote delivery caused a very long error message to be
4238 recorded - for instance if somebody sent an entire SpamAssassin report back
4239 as a large number of 550 error lines. This bug was coincidentally fixed by
4240 increasing the size of one of Exim's internal buffers (big_buffer) that
4241 happened as part of the Exiscan merge. However, to be on the safe side, I
4242 have made the code more robust (and fixed the comments that describe what
4245 33. Some experimental protocols are using DNS PTR records for new purposes. The
4246 keys for these records are domain names, not reversed IP addresses. The
4247 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4248 leaves it alone. Component reversal etc. now happens only for IP addresses.
4251 34. The host_aton() function is supposed to be passed a string that is known
4252 to be a valid IP address. However, in the case of IPv6 addresses, it was
4253 not checking this. This is a hostage to fortune. Exim now panics and dies
4254 if the condition is not met. A case was found where this could be provoked
4255 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4256 components; fortuitously, this particular loophole had already been fixed
4257 by change 4.50/55 or 4.44/33 above.
4259 If there are any other similar loopholes, the new check in host_aton()
4260 itself should stop them being exploited. The report I received stated that
4261 data on the command line could provoke the exploit when Exim was running as
4262 exim, but did not say which command line option was involved. All I could
4263 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4264 running as the user.
4267 35. There was a buffer overflow vulnerability in the SPA authentication code
4268 (which came originally from the Samba project). I have added a test to the
4269 spa_base64_to_bits() function which I hope fixes it.
4272 36. The daemon start-up calls getloadavg() while still root for those OS that
4273 need the first call to be done as root, but it missed one case: when
4274 deliver_queue_load_max is set with deliver_drop_privilege. This is
4275 necessary for the benefit of the queue runner, because there is no re-exec
4276 when deliver_drop_privilege is set.
4278 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4279 was in use, was not putting the data itself into the right store pool;
4280 consequently, it could be overwritten for a subsequent message in the same
4281 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4284 38. Sometimes the final signoff response after QUIT could fail to get
4285 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4286 < 0 before doing a fflush(). This bug looks as though it goes back to the
4287 introduction of TLS in release 3.20, but "sometimes" must have been rare
4288 because the tests only now provoked it.
4294 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4295 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4296 Now it returns an integer. A number of calls were still expecting a T/F
4297 return. Fortuitously, in all cases, the tests worked in OK situations,
4298 which is the norm. However, things would have gone wrong on any write
4299 failures on the smtp file descriptor. This function is used when sending
4300 messages over SMTP and also when doing verify callouts.
4302 2. When Exim is called to do synchronous delivery of a locally submitted
4303 message (the -odf or -odi options), it no longer closes stderr before doing
4306 3. Implemented the mua_wrapper option.
4308 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4310 5. Implemented the functions header_remove(), header_testname(),
4311 header_add_at_position(), and receive_remove_recipient(), and exported them
4314 6. If an ACL "warn" statement specified the addition of headers, Exim already
4315 inserted X-ACL-Warn: at the start if there was no header name. However, it
4316 was not making this test for the second and subsequent header lines if
4317 there were newlines in the string. This meant that an invalid header could
4318 be inserted if Exim was badly configured.
4320 7. Allow an ACL "warn" statement to add header lines at the start or after all
4321 the Received: headers, as well as at the end.
4323 8. Added the rcpt_4xx retry error code.
4325 9. Added postmaster_mailfrom=xxx to callout verification option.
4327 10. Added mailfrom=xxxx to the callout verification option, for verify=
4330 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4331 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4333 12. Inserted some casts to stop certain compilers warning when using pointer
4334 differences as field lengths or precisions in printf-type calls (mostly
4335 affecting debugging statements).
4337 13. Added optional readline() support for -be (dynamically loaded).
4339 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4340 same clock tick as a message's arrival, so that its received time was the
4341 same as the "first fail" time on the retry record, and that message
4342 remained on the queue past the ultimate address timeout, every queue runner
4343 would try a delivery (because it was past the ultimate address timeout) but
4344 after another failure, the ultimate address timeout, which should have then
4345 bounced the address, did not kick in. This was a "< instead of <=" error;
4346 in most cases the first failure would have been in the next clock tick
4347 after the received time, and all would be well.
4349 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4350 being recognized when the domain list was tested by the match_domain
4351 condition in an expansion string.
4353 16. Added the ${str2b64: operator.
4355 17. Exim was always calling setrlimit() to set a large limit for the number of
4356 processes, without checking whether the existing limit was already
4357 adequate. (It did check for the limit on file descriptors.) Furthermore,
4358 errors from getrlimit() and setrlimit() were being ignored. Now they are
4359 logged to the main and panic logs, but Exim does carry on, to try to do its
4360 job under whatever limits there are.
4362 18. Imported PCRE 5.0.
4364 19. Trivial typo in log message " temporarily refused connection" (the leading
4367 20. If the log selector return_path_on_delivery was set and an address was
4368 redirected to /dev/null, the delivery process crashed because it assumed
4369 that a return path would always be set for a "successful" delivery. In this
4370 case, the whole delivery is bypassed as an optimization, and therefore no
4373 21. Internal re-arrangement: the function for sending a challenge and reading
4374 a response while authentication was assuming a zero-terminated challenge
4375 string. It's now changed to take a pointer and a length, to allow for
4376 binary data in such strings.
4378 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4380 23. Exim was not respecting finduser_retries when seeking the login of the
4381 uid under which it was called; it was always trying 10 times. (The default
4382 setting of finduser_retries is zero.) Also, it was sleeping after the final
4383 failure, which is pointless.
4385 24. Implemented tls_on_connect_ports.
4387 25. Implemented acl_smtp_predata.
4389 26. If the domain in control=submission is set empty, Exim assumes that the
4390 authenticated id is a complete email address when it generates From: or
4391 Sender: header lines.
4393 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4394 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4395 chown and chgrp in /bin and hostname in /usr/bin.
4397 28. Exim was keeping the "process log" file open after each use, just as it
4398 does for the main log. This opens the possibility of it remaining open for
4399 long periods when the USR1 signal hits a daemon. Occasional processlog
4400 errors were reported, that could have been caused by this. Anyway, it seems
4401 much more sensible not to leave this file open at all, so that is what now
4404 29. The long-running daemon process does not normally write to the log once it
4405 has entered its main loop, and it closes the log before doing so. This is
4406 so that log files can straightforwardly be renamed and moved. However,
4407 there are a couple of unusual error situations where the daemon does write
4408 log entries, and I had neglected to close the log afterwards.
4410 30. The text of an SMTP error response that was received during a remote
4411 delivery was being truncated at 512 bytes. This is too short for some of
4412 the long messages that one sometimes sees. I've increased the limit to
4415 31. It is now possible to make retry rules that apply only when a message has a
4416 specific sender, in particular, an empty sender.
4418 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4419 it possible to be selective about when SMTP synchronization is enforced.
4421 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4423 32. Implemented hosts_connection_nolog.
4425 33. Added an ACL for QUIT.
4427 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4430 35. Added mailbox_size and mailbox_filecount to appendfile.
4432 36. Added control = no_multiline_responses to ACLs.
4434 37. There was a bug in the logic of the code that waits for the clock to tick
4435 in the case where the clock went backwards by a substantial amount such
4436 that the microsecond fraction of "now" was more than the microsecond
4437 fraction of "then" (but the whole seconds number was less).
4439 38. Added support for the libradius Radius client library this is found on
4440 FreeBSD (previously only the radiusclient library was supported).
4446 1. When certain lookups returned multiple values in the form name=value, the
4447 quoting of the values was not always being done properly. Specifically:
4448 (a) If the value started with a double quote, but contained no whitespace,
4450 (b) If the value contained whitespace other than a space character (i.e.
4451 tabs or newlines or carriage returns) it was not quoted.
4452 This fix has been applied to the mysql and pgsql lookups by writing a
4453 separate quoting function and calling it from the lookup code. The fix
4454 should probably also be applied to nisplus, ibase and oracle lookups, but
4455 since I cannot test any of those, I have not disturbed their existing code.
4457 2. A hit in the callout cache for a specific address caused a log line with no
4458 reason for rejecting RCPT. Now it says "Previous (cached) callout
4459 verification failure".
4461 3. There was an off-by-one bug in the queryprogram router. An over-long
4462 return line was truncated at 256 instead of 255 characters, thereby
4463 overflowing its buffer with the terminating zero. As well as fixing this, I
4464 have increased the buffer size to 1024 (and made a note to document this).
4466 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4467 when Exim is waiting for an SMTP response from a remote server, Exim
4468 restarts its select() call on the socket, thereby resetting its timeout.
4469 This is not a problem when such interrupts are rare. Somebody set up a cron
4470 job to run exiwhat every 2 minutes, which is less than the normal select()
4471 timeout (5 or 10 minutes). This meant that the select() timeout never
4472 kicked in because it was always reset. I have fixed this by comparing the
4473 time when an interrupt arrives with the time at the start of the first call
4474 to select(). If more time than the timeout has elapsed, the interrupt is
4475 treated as a timeout.
4477 5. Some internal re-factoring in preparation for the addition of Sieve
4478 extensions (by MH). In particular, the "personal" test is moved to a
4479 separate function, and given an option for scanning Cc: and Bcc: (which is
4480 not set for Exim filters).
4482 6. When Exim created an email address using the login of the caller as the
4483 local part (e.g. when creating a From: or Sender: header line), it was not
4484 quoting the local part when it contained special characters such as @.
4486 7. Installed new OpenBSD configuration files.
4488 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4489 try to make them clearer.
4491 9. Callout options, other than the timeout value, were being ignored when
4492 verifying sender addresses in header lines. For example, when using
4494 verify = header_sender/callout=no_cache
4496 the cache was (incorrectly) being used.
4498 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4499 only the Cygwin environment.
4501 11. When return_path_on_delivery was set as a log selector, if different remote
4502 addresses in the same message used different return paths and parallel
4503 remote delivery occurred, the wrong values would sometimes be logged.
4504 (Whenever a remote delivery process finished, the return path value from
4505 the most recently started remote delivery process was logged.)
4507 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4508 lines when AUTH and/or TLS are in use. This is the "received protocol"
4509 field. Exim used to use "asmtp" for authenticated SMTP, without any
4510 indication (in the protocol name) for TLS use. Now it follows the RFC and
4511 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4512 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4513 names appear in log lines as well as in Received: header lines.
4515 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4516 extensions, and comparison tests, and to fix some bugs.
4518 14. Changes to the "personal" filter test:
4520 (1) The test was buggy in that it was just doing the equivalent of
4521 "contains" tests on header lines. For example, if a user's address was
4522 anne@some.where, the "personal" test would incorrectly be true for
4524 To: susanne@some.where
4526 This test is now done by extracting each address from the header in turn,
4527 and checking the entire address. Other tests that are part of "personal"
4528 are now done using regular expressions (for example, to check local parts
4529 of addresses in From: header lines).
4531 (2) The list of non-personal local parts in From: addresses has been
4532 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4533 taken from the Sieve specification recommendations.
4535 (3) If the message contains any header line starting with "List-" it is
4536 treated as non-personal.
4538 (4) The test for "circular" in the Subject: header line has been removed
4539 because it now seems ill-conceived.
4541 15. Minor typos in src/EDITME comments corrected.
4543 16. Installed latest exipick from John Jetmore.
4545 17. If headers_add on a router specified a text string that was too long for
4546 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4547 of string_sprintf() is now avoided.
4549 18. $message_body_size was not set (it was always zero) when running the DATA
4550 ACL and the local_scan() function.
4552 19. For the "mail" command in an Exim filter, no default was being set for
4553 the once_repeat time, causing a random time value to be used if "once" was
4554 specified. (If the value happened to be <= 0, no repeat happened.) The
4555 default is now 0s, meaning "never repeat". The "vacation" command was OK
4556 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4557 (I found it when inspecting the code).
4559 20. There is now an overall timeout for performing a callout verification. It
4560 defaults to 4 times the callout timeout, which applies to individual SMTP
4561 commands during the callout. The overall timeout applies when there is more
4562 than one host that can be tried. The timeout is checked before trying the
4563 next host. This prevents very long delays if there are a large number of
4564 hosts and all are timing out (e.g. when the network connections are timing
4565 out). The value of the overall timeout can be changed by specifying an
4566 additional sub-option for "callout", called "maxwait". For example:
4568 verify = sender/callout=5s,maxwait=20s
4570 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4571 to the end before writing, but this should make it even safer).
4573 22. Exim was forgetting that it had advertised PIPELINING for the second and
4574 subsequent messages on an SMTP connection. It was also not resetting its
4575 memory on STARTTLS and an internal HELO.
4577 23. When Exim logs an SMTP synchronization error within a session, it now
4578 records whether PIPELINING has been advertised or not.
4580 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4581 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4582 rather than long int.
4584 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4586 26. Added the never_mail option to autoreply.
4592 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4593 crash if the getsockname() call failed; this can happen if a connection is
4594 closed very soon after it is established. The problem was simply in the
4595 order in which certain operations were done, causing Exim to try to write
4596 to the SMTP stream before it had set up the file descriptor. The bug has
4597 been fixed by making things happen in the correct order.
4603 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4604 before the connection was closed, thus losing the rejection response.
4606 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4607 some early Solaris releases, but causes trouble in current releases where
4608 socklen_t is defined.
4610 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4613 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4616 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4617 that releasing the top of it at the end releases what was used for sub-
4618 expansions (unless the block got too big). However, discard this block if
4619 the first thing is a variable or header, so that we can use its block when
4620 it is dynamic (useful for very large $message_headers, for example).
4622 6. Lookups now cache *every* query, not just the most recent. A new, separate
4623 store pool is used for this. It can be recovered when all lookup caches are
4624 flushed. Lookups now release memory at the end of their result strings.
4625 This has involved some general refactoring of the lookup sources.
4627 7. Some code has been added to the store_xxx() functions to reduce the amount
4628 of flapping under certain conditions.
4630 8. log_incoming_interface used to affect only the <= reception log lines. Now
4631 it causes the local interface and port to be added to several more SMTP log
4632 lines, for example "SMTP connection from", and rejection lines.
4634 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4636 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4638 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4639 could be overwritten at the end of the current message (or the start of a
4640 new message if it was set in a HELO ACL). The value is now preserved for
4641 the duration of the SMTP connection.
4643 12. If a transport had a headers_rewrite setting, and a matching header line
4644 contained an unqualified address, that address was qualified, even if it
4645 did not match any rewriting rules. The underlying bug was that the values
4646 of the flags that permit the existence of unqualified sender and recipient
4647 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4648 non-local messages, and by -bnq for local messages) were not being
4649 preserved with the message after it was received.
4651 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4652 "next input=" as part of the text comprising the host identity instead of
4653 the correct text. The code was using the same buffer for two different
4654 strings. However, depending on which order the printing function evaluated
4655 its arguments, the bug did not always show up. Under Linux, for example, my
4656 test suite worked just fine.
4658 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4659 doesn't work with some older versions of Perl. It has been changed to "my",
4660 which in any case is probably the better facility to use.
4662 15. A really picky compiler found some instances of statements for creating
4663 error messages that either had too many or two few arguments for the format
4666 16. The size of the buffer for calls to the DNS resolver has been increased
4667 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4668 for addresses that have a lot of PTR records. This alleviates a problem; it
4669 does not fully solve it.
4671 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4672 buffer now truncates the list and logs the incident, which is the same
4673 action as happens when Exim is looking up a host name and its aliases.
4674 Previously in this situation something unpredictable would happen;
4675 sometimes it was "internal error: store_reset failed".
4677 18. If a server dropped the connection unexpectedly when an Exim client was
4678 using GnuTLS and trying to read a response, the client delivery process
4679 crashed while trying to generate an error log message.
4681 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4682 string, for example:
4684 warn message = H1: something\nH2: something
4686 the text was added as a single header line from Exim's point of view
4687 though it ended up OK in the delivered message. However, searching for the
4688 second and subsequent header lines using $h_h2: did not work. This has been
4689 fixed. Similarly, if a system filter added multiple headers in this way,
4690 the routers could not see them.
4692 20. Expanded the error message when iplsearch is called with an invalid key to
4693 suggest using net-iplsearch in a host list.
4695 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4696 ACLs are no longer actually imposed (and a message to that effect is
4699 22. If a "gecos" field in a passwd entry contained escaped characters, in
4700 particular, if it contained a \" sequence, Exim got it wrong when building
4701 a From: or a Sender: header from that name. A second bug also caused
4702 incorrect handling when an unquoted " was present following a character
4703 that needed quoting.
4705 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4706 was not being matched caselessly.
4708 24. Arranged for all hyphens in the exim.8 source to be escaped with
4711 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4712 properly. Recipient callout cache records were still being keyed to include
4713 the sender, even when use_sender was set false. This led to far more
4714 callouts that were necessary. The sender is no longer included in the key
4715 when use_sender is false.
4717 26. Added "control = submission" modifier to ACLs.
4719 27. Added the ${base62d: operator to decode base 62 numbers.
4721 28. dnsdb lookups can now access SRV records.
4723 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4724 the configuration file.
4726 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4727 (-v) mode. This makes the output for a verbose queue run more intelligible.
4729 31. Added a use_postmaster feature to recipient callouts.
4731 32. Added the $body_zerocount variable, containing the number of binary zero
4732 bytes in the message body.
4734 33. The time of last modification of the "new" subdirectory is now used as the
4735 "mailbox time last read" when there is a quota error for a maildir
4738 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4740 35. Added +ignore_unknown as a special item in host lists.
4742 36. Code for decoding IPv6 addresses in host lists is now included, even if
4743 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4744 address was recognized as an IP address, but was then not correctly decoded
4745 into binary, causing unexpected and incorrect effects when compared with
4752 1. Very minor rewording of debugging text in manualroute to say "list of
4753 hosts" instead of "hostlist".
4755 2. If verify=header_syntax was set, and a header line with an unqualified
4756 address (no domain) and a large number of spaces between the end of the
4757 name and the colon was received, the reception process suffered a buffer
4758 overflow, and (when I tested it) crashed. This was caused by some obsolete
4759 code that should have been removed. The fix is to remove it!
4761 3. When running in the test harness, delay a bit after writing a bounce
4762 message to get a bit more predictability in the log output.
4764 4. Added a call to search_tidyup() just before forking a reception process. In
4765 theory, someone could use a lookup in the expansion of smtp_accept_max_
4766 per_host which, without the tidyup, could leave open a database connection.
4768 5. Added the variables $recipient_data and $sender_data which get set from a
4769 lookup success in an ACL "recipients" or "senders" condition, or a router
4770 "senders" option, similar to $domain_data and $local_part_data.
4772 6. Moved the writing of debug_print from before to after the "senders" test
4775 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4776 problems for message scanning, either using a data ACL, or using
4777 local_scan() because the Received: header was not generated till after they
4778 were called (in order to set the time as the time of reception completion).
4779 I have revised the way this works. The header is now generated after the
4780 body is received, but before the ACL or local_scan() are called. After they
4781 are run, the timestamp in the header is updated.
4787 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4788 before starting a queue runner without re-exec. This happened only when
4789 deliver_drop_privilege was set or when the Exim user was set to root. The
4790 effect of the bug was that timeouts during subsequent deliveries caused
4791 crashes instead of being properly handled. The handler is now left at its
4792 default (and expected) setting.
4794 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4795 message, again when deliver_drop_privilege is set or Exim is run as root.
4796 The bug described in (1) was not present in this case, but the tidying up
4797 of the other signals was missing. I have made the two cases consistent.
4799 3. The ignore_target_hosts setting on a manualroute router was being ignored
4800 for hosts that were looked up using the /MX notation.
4802 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4805 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4806 operated on the sender address. After changing the $sender_address to <>
4807 for the sender address verify, Exim was re-instated it as the original
4808 (before rewriting) address, but remembering that it had rewritten it, so it
4809 wasn't rewriting it again. This bug also had the effect of breaking the
4810 sender address verification caching when the sender address was rewritten.
4812 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4813 This has been changed so that if the ip literal address matches
4814 ignore_target_hosts, the router declines.
4816 7. Added expansion conditions match_domain, match_address, and match_local_
4817 part (NOT match_host).
4819 8. The placeholder for the Received: header didn't have a length field set.
4821 9. Added code to Exim itself and to exim_lock to test for a specific race
4822 condition that could lead to file corruption when using MBX delivery. The
4823 issue is with the lockfile that is created in /tmp. If this file is removed
4824 after a process has opened it but before that process has acquired a lock,
4825 there is the potential for a second process to recreate the file and also
4826 acquire a lock. This could lead to two Exim processes writing to the file
4827 at the same time. The added code performs the same test as UW imapd; it
4828 checks after acquiring the lock that its file descriptor still refers to
4829 the same named file.
4831 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4832 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4835 11. Added the smtp_active_hostname option. If used, this will typically be made
4836 to depend on the incoming interface address. Because $interface_address is
4837 not set up until the daemon has forked a reception process, error responses
4838 that can happen earlier (such as "too many connections") no longer contain
4841 12. If an expansion in a condition on a "warn" statement fails because a lookup
4842 defers, the "warn" statement is abandoned, and the next ACL statement is
4843 processed. Previously this caused the whole ACL to be aborted.
4845 13. Added the iplsearch lookup type.
4847 14. Added ident_timeout as a log selector.
4849 15. Added tls_certificate_verified as a log selector.
4851 16. Added a global option tls_require_ciphers (compare the smtp transport
4852 option of the same name). This controls incoming TLS connections.
4854 17. I finally figured out how to make tls_require_ciphers do a similar thing
4855 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4856 before starting the TLS session.
4858 18. Tabs are now shown as \t in -bP output.
4860 19. If the log selector return_path_on_delivery was set, Exim crashed when
4861 bouncing a message because it had too many Received: header lines.
4863 20. If two routers both had headers_remove settings, and the first one included
4864 a superfluous trailing colon, the final name in the first list and the
4865 first name in the second list were incorrectly joined into one item (with a
4866 colon in the middle).
4872 1. Added -C and -D options to the exinext utility, mainly to make it easier
4873 to include in the automated testing, but these could be helpful when
4874 multiple configurations are in use.
4876 2. The exinext utility was not formatting the output nicely when there was
4877 an alternate port involved in the retry record key, nor when there was a
4878 message id as well (for retries that were specific to a specific message
4879 and a specific host). It was also confused by IPv6 addresses, because of
4880 the additional colons they contain. I have fixed the IPv4 problem, and
4881 patched it up to do a reasonable job for IPv6.
4883 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4884 delivery, the log line now contains "pipelined" if PIPELINING was used.
4886 4. An SMTP transport process used to panic and die if the bind() call to set
4887 an explicit outgoing interface failed. This has been changed; it is now
4888 treated in the same way as a connect() failure.
4890 5. A reference to $sender_host_name in the part of a conditional expansion
4891 that was being skipped was still causing a DNS lookup. This no longer
4894 6. The def: expansion condition was not recognizing references to header lines
4895 that used bh_ and bheader_.
4897 7. Added the _cache feature to named lists.
4899 8. The code for checking quota_filecount in the appendfile transport was
4900 allowing one more file than it should have been.
4902 9. For compatibility with Sendmail, the command line option
4910 and sets the incoming protocol and host name (for trusted callers). The
4911 host name and its colon can be omitted when only the protocol is to be set.
4912 Note the Exim already has two private options, -pd and -ps, that refer to
4913 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4914 "s", but I don't think that's a major issue.
4916 10. A number of refactoring changes to the code, none of which should affect
4919 (a) The number of logging options was getting close to filling up the
4920 32-bit word that was used as a bit map. I have split them into two classes:
4921 those that are passed in the argument to log_write(), and those that are
4922 only ever tested independently outside of that function. These are now in
4923 separate 32-bit words, so there is plenty of room for expansion again.
4924 There is no change in the user interface or the logging behaviour.
4926 (b) When building, for example, log lines, the code previously used a
4927 macro that called string_cat() twice, in order to add two strings. This is
4928 not really sufficiently general. Furthermore, there was one instance where
4929 it was actually wrong because one of the argument was used twice, and in
4930 one call a function was used. (As it happened, calling the function twice
4931 did not affect the overall behaviour.) The macro has been replaced by a
4932 function that can join an arbitrary number of extra strings onto a growing
4935 (c) The code for expansion conditions now uses a table and a binary chop
4936 instead of a serial search (which was left over from when there were very
4937 few conditions). Also, it now recognizes conditions like "pam" even when
4938 the relevant support is not compiled in: a suitably worded error message is
4939 given if an attempt is made to use such a condition.
4941 11. Added ${time_interval:xxxxx}.
4943 12. A bug was causing one of the ddress fields not to be passed back correctly
4944 from remote delivery subprocesses. The field in question was not being
4945 subsequently used, so this caused to problems in practice.
4947 13. Added new log selectors queue_time and deliver_time.
4949 14. Might have fixed a bug in maildirsizefile handling that threw up
4950 "unexpected character" debug warnings, and recalculated the data
4951 unnecessarily. In any case, I expanded the warning message to give more
4954 15. Added the message "Restricted characters in address" to the statements in
4955 the default ACL that block characters like @ and % in local parts.
4957 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4958 Three changes have been made:
4960 (a) There was a serious bug; a negative response to MAIL caused the whole
4961 recipient domain to be cached as invalid, thereby blocking all messages
4962 to all local parts at the same domain, from all senders. This bug has
4963 been fixed. The domain is no longer cached after a negative response to
4964 MAIL if the sender used is not empty.
4966 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4969 (c) A new callout option, "use_sender" has been added for people who want
4970 the modified behaviour.
4976 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4979 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4980 indeed breaks things for older releases.
4982 3. Added additional logging to the case where there is a problem reading data
4983 from a filter that is running in a subprocess using a pipe, in order to
4984 try to track down a specific problem.
4986 4. Testing facility fudge: when running in the test harness and attempting
4987 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4988 getting "No route to host". Convert this to a timeout.
4990 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4993 6. Some OS don't have socklen_t but use size_t instead. This affects the
4994 fifth argument of getsockopt() amongst other things. This is now
4995 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4996 can be set for individual OS. I have set it for SunOS5, OSF1, and
4997 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4998 some earlier ones do not.
5000 7. Change 4.30/15 was not doing the test caselessly.
5002 8. The standard form for an IPv6 address literal was being rejected by address
5003 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5004 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5005 this, as well as the form without the "IPv6" on the front (but only when
5006 address literals are enabled, of course).
5008 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5010 10. Exim crashed if a message with an empty sender address specified by -f
5011 encountered a router with an errors_to setting. This could be provoked only
5012 by a command such as
5016 where an empty string was supplied; "<>" did not hit this bug.
5018 11. Installed PCRE release 4.5.
5020 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5021 remained set. It is now erased.
5023 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5024 times from message ids (which are base 36 rather than the normal 62).
5026 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5027 were being counted as actual protocol errors, and logged if the log
5028 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5029 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5030 and DATA following a set of rejected RCPTs do not count as protocol errors.
5031 In other words, Exim assumes they were pipelined, though this may not
5032 actually be the case. Of course, in all cases the client gets an
5033 appropriate error code.
5035 15. If a lookup fails in an ACL condition, a message about the failure may
5036 be available; it is used if testing the ACL cannot continue, because most
5037 such messages specify what the cause of the deferral is. However, some
5038 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5039 that caused an old message to be retained and used if a later statement
5040 caused a defer, replacing the real cause of the deferral.
5042 16. If an IP address had so many PTR records that the DNS lookup buffer
5043 was not large enough to hold them, Exim could crash while trying to process
5044 the truncated data. It now detects and logs this case.
5046 17. Further to 4.21/58, another change has been made: if (and only if) the
5047 first line of a message (the first header line) ends with CRLF, a bare LF
5048 in a subsequent header line has a space inserted after it, so as not to
5049 terminate the header.
5051 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5052 unnecessarily, used atoi() instead of strtol(), and didn't check the
5053 termination when getting file sizes from file names by regex.
5055 19. Completely re-implemented the support for maildirsize files, in the light
5056 of a number of problems with the previous contributed implementation
5057 (4.30/29). In particular:
5059 . If the quota is zero, the maildirsize file is maintained, but no quota is
5062 . If the maildir directory does not exist, it is created before any attempt
5063 to write a maildirsize file.
5065 . The quota value in the file is just a cache; if the quota is changed in
5066 the transport, the new value overrides.
5068 . A regular expression is available for excluding directories from the
5071 20. The autoreply transport checks the characters in options that define the
5072 message's headers; it allows continued headers, but it was checking with
5073 isspace() after an embedded newline instead of explicitly looking for a
5076 21. If all the "regular" hosts to which an address was routed had passed their
5077 expiry times, and had not reached their retry times, the address was
5078 bounced, even if fallback hosts were defined. Now Exim should go on to try
5081 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5082 equivalent code in the SMTP transport. Some hosts send humungous responses
5083 to HELO/EHLO, more than 1024 it seems.
5085 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5086 gives compiler warnings in some environments. I've now done it "properly",
5089 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5090 (because of problems with the built-in one) was declared to return uschar *
5091 instead of char *, causing compiler failure.
5093 25. Fixed a file descriptor leak when processing alias/forward files.
5095 26. Fixed a minor format string issue in dbfn.c.
5097 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5099 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5100 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5103 29. When a sender address is verified, it is cached, to save repeating the test
5104 when there is more than one recipient in a message. However, when the
5105 verification involves a callout, it is possible for different callout
5106 options to be set for different recipients. It is too complicated to keep
5107 track of this in the cache, so now Exim always runs a verification when a
5108 callout is required, relying on the callout cache for the optimization.
5109 The overhead is duplication of the address routing, but this should not be
5112 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5113 to be verified with callout=postmaster, and the main callout worked but the
5114 postmaster check failed, the verification correctly failed. However, if a
5115 subsequent RCPT command asked for sender verification *without* the
5116 postmaster check, incorrect caching caused this verification also to fail,
5119 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5120 it was not caching the DNS options (qualify_single, search_parents) that
5121 were used when the lookup failed. A subsequent lookup with different
5122 options therefore always gave the same answer, though there were cases
5123 where it should not have. (Example: a "domains = !$mx_any" option on a
5124 dnslookup router: the "domains" option is always processed without any
5125 widening, but the router might have qualify_single set.) Now Exim uses the
5126 cached value only when the same options are set.
5128 32. Added John Jetmore's "exipick" utility to the distribution.
5130 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5131 than a timeout (e.g. a certificate is required, and is not provided), an
5132 Exim server now closes the connection immediately. Previously it waited for
5133 the client to close - but if the client is SSL, it seems that they each
5134 wait for each other, leading to a delay before one of them times out.
5136 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5137 maintained 0.8.x compatibility because I don't think many are using it, and
5138 it is clearly obsolete.
5140 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5143 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5144 client certificate was expired. A simple patch fixes this, though I don't
5145 understand the full logic of why the verify callback is called multiple
5148 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5149 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5150 which causes problems with some clients (such as the Certicom SSL Plus
5151 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5152 disables the coutermeasure allowing Eudora to connect."
5154 38. Exim was not checking that a write() to a log file succeeded. This could
5155 lead to Bad Things if a log got too big, in particular if it hit a file
5156 size limit. Exim now panics and dies if it cannot write to a log file, just
5157 as it does if it cannot open a log file.
5159 39. Modified OS/Makefile-Linux so that it now contains
5161 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5163 The two -D definitions ensure that Exim is compiled with large file
5164 support, which makes it possible to handle log files that are bigger than
5167 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5168 instance) a domain was checked against a named list that involved a lookup,
5169 causing $domain_data to be set, then another domain was checked against the
5170 same list, then the first domain was re-checked, the value of $domain_data
5171 after the final check could be wrong. In particular, if the second check
5172 failed, it could be set empty. This bug probably also applied to
5175 41. The strip_trailing_dot option was not being applied to the address given
5176 with the -f command-line option.
5178 42. The code for reading a message's header from the spool was incrementing
5179 $received_count, but never initializing it. This meant that the value was
5180 incorrect (doubled) while delivering a message in the same process in which
5181 it was received. In the most common configuration of Exim, this never
5182 happens - a fresh exec is done - but it can happen when
5183 deliver_drop_privilege is set.
5185 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5186 it now includes up to 150 characters of the unexpected data in the log
5189 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5190 and building data strings. The size of both of these buffers was 10 000
5191 bytes - far larger than anybody would *ever* want, thought I. Needless to
5192 say, somebody hit the limit. I have increased the maximum line length to
5193 20 000 and the maximum data length of concatenated lines to 100 000. I have
5194 also fixed two bugs, because there was no checking on these buffers. Tsk,
5195 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5196 buffer is too small.
5198 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5199 lsearch lookups. Now it does.
5201 46. When parsing a route_list item in a manualroute router, a fixed-length
5202 buffer was used for the list of hosts. I made this 1024 bytes long,
5203 thinking that nobody would ever have a list of hosts that long. Wrong.
5204 Somebody had a whole pile of complicated expansion conditions, and the
5205 string was silently truncated, leading to an expansion error. It turns out
5206 that it is easier to change to an unlimited length (owing to other changes
5207 that have happened since this code was originally written) than to build
5208 structure for giving a limitation error. The length of the item that
5209 expands into the list of hosts is now unlimited.
5211 47. The lsearch lookup could not handle data where the length of text line was
5212 more than 4095 characters. Such lines were truncated, leading to shortened
5213 data being returned. It should now handle lines of any length.
5215 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5216 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5219 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5220 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5221 no longer happens in comment lines. A list of replacements is now placed
5222 at the head of all of the source files, except those whose only change is
5223 to replace PERL_COMMAND in the very first #! line.
5225 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5226 messages on the queue, with a bottom-up merge sort, using code contributed
5227 by Michael Haardt. This should make operations like -bp somewhat faster on
5228 large queues. It won't affect queue runners, except when queue_run_in_order
5231 51. Installed eximstats 1.31 in the distribution.
5233 52. Added support for SRV lookups to the dnslookup router.
5235 53. If an ACL referred to $message_body or $message_body_end, the value was not
5236 reset for any messages that followed in the same SMTP session.
5238 54. The store-handling optimization for building very long strings was not
5239 differentiating between the different store pools. I don't think this
5240 actually made any difference in practice, but I've tidied it.
5242 55. While running the routers to verify a sender address, $sender_address
5243 was still set to the sender address. This is wrong, because when routing to
5244 send a bounce to the sender, it would be empty. Therefore, I have changed
5245 it so that, while verifying a sender address, $sender_address is set to <>.
5246 (There is no change to what happens when verifying a recipient address.)
5248 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5249 target A or AAAA records (if not already returned) without resetting the
5250 qualify_single or search_parents options of the DNS resolver. These are
5251 inappropriate in this case because the targets of MX and SRV records must
5252 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5253 target that, when qualified, matched something in the local domain. These
5254 two options are now turned off when doing these lookups.
5256 57. It seems that at least some releases of Reiserfs (which does not have the
5257 concept of a fixed number of inodes) returns zero and not -1 for the
5258 number of available inodes. This interacted badly with check_spool_inodes,
5259 which assumed that -1 was the "no such thing" setting. What I have done is
5260 to check that the total number of inodes is greater than zero before doing
5261 the test of how many are available.
5263 58. When a "warn" ACL statement has a log_message modifier, the message is
5264 remembered, and not repeated. This is to avoid a lot of repetition when a
5265 message has many recipients that cause the same warning to be written.
5266 Howewer, Exim was preserving the list of already written lines for an
5267 entire SMTP session, which doesn't seem right. The memory is now reset if a
5268 new message is started.
5270 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5271 debugging output unless log_rewrite was also set.
5273 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5274 of (int)(handle) when we know that handle contains (void *)(-1).
5276 61. The Exim daemon panic-logs an error return when it closes the incoming
5277 connection. However "connection reset by peer" seems to be common, and
5278 isn't really an error worthy of noting specially, so that particular error
5281 62. When Exim is trying to find all the local interfaces, it used to panic and
5282 die if the ioctl to get the interface flags failed. However, it seems that
5283 on at least one OS (Solaris 9) it is possible to have an interface that is
5284 included in the list of interfaces, but for which you get a failure error
5285 for this call. This happens when the interface is not "plumbed" into a
5286 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5287 failure of the "get flags" call assumes that the interface is down.
5289 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5290 makes life easier for people who are doing arithmetic on fields extracted
5291 from dates, where you often get leading zeros that should not be
5292 interpreted as octal.
5294 64. Added qualify_domain to the redirect router, to override the global
5297 65. If a pathologically long header line contained very many addresses (the
5298 report of this problem mentioned 10 000) and each of them was rewritten,
5299 Exim could use up a very large amount of memory. (It kept on making new
5300 copies of the header line as it rewrote, and never released the old ones.)
5301 At the expense of a bit more processing, the header rewriting function has
5302 been changed so that it no longer eats memory in this way.
5304 66. The generation of the Received: header has been moved from the time that a
5305 message starts to be received, to the time that it finishes. The timestamp
5306 in the Received: header should now be very close to that of the <= log
5307 line. There are two side-effects of this change:
5309 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5310 logged header lines no longer include the local Received: line, because
5311 it has not yet been created. The same applies to a copy of the message
5312 that is returned to a non-SMTP sender when a message is rejected.
5314 (b) When a filter file is tested using -bf, no additional Received: header
5315 is added to the test message. After some thought, I decided that this
5318 This change does not affect the value of $received_for. It is still set
5319 after address rewriting, but before local_scan() is called.
5321 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5323 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5324 gave an unhelpful panic error message, and a defer error. I have managed to
5325 change this behaviour so that it now rejects any supplied certificate,
5326 which seems right, as the list of acceptable certificates is empty.
5328 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5329 gave an unhelpful defer error. I have not managed to make this reject any
5330 supplied certificates, but the error message it gives is "no certificate
5331 supplied", which is not helpful.
5333 70. exigrep's output now also includes lines that are not associated with any
5334 message, but which match the given pattern. Implemented by a patch from
5335 Martin Sluka, which also tidied up the Perl a bit.
5337 71. Recipient callout verification, like sender verification, was using <> in
5338 the MAIL FROM command. This isn't really the right thing, since the actual
5339 sender may affect whether the remote host accepts the recipient or not. I
5340 have changed it to use the actual sender in the callout; this means that
5341 the cache record is now keyed on a recipient/sender pair, not just the
5342 recipient address. There doesn't seem to be a real danger of callout loops,
5343 since a callout by the remote host to check the sender would use <>.
5344 [SEE ABOVE: changed after hitting problems.]
5346 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5347 temporary errors. However, in the case of such a code being given after
5348 the end of a data transmission (i.e. after ".") Exim was failing to write
5349 a retry record for the message. (Yes, there was some broken host that was
5350 actually sending 8xx at this point.)
5352 73. An unknown lookup type in a host list could cause Exim to panic-die when
5353 the list was checked. (An example that provoked this was putting <; in the
5354 middle of a list instead of at the start.) If this happened during a DATA
5355 ACL check, a -D file could be left lying around. This kind of configuration
5356 error no longer causes Exim to die; instead it causes a defer errror. The
5357 incident is still logged to the main and panic logs.
5359 74. Buglet left over from Exim 3 conversion. The message "too many messages
5360 in one connection" was written to the rejectlog but not the mainlog, except
5361 when address rewriting (yes!) was being logged.
5363 75. Added write_rejectlog option.
5365 76. When a system filter was run not as root (that is, when system_filter_user
5366 was set), the values of the $n variables were not being returned to the
5367 main process; thus, they were not subsequently available in the $sn
5370 77. Added +return_path_on_delivery log selector.
5372 78. A connection timeout was being treated differently from recipients deferred
5373 when testing hosts_max_try with a message that was older than the host's
5374 retry timeout. (The host should not be counted, thus allowing all hosts to
5375 be tried at least once before bouncing.) This may have been the cause of an
5376 occasionally reported bug whereby a message would remain on the queue
5377 longer than the retry timeout, but would be bounced if a delivery was
5378 forced. I say "may" because I never totally pinned down the problem;
5379 setting up timeout/retry tests is difficult. See also the next item.
5381 79. The ultimate address timeout was not being applied to errors that involved
5382 a combination of host plus message (for example, a timeout on a MAIL
5383 command). When an address resolved to a number of possible hosts, and they
5384 were not all tried for each delivery (e.g. because of hosts_max_try), a
5385 message could remain on the queue longer than the retry timeout.
5387 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5390 81. Fixed an obscure SMTP outgoing bug which required at least the following
5391 conditions: (a) there was another message waiting for the same server;
5392 (b) the server returned 5xx to all RCPT commands in the first message so
5393 that the message was not completed; (c) the server dropped the connection
5394 or gave a negative response to the RSET that Exim sends to abort the
5395 transaction. The observed case was a dropped connection after DATA that had
5396 been sent in pipelining mode. That is, the server had advertised PIPELINING
5397 but was not implementing it correctly. The effect of the bug was incorrect
5398 behaviour, such as trying another host, and this could lead to a crash.
5404 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5405 and daemon.c were passed as pointers to ints; they should have been
5406 pointers to socklen_t variables (which are typically unsigned ints).
5408 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5411 3. Fixed a really odd bug that affected only the testing scheme; patching a
5412 certain fixed string in the binary changed the value of another string that
5413 happened to be identical to the end of the original first string.
5415 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5416 name", it returns that address as the IP address. On some operating
5417 systems (e.g. Solaris), it also passes back the IP address string as the
5418 "host name". However, on others (e.g. Linux), it passes back an empty
5419 string. Exim wasn't checking for this, and was changing the host name to an
5420 empty string, assuming it had been canonicized.
5422 5. Although rare, it is permitted to have more than one PTR record for a given
5423 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5424 all the names associated with an address, because they do in Solaris.
5425 However, it seems that they do not in Linux for data that comes from the
5426 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5427 I found this out when I moved to a new Linux workstation and tried to run
5428 the Exim test suite.
5430 To get round this problem I have changed the code so that it now does its
5431 own call to the DNS to look up PTR records when searching for a host name.
5432 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5433 addresses that are only in /etc/hosts are still found.
5435 This behaviour is, however, controlled by an option called host_lookup_
5436 order, which defaults to "bydns:byaddr". If people want to use the other
5437 order, or indeed, just use one or the other means of lookup, they can
5438 specify it in this variable.
5440 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5441 some operating systems, this comes back from gethostbyaddr() as an empty
5442 string, and this is what Exim used to test for. However, it seems that in
5443 other systems, "." is yielded. Exim now tests for this case too.
5445 7. The values of check_spool_space and check_log_space are now held internally
5446 as a number of kilobytes instead of an absolute number of bytes. If a
5447 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5448 kilobyte. This means that much larger values can be stored.
5450 8. Exim monitor: an attempt to get the action menu when not actually pointing
5451 at a message produces an empty menu entitled "No message selected". This
5452 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5453 no entries in it ("Shell widget menu has zero width and/or height"). So I
5454 have added a single, blank menu entry in this case.
5456 9. Added ${quote_local_part.
5458 10. MIME decoding is now applied to the contents of Subject: header lines when
5461 11. Now that a reference to $sender_host_address automatically causes a reverse
5462 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5463 host lookup before query-style lookups in lists that might use this
5464 variable. This has therefore been abolished, and the "net-" prefix is no
5465 longer necessary for query-style lookups.
5467 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5468 have been a typo for LFLAGS, so it has been changed.
5470 13. The install script calls Exim with "-C /dev/null" in order to find the
5471 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5472 to be output. Howeve, since Exim outputs its version number before the
5473 error, it didn't break the script. It just looked ugly. I fixed this by
5474 always allowing "-C /dev/null" if the caller is root.
5476 14. Ignore overlarge ACL variable number when reading spool file - insurance
5477 against a later release with more variables having written the file.
5479 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5480 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5481 this, as well as the form without the "IPv6" on the front.
5483 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5484 OS/Makefile-Darwin file.
5486 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5487 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5489 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5490 variable. However, some OS use ints for inodes. I've added cast to long int
5491 to get rid of the compiler warning.
5493 19. I had forgotten to lock out "/../" in configuration file names when
5494 ALT_CONFIG_PREFIX was set.
5496 20. Routers used for verification do not need to specify transports. However,
5497 if such a router generated a host list, and callout was configured, Exim
5498 crashed, because it could not find a port number from the (non-existent)
5499 transport. It now assumes port 25 in this circumstance.
5501 21. Added the -t option to exigrep.
5503 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5504 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5505 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5506 them is set, LOOKUP_LSEARCH is forced.
5508 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5510 24. Added sender and host information to the "rejected by local_scan()" log
5511 line; previously there was no indication of these.
5513 25. Added .include_if_exists.
5515 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5516 Linux. It turns out that not all file systems support this. Apparently some
5517 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5518 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5519 not supported on the file descriptor, is now ignored when Exim is trying to
5520 sync a directory. This applies only to Linux.
5522 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5524 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5525 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5528 29. Added support for maildirsize files from supplied patch (modified a bit).
5530 30. The use of :fail: followed by an empty string could lead Exim to respond to
5531 sender verification failures with (e.g.):
5533 550 Verification failed for <xxx>
5534 550 Sender verify failed
5536 where the first response line was missing the '-' that indicates it is not
5537 the final line of the response.
5539 31. The loop for finding the name of the user that called Exim had a hardwired
5540 limit of 10; it now uses the value of finduser_retries, which is used for
5541 all other user lookups.
5543 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5546 33. Exim was neglecting to zero errno before one call of strtol() when
5547 expanding a string and expecting an integer value. On some systems this
5548 resulted in spurious "integer overflow" errors. Also, it was casting the
5549 result into an int without checking.
5551 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5552 did not work. The code looks as if it has *never* worked, though it appears
5553 to have been documented since at least releast 1.62. I have made it work.
5555 35. The "timeout_DNS" error in retry rules, also documented since at least
5556 1.62, also never worked. As it isn't clear exactly what this means, and
5557 clearly it isn't a major issue, I have abolished the feature by treating it
5558 as "timeout", and writing a warning to the main and panic logs.
5560 36. The display of retry rules for -brt wasn't always showing the error code
5563 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5564 timeout_connect_A, timeout_connect_MX.
5566 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5567 to the empty sender.
5569 39. The daemon was not analysing the content of -oX till after it had closed
5570 stderr and disconnected from the controlling terminal. This meant that any
5571 syntax errors were only noted on the panic log, and the return code from
5572 the command was 0. By re-arranging the code a little, I've made the
5573 decoding happen first, so such errors now appear on stderr, and the return
5574 code is 1. However, the actual setting up of the sockets still happens in
5575 the disconnected process, so errors there are still only recorded on the
5578 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5579 connections (as happens on some IP stacks) was logged at start up time as
5580 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5581 it from "IPv6 and IPv4", which means that two separate sockets are being
5584 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5585 says whether AF_INET or AF_INET6 was passed as an argument.
5587 42. Exiwhat output was messed up when time zones were included in log
5590 43. Exiwhat now gives more information about the daemon's listening ports,
5591 and whether -tls-on-connect was used.
5593 44. The "port" option of the smtp transport is now expanded.
5595 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5596 silently ignored. Now an error message is written to the main and panic
5599 46. There's a new ACL modifier called "logwrite" which writes to a log file
5600 as soon as it is encountered.
5602 47. Added $local_user_uid and $local_user_gid at routing time.
5604 48. Exim crashed when trying to verify a sender address that was being
5607 49. Exim was recognizing only a space character after ".include". It now also
5608 recognizes a tab character.
5610 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5611 extracting the relevant information from the specification. The man page no
5612 longer contains scrambled data for the -d option, and I've added a section
5613 at the front about calling Exim under different names.
5615 51. Added "extra_headers" argument to the "mail" command in filter files.
5617 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5620 53. Installed eximstats 1.29.
5622 54. Added transport_filter_timeout as a generic transport option.
5624 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5625 Cc: header lines. This was required by RFC 822, but it not required by RFC
5628 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5629 incoming messages that did not have them. Now it does so only if the
5630 message originates locally, that is, if there is no associated remote host
5631 address. When Resent- header lines are present, this applies to the Resent-
5632 lines rather than the non-Resent- lines.
5634 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5635 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5637 58. Messages for configuration errors now include the name of the main
5638 configuration file - useful now that there may be more than one file in a
5639 list (.included file names were always shown).
5641 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5642 for those rare installations that do not start the daemon as root or run it
5643 setuid root. I've cut out the call to initgroups() if the daemon is not
5646 60. The Exim user and group can now be bound into the binary as text strings
5647 that are looked up at the start of Exim's processing.
5649 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5651 62. Added $mailstore_basename variable.
5653 63. Installed patch to sieve.c from Michael Haardt.
5655 64. When Exim failed to open the panic log after failing to open the main log,
5656 the original message it was trying to log was written to stderr and debug
5657 output, but if they were not available (the usual case in production), it
5658 was lost. Now it is written to syslog before the two lines that record the
5659 failures to open the logs.
5661 65. Users' Exim filters run in subprocesses under the user's uid. It is
5662 possible for a "deliver" command or an alias in a "personal" command to
5663 provoke an address rewrite. If logging of address rewriting is configured,
5664 this fails because the process is not running as root or exim. There may be
5665 a better way of dealing with this, but for the moment (because 4.30 needs
5666 to be released), I have disabled address rewrite logging when running a
5667 filter in a non-root, non-exim process.
5673 1. The buildconfig auxiliary program wasn't quoting the value set for
5674 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5675 not defined. This bug was masked in 4.22 by the effect that was fixed in
5678 2. Some messages that were rejected after a message id was allocated were
5679 shown as "incomplete" by exigrep. It no longer does this for messages that
5680 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5682 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5683 have allow_domain_literals set, the ID did not get logged in the <= line.
5684 Domain literals are now always recognized in Message-ID: header lines.
5686 4. The first argument for a ${extract expansion item is the key name or field
5687 number. Leading and trailing spaces in this item were not being ignored,
5688 causing some misleading effects.
5690 5. When deliver_drop_privilege was set, single queue runner processes started
5691 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5692 same command in the process it spins off) were not dropping privilege.
5694 6. When the daemon running as "exim" started a queue runner, it always
5695 re-executed Exim in the spun-off process. This is a waste of effort when
5696 deliver_drop_privilege is set. The new process now just calls the
5697 queue-runner function directly.
5703 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5706 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5707 ignored. Though the use of -oP was forcing the writing of a pid file, it
5708 was always written to the default place.
5710 3. If the message "no IP address found for host xxxx" is generated during
5711 incoming verification, it is now followed by identification of the incoming
5712 connection (so you can more easily find what provoked it).
5714 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5716 5. Added some features to "harden" Exim a bit more against certain attacks:
5718 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5719 be put in Local/Makefile. This is like the never_users runtime option,
5720 but it cannot be overridden. The default setting is "root".
5722 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5723 prefix string with which any file named in a -C command line option
5726 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5727 is retained for -C and -D only if the caller of Exim is root. Without
5728 it, the exim user may also use -C and -D and retain privilege.
5730 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5731 command line option is disabled.
5733 6. Macro names set by the -D option must start with an upper case letter, just
5734 like macro names defined in the configuration file.
5736 7. Added "dereference=" facility to LDAP.
5738 8. Two instances of the typo "uknown" in the source files are fixed.
5740 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5741 the Configure-Makefile script screwed up while processing it.
5743 10. Incorporated PCRE 4.4.
5745 11. The SMTP synchronization check was not operating right at the start of an
5746 SMTP session. For example, it could not catch a HELO sent before the client
5747 waited for the greeting. There is now a check for outstanding input at the
5748 point when the greeting is written. Because of the duplex, asynchronous
5749 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5750 way, but not yet received, when the check is performed.
5752 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5753 on TCP/IP sockets, because this apparently causes some broken clients to
5756 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5757 unchanged) from the Cygwin maintainer.
5759 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5760 was supported instead of testing for mbx. Effectively a typo.
5762 15. The spa authenticator server code was not checking that the input it
5763 received was valid base64.
5765 16. The debug output line for the "set" modifier in ACLs was not showing the
5766 name of the variable that was being set.
5768 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5770 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5771 lookup on its own. Something else was needed to trigger the lookup. For
5772 example, a match in host_lookup or the need for a host name in a host list.
5773 Now, if $sender_host_name is referenced and the host name has not yet been
5774 looked up, a lookup is performed. If the lookup fails, the variable remains
5775 empty, and $host_lookup_failed is set to "1".
5777 19. Added "eqi" as a case-independent comparison operator.
5779 20. The saslauthd authentication condition could segfault if neither service
5780 nor realm was specified.
5782 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5783 error message that was logged was misleading, and incoming SMTP
5784 connections were dropped. The message is now more accurate, and temporary
5785 errors are given to SMTP connections.
5787 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5788 (except RSET) with a 421 error, until QUIT is received. However, it was
5789 failing to send a response to QUIT.
5791 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5792 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5793 ACL is now run after the helo_try_verify_hosts code.
5795 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5796 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5797 case-independent, so other case variants are also recognized). Apparently
5798 some systems use these upper case variants.
5800 25. If more than two messages were waiting for the same host, and a transport
5801 filter was specified for the transport, Exim sent two messages over the
5802 same TCP/IP connection, and then failed with "socket operation on non-
5803 socket" when it tried to send the third.
5805 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5807 27. The extern definition of crypt16() in expand.c was not being excluded when
5808 the OS had its own crypt16() function.
5810 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5811 as a preferred synonym for return_size_limit, both as an option and as an
5814 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5816 30. Changed the default configuration ACL to relax the local part checking rule
5817 for addresses that are not in any local domains. For these addresses,
5818 slashes and pipe symbols are allowed within local parts, but the sequence
5819 /../ is explicitly forbidden.
5821 31. SPA server authentication was not clearing the challenge buffer before
5824 32. log_message in a "warn" ACL statement was writing to the reject log as
5825 well as to the main log, which contradicts the documentation and doesn't
5826 seem right (because no rejection is happening). So I have stopped it.
5828 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5829 However, I am unable to do any testing of this.
5831 34. Fixed an infelicity in the appendfile transport. When checking directories
5832 for a mailbox, to see if any needed to be created, it was accidentally
5833 using path names with one or more superfluous leading slashes; tracing
5834 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5836 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5837 contents are added to the log line that is written for every discarded
5838 recipient. (Previously a log_message setting was ignored.)
5840 36. The ${quote: operator now quotes the string if it is empty.
5842 37. The install script runs exim in order to find its version number. If for
5843 some reason other than non-existence or emptiness, which it checks, it
5844 could not run './exim', it was installing it with an empty version number,
5845 i.e. as "exim-". This error state is now caught, and the installation is
5848 38. An argument was missing from the function that creates an error message
5849 when Exim fails to connect to the socket for saslauthd authentication.
5850 This could cause Exim to crash, or give a corrupted message.
5852 39. Added isip, isip4, and isip6 to ${if conditions.
5854 40. The ACL variables $acl_xx are now saved with the message, and can be
5855 accessed later in routers, transports, and filters.
5857 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5858 strings in the file are not string-expanded.
5860 42. If a MAIL command specified a SIZE value that was too large to fit into an
5861 int variable, the check against message_size_limit failed. Such values are
5862 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5863 day this will have to be increased, but I don't think I want to be around
5864 when emails are that large.
5871 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5872 iconv() is not standard in FreeBSD.
5874 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5875 IPv6 enabled. The observed symptom was a segmentation fault on return from
5876 the function os_common_find_running_interfaces() in src/os.c.
5878 3. In the check_special_case() function in daemon.c I had used "errno" as an
5879 argument name, which causes warnings on some systems. This was basically a
5880 typo, since it was named "eno" in the comments!
5882 4. The code that waits for the clock to tick (at a resolution of some fraction
5883 of a second) so as to ensure message-id uniqueness was always waiting for
5884 at least one whole tick, when it could have waited for less. [This is
5885 almost certainly not relevant at current processor speeds, where it is
5886 unlikely to ever wait at all. But we try to future-proof.]
5888 5. The function that sleeps for a time interval that includes fractions of a
5889 second contained a race. It did not block SIGALRM between setting the
5890 timer, and suspending (a couple of lines later). If the interval was short
5891 and the sigsuspend() was delayed until after it had expired, the suspension
5892 never ended. On busy systems this could lead to processes getting stuck for
5895 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5896 process, before it forks any delivery processes. The open lookup caching
5897 mechanism meant that the open file or database connection was passed into
5898 the delivery process. The problem was that delivery processes always tidy
5899 up cached lookup data. This could cause a problem for the next delivery
5900 process started by the queue runner, because the external queue runner
5901 process does not know about the closure. So the next delivery process
5902 still has data in the lookup cache. In the case of a file lookup, there was
5903 no problem because closing a file descriptor in a subprocess doesn't affect
5904 the parent. However, if the lookup was caching a connection to a database,
5905 the connection was closed, and the second delivery process was likely to
5906 see errors such as "PGSQL: query failed: server closed the connection
5907 unexpectedly". The problem has been fixed by closing all cached lookups
5908 in a queue runner before running a delivery process.
5910 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5911 seem to have the "const" qualifier which it has on other OS. I've
5914 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5915 *of the same type* (client or server) with the same public name that an
5916 error should be diagnosed.
5918 9. When Exim looked up a host name for an IP address, but failed to find the
5919 original IP address when looking up the host name (a safety check), it
5920 output the message "<ip address> does not match any IP for NULL", which was
5921 confusing, to say the least. The bug was that the host name should have
5922 appeared instead of "NULL".
5924 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5925 user that is built into the binary, and the -C or -D options is used, root
5926 privilege is dropped before the configuration file is read. In addition,
5927 logging is switched to stderr instead of the normal log files. If the
5928 configuration then re-defines the Exim user, the unprivileged environment
5929 is probably not what is expected, so Exim logs a panic warning message (but
5932 However, if deliver_drop_privilege is set, the unprivileged state may well
5933 be exactly what is intended, so the warning has been cut out in that case,
5934 and Exim is allowed to try to write to its normal log files.
5940 1. smtp_return_error_details was not giving details for temporary sender
5941 or receiver verification errors.
5943 2. Diagnose a configuration error if two authenticators have the same public
5946 3. Exim used not to create the message log file for a message until the first
5947 delivery attempt. This could be confusing when incoming messages were held
5948 for policy or load reasons. The message log file is now created at the time
5949 the message is received, and an initial "Received" line is written to it.
5951 4. The automatically generated man page for command line options had a minor
5952 bug that caused no ill effects; however, a more serious problem was that
5953 the procedure for building the man page automatically didn't always
5954 operate. Consequently, release 4.20 contains an out-of-date version. This
5955 shouldn't happen again.
5957 5. When building Exim with embedded Perl support, the script that builds the
5958 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5959 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5961 6. The freeze_tell option was not being used for messages that were frozen on
5962 arrival, either by an ACL or by local_scan().
5964 7. Added the smtp_incomplete_transaction log selector.
5966 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5967 was accepting AUTH without a new EHLO.
5969 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5970 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5973 10. Logging of TCP/IP connections (when configured) now happens in the main
5974 daemon process instead of the child process, so that the TCP/IP connection
5975 count is more accurate (but it can never be perfect).
5977 11. The use of "drop" in a nested ACL was not being handled correctly in the
5978 outer ACL. Now, if condition failure induced by the nested "drop" causes
5979 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5980 or "require"), the connection is dropped.
5982 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5983 that yield "discard" can now be used with an "accept" or a "discard" verb,
5984 but an error is generated for any others (because I can't see a useful way
5985 to define what should happen).
5987 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5988 are now processed in the same way as lines in the Exim configuration file.
5989 In particular, continuation lines are supported.
5991 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5993 15. Added -ti meaning -t -i.
5995 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5996 domains, and warn by logging if others are found.
5998 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5999 returned from GIFCONF when Exim is trying to find the list of interfaces on
6000 a host. The code in os.c has been modified to copy each ifreq to an aligned
6001 structure in all cases.
6003 Also, in some cases, the returned ifreq's were being copied to a 'struct
6004 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6005 means the last couple of bytes of an IPv6 address could be chopped if the
6006 ifreq contained only a normal sockaddr (14 bytes storage).
6008 18. Named domain lists were not supported in the hosts_treat_as_local option.
6009 An entry such as +xxxx was not recognized, and was treated as a literal
6012 19. Ensure that header lines added by a DATA ACL are included in the reject log
6013 if the ACL subsequently rejects the message.
6015 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6016 MD5 (which is deprecated).
6018 21. When testing a filter file using -bf, Exim was writing a message when it
6019 took the sender from a "From " line in the message, but it was not doing so
6020 when it took $return_path from a Return-Path: header line. It now does.
6022 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6023 with a valid header line field name (a series of printing characters
6024 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6026 23. Changed "disc" in the source to "disk" to conform to the documentation and
6027 the book and for uniformity.
6029 24. Ignore Sendmail's -Ooption=value command line item.
6031 25. When execve() failed while trying to run a command in a pipe transport,
6032 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6033 could be confused with a return value of 69 from the command itself. This
6034 has been changed to 127, the value the shell returns if it is asked to run
6035 a non-existent command. The wording for the related log line suggests a
6036 non-existent command as the problem.
6038 26. If received_header_text expands to an empty string, do not add a Received:
6039 header line to the message. (Well, it adds a token one on the spool, but
6040 marks it "old" so that it doesn't get used or transmitted.)
6042 27. Installed eximstats 1.28 (addition of -nt option).
6044 28. There was no check for failure on the call to getsockname() in the daemon
6045 code. This can fail if there is a shortage of resources on the system, with
6046 ENOMEM, for example. A temporary error is now given on failure.
6048 29. Contrary to the C standard, it seems that in some environments, the
6049 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6050 program. Exim now does this explicitly; it affects the formatting of
6051 timestamps using strftime().
6053 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6054 complaints. I've now initialized all the variables, to avoid this.
6056 32. Header lines added by a system filter were not being "seen" during
6057 transport-time rewrites.
6059 33. The info_callback() function passed to OpenSSL is set up with type void
6060 (*)(SSL *, int, int), as described somewhere. However, when calling the
6061 function (actually a macro) that sets it up, the type void(*)() is
6062 expected. I've put in a cast to prevent warnings from picky compilers.
6064 34. If a DNS black list lookup found a CNAME record, but there were no A
6065 records associated with the domain it pointed at, Exim crashed.
6067 35. If a DNS black list lookup returned more than one A record, Exim ignored
6068 all but the first. It now scans all returned addresses if a particular IP
6069 value is being sought. In this situation, the contents of the
6070 $dnslist_value variable are a list of all the addresses, separated by a
6073 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6074 to accept a host name and all its aliases if the forward lookup for any of
6075 them yielded the IP address of the incoming connection. Now it accepts only
6076 those names whose forward lookup yields the correct IP address. Any other
6077 names are discarded. This closes a loophole whereby a rogue DNS
6078 administrator could create reverse DNS records to break through a
6079 wildcarded host restriction in an ACL.
6081 37. If a user filter or a system filter that ran in a subprocess used any of
6082 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6083 the wrong values were passed to the pipe command ($thisaddress had the
6084 value of $0, $0 had the value of $1, etc). This bug was introduced by
6085 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6087 38. Improved the line breaking for long SMTP error messages from ACLs.
6088 Previously, if there was no break point between 40 and 75 characters, Exim
6089 left the rest of the message alone. Two changes have been made: (a) I've
6090 reduced the minimum length to 35 characters; (b) if it can't find a break
6091 point between 35 and 75 characters, it looks ahead and uses the first one
6092 that it finds. This may give the occasional overlong line, but at least the
6093 remaining text gets split now.
6095 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6096 file descriptors might be low, and that setting 1000 would always raise it.
6097 It turns out that in some environments, the limit is already over 1000 and
6098 that lowering it causes trouble. So now Exim takes care not to decrease it.
6100 40. When delivering a message, the value of $return_path is set to $sender_
6101 address at the start of routing (routers may change the value). By an
6102 oversight, this default was not being set up when an address was tested by
6103 -bt or -bv, which affected the outcome if any router or filter referred to
6106 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6107 log only when "message" or "log_message" are set. However, if one of the
6108 conditions was an address verification, or a call to a nested ACL, the
6109 messages generated by the underlying test were being passed through. This
6110 no longer happens. The underlying message is available in $acl_verify_
6111 message for both "message" and "log_message" expansions, so it can be
6112 passed through if needed.
6114 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6115 new expansion $bh_ to give the encoded byte string without charset
6116 translation. Translation happens only if iconv() is available; HAVE_ICONV
6117 indicates this at build time. HEADERS_CHARSET gives the charset to
6118 translate to; headers_charset can change it in the configuration, and
6119 "headers charset" can change it in an individual filter file.
6121 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6122 that creates RFC 2047 encoded "words" labels them as that charset instead
6123 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6124 expansion operator; (ii) when Exim creates a From: line for a local
6125 message; (iii) when a header line is rewritten to include a "phrase" part.
6127 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6128 buggy, causing it to skip the first lines of messages whose message ID
6129 ended in 'D'. This would not have bitten before Exim release 4.14, because
6130 message IDs were unlikely to end in 'D' before then. The effect was to have
6131 incorrect size information for certain domains.
6133 45. #include "config.h" was missing at the start of the crypt16.c module. This
6134 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6137 46. If there was a timeout during a "random" callout check, Exim treated it as
6138 a failure of the random address, and carried on sending RSET and the real
6139 address. If the delay was just some slowness somewhere, the response to the
6140 original RCPT would be taken as a response to RSET and so on, causing
6141 mayhem of various kinds.
6143 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6144 when I implemented it. It didn't allow for the fact that some option values
6145 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6146 the right test for positive values.
6148 48. Domain names in DNS records are case-independent. Exim always looks them up
6149 in lower case. Some resolvers return domain names in exactly the case they
6150 appear in the zone file, that is, they may contain uppercase letters. Not
6151 all resolvers do this - some return always lower case. Exim was treating a
6152 change of case by a resolver as a change of domain, similar to a widening
6153 of a domain abbreviation. This triggered its re-routing code and so it was
6154 trying to route what was effectively the same domain again. This normally
6155 caused routing to fail (because the router wouldn't handle the domain
6156 twice). Now Exim checks for this case specially, and just changes the
6157 casing of the domain that it ultimately uses when it transmits the message
6160 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6163 50. If a filter generated a file delivery with a non-absolute name (possible if
6164 no home directory exists for the router), the forbid_file option was not
6167 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6168 the existing equality matching.
6170 52. Exim was using ints instead of ino_t variables in some places where it was
6171 dealing with inode numbers.
6173 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6174 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6175 called TMPDIR, and if it finds it is different, it changes its value.
6177 54. The smtp_printf() function is now made available to local_scan() so
6178 additional output lines can be written before returning. There is also an
6179 smtp_fflush() function to enable the detection of a dropped connection.
6180 The variables smtp_input and smtp_batched_input are exported to
6183 55. Changed the default runtime configuration: the message "Unknown user"
6184 has been removed from the ACL, and instead placed on the localuser router,
6185 using the cannot_route_message feature. This means that any verification
6186 failures that generate their own messages won't get overridden. Similarly,
6187 the "Unrouteable address" message that was in the ACL for unverifiable
6188 relay addresses has also been removed.
6190 56. Added hosts_avoid_esmtp to the smtp transport.
6192 57. The exicyclog script was not checking for the esoteric option
6193 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6194 will work only if exicyclog is run under the appropriate euid.
6196 58. Following a discussion on the list, the rules by which Exim recognises line
6197 endings on incoming messages have been changed. The -dropcr and drop_cr
6198 options are now no-ops, retained only for backwards compatibility. The
6199 following line terminators are recognized: LF CRLF CR. However, special
6200 processing applies to CR:
6202 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6203 nor a local message in the state where . is a terminator.
6205 (ii) If a bare CR is encountered in a header line, an extra space is added
6206 after the line terminator so as not to end the header. The reasoning
6207 behind this is that bare CRs in header lines are most likely either
6208 to be mistakes, or people trying to play silly games.
6210 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6211 was being incorrectly given as 18 bytes larger than it should have been.
6212 This is a VOB (very old bug).
6214 60. This may never have affected anything current, but just in case it has:
6215 When the local host is found other than at the start of a list of hosts,
6216 the local host, those with the same MX, and any that follow, are discarded.
6217 When the list in question was part of a longer list of hosts, the following
6218 hosts (not currently being processed) were also being discarded. This no
6219 longer happens. I'm not sure if this situation could ever has previously
6222 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6225 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6226 header. This is something that is recommended in a new Internet Draft, and
6227 is something that is documented as being done by Sendmail. There are two
6228 possible values. For messages generated by the autoreply transport, Exim
6231 Auto-Submitted: auto-replied
6233 whereas for all other generated messages (e.g. bounces) it adds
6235 Auto-Submitted: auto-generated
6237 63. The "personal" condition in filters now includes a test for the
6238 Auto-Submitted: header. If it contains the string "auto-" the message it
6239 not considered personal.
6241 64. Added rcpt_include_affixes as a generic transport option.
6243 65. Added queue_only_override (default true).
6245 66. Added the syslog_duplication option.
6247 67. If what should have been the first header line of a message consisted of
6248 a space followed by a colon, Exim was mis-interpreting it as a header line.
6249 It isn't of course - it is syntactically invalid and should therefore be
6250 treated as the start of the message body. The misbehaviour could have
6251 caused a number of strange effects, including loss of data in subsequent
6252 header lines, and spool format errors.
6254 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6255 client host had authenticated. This control can now be exercised by an ACL
6256 for more flexibility.
6258 69. By default, callouts do not happen when testing with -bh. There is now a
6259 variant, -bhc, which does actually run the callout code, including
6260 consulting and updating the callout cache.
6262 70. Added support for saslauthd authentication, courtesy of Alexander
6265 71. If statvfs() failed on the spool or log directories while checking their
6266 size for availability, Exim confusingly gave the error "space shortage".
6267 Furthermore, in debugging mode it crashed with a floating point exception.
6268 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6269 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6270 serious problem, Exim now writes to the main and panic logs when this
6271 happens, with details of the failure. It then refuses to accept the
6272 incoming message, giving the message "spool directory problem" or "log
6273 directory problem" with a 421 code for SMTP messages.
6275 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6276 0, 1, and 2 exist, because some OS complain for execs without them (see
6277 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6278 descriptors. However, the code omitted to check that the open succeeded,
6279 causing mysterious errors if for some reason the permissions on /dev/null
6280 got screwed. Now Exim writes a message to the main and panic logs, and
6281 bombs out if it can't open /dev/null.
6283 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6284 interact so that it is all more flexible. It is supposed to remain
6285 backwards compatible. Also added extra_local_interfaces.
6287 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6288 to bomb out with an assertion failure - to the client this appears as a
6289 connection drop. This problem occurs in the part of the code that was taken
6290 from the Samba project. Fortunately, the assertion is in a very simple
6291 function, so I have fixed this by reproducing the function inline in the
6292 one place where it is called, and arranging for authentication to fail
6293 instead of killing the process with assert().
6295 75. The SPA client code was not working when the server requested OEM rather
6296 than Unicode encoding.
6298 76. Added code to make require_files with a specific uid setting more usable in
6299 the case where statting the file as root fails - usually a non-root-mounted
6300 NFS file system. When this happens and the failure is EACCES, Exim now
6301 forks a subprocess and does the per-uid checking as the relevant uid.
6303 77. Added process_log_path.
6305 78. If log_file_path was not explicitly set, a setting of check_log_space or
6306 check_log_inodes was ignored.
6308 79. If a space check for the spool or log partitions fails, the incident is now
6309 logged. Of course, in the latter case the data may get lost...
6311 80. Added the %p formatting code to string_format() so that it can be used to
6312 print addresses in debug_print(). Adjusted all the address printing in the
6313 debugging in store.c to use %p rather than %d.
6315 81. There was a concern that a line of code in smtp_in.c could overflow a
6316 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6317 initially expressed, the concern was not well-founded, because trailing
6318 spaces are removed early. However, if the trailing spaces were followed by
6319 a NULL, they did not get removed, so the overflow was possible. Two fixes
6322 (a) I re-wrote the offending code in a cleaner fashion.
6323 (b) If an incoming SMTP command contains a NULL character, it is rejected
6326 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6327 runs initgroups(), so that if the Exim user is in any additional groups,
6328 they will be used during message reception.
6334 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob