1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
10 JH/02 The smtp transport option "multi_domain" is now expanded.
12 JH/03 The smtp transport now requests PRDR by default, if the server offers
15 JH/04 Certificate name checking on server certificates, when exim is a client,
16 is now done by default. The transport option tls_verify_cert_hostnames
17 can be used to disable this per-host. The build option
18 EXPERIMENTAL_CERTNAMES is withdrawn.
20 JH/05 The value of the tls_verify_certificates smtp transport and main options
21 default to the word "system" to access the system default CA bundle.
22 For GnuTLS, only version 3.0.20 or later.
24 JH/06 Verification of the server certificate for a TLS connection is now tried
25 (but not required) by default. The verification status is now logged by
26 default, for both outbound TLS and client-certificate supplying inbound
29 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
32 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
33 Status Notification (bounce) messages are now MIME format per RFC 3464.
34 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
35 under the control of the dsn_advertise_hosts option, and routers may
36 have a dsn_lasthop option.
38 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
39 default, modifiable by a malware= option. The list separator for
40 the options can now be changed in the usual way. Bug 68.
42 JH/10 The smtp_receive_timeout main option is now expanded before use.
44 JH/11 The incoming_interface log option now also enables logging of the
45 local interface on delivery outgoing connections.
47 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
48 if the interface and destination host and port all match.
50 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
53 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
54 Patch from Andrew Lewis.
56 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
57 now supports optional time-restrictions, weighting, and priority
58 modifiers per server. Patch originally by <rommer@active.by>.
60 JH/16 The spamd_address main option now supports a mixed list of local
61 and remote servers. Remote servers can be IPv6 addresses, and
64 JH/17 Bug 68: The spamd_address main option now supports an optional
65 timeout value per server.
67 JH/18 Bug 1581: Router and transport options headers_add/remove can
68 now have the list separator specified.
70 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
73 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
76 JH/21 Support for the A6 type of dns record is withdrawn.
78 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
79 rather than the verbs used.
81 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
82 from 255 to 1024 chars.
84 JH/24 Verification callouts now attempt to use TLS by default.
86 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
87 are generic router options now. The defaults didn't change.
89 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
90 Original patch from Alexander Shikoff, worked over by JH.
92 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
93 files if ZCAT_COMMAND is not executable.
95 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
97 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
99 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
100 Normally benign, it bites when the pair was led to by a CNAME;
101 modern usage is to not canoicalize the domain to a CNAME target
102 (and we were inconsistent anyway for A-only vs AAAA+A).
104 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
106 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
107 when evaluating $sender_host_dnssec.
109 JH/31 Check the HELO verification lookup for DNSSEC, adding new
110 $sender_helo_dnssec variable.
112 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
114 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
116 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
118 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
119 documented as working, but never had. Support all but $spam_report.
124 TL/01 When running the test suite, the README says that variables such as
125 no_msglog_check are global and can be placed anywhere in a specific
126 test's script, however it was observed that placement needed to be near
127 the beginning for it to behave that way. Changed the runtest perl
128 script to read through the entire script once to detect and set these
129 variables, reset to the beginning of the script, and then run through
130 the script parsing/test process like normal.
132 TL/02 The BSD's have an arc4random API. One of the functions to induce
133 adding randomness was arc4random_stir(), but it has been removed in
134 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
135 function when detected.
137 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
138 cause callback expansion.
140 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
141 syntax errors in an expansion can be treated as a string instead of
142 logging or causing an error, due to the internal use of bool_lax
143 instead of bool when processing it.
145 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
146 server certificates when making smtp deliveries.
148 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
150 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
152 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
154 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
155 Merged patch from Sebastian Wiedenroth.
157 JH/05 Fix results-pipe from transport process. Several recipients, combined
158 with certificate use, exposed issues where response data items split
159 over buffer boundaries were not parsed properly. This eventually
160 resulted in duplicates being sent. This issue only became common enough
161 to notice due to the introduction of conection certificate information,
162 the item size being so much larger. Found and fixed by Wolfgang Breyha.
164 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
165 size buffer was used, resulting in syntax errors when an expansion
168 JH/07 Add support for directories of certificates when compiled with a GnuTLS
169 version 3.3.6 or later.
171 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
172 is EXPERIMENTAL_EVENT, the main-configuration and transport options
173 both become "event_action", the variables become $event_name, $event_data
174 and $event_defer_errno. There is a new variable $verify_mode, usable in
175 routers, transports and related events. The tls:cert event is now also
176 raised for inbound connections, if the main configuration event_action
179 TL/06 In test suite, disable OCSP for old versions of openssl which contained
180 early OCSP support, but no stapling (appears to be less than 1.0.0).
182 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
183 server certificate names available under the smtp transport option
184 "tls_verify_cert_hostname" now do not permit multi-component wildcard
187 JH/10 Time-related extraction expansions from certificates now use the main
188 option "timezone" setting for output formatting, and are consistent
189 between OpenSSL and GnuTLS compilations. Bug 1541.
191 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
192 encoded parameter in the incoming message. Bug 1558.
194 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
195 include certificate info, eximon was claiming there were spoolfile
198 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
200 JH/14 Log delivery-related information more consistently, using the sequence
201 "H=<name> [<ip>]" wherever possible.
203 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
204 are problematic for Debian distribution, omit them from the release
207 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
209 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
212 JH/17 Fix a null-indirection in certextract expansions when a nondefault
213 output list separator was used.
218 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
219 checkers that were complaining about end of non-void function with no
222 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
223 This was a regression intruduced in 4.83 by another bugfix.
225 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
227 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
228 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
234 TF/01 Correctly close the server side of TLS when forking for delivery.
236 When a message was received over SMTP with TLS, Exim failed to clear up
237 the incoming connection properly after forking off the child process to
238 deliver the message. In some situations the subsequent outgoing
239 delivery connection happened to have the same fd number as the incoming
240 connection previously had. Exim would try to use TLS and fail, logging
241 a "Bad file descriptor" error.
243 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
244 utilities have not been installed.
246 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
247 temporary space as the ACL may create new global variables.
249 TL/01 LDAP support uses per connection or global context settings, depending
250 upon the detected version of the libraries at build time.
252 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
253 to extract and use the src ip:port in logging and expansions as if it
254 were a direct connection from the outside internet. PPv2 support was
255 updated based on HAProxy spec change in May 2014.
257 JH/02 Add ${listextract {number}{list}{success}{fail}}.
259 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
260 Properly escape header and check for NULL return.
262 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
265 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
267 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
268 characters in header names, implemented as a verify condition.
269 Contributed by Michael Fischer v. Mollard.
271 TL/05 Rename SPF condition results err_perm and err_temp to standardized
272 results permerror and temperror. Previous values are deprecated but
273 still accepted. In a future release, err_perm and err_temp will be
274 completely removed, which will be a backward incompatibility if the
275 ACL tests for either of these two old results. Patch contributed by
276 user bes-internal on the mailing list.
278 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
280 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
281 selectors, in both main and reject logs.
283 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
286 JH/07 Add malware type "sock" for talking to simple daemon.
288 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
290 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
291 routers/transports under cutthrough routing.
293 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
294 numbers. Touch up "bool" conditional to keep the same definition.
296 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
298 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
300 JH/12 Expand items in router/transport headers_add or headers_remove lists
301 individually rather than the list as a whole. Bug 1452.
303 Required for reasonable handling of multiple headers_ options when
304 they may be empty; requires that headers_remove items with embedded
305 colons must have them doubled (or the list-separator changed).
307 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
308 view the policy declared in the DMARC record. Currently, $dmarc_status
309 is a combined value of both the record presence and the result of the
312 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
314 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
315 dnslookup router and the smtp transport (applying to the forward
318 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
319 of ldap servers used for a specific lookup. Patch provided by Heiko
322 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
323 New variable $lookup_dnssec_authenticated for observability.
325 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
326 Patch submitted by Lars Timman.
328 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
330 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
331 Requires trusted mode and valid format message id, aborts otherwise.
332 Patch contributed by Heiko Schlichting.
334 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
335 certextract with support for various fields. Bug 1358.
337 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
338 is requested by default, modifiable by smtp transport option
341 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
342 operate on certificate variables to give certificate fingerprints
343 Also new ${sha256:cert_variable}.
345 JH/23 The PRDR feature is moved from being Experimental into the mainline.
347 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
348 Christian Aistleitner.
350 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
352 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
353 file. Patch from Wolfgang Breyha.
355 JH/25 Expand the coverage of the delivery $host and $host_address to
356 client authenticators run in verify callout. Bug 1476.
358 JH/26 Port service names are now accepted for tls_on_connect_ports, to
359 align with daemon_smtp_ports. Bug 72.
361 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
362 support and error reporting did not work properly.
364 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
365 and is readable. Patch from Andrew Colin Kissa.
367 TL/14 Enhance documentation of ${run expansion and how it parses the
368 commandline after expansion, particularly in the case when an
369 unquoted variable expansion results in an empty value.
371 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
373 PP/02 Fix internal collision of T_APL on systems which support RFC3123
374 by renaming away from it. Addresses GH issue 15, reported by
377 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
379 TL/15 SECURITY: prevent double expansion in math comparison functions
380 (can expand unsanitized data). Not remotely exploitable.
387 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
389 PP/02 Make -n do something, by making it not do something.
390 When combined with -bP, the name of an option is not output.
392 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
395 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
396 $sender_host_name and config options to manage this, and basic check
399 PP/05 DSCP support for outbound connections and control modifier for inbound.
401 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
402 (Only plugin which currently uses this is kerberos4, which nobody should
403 be using, but we should make it available and other future plugins might
404 conceivably use it, even though it would break NAT; stuff *should* be
405 using channel bindings instead).
407 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
408 name; added for Sendmail compatibility; requires admin caller.
409 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
410 just ignore it); requires trusted caller.
411 Also parse but ignore: -Ac -Am -X<logfile>
414 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
416 TL/02 Add +smtp_confirmation as a default logging option.
418 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
419 Patch by Magnus Holmgren from 2007-02-20.
421 TL/04 Bugzilla 1281 - Spec typo.
422 Bugzilla 1283 - Spec typo.
423 Bugzilla 1290 - Spec grammar fixes.
425 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
427 TL/06 Add Experimental DMARC support using libopendmarc libraries.
429 TL/07 Fix an out of order global option causing a segfault. Reported to dev
430 mailing list by by Dmitry Isaikin.
432 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
434 JH/02 Support "G" suffix to numbers in ${if comparisons.
436 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
438 NM/01 Bugzilla 1197 - Spec typo
439 Bugzilla 1196 - Spec examples corrections
441 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
443 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
444 gnutls_enable_pkcs11, but renamed to more accurately indicate its
447 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
448 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
450 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
451 "acl {{name}{arg}...}", and optional args on acl condition
454 JH/05 Permit multiple router/transport headers_add/remove lines.
456 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
458 JH/07 Avoid using a waiting database for a single-message-only transport.
459 Performance patch from Paul Fisher. Bugzilla 1262.
461 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
464 JH/09 Add $headers_added variable, with content from use of ACL modifier
465 add_header (but not yet added to the message). Bugzilla 199.
467 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
468 Pulled from Bugzilla 817 by Wolfgang Breyha.
470 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
472 (nb: this is the same fix as in Exim 4.80.1)
474 JH/11 Add A= logging on delivery lines, and a client_set_id option on
477 JH/12 Add optional authenticated_sender logging to A= and a log_selector
480 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
482 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
483 advertise SMTP AUTH mechanism to us, instead of a generic
484 protocol violation error. Also, make Exim more robust to bad
485 data from the Dovecot auth socket.
487 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
489 When a queue runner is handling a message, Exim first routes the
490 recipient addresses, during which it prunes them based on the retry
491 hints database. After that it attempts to deliver the message to
492 any remaining recipients. It then updates the hints database using
495 So if a recipient address works intermittently, it can get repeatedly
496 deferred at routing time. The retry hints record remains fresh so the
497 address never reaches the final cutoff time.
499 This is a fairly common occurrence when a user is bumping up against
500 their storage quota. Exim had some logic in its local delivery code
501 to deal with this. However it did not apply to per-recipient defers
502 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
504 This change adds a proper retry rule check during routing so that the
505 final cutoff time is checked against the message's age. We only do
506 this check if there is an address retry record and there is not a
507 domain retry record; this implies that previous attempts to handle
508 the address had the retry_use_local_parts option turned on. We use
509 this as an approximation for the destination being like a local
510 delivery, as in LMTP.
512 I suspect this new check makes the old local delivery cutoff check
513 redundant, but I have not verified this so I left the code in place.
515 TF/02 Correct gecos expansion when From: is a prefix of the username.
517 Test 0254 submits a message to Exim with the header
521 When I ran the test suite under the user fanf2, Exim expanded
522 the header to contain my full name, whereas it should have added
523 a Resent-Sender: header. It erroneously treats any prefix of the
524 username as equal to the username.
526 This change corrects that bug.
528 GF/01 DCC debug and logging tidyup
529 Error conditions log to paniclog rather than rejectlog.
530 Debug lines prefixed by "DCC: " to remove any ambiguity.
532 TF/03 Avoid unnecessary rebuilds of lookup-related code.
534 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
535 Bug spotted by Jeremy Harris; was flawed since initial commit.
536 Would have resulted in OCSP responses post-SNI triggering an Exim
537 NULL dereference and crash.
539 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
541 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
542 Bug detection, analysis and fix by Samuel Thibault.
543 Bugzilla 1331, Debian bug #698092.
545 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
547 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
548 Server implementation by Todd Lyons, client by JH.
549 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
550 config variable "prdr_enable" controls whether the server
551 advertises the facility. If the client requests PRDR a new
552 acl_data_smtp_prdr ACL is called once for each recipient, after
553 the body content is received and before the acl_smtp_data ACL.
554 The client is controlled by bolth of: a hosts_try_prdr option
555 on the smtp transport, and the server advertisement.
556 Default client logging of deliveries and rejections involving
557 PRDR are flagged with the string "PRDR".
559 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
560 fclose(). Diagnosis by Todd Lyons.
562 PP/17 Update configure.default to handle IPv6 localhost better.
563 Patch by Alain Williams (plus minor tweaks).
566 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
567 This is now consistent with GnuTLS, and is now documented: the
568 previous undocumented portable approach to treating the option as
569 unset was to force an expansion failure. That still works, and
570 an empty string is now equivalent.
572 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
573 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
574 not performing validation itself.
576 PP/20 Added force_command boolean option to pipe transport.
577 Patch from Nick Koston, of cPanel Inc.
579 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
582 TF/04 Added udpsend ACL modifer and hexquote expansion operator
584 PP/21 Fix eximon continuous updating with timestamped log-files.
585 Broken in a format-string cleanup in 4.80, missed when I repaired the
586 other false fix of the same issue.
587 Report and fix from Heiko Schlichting.
590 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
591 Report from Prashanth Katuri.
593 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
594 It's SecureTransport, so affects any MacOS clients which use the
595 system-integrated TLS libraries, including email clients.
597 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
598 using a MIME ACL for non-SMTP local injection.
599 Report and assistance in diagnosis by Warren Baker.
601 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
603 JH/16 Fix comparisons for 64b. Bugzilla 1385.
605 TL/09 Add expansion variable $authenticated_fail_id to keep track of
606 last id that failed so it may be referenced in subsequent ACL's.
608 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
611 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
612 ldap library initialization, allowing self-signed CA's to be
613 used. Also properly sets require_cert option later in code by
614 using NULL (global ldap config) instead of ldap handle (per
615 session). Bug diagnosis and testing by alxgomz.
617 TL/12 Enhanced documentation in the ratelimit.pl script provided in
618 the src/util/ subdirectory.
620 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
621 renamed to Transport Post Delivery Action by Jeremy Harris, as
624 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
625 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
626 redis_servers = needs to be configured which will be used by the redis
627 lookup. Patch from Warren Baker, of The Packet Hub.
629 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
631 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
632 hostname or reverse DNS when processing a host list. Used suggestions
633 from multiple comments on this bug.
635 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
637 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
638 Missed a few lines, added it to make the runtest require no keyboard
641 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
642 contains upper case chars. Make router use caseful_local_part.
644 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
645 support when GnuTLS has been built with p11-kit.
651 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
653 This, or similar/improved, will also be change PP/11 of 4.82.
659 PP/01 Handle short writes when writing local log-files.
660 In practice, only affects FreeBSD (8 onwards).
661 Bugzilla 1053, with thanks to Dmitry Isaikin.
663 NM/01 Bugzilla 949 - Documentation tweak
665 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
668 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
670 PP/02 Implemented gsasl authenticator.
672 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
674 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
675 `pkg-config foo` for cflags/libs.
677 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
678 with rest of GSASL and with heimdal_gssapi.
680 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
681 `pkg-config foo` for cflags/libs for the TLS implementation.
683 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
684 properties get this fed in as external SSF. A number of robustness
685 and debugging improvements to the cyrus_sasl authenticator.
687 PP/08 cyrus_sasl server now expands the server_realm option.
689 PP/09 Bugzilla 1214 - Log authentication information in reject log.
690 Patch by Jeremy Harris.
692 PP/10 Added dbmjz lookup type.
694 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
696 PP/12 MAIL args handles TAB as well as SP, for better interop with
697 non-compliant senders.
698 Analysis and variant patch by Todd Lyons.
700 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
701 Bug report from Lars Müller <lars@samba.org> (via SUSE),
702 Patch from Dirk Mueller <dmueller@suse.com>
704 PP/13 tls_peerdn now print-escaped for spool files.
705 Observed some $tls_peerdn in wild which contained \n, which resulted
706 in spool file corruption.
708 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
709 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
710 or write after TLS renegotiation, which otherwise led to messages
713 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
714 as a tracking header (ie: a signed header comes before the signature).
715 Patch from Wolfgang Breyha.
717 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
718 comma-sep list; embedded commas doubled.
720 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
722 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
724 Report and patch from Dmitry Banschikov.
726 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
727 Removed SSL_clear() after SSL_new() which led to protocol negotiation
728 failures. We appear to now support TLS1.1+ with Exim.
730 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
731 lets Exim select keys and certificates based upon TLS SNI from client.
732 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
733 before an outbound SMTP session. New log_selector, +tls_sni.
735 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
736 NULL dereference. Report and patch from Alun Jones.
738 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
739 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
740 Not seeing resolver debug output on NetBSD, but suspect this is a
741 resolver implementation change.
743 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
744 Left warnings. Added "eximon gdb" invocation mode.
746 PP/21 Defaulting "accept_8bitmime" to true, not false.
748 PP/22 Added -bw for inetd wait mode support.
750 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
751 locate the relevant includes and libraries. Made this the default.
753 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
754 Bugzilla 1246, report and most of solution from Tomasz Kusy.
756 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
757 This may cause build issues on older platforms.
759 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
760 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
761 gnutls_require_mac & gnutls_require_protocols (no longer supported).
762 Added SNI support via GnuTLS too.
763 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
765 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
767 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
768 Applied second patch from Janne, implementing suggestion to default
769 multiple-strings-in-record handling to match SPF spec.
771 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
773 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
774 read-only, out of scope).
775 Patch from Wolfgang Breyha, report from Stuart Northfield.
777 PP/29 Fix three issues highlighted by clang analyser static analysis.
778 Only crash-plausible issue would require the Cambridge-specific
779 iplookup router and a misconfiguration.
780 Report from Marcin Mirosław.
782 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
784 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
785 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
786 As part of this, removing so much warning spew let me fix some minor
787 real issues in debug logging.
789 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
790 assignment on my part. Fixed.
792 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
793 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
794 Janne Snabb (who went above and beyond: thank you).
796 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
797 string otherwise requires a connection and a bunch more work and it's
798 relatively easy to get wrong. Should also expose TLS library linkage
801 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
802 64-bit ${eval} (JH/03).
804 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
805 GNU libc to support some of the 64-bit stuff, should not lead to
806 conflicts. Defined before os.h is pulled in, so if a given platform
807 needs to override this, it can.
809 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
810 protection layer was required, which is not implemented.
811 Bugzilla 1254, patch from Wolfgang Breyha.
813 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
814 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
815 tls_dhparam take prime identifiers. Also unbreak combination of
816 OpenSSL+DH_params+TLSSNI.
818 PP/39 Disable SSLv2 by default in OpenSSL support.
824 PP/01 Solaris build fix for Oracle's LDAP libraries.
825 Bugzilla 1109, patch from Stephen Usher.
827 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
829 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
832 TF/02 Fix a couple more cases where we did not log the error message
833 when unlink() failed. See also change 4.74-TF/03.
835 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
836 lock up or crash if it happened to be inside a call to libc when it
837 got a SIGUSR1 from exiwhat.
839 The SIGUSR1 handler appends the current process status to the process
840 log which is later printed by exiwhat. It used to use the general
841 purpose logging code to do this, but several functions it calls are
842 not safe for signals.
844 The new output code in the SIGUSR1 handler is specific to the process
845 log, and simple enough that it's easy to inspect for signal safety.
846 Removing some special cases also simplifies the general logging code.
847 Removing the spurious timestamps from the process log simplifies
850 TF/04 Improved ratelimit ACL condition.
852 The /noupdate option has been deprecated in favour of /readonly which
853 has clearer semantics. The /leaky, /strict, and /readonly update modes
854 are mutually exclusive. The update mode is no longer included in the
855 database key; it just determines when the database is updated. (This
856 means that when you upgrde Exim will forget old rate measurements.)
858 Exim now checks that the per_* options are used with an update mode that
859 makes sense for the current ACL. For example, when Exim is processing a
860 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
861 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
862 must specify per_mail/readonly. If you omit the update mode it defaults to
863 /leaky where that makes sense (as before) or /readonly where required.
865 The /noupdate option is now undocumented but still supported for
866 backwards compatibility. It is equivalent to /readonly except that in
867 ACLs where /readonly is required you may specify /leaky/noupdate or
868 /strict/noupdate which are treated the same as /readonly.
870 A useful new feature is the /count= option. This is a generalization
871 of the per_byte option, so that you can measure the throughput of other
872 aggregate values. For example, the per_byte option is now equivalent
873 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
875 The per_rcpt option has been generalized using the /count= mechanism
876 (though it's more complicated than the per_byte equivalence). When it is
877 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
878 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
879 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
880 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
881 Note that using per_rcpt with a non-readonly update mode in more than
882 one ACL will cause the recipients to be double-counted. (The per_mail
883 and per_byte options don't have this problem.)
885 The handling of very low rates has changed slightly. If the computed rate
886 is less than the event's count (usually one) then this event is the first
887 after a long gap. In this case the rate is set to the same as this event's
888 count, so that the first message of a spam run is counted properly.
890 The major new feature is a mechanism for counting the rate of unique
891 events. The new per_addr option counts the number of different
892 recipients that someone has sent messages to in the last time period. It
893 behaves like per_rcpt if all the recipient addresses are different, but
894 duplicate recipient addresses do not increase the measured rate. Like
895 the /count= option this is a general mechanism, so the per_addr option
896 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
897 example, measure the rate that a client uses different sender addresses
898 with the options per_mail/unique=$sender_address. There are further
899 details in the main documentation.
901 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
903 TF/06 Removed a few PCRE remnants.
905 TF/07 Automatically extract Exim's version number from tags in the git
906 repository when doing development or release builds.
908 PP/02 Raise smtp_cmd_buffer_size to 16kB.
909 Bugzilla 879. Patch from Paul Fisher.
911 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
912 Heavily based on revision 40f9a89a from Simon Arlott's tree.
915 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
917 PP/05 Variable $av_failed, true if the AV scanner deferred.
918 Bugzilla 1078. Patch from John Horne.
920 PP/06 Stop make process more reliably on build failure.
921 Bugzilla 1087. Patch from Heiko Schlittermann.
923 PP/07 Make maildir_use_size_file an _expandable_ boolean.
924 Bugzilla 1089. Patch from Heiko Schlittermann.
926 PP/08 Handle ${run} returning more data than OS pipe buffer size.
927 Bugzilla 1131. Patch from Holger Weiß.
929 PP/09 Handle IPv6 addresses with SPF.
930 Bugzilla 860. Patch from Wolfgang Breyha.
932 PP/10 GnuTLS: support TLS 1.2 & 1.1.
934 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
937 PP/11 match_* no longer expand right-hand-side by default.
938 New compile-time build option, EXPAND_LISTMATCH_RHS.
939 New expansion conditions, "inlist", "inlisti".
941 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
943 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
945 PP/14 fix log_write() format string regression from TF/03.
946 Bugzilla 1152. Patch from Dmitry Isaikin.
952 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
954 PP/02 Harmonised TLS library version reporting; only show if debugging.
955 Layout now matches that introduced for other libraries in 4.74 PP/03.
957 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
959 PP/04 New "dns_use_edns0" global option.
961 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
964 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
965 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
967 TK/01 Updated PolarSSL code to 0.14.2.
968 Bugzilla 1097. Patch from Andreas Metzler.
970 PP/07 Catch divide-by-zero in ${eval:...}.
973 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
976 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
977 format-string attack -- SECURITY: remote arbitrary code execution.
979 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
980 time unintentionally subject to list matching rules, letting the header
981 cause arbitrary Exim lookups (of items which can occur in lists, *not*
982 arbitrary string expansion). This allowed for information disclosure.
984 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
985 INT_MIN/-1 -- value coerced to INT_MAX.
991 NM/01 Workround for PCRE version dependency in version reporting
994 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
995 This fixes portability to compilers other than gcc, notably
996 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
998 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
999 makefiles for portability to HP-UX and POSIX correctness.
1001 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1002 Also via indented variable definition in the Makefile.
1003 (Debugging by Oliver Heesakkers).
1005 PP/02 Restore caching of spamd results with expanded spamd_address.
1006 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1008 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1009 Improves build reliability. Fix from: Frank Elsner
1011 NM/02 Fix wide character breakage in the rfc2047 coding
1012 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1014 NM/03 Allow underscore in dnslist lookups
1015 Fixes bug 1026. Patch from Graeme Fowler
1017 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1018 Code patches from Adam Ciarcinski of NetBSD.
1020 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1023 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1024 is logged, to avoid truncation. Patch from John Horne.
1026 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1027 Patch from Jakob Hirsch.
1029 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1030 SQL string expansion failure details.
1031 Patch from Andrey Oktyabrski.
1033 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1034 Patch from Simon Arlott.
1036 PP/09 New lookups functionality failed to compile on old gcc which rejects
1037 extern declarations in function scope.
1038 Patch from Oliver Fleischmann
1040 PP/10 Use sig_atomic_t for flags set from signal handlers.
1041 Check getgroups() return and improve debugging.
1042 Fixed developed for diagnosis in bug 927 (which turned out to be
1045 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1046 Patch from Mark Zealey.
1048 PP/12 Bugzilla 1056: Improved spamd server selection.
1049 Patch from Mark Zealey.
1051 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1052 Based on patch from Heiko Schlittermann.
1054 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1055 Patch from Uwe Doering, sign-off by Michael Haardt.
1057 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1058 variable declaration deep within a block. Bug and patch from
1061 PP/15 lookups-Makefile IRIX compatibilty coercion.
1063 PP/16 Make DISABLE_DKIM build knob functional.
1065 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1066 Patch by Simon Arlott
1068 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1069 variable argument macros. Our copy now differs from upstream.
1075 TF/01 Failure to get a lock on a hints database can have serious
1076 consequences so log it to the panic log.
1078 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1079 controlled using the smtp_confirmation log selector.
1081 TF/03 Include the error message when we fail to unlink a spool file.
1083 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1084 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1085 for maintaining out-of-tree patches for some time.
1087 PP/01 Bugzilla 139: Documentation and portability issues.
1088 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1089 Handle per-OS dynamic-module compilation flags.
1091 PP/02 Let /dev/null have normal permissions.
1092 The 4.73 fixes were a little too stringent and complained about the
1093 permissions on /dev/null. Exempt it from some checks.
1094 Reported by Andreas M. Kirchwitz.
1096 PP/03 Report version information for many libraries, including
1097 Exim version information for dynamically loaded libraries. Created
1098 version.h, now support a version extension string for distributors
1099 who patch heavily. Dynamic module ABI change.
1101 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1102 privilege escalation vulnerability whereby the Exim run-time user
1103 can cause root to append content of the attacker's choosing to
1106 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1109 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1110 If dropping privileges for untrusted macros, we disabled normal logging
1111 on the basis that it would fail; for the Exim run-time user, this is not
1112 the case, and it resulted in successful deliveries going unlogged.
1113 Fixed. Reported by Andreas Metzler.
1119 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1120 only prepend for the Resent-* case. Fixes regression introduced in
1121 Exim 4.70 by NM/22 for Bugzilla 607.
1123 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1124 increasing numbers of administrators be bitten by this.
1126 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1128 PP/03 Bugzilla 994: added openssl_options main configuration option.
1130 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1132 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1134 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1136 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1137 without a peer certificate, leading to a segfault because of an
1138 assumption that peers always have certificates. Be a little more
1139 paranoid. Problem reported by Martin Tscholak.
1141 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1142 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1143 NB: ClamAV planning to remove STREAM in "middle of 2010".
1144 CL also introduces -bmalware, various -d+acl logging additions and
1145 more caution in buffer sizes.
1147 PP/09 Implemented reverse_ip expansion operator.
1149 PP/10 Bugzilla 937: provide a "debug" ACL control.
1151 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1153 PP/12 Bugzilla 973: Implement --version.
1155 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1157 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1159 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1161 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1162 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1163 ignore trailing whitespace.
1165 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1167 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1170 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1171 Notification from Dr Andrew Aitchison.
1173 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1174 ExtendedDetectionInfo response format.
1175 Notification from John Horne.
1177 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1180 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1181 XSL and documented dependency on system catalogs, with examples of how
1184 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1187 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1188 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1189 configuration file which is writeable by the Exim user or group.
1191 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1192 of configuration files to cover files specified with the -C option if
1193 they are going to be used with root privileges, not just the default
1196 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1197 option (effectively making it always true).
1199 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1200 files to be used while preserving root privileges.
1202 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1203 that rogue child processes cannot use them.
1205 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1206 run-time user, instead of root.
1208 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1209 Exim run-time user without dropping privileges.
1211 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1212 result string, instead of calling string_vformat() twice with the same
1215 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1216 for other users. Others should always drop root privileges if they use
1217 -C on the command line, even for a whitelisted configure file.
1219 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1221 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1227 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1228 $data_path, and $header_path variables; fixed documentation bugs and
1231 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1232 exipick to access non-standard spools, including the "frozen" queue
1235 NM/01 Bugzilla 965: Support mysql stored procedures.
1236 Patch from Alain Williams
1238 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1240 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1241 Patch from Andreas Metzler
1243 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1244 Patch from Kirill Miazine
1246 NM/05 Bugzilla 671: Added umask to procmail example.
1248 JJ/03 installed exipick 20100323.0, fixing doc bug
1250 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1251 directory. Notification and patch from Dan Rosenberg.
1253 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1255 TK/02 Improve log output when DKIM signing operation fails.
1257 MH/01 Treat the transport option dkim_domain as a colon separated
1258 list, not as a single string, and sign the message with each element,
1259 omitting multiple occurences of the same signer.
1261 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1262 Bugzilla 985, 986. Patch by Simon Arlott
1264 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1265 Patch by Simon Arlott
1267 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1268 MBX locking. Notification from Dan Rosenberg.
1274 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1276 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1278 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1280 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1282 NM/04 Bugzilla 915: Flag broken perl installation during build.
1288 TK/01 Added patch by Johannes Berg that expands the main option
1289 "spamd_address" if it starts with a dollar sign.
1291 TK/02 Write list of recipients to X-Envelope-Sender header when building
1292 the mbox-format spool file for content scanning (suggested by Jakob
1295 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1296 (http://www.dcc-servers.net/) support via dccifd. Activated by
1297 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1299 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1300 by Mark Daniel Reidel <mr@df.eu>.
1302 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1303 When building exim an external PCRE library is now needed -
1304 PCRE is a system library on the majority of modern systems.
1305 See entry on PCRE_LIBS in EDITME file.
1307 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1308 conversation. Added nologin parameter to request.
1309 Patch contributed by Kirill Miazine.
1311 TF/01 Do not log submission mode rewrites if they do not change the address.
1313 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1315 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1316 log files in place. Contributed by Roberto Lima.
1318 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1320 TF/03 Bugzilla 615: When checking the local_parts router precondition
1321 after a local_part_suffix or local_part_prefix option, Exim now
1322 does not use the address's named list lookup cache, since this
1323 contains cached lookups for the whole local part.
1325 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1326 Robert Millan. Documentation is in experimental-spec.txt.
1328 TF/04 Bugzilla 668: Fix parallel build (make -j).
1330 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1332 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1333 Patch provided by Jan Srzednicki.
1335 TF/05 Leading white space used to be stripped from $spam_report which
1336 wrecked the formatting. Now it is preserved.
1338 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1339 that they are available at delivery time.
1341 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1343 TF/08 TLS error reporting now respects the incoming_interface and
1344 incoming_port log selectors.
1346 TF/09 Produce a more useful error message if an SMTP transport's hosts
1347 setting expands to an empty string.
1349 NM/06 Bugzilla 744: EXPN did not work under TLS.
1350 Patch provided by Phil Pennock.
1352 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1353 Patch provided by Richard Godbee.
1355 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1356 acl_smtp_notquit, added index entry.
1358 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1359 Patch provided by Eugene Bujak.
1361 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1362 accept(). Patch provided by Maxim Dounin.
1364 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1365 Patch provided by Phil Pennock.
1367 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1369 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1370 Patch provided by Brad "anomie" Jorsch.
1372 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1373 Patch provided by Dean Brooks.
1375 TK/05 Add native DKIM support (does not depend on external libraries).
1377 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1378 Patch provided by Graeme Fowler.
1380 NM/16 Bugzilla 851: Documentation example syntax fix.
1382 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1384 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1387 NM/19 Bugzilla 745: TLS version reporting.
1388 Patch provided by Phil Pennock.
1390 NM/20 Bugzilla 167: bool: condition support.
1391 Patch provided by Phil Pennock.
1393 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1394 clients. Patch provided by Phil Pennock.
1396 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1397 Patch provided by Brad "anomie" Jorsch.
1399 NM/23 Bugzilla 687: Fix misparses in eximstats.
1400 Patch provided by Heiko Schlittermann.
1402 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1403 Patch provided by Heiko Schlittermann.
1405 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1406 plus update to original patch.
1408 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1410 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1411 Patch provided by David Brownlee.
1413 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1415 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1417 NM/30 Bugzilla 888: TLS documentation bugfixes.
1419 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1421 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1422 Unlike the original bugzilla I have changed all shell scripts in src tree.
1424 NM/33 Bugzilla 898: Transport filter timeout fix.
1425 Patch by Todd Rinaldo.
1427 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1428 Patch by Serge Demonchaux.
1430 NM/35 Bugzilla 39: Base64 decode bug fixes.
1431 Patch by Jakob Hirsch.
1433 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1435 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1437 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1439 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1445 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1446 ALT-N's libdkim that I have put here:
1447 http://duncanthrax.net/exim-experimental/
1449 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1450 were called 'true' and it seems that C99 defines that as a reserved
1451 keyword to be used with 'bool' variable types. That means you could
1452 not include C99-style headers which use bools without triggering
1453 build errors in sieve.c.
1455 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1456 as mailq or other aliases. Changed the --help handling significantly
1457 to do whats expected. exim_usage() emits usage/help information.
1459 SC/01 Added the -bylocaldomain option to eximstats.
1461 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1463 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1465 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1471 PH/01 Another patch from the Sieve maintainer.
1473 PH/02 When an IPv6 address is converted to a string for single-key lookup
1474 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1475 dots are used instead of colons so that keys in lsearch files need not
1476 contain colons. This was done some time before quoting was made available
1477 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1478 (notated using the quote facility) so as to distinguish them from IPv4
1479 keys. This meant that lookups for IP addresses in host lists did not work
1480 for iplsearch lookups.
1482 This has been fixed by arranging for IPv6 addresses to be expressed with
1483 colons if the lookup type is iplsearch. This is not incompatible, because
1484 previously such lookups could never work.
1486 The situation is now rather anomolous, since one *can* have colons in
1487 ordinary lsearch keys. However, making the change in all cases is
1488 incompatible and would probably break a number of configurations.
1490 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1493 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1494 conversion specification without a maximum field width, thereby enabling
1495 a rogue spamd server to cause a buffer overflow. While nobody in their
1496 right mind would setup Exim to query an untrusted spamd server, an
1497 attacker that gains access to a server running spamd could potentially
1498 exploit this vulnerability to run arbitrary code as the Exim user.
1500 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1501 $primary_hostname instead of what libspf2 thinks the hosts name is.
1503 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1504 a directory entry by the name of the lookup key. Previously, if a
1505 symlink pointed to a non-existing file or a file in a directory that
1506 Exim lacked permissions to read, a lookup for a key matching that
1507 symlink would fail. Now it is enough that a matching directory entry
1508 exists, symlink or not. (Bugzilla 503.)
1510 PH/03 The body_linecount and body_zerocount variables are now exported in the
1513 PH/04 Added the $dnslist_matched variable.
1515 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1516 This means they are set thereafter only if the connection becomes
1519 PH/06 Added the client_condition to authenticators so that some can be skipped
1520 by clients under certain conditions.
1522 PH/07 The error message for a badly-placed control=no_multiline_responses left
1523 "_responses" off the end of the name.
1525 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1527 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1528 (without spaces) instead of just copying the configuration text.
1530 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1532 PH/11 Added $max_received_linelength.
1534 PH/12 Added +ignore_defer and +include_defer to host lists.
1536 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1537 way in which PCRE > 7.0 is built.
1539 PH/14 Implemented queue_only_load_latch.
1541 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1542 MAIL command. The effect was to mangle the value on 64-bit systems.
1544 PH/16 Another patch from the Sieve maintainer.
1546 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1548 PH/18 If a system quota error occurred while trying to create the file for
1549 a maildir delivery, the message "Mailbox is full" was not appended to the
1550 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1551 only to a quota excession during the actual writing of the file.
1553 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1554 characters?) which causes problems in log lines. The DN values are now
1555 passed through string_printing() before being added to log lines.
1557 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1558 and InterBase are left for another time.)
1560 PH/21 Added message_body_newlines option.
1562 PH/22 Guard against possible overflow in moan_check_errorcopy().
1564 PH/23 POSIX allows open() to be a macro; guard against that.
1566 PH/24 If the recipient of an error message contained an @ in the local part
1567 (suitably quoted, of course), incorrect values were put in $domain and
1568 $local_part during the evaluation of errors_copy.
1574 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1575 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1578 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1579 issue a MAIL command.
1581 PH/02 In an ACL statement such as
1583 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1585 if a client was not listed at all, or was listed with a value other than
1586 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1587 the condition was not true (as it should be), so access was not denied.
1588 The bug was that the ! inversion was incorrectly passed on to the second
1589 item. This has been fixed.
1591 PH/03 Added additional dnslists conditions == and =& which are different from
1592 = and & when the dns lookup returns more than one IP address.
1594 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1595 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1597 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1598 FSYNC, which compiles an option called disable_fsync that allows for
1599 bypassing fsync(). The documentation is heavily laced with warnings.
1601 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1603 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1604 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1605 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1606 including adding "make clean"; (3) Added -fPIC when compiling the test
1607 dynamically loaded module, to get rid of a warning.
1609 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1610 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1611 = 0s. The bug is otherwise harmless.
1613 PH/07 There was a bug in the dovecot authenticator such that the value of
1614 $auth1 could be overwritten, and so not correctly preserved, after a
1615 successful authentication. This usually meant that the value preserved by
1616 the server_setid option was incorrect.
1618 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1620 PH/09 Installed PCRE release 7.0.
1622 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1623 run for batched SMTP input. It is now run at the start of every message
1624 in the batch. While fixing this I discovered that the process information
1625 (output by running exiwhat) was not always getting set for -bs and -bS
1626 input. This is fixed, and it now also says "batched" for BSMTP.
1628 PH/11 Added control=no_pipelining.
1630 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1631 patch, slightly modified), and move the expansion of helo_data till after
1632 the connection is made in the smtp transport (so it can use these
1635 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1637 PH/14 Added log_selector = +pid.
1639 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1641 PH/16 Add ${if forany and ${if forall.
1643 PH/17 Added dsn_from option to vary the From: line in DSNs.
1645 PH/18 Flush SMTP output before performing a callout, unless control =
1646 no_callout_flush is set.
1648 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1649 was true (the default) a successful delivery failed to delete the retry
1650 item, thus causing premature timeout of the address. The bug is now
1653 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1655 PH/21 Long custom messages for fakedefer and fakereject are now split up
1656 into multiline reponses in the same way that messages for "deny" and
1657 other ACL rejections are.
1659 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1660 with slight modification.
1662 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1663 draft, changing the syntax and factoring some duplicate code".
1665 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1666 for deliveries of the second and subsequent messages over the same SMTP
1669 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1670 ${reduce, with only minor "tidies".
1672 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1674 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1675 expansion side effects.
1677 PH/27 When a message times out after an over-quota error from an Exim-imposed
1678 quota, the bounce message says "mailbox is full". This message was not
1679 being given when it was a system quota that was exceeded. It now should
1682 MH/03 Made $recipients available in local_scan(). local_scan() already has
1683 better access to the recipient list through recipients_list[], but
1684 $recipients can be useful in postmaster-provided expansion strings.
1686 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1687 in the case of a MAIL command with additional options following the
1688 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1689 were accidentally chopped off.
1691 PH/29 SMTP synchronization checks are implemented when a command is read -
1692 there is a check that no more input is waiting when there shouldn't be
1693 any. However, for some commands, a delay in an ACL can mean that it is
1694 some time before the response is written. In this time, more input might
1695 arrive, invalidly. So now there are extra checks after an ACL has run for
1696 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1697 pipelining has not been advertised.
1699 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1701 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1702 returned in the SMTP response when smtp_return_error_details was false.
1703 This has been fixed.
1705 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1706 instead of the C I/O that was originally supplied, because problems were
1707 reported on Solaris.
1709 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1710 Exim which did not show up earlier: it was assuming that a call to
1711 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1712 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1713 was a macro that became an assignment, so it seemed to work. This has
1714 changed to a proper function call with a void return, hence the compile
1715 error. Exim's code has been fixed.
1717 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1720 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1722 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1724 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1725 because it turns out to be too big or there is a timeout) there is no
1726 "Completed" line in the log. When some messages of this type were
1727 selected by exigrep, they were listed as "not completed". Others were
1728 picked up by some special patterns. I have improved the selection
1729 criteria to be more general.
1731 PH/38 The host_find_failed option in the manualroute router can now be set
1732 to "ignore", to completely ignore a host whose IP address cannot be
1733 found. If all hosts are ignored, the behaviour is controlled by the new
1734 host_all_ignored option.
1736 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1737 homing or because of multiple MX records with /mx) generated more than
1738 one IP address, and the following item turned out to be the local host,
1739 all the secondary addresses of the first item were incorrectly removed
1740 from the list, along with the local host and any following hosts (which
1741 is what is supposed to happen).
1743 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1744 whoever called Exim into the -H file. In the case of the daemon it was
1745 behaving confusingly. When first started, it used values for whoever
1746 started the daemon, but after a SIGHUP it used the Exim user (because it
1747 calls itself on a restart). I have changed the code so that it now always
1750 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1751 message are rejected with the same error (e.g. no authentication or bad
1752 sender address), and a DATA command is nevertheless sent (as can happen
1753 with PIPELINING or a stupid MUA), the error message that was given to the
1754 RCPT commands is included in the rejection of the DATA command. This is
1755 intended to be helpful for MUAs that show only the final error to their
1758 PH/42 Another patch from the Sieve maintainer.
1760 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1761 Eximstats - Fixed some broken HTML links and added missing column headers
1763 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1764 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1766 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1768 PH/43 Yet another patch from the Sieve maintainer.
1770 PH/44 I found a way to check for a TCP/IP connection going away before sending
1771 the response to the final '.' that terminates a message, but only in the
1772 case where the client has not sent further data following the '.'
1773 (unfortunately, this is allowed). However, in many cases there won't be
1774 any further data because there won't be any more messages to send. A call
1775 to select() can be used: if it shows that the input is "ready", there is
1776 either input waiting, or the socket has been closed. An attempt to read
1777 the next input character can distinguish the two cases. Previously, Exim
1778 would have sent an OK response which the client would never have see.
1779 This could lead to message repetition. This fix should cure that, at
1780 least in a lot of common cases.
1782 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1783 advertised in response to EHLO.
1789 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1790 fixed by 4.65/MH/01 (is this a record?) are fixed:
1792 (i) An empty string was always treated as zero by the numeric comparison
1793 operators. This behaviour has been restored.
1795 (ii) It is documented that the numeric comparison operators always treat
1796 their arguments as decimal numbers. This was broken in that numbers
1797 starting with 0 were being interpreted as octal.
1799 While fixing these problems I realized that there was another issue that
1800 hadn't been noticed. Values of message_size_limit (both the global option
1801 and the transport option) were treated as octal if they started with 0.
1802 The documentation was vague. These values are now always treated as
1803 decimal, and I will make that clear in the documentation.
1809 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1810 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1813 MH/01 Don't check that the operands of numeric comparison operators are
1814 integers when their expansion is in "skipping" mode (fixes bug
1815 introduced by 4.64-PH/07).
1817 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1818 child addresses, Exim now panics and dies. Previously, because the count
1819 is held in a short int, deliveries were likely to be lost. As such a
1820 large number of recipients for a single message is ridiculous
1821 (performance will be very, very poor), I have chosen to impose a limit
1822 rather than extend the field.
1828 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1829 leftover -K file (the existence of which was triggered by #402).
1830 While we were at it, introduced process PID as part of the -K
1831 filename. This should rule out race conditions when creating
1834 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1835 processing considerably. Previous code took too long for large mails,
1836 triggering a timeout which in turn triggers #401.
1838 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1839 in the DK code in transports.c. sendfile() is not really portable,
1840 hence the _LINUX specificness.
1842 TF/01 In the add_headers option to the mail command in an Exim filter,
1843 there was a bug that Exim would claim a syntax error in any
1844 header after the first one which had an odd number of characters
1847 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1848 callout verification, Exim cached a "reject" for the entire domain. This
1849 is correct for most verifications, but it is not correct for a recipient
1850 verification with use_sender or use_postmaster set, because in that case
1851 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1852 case of MAIL FROM:<> rejection from other early rejections (e.g.
1853 rejection of HELO). When verifying a recipient using a non-null MAIL
1854 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1855 Whatever the result of the callout, the value of the domain cache is
1856 left unchanged (for any other kind of callout, getting as far as trying
1857 RCPT means that the domain itself is ok).
1859 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1862 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1863 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1864 session, the daemon ignores ECONNECT errors and logs others; it now
1865 ignores EPIPE as well.
1867 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1868 (quoted-printable decoding).
1870 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1871 later the small subsequent patch to fix an introduced bug.
1873 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1875 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1877 PH/08 An error is now given if message_size_limit is specified negative.
1879 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1880 to be given (somewhat) arbitrary names.
1882 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1885 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1886 miscellaneous code fixes
1888 PH/10 Added the log_reject_target ACL modifier to specify where to log
1891 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1892 hostname. This is wrong, because it relates to the incoming message (and
1893 probably the interface on which it is arriving) and not to the outgoing
1894 callout (which could be using a different interface). This has been
1895 changed to use the value of the helo_data option from the smtp transport
1896 instead - this is what is used when a message is actually being sent. If
1897 there is no remote transport (possible with a router that sets up host
1898 addresses), $smtp_active_hostname is used.
1900 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1901 tweaks were necessary in order to get it to work (see also 21 below):
1902 (a) The code assumed that strncpy() returns a negative number on buffer
1903 overflow, which isn't the case. Replaced with Exim's string_format()
1905 (b) There were several signed/unsigned issues. I just did the minimum
1906 hacking in of casts. There is scope for a larger refactoring.
1907 (c) The code used strcasecmp() which is not a standard C function.
1908 Replaced with Exim's strcmpic() function.
1909 (d) The code set only $1; it now sets $auth1 as well.
1910 (e) A simple test gave the error "authentication client didn't specify
1911 service in request". It would seem that Dovecot has changed its
1912 interface. Fortunately there's a specification; I followed it and
1913 changed what the client sends and it appears to be working now.
1915 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1918 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1919 address A is aliased to B and C, where B exists and C does not. Without
1920 -v the output is "A verified" because verification stops after a
1921 successful redirection if more than one address is generated. However,
1922 with -v the child addresses are also verified. Exim was outputting "A
1923 failed to verify" and then showing the successful verification for C,
1924 with its parentage. It now outputs "B failed to verify", showing B's
1925 parentage before showing the successful verification of C.
1927 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1928 look up a TXT record in a specific list after matching in a combined
1931 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1932 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1933 they consult the DNS. I had assumed they would set it the way they
1934 wanted; and indeed my experiments on Linux seem to show that in some
1935 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1936 To be on the safe side, however, I have now made the interface to
1937 host_find_byname() similar to host_find_bydns(), with an argument
1938 containing the DNS resolver options. The host_find_byname() function now
1939 sets these options at its start, just as host_find_bydns() does. The smtp
1940 transport options dns_qualify_single and dns_search_parents are passed to
1941 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1942 of host_find_byname() use the default settings of RES_DEFNAMES
1943 (qualify_single) but not RES_DNSRCH (search_parents).
1945 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1946 spool_read_header() do less string testing, by means of a preliminary
1947 switch on the second character of optional "-foo" lines. (This is
1948 overdue, caused by the large number of possibilities that now exist.
1949 Originally there were few.) While I was there, I also converted the
1950 str(n)cmp tests so they don't re-test the leading "-" and the first
1951 character, in the hope this might squeeze out yet more improvement.
1953 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1954 flag allowing group syntax was set by the header_syntax check but not
1955 turned off, possible causing trouble later; (2) The flag was not being
1956 set at all for the header_verify test, causing "group"-style headers to
1957 be rejected. I have now set it in this case, and also caused header_
1958 verify to ignore an empty address taken from a group. While doing this, I
1959 came across some other cases where the code for allowing group syntax
1960 while scanning a header line wasn't quite right (mostly, not resetting
1961 the flag correctly in the right place). These bugs could have caused
1962 trouble for malformed header lines. I hope it is now all correct.
1964 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1965 with the "reply" argument non-NULL. The code, however (which originally
1966 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1967 but it didn't always do it. This confused somebody who was copying the
1968 code for some other use. I have removed all the tests.
1970 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1971 feature that was used to support insecure browsers during the U.S. crypto
1972 embargo. It requires special client support, and Exim is probably the
1973 only MTA that supported it -- and would never use it because real RSA is
1974 always available. This code has been removed, because it had the bad
1975 effect of slowing Exim down by computing (never used) parameters for the
1976 RSA_EXPORT functionality.
1978 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1979 authenticator to fail if there's a tab character in the incoming data
1980 (there should never be unless someone is messing about, as it's supposed
1981 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1982 if the connection is using TLS or if the remote IP is the same as the
1983 local IP, and the "valid-client-cert option" if a client certificate has
1986 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1987 authenticators. This can be used for authorization after authentication
1988 succeeds. (In the case of plaintext, it servers for both authentication
1991 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1992 if any retry times were supplied.
1994 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1995 connection, where there is no client IP address to check. In this
1996 situation, the verify now always succeeds.
1998 PH/25 Applied John Jetmore's -Mset patch.
2000 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2002 PH/27 In a string expansion for a processed (not raw) header when multiple
2003 headers of the same name were present, leading whitespace was being
2004 removed from all of them, but trailing whitespace was being removed only
2005 from the last one. Now trailing whitespace is removed from each header
2006 before concatenation. Completely empty headers in a concatenation (as
2007 before) are ignored.
2009 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2010 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2012 PH/29 [Removed. This was a change that I later backed out, and forgot to
2013 correct the ChangeLog entry (that I had efficiently created) before
2014 committing the later change.]
2016 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2017 address errors (4xx response to RCPT) over the same connection as other
2018 messages routed to the same hosts. Such deliveries are always "forced",
2019 so retry times are not inspected. This resulted in far too many retries
2020 for the affected addresses. The effect occurred only when there were more
2021 hosts than the hosts_max_try setting in the smtp transport when it had
2022 the 4xx errors. Those hosts that it had tried were not added to the list
2023 of hosts for which the message was waiting, so if all were tried, there
2024 was no problem. Two fixes have been applied:
2026 (i) If there are any address or message errors in an SMTP delivery, none
2027 of the hosts (tried or untried) are now added to the list of hosts
2028 for which the message is waiting, so the message should not be a
2029 candidate for sending over the same connection that was used for a
2030 successful delivery of some other message. This seems entirely
2031 reasonable: after all the message is NOT "waiting for some host".
2032 This is so "obvious" that I'm not sure why it wasn't done
2033 previously. Hope I haven't missed anything, but it can't do any
2034 harm, as the worst effect is to miss an optimization.
2036 (ii) If, despite (i), such a delivery is accidentally attempted, the
2037 routing retry time is respected, so at least it doesn't keep
2038 hammering the server.
2040 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2041 in ${readsocket because some servers need this prod.
2043 PH/32 Added some extra debug output when updating a wait-xxx database.
2045 PH/33 The hint "could be header name not terminated by colon", which has been
2046 given for certain expansion errors for a long time, was not being given
2047 for the ${if def:h_colon_omitted{... case.
2049 PH/34 The spec says: "With one important exception, whenever a domain list is
2050 being scanned, $domain contains the subject domain." There was at least
2051 one case where this was not true.
2053 PH/35 The error "getsockname() failed: connection reset by peer" was being
2054 written to the panic log as well as the main log, but it isn't really
2055 panic-worthy as it just means the connection died rather early on. I have
2056 removed the panic log writing for the ECONNRESET error when getsockname()
2059 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2060 runs only) independently of the message's sender address. This meant
2061 that, if the 4xx error was in fact related to the sender, a different
2062 message to the same recipient with a different sender could confuse
2063 things. In particualar, this can happen when sending to a greylisting
2064 server, but other circumstances could also provoke similar problems.
2065 I have changed the default so that the retry time for these errors is now
2066 based a combination of the sender and recipient addresses. This change
2067 can be overridden by setting address_retry_include_sender=false in the
2070 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2071 remote server are returned as part of bounce messages. This was not
2072 happening for LMTP over a pipe (the lmtp transport), but now it is the
2073 same for both kinds of LMTP.
2075 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2076 in header lines that were in fact CNAMEs. This is no longer the case.
2078 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2079 and queue runs started by the daemon processed all messages. This has
2080 been fixed so that -R and -S can now usefully be given with -q<time>.
2082 PH/40 Import PCRE release 6.7 (fixes some bugs).
2084 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2086 PH/42 Give an error if -q is specified more than once.
2088 PH/43 Renamed the variables $interface_address and $interface_port as
2089 $received_ip_address and $received_port, to make it clear that these
2090 values apply to message reception, and not to the outgoing interface when
2091 a message is delivered. (The old names remain recognized, of course.)
2093 PH/44 There was no timeout on the connect() call when using a Unix domain
2094 socket in the ${readsocket expansion. There now is.
2096 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2097 be meaningful with "accept".
2099 SC/01 Eximstats V1.43
2100 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2102 SC/02 Eximstats V1.44
2103 Use a glob alias rather than an array ref in the generated
2104 parser. This improves both readability and performance.
2106 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2107 Collect SpamAssassin and rejection statistics.
2108 Don't display local sender or destination tables unless
2109 there is data to show.
2110 Added average volumes into the top table text output.
2112 SC/04 Eximstats V1.46
2113 Collect data on the number of addresses (recipients)
2114 as well as the number of messages.
2116 SC/05 Eximstats V1.47
2117 Added 'Message too big' to the list of mail rejection
2118 reasons (thanks to Marco Gaiarin).
2120 SC/06 Eximstats V1.48
2121 Mainlog lines which have GMT offsets and are too short to
2122 have a flag are now skipped.
2124 SC/07 Eximstats V1.49 (Alain Williams)
2125 Added the -emptyok flag.
2127 SC/08 Eximstats V1.50
2128 Fixes for obtaining the IP address from reject messages.
2130 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2131 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2132 whitesspace changes from 4.64-PH/27
2134 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2137 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2138 are found, allow negative numbers in numeric criteria)
2140 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2142 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2143 to match changes made in 4.64-PH/43
2145 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2147 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2148 "domain" is actually an IP address, and force a failure. This locks out
2149 those revolvers/nameservers that support "A-for-A" lookups, in
2150 contravention of the specifications.
2152 PH/48 When a host name was looked up from an IP address, and the subsequent
2153 forward lookup of the name timed out, the host name was left in
2154 $sender_host_name, contrary to the specification.
2156 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2157 restricted to single-key lookups, Exim was not diagnosing an error if
2158 * or *@ was used with a query-style lookup.
2160 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2162 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2163 long ago, but noone interested enough thought of it. Let's just say that
2164 the "1.1" means that there are some new functions that weren't there at
2165 some point in the past.
2167 PH/51 Error processing for expansion failure of helo_data from an smtp
2168 transport during callout processing was broken.
2170 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2171 tested/used via the -bh/-bhc/-bs options.
2173 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2174 bug, fixed in subsequent PCRE releases).
2176 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2177 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2179 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2185 SC/01 Use a glob alias rather than an array ref in eximstats generated
2186 parser. This improves both readability and performance.
2188 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2189 Don't display local sender or destination tables in eximstats unless
2190 there is data to show.
2191 Added average volumes into the eximstats top table text output.
2193 SC/03 Collect data on the number of addresses (recipients) as well
2194 as the number of messages in eximstats.
2196 TF/01 Correct an error in the documentation for the redirect router. Exim
2197 does (usually) call initgroups() when daemonizing.
2199 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2200 with consistent privilege compared to when running as a daemon.
2202 TF/03 Note in the spec that $authenticated_id is not set for local
2203 submissions from trusted users.
2205 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2206 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2208 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2209 by adding some example configuration directives to the default
2210 configuration file. A little bit of work is required to uncomment the
2211 directives and define how usernames and passwords are checked, but
2212 there is now a framework to start from.
2214 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2215 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2216 without this. I don't know how relevant this is to other LDAP libraries.
2218 PH/02 Add the verb name to the "unknown ACL verb" error.
2220 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2222 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2224 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2225 directory not expanded when it should be if an expanded home directory
2226 was set for the address (which is overridden by the transport).
2228 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2231 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2232 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2233 because it is too late at that time, and has no effect.
2235 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2236 security issue with \' (bugzilla #107). I could not use the
2237 PQescapeStringConn() function, because it needs a PGconn value as one of
2240 PH/08 When testing addresses using -bt, indicate those final addresses that
2241 are duplicates that would not cause an additional delivery. At least one
2242 person was confused, thinking that -bt output corresponded to deliveries.
2243 (Suppressing duplicates isn't a good idea as you lose the information
2244 about possibly different redirections that led to the duplicates.)
2246 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2247 systems where poll() doesn't work, in particular OS X.
2249 PH/10 Added more information to debugging output for retry time not reached.
2251 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2252 operations in malware.c.
2254 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2257 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2258 syslog_duplication set false, log lines that would normally be written
2259 both the the main log and to the reject log were not written to syslog at
2262 PH/14 In the default configuration, change the use of "message" in ACL warn
2263 statements to "add_header".
2265 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2266 not followed by a command (e.g. "seen endif").
2268 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2269 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2272 PH/17 Added extra conditions to the default value of delay_warning_condition
2276 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2277 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2278 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2281 The Auto-Submitted: and various List- headers are standardised, whereas I
2282 don't think Precedence: ever was.
2284 PH/18 Refactored debugging code in route_finduser() to show more information,
2285 in particular, the error code if getpwnam() issues one.
2287 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2288 This is apparently needed in addition to the PH/07 change above to avoid
2289 any possible encoding problems.
2291 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2292 but not after initializing Perl.
2294 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2295 output them only if debugging. By default they are written stderr,
2296 apparently, which is not desirable.
2298 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2301 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2304 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2306 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2307 authenticated or an ident call has been made. Suppress the default
2308 values for $authenticated_id and $authenticated_sender (but permit -oMai
2309 and -oMas) when testing with -bh.
2311 PH/24 Re-jigged the order of the tests in the default configuration so that the
2312 tests for valid domains and recipients precede the DNS black list and CSA
2313 tests, on the grounds that those ones are more expensive.
2315 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2316 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2317 This bug exists in every version of Exim that I still have, right back to
2320 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2321 However, an attempt to turn on case-sensitivity in a regex key by
2322 including (?-i) didn't work because the subject string was already
2323 lowercased, and the effects were non-intuitive. It turns out that a
2324 one-line patch can be used to allow (?-i) to work as expected.
2330 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2331 other effects) broke the use of negated acl sub-conditions.
2333 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2336 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2337 "Deny" causes Exim to reject the incoming connection with a 554 error.
2338 Unfortunately, if there is a major crisis, such as a disk failure,
2339 tcp-wrappers gives "deny", whereas what one would like would be some
2340 kind of temporary error. A kludge has been added to help with this.
2341 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2342 554 error is used if errno is still zero or contains ENOENT (which occurs
2343 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2346 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2348 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2349 errors. Otherwise a message that provokes a temporary error (when other
2350 messages do not) can cause a whole host to time out.
2352 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2353 addresses were routed directly to files or pipes from a redirect router.
2354 File deliveries just didn't batch; pipe deliveries might have suffered
2357 PH/06 A failure to get a lock for a hints database would erroneously always say
2358 "Failed to get write lock", even when it was really a read lock.
2360 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2361 of 0600. This has been changed to use the value of the lockfile_mode
2362 option (which defaults to 0600).
2364 PH/08 Applied small patch from the Sieve maintainer.
2366 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2367 folder from quota calculations, a direct delivery into this folder messed
2368 up the contents of the maildirsize file. This was because the regex was
2369 used only to exclude .Trash (or whatever) when the size of the mailbox
2370 was calculated. There was no check that a delivery was happening into an
2371 excluded directory. This bug has been fixed by ignoring all quota
2372 processing for deliveries into excluded directories.
2374 PH/10 Added the maildirfolder_create_regex option to appendfile.
2380 PH/01 The code for finding all the local interface addresses on a FreeBSD
2381 system running IPv6 was broken. This may well have applied to all BSD
2382 systems, as well as to others that have similar system calls. The broken
2383 code found IPv4 interfaces correctly, but gave incorrect values for the
2384 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2385 that it would not match correctly against @[] and not recognize the IPv6
2388 PH/02 The ipliteral router was not recognizing addresses of the form user@
2389 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2391 PH/03 Added disable_ipv6.
2393 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2394 decoded form, because it is most often used to construct To: headers
2395 lines in autoreplies, and the decoded form may well be syntactically
2396 invalid. However, $reply_address has leading white space removed, and all
2397 newlines turned into spaces so that the autoreply transport does not
2400 PH/05 If group was specified without a user on a router, and no group or user
2401 was specified on a transport, the group from the router was ignored.
2403 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2404 for visible compile-time settings that can be used to change these
2405 numbers, for those that want even more. Backwards compatibility with old
2406 spool files has been maintained. However, going back to a previous Exim
2407 release will lost any variables that are in spool files.
2409 PH/07 Two small changes when running in the test harness: increase delay when
2410 passing a TCP/IP connection to a new process, in case the original
2411 process has to generate a bounce, and remove special handling of
2412 127.0.0.2 (sic), which is no longer necessary.
2414 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2415 be the same on different OS.
2417 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2420 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2421 whether --show-vars was specified or not
2423 JJ/02 exipick: Added support for new ACL variable spool format introduced
2426 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2427 syntactically invalid From: or Reply-to: line, and a filter used this to
2428 generate an autoreply, and therefore failed to obtain an address for the
2429 autoreply, Exim could try to deliver to a non-existent relative file
2430 name, causing unrelated and misleading errors. What now happens is that
2431 it logs this as a hard delivery error, but does not attempt to create a
2434 PH/11 The exinext utility has a -C option for testing purposes, but although
2435 the given file was scanned by exinext itself; it wasn't being passed on
2436 when Exim was called.
2438 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2439 an end-of-file indication when reading a command response.
2441 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2442 compiled. In many other places in Exim, IPv6 addresses are always
2443 recognized, so I have changed this. It also means that IPv4 domain
2444 literals of the form [IPV4:n.n.n.n] are now always recognized.
2446 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2447 used if the router is not running as root, for example, when verifying at
2448 ACL time, or when using -bh. The debugging output from this situation was
2449 non-existent - all you got was a failure to exec. I have made two
2452 (a) Failures to set uid/gid, the current directory, or a process leader
2453 in a subprocess such as that created by queryprogram now generate
2454 suitable debugging ouput when -d is set.
2456 (b) The queryprogram router detects when it is not running as root,
2457 outputs suitable debugging information if -d is set, and then runs
2458 the subprocess without attempting to change uid/gid.
2460 PH/15 Minor change to Makefile for building test_host (undocumented testing
2463 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2464 additional section of a DNS packet that returns MX or SRV records.
2465 Instead, it always explicitly searches for A/AAAA records. This avoids
2466 major problems that occur when a DNS server includes only records of one
2467 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2468 fixed another bug: if SRV records were looked up and the corresponding
2469 address records were *not* found in the additional section, the port
2470 values from the SRV records were lost.
2472 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2473 using the correct key (the original address) when searching the retry
2474 rules in order to find which one to use for generating the retry hint.
2476 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2477 adding the default one. Similarly, if it contains a Reply-To: header, the
2478 errors_reply_to option, if set, is not used.
2480 PH/19 When calculating a retry time, Exim used to measure the "time since
2481 failure" by looking at the "first failed" field in the retry record. Now
2482 it does not use this if it is later than than the arrival time of the
2483 message. Instead it uses the arrival time. This makes for better
2484 behaviour in cases where some deliveries succeed, thus re-setting the
2485 "first failed" field. An example is a quota failure for a huge message
2486 when small messages continue to be delivered. Without this change, the
2487 "time since failure" will always be short, possible causing more frequent
2488 delivery attempts for the huge message than are intended.
2489 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2491 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2492 $1, $2, $3) because the numerical variables can be reset during some
2493 expansion items (e.g. "match"), thereby losing the authentication data.
2495 PH/21 Make -bV show the size of off_t variables so that the test suite can
2496 decide whether to run tests for quotas > 2G.
2498 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2499 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2500 filecount value is greater than 2G or if a quota value is greater than 2G
2501 on a system where the size of off_t is not greater than 4, a panic error
2504 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2505 never match. The debug and -bh output now contains an explicit error
2506 message indicating a malformed IPv4 address or mask.
2508 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2509 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2510 PH/23 above applies.
2512 PH/25 Do not write to syslog when running in the test harness. The only
2513 occasion when this arises is a failure to open the main or panic logs
2514 (for which there is an explicit test).
2516 PH/26 Added the /no_tell option to "control=freeze".
2518 PH/27 If a host name lookup failed very early in a connection, for example, if
2519 the IP address matched host_lookup and the reverse lookup yielded a name
2520 that did not have a forward lookup, an error message of the form "no IP
2521 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2522 could be logged. Now it outputs the IP address instead of "NULL".
2524 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2525 allows the sender and the authenticated sender to be set when
2526 submitting a message from within Exim. Since child_open_exim() is
2527 documented for local_scan(), the new function should be too.
2529 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2530 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2531 results in an empty string is now treated as unset.
2533 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2535 PH/31 Added sender_verify_fail logging option.
2537 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2538 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2539 tidied the source and removed it altogether.
2541 PH/33 When a queue run was abandoned because the load average was too high, a
2542 log line was always written; now it is written only if the queue_run log
2543 selector is set. In addition, the log line for abandonment now contains
2544 information about the queue run such as the pid. This is always present
2545 in "start" and "stop" lines but was omitted from the "abandon" line.
2547 PH/34 Omit spaces between a header name and the colon in the error message that
2548 is given when verify = headers_syntax fails (if there are lots of them,
2549 the message gets confusing).
2551 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2552 names, as there are now some PTR records that contain slashes. This check
2553 is only to protect against broken name servers that fall over on strange
2554 characters, so the fact that it applies to all lookups doesn't matter.
2556 PH/36 Now that the new test suite is complete, we can remove some of the
2557 special code in Exim that was needed for the old test suite. For example,
2558 sorting DNS records because real resolvers return them in an arbitrary
2559 order. The new test suite's fake resolver always returns records in the
2562 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2563 bounces) except when queue_only is set, to avoid logging races between
2564 the different processes.
2566 PH/38 Panic-die if .include specifies a non-absolute path.
2568 PH/39 A tweak to the "H" retry rule from its user.
2570 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2571 a label. They prevented compilation on older perls.
2573 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2574 a warning to be raised on newish perls.
2576 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2577 on queue. Changes to match documented behaviour of showing count of
2578 messages matching specified criteria.
2580 PH/40 Changed the default ident timeout from 30s to 5s.
2582 PH/41 Added support for the use of login_cap features, on those BSD systems
2583 that have them, for controlling the resources used by pipe deliveries.
2585 PH/42 The content-scanning code uses fopen() to create files in which to put
2586 message data. Previously it was not paying any attention to the mode of
2587 the files. Exim runs with umask(0) because the rest of the code creates
2588 files with open(), and sets the required mode explicitly. Thus, these
2589 files were ending up world-writeable. This was not a big issue, because,
2590 being within the spool directory, they were not world-accessible. I have
2591 created a function called modefopen, which takes an additional mode
2592 argument. It sets umask(777), creates the file, chmods it to the required
2593 mode, then resets the umask. All the relevant calls to fopen() in the
2594 content scanning code have been changed to use this function.
2596 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2597 to 24 hours. This avoids potential overflow problems when processing G
2598 and H retry rules. I suspect nobody ever tinkers with this value.
2600 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2602 PH/45 When the plaintext authenticator is running as a client, the server's
2603 challenges are checked to ensure they are valid base64 strings. By
2604 default, the authentication attempt is cancelled if an invalid string is
2605 received. Setting client_ignore_invalid_base64 true ignores these errors.
2606 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2607 they are received. Thus, the responses can be made to depend on the
2608 challenges. If an invalid string is ignored, an empty string is placed in
2611 PH/46 Messages that are created by the autoreply transport now contains a
2612 References: header, in accordance with RFCs 2822 and 3834.
2614 PH/47 Added authenticated_sender_force to the smtp transport.
2616 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2618 PH/49 Installed latest patch from the Sieve maintainer.
2620 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2621 was also set, the appendfile transport was unnecessarily scanning a
2622 directory of message files (e.g. for maildir delivery) to find the count
2623 of files (along with the size), even though it did not need this
2624 information. It now does the scan only if it needs to find either the
2625 size of the count of files.
2627 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2629 PH/52 Two bugs concerned with error handling when the smtp transport is
2632 (i) Exim was not creating retry information for temporary errors given
2633 for individual recipients after the DATA command when the smtp transport
2634 was used in LMTP mode. This meant that they could be retried too
2635 frequently, and not timed out correctly.
2637 (ii) Exim was setting the flag that allows error details to be returned
2638 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2639 recipients that were returned after the DATA command.
2641 PH/53 This is related to PH/52, but is more general: for any failing address,
2642 when detailed error information was permitted to be returned to the
2643 sender, but the error was temporary, then after the final timeout, only
2644 "retry timeout exceeded" was returned. Now it returns the full error as
2645 well as "retry timeout exceeded".
2647 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2648 do this, and (what is worse) MTAs that accept it.
2650 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2651 will now be deprecated.
2653 PH/56 New os.c-cygwin from the Cygwin maintainer.
2655 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2656 formats (previously only available in exim formats via -bpr, -bpru,
2657 and -bpra. Now also available in native and exiqgrep formats)
2659 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2660 with very large, slow to parse queues
2662 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2664 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2666 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2667 responses to authentication challenges, though it was showing the
2668 challenges; (ii) I've removed the CR characters from the debug output for
2671 PH/58 Allow for the insertion of a newline as well as a space when a string
2672 is turned into more than one encoded-word during RFC 2047 encoding. The
2673 Sieve code now uses this.
2675 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2676 data_4xx, lost_connection, tls_required.
2678 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2679 message was being sent as an SMTP response.
2681 PH/61 Add -l and -k options to exicyclog.
2683 PH/62 When verifying, if an address was redirected to one new address, so that
2684 verification continued, and the new address failed or deferred after
2685 having set something in $address_data, the value of $address_data was not
2686 passed back to the ACL. This was different to the case when no
2687 redirection occurred. The value is now passed back in both cases.
2689 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2690 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2691 use login_cap.h, so on its own it isn't the distinguishing feature. The
2692 new name refers directly to the setclassresources() function.
2694 PH/65 Added configuration files for NetBSD3.
2696 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2698 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2699 is preferred over IPv4.
2701 PH/68 The bounce_return_message and bounce_return_body options were not being
2702 honoured for bounces generated during the reception of non-SMTP messages.
2703 In particular, this applied to messages rejected by the ACL. This bug has
2704 been fixed. However, if bounce_return_message is true and bounce_return_
2705 body is false, the headers that are returned for a non-SMTP message
2706 include only those that have been read before the error was detected.
2707 (In the case of an ACL rejection, they have all been read.)
2709 PH/69 The HTML version of the specification is now built in a directory called
2710 spec_html instead of spec.html, because the latter looks like a path with
2711 a MIME-type, and this confuses some software.
2713 PH/70 Catch two compiler warnings in sieve.c.
2715 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2716 function verify_get_ident() calls ip_connect() to connect a socket, but
2717 if the "connect()" function timed out, ip_connect() used to close the
2718 socket. However, verify_get_ident() also closes the socket later, and in
2719 between Exim writes to the log, which may get opened at this point. When
2720 the socket was closed in ip_connect(), the log could get the same file
2721 descriptor number as the socket. This naturally causes chaos. The fix is
2722 not to close the socket in ip_connect(); the socket should be closed by
2723 the function that creates it. There was only one place in the code where
2724 this was missing, in the iplookup router, which I don't think anybody now
2725 uses, but I've fixed it anyway.
2727 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2728 well as to direct DNS lookups. Otherwise the handling of names in host
2729 lists is inconsistent and therefore confusing.
2735 PH/01 Two changes to the default runtime configuration:
2737 (1) Move the checks for relay_from_hosts and authenticated clients from
2738 after to before the (commented out) DNS black list checks.
2740 (2) Add control=submission to the relay_from_hosts and authenticated
2741 clients checks, on the grounds that messages accepted by these
2742 statements are most likely to be submissions.
2744 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2746 (1) Generate an error if the third argument for the ${prvs expansion is
2749 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2752 (3) Reset the variables that are obtained from the first argument of
2753 ${prvscheck and used in the second argument before leaving the code,
2754 because their memory is reclaimed, so using them afterwards may do
2757 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2758 one (it's much easier than Tom thought :-).
2760 (5) Because of (4), we can now allow for the use of $prvscheck_result
2761 inside the third argument.
2763 PH/03 For some reason, the default setting of PATH when running a command from
2764 a pipe transport was just "/usr/bin". I have changed it to
2767 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2768 anything to be listed in the output from -bV.
2770 PH/05 When a filter generated an autoreply, the entire To: header line was
2771 quoted in the delivery log line, like this:
2773 => >A.N.Other <ano@some.domain> <original@ddress> ...
2775 This has been changed so that it extracts the operative address. There
2776 may be more than one such address. If so, they are comma-separated, like
2779 => >ano@some.domain,ona@other.domain <original@ddress> ...
2781 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2782 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2783 not being looked up in the rDNS to get a host name, Exim was showing the
2784 IP address twice in Received: lines, even though the IP addresses were
2785 identical. For example:
2787 Received: from [1.2.3.4] (helo=[1.2.3.4])
2789 However, if the real host name was known, it was omitting the HELO data
2790 if it matched the actual IP address. This has been tidied up so that it
2791 doesn't show the same IP address twice.
2793 PH/07 When both +timestamp and +memory debugging was on, the value given by
2794 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2795 called by the debug printing, thereby overwriting the timestamp buffer.
2796 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2798 PH/08 When the original message was included in an autoreply transport, it
2799 always said "this is a copy of the message, including all the headers",
2800 even if body_only or headers_only was set. It now gives an appropriate
2803 PH/09 Applied a patch from the Sieve maintainer which:
2805 o fixes some comments
2806 o adds the (disabled) notify extension core
2807 o adds some debug output for the result of if/elsif tests
2808 o points to the current vacation draft in the documentation
2809 and documents the missing references header update
2813 o fixes a bug in processing the envelope test (when testing
2814 multiple envelope elements, the last element determinted the
2817 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2818 Electronic Mail") by including:
2820 Auto-submitted: auto-generated
2822 in the messages that it generates (bounce messages and others, such as
2823 warnings). In the case of bounce messages for non-SMTP mesages, there was
2824 also a typo: it was using "Auto_submitted" (underscore instead of
2825 hyphen). Since every message generated by Exim is necessarily in response
2826 to another message, thes have all been changed to:
2828 Auto-Submitted: auto-replied
2830 in accordance with these statements in the RFC:
2832 The auto-replied keyword:
2834 - SHOULD be used on messages sent in direct response to another
2835 message by an automatic process,
2837 - MUST NOT be used on manually-generated messages,
2839 - MAY be used on Delivery Status Notifications (DSNs) and Message
2840 Disposition Notifications (MDNs),
2842 - MUST NOT be used on messages generated by automatic or periodic
2843 processes, except for messages which are automatic responses to
2846 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2847 to the default Received: header definition.
2849 PH/12 Added log selector acl_warn_skipped (default on).
2851 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2852 variables because (a) they are in the wrong storage pool and (b) even if
2853 they were copied, it wouldn't work properly because of the caching.
2855 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2856 checking when decoding. Apparently there are clients that generate
2857 overlong encoded strings. Why am I not surprised?
2859 PH/15 If the first argument of "${if match_address" was not empty, but did not
2860 contain an "@" character, Exim crashed. Now it writes a panic log message
2861 and treats the condition as false.
2863 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2865 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2866 extension "envelope-auth". The code is finished and in agreement with
2867 other implementations, but there is no documentation so far and in fact,
2868 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2869 not changing the active code.
2871 Print executed "if" and "elsif" statements when debugging is used. This
2872 helps a great deal to understand what a filter does.
2874 Document more things not specified clearly in RFC3028. I had all this
2875 sorted out, when out of a sudden new issues came to my mind. Oops."
2877 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2880 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2881 canonical form (as documented). However, after a host name lookup from
2882 the IP address, check_host() was doing a simple string comparison with
2883 addresses acquired from the DNS when checking that the found name did
2884 have the original IP as one of its addresses. Since any found IPv6
2885 addresses are likely to be in abbreviated form, the comparison could
2886 fail. Luckily, there already exists a function for doing the comparison
2887 by converting both addresses to binary, so now that is used instead of
2888 the text comparison.
2890 PH/20 There was another similar case to PH/19, when a complete host name was
2891 given in a host list; looking up its IP address could give an abbreviated
2892 form, whereas the current host's name might or might not be abbreviated.
2893 The same fix has been applied.
2899 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2900 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2903 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2904 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2906 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2908 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2909 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2910 Also, for "Auto-Submitted", treat anything other than "no" as
2911 non-personal, in accordance with RFC 3834. (Previously it treated
2912 anything starting "auto-" as non-personal.)
2914 TF/01 The control=submission/name=... option had a problem with syntax
2915 errors if the name included a slash character. The /name= option
2916 now slurps the rest of the string, so it can include any characters
2917 but it must come last in the list of options (after /sender_retain
2920 PH/05 Some modifications to the interface to the fake nameserver for the new
2928 TK/01 Added the "success_on_redirect" address verification option. See
2929 NewStuff for rationale and an example.
2931 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2933 PH/02 Patch to exigrep to allow it to work on syslog lines.
2935 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2936 fread() to skip over the body file's header line, because in Cygwin the
2937 header line is locked and is inaccessible.
2939 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2940 co-exist for some time) to make it clear that it is the Exim ID that is
2941 referenced, not the Message-ID: header line.
2943 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2944 string_format() function, because snprintf() does not exist on all
2947 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2948 ${stat: expansion item.
2950 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2951 protocol synchronization error", to keep the pedants happy.
2953 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2954 well as for IRIX systems, when gcc is being used. See the host.c source
2957 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2959 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2962 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2963 transport and to the smtp transport in LMTP mode.
2965 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2967 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2968 run a filter in a subprocess. This could lead to confusion in subsequent
2969 lookups in the parent process. There should also be a search_tidyup() at
2970 the end of the subprocess.
2972 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2973 only if the host matched helo_try_verify_hosts, which caused the
2974 verification to occur when the EHLO/HELO command was issued. The ACL just
2975 tested the remembered result. Now, if a previous verification attempt has
2976 not happened, "verify = helo" does it there and then.
2978 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2980 TK/03 Fix log output including CR from clamd.
2982 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2983 exist provoked a memory error which could cause a segfault.
2985 PH/15 Installed PCRE 6.2
2987 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2989 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2990 of the problem. Specifically, suggested +O2 rather than +O1 for the
2993 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2995 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2996 record, the port was not correctly logged when the outgoing_port log
2997 selector was set (it logged the transort's default port).
2999 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3000 fallback_hosts, and "hosts" in the smtp transport.
3002 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3003 host errors such as "Connection refused".
3005 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3006 authentication with radiusclient 0.4.9:
3008 - Error returned from rc_read_config was caught wrongly
3009 - Username/password not passed on to radius server due to wrong length.
3011 The presumption is that some radiusclient API changes for 4.51/PH/17
3012 were not taken care of correctly. The code is still untested by me (my
3013 Linux distribution still has 0.3.2 of radiusclient), but it was
3014 contributed by a Radius user.
3016 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3017 expanding the "port" option of the smtp transport.
3019 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3020 while reading a MIME header. Thanks to Tom Hughes for a patch.
3022 PH/24 Include config.h inside local_scan.h so that configuration settings are
3025 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3026 that in an ACL for RCPT (for example), you can examine exactly what was
3029 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3030 commands, but it was not correctly comparing the address with the actual
3031 client host address. Thus, it would show the EHLO address in Received:
3032 header lines when this was not necessary.
3034 PH/27 Added the % operator to ${eval:}.
3036 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3037 it should be ignoring failures (because with -C, for example, it has lost
3038 privilege). It wasn't ignoring creation failures other than "already
3041 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3044 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3045 because an input line was too long, either on its own, or by virtue of
3046 too many continuations, the temporary file was not being removed, and the
3047 return code was incorrect.
3049 PH/31 Missing "BOOL" in function definition in filtertest.c.
3051 PH/32 Applied Sieve patches from the maintainer.
3053 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3055 PH/33 Added "verify = not_blind".
3057 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3058 Local/Makefile (with some defaults set). These are used in built scripts
3059 such as exicyclog, but they have never been used in the exim_install
3060 script (though there are many overriding facilities there). I have
3061 arranged that the exim_install script now takes note of these two
3064 PH/35 Installed configuration files for Dragonfly.
3066 PH/36 When a locally submitted message by a trusted user did not contain a
3067 From: header, and the sender address was obtained from -f or from an SMTP
3068 MAIL command, and the trusted user did not use -F to supply a sender
3069 name, $originator_name was incorrectly used when constructing a From:
3070 header. Furthermore, $originator_name was used for submission mode
3071 messages from external hosts without From: headers in a similar way,
3072 which is clearly wrong.
3074 PH/37 Added control=suppress_local_fixups.
3076 PH/38 When log_selector = +received_sender was set, and the addition of the
3077 sender made the log line's construction buffer exactly full, or one byte
3078 less than full, an overflow happened when the terminating "\n" was
3081 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3082 when the result of a list match is failure because a DNS lookup failed.
3084 PH/40 RM_COMMAND is now used in the building process.
3086 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3087 the "build-* directories that it finds.
3089 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3090 address in a domain literal was a prefix of an interface address.
3092 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3093 when verifying a sender address, unless rewrite_headers is false.
3095 PH/44 Wrote a long comment about why errors_to addresses are verified as
3096 recipients, not senders.
3098 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3099 the ratelimit ACL was added.
3101 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3103 PH/46 Added extra information about PostgreSQL errors to the error string.
3105 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3106 avoiding the need to install special zones in a real server. This is
3107 backwards compatible; if it can't find the fake resolver, it drops back.
3108 Thus, both old and new test suites can be run.
3110 TF/02 Added util/ratelimit.pl
3112 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3113 clock is set back in time.
3115 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3116 Candler <B.Candler@pobox.com>.
3118 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3119 OK for addresses that are the result of redirections.
3121 PH/48 A number of further additions for the benefit of the new test suite,
3122 including a fake gethostbyname() that interfaces to the fake DNS resolver
3125 TF/06 The fix for widen_domains has also been applied to qualify_single and
3126 search_parents which are the other dnslookup options that can cause
3129 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3132 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3134 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3135 service" if it encountered an MX record with an empty target hostname.
3136 The message is now "an MX or SRV record indicated no SMTP service".
3138 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3139 if the DNS of the sending site is misconfigured. This is quite a
3140 common situation. This change restores the behaviour of treating a
3141 helo verification defer as a failure.
3143 PH/51 If self=fail was set on a router, the bounce message did not include the
3144 actual error message.
3150 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3152 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3153 command itself ended in error, the underlying message about the transport
3154 filter timeout was being overwritten with the pipe command error. Now the
3155 underlying error message should be appended to the second error message.
3157 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3159 PH/02 Reduce the amount of output that "make" produces by default. Full output
3160 can still be requested.
3162 PH/03 The warning log line about a condition test deferring for a "warn" verb
3163 was being output only once per connection, rather than after each
3164 occurrence (because it was using the same function as for successful
3165 "warn" verbs). This seems wrong, so I have changed it.
3167 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3168 it should not have, which might have caused a crash in the right
3169 circumstances, but probably never did.
3171 PH/04 Installed a modified version of Tony Finch's patch to make submission
3172 mode fix the return path as well as the Sender: header line, and to
3173 add a /name= option so that you can make the user's friendly name appear
3176 TF/03 Added the control = fakedefer ACL modifier.
3178 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3179 Mark Lowes for thorough testing.
3181 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3183 TK/03 Merged latest SRS patch from Miles Wilton.
3185 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3186 with the definition in sysexits.h (which is #included earlier).
3187 Fortunately, Exim does not actually use EX_OK. The code used to try to
3188 preserve the sysexits.h value, by assumimg that macro definitions were
3189 scanned for macro replacements. I have been disabused of this notion,
3190 so now the code just undefines EX_OK before #including unistd.h.
3192 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3193 in the smtp transport. When a block could not be written in a single
3194 write() function, the timeout was being re-applied to each part-write.
3195 This seems wrong - if the receiver was accepting one byte at a time it
3196 would take for ever. The timeout is now adjusted when this happens. It
3197 doesn't have to be particularly precise.
3199 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3200 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3202 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3203 without a domain if the check to <postmaster@domain> fails.
3205 SC/01 Eximstats: added -xls and the ability to specify output files
3206 (patch written by Frank Heydlauf).
3208 SC/02 Eximstats: use FileHandles for outputing results.
3210 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3212 SC/04 Eximstats: fixed display of large numbers with -nvr option
3214 SC/05 Eximstats: fixed merging of reports with empty tables.
3216 SC/06 Eximstats: added the -include_original_destination flag
3218 SC/07 Eximstats: removed tabs and trailing whitespace.
3220 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3222 TK/06 MBOX spool code: Add real "From " MBOX separator line
3223 so the .eml file is really in mbox format (even though
3224 most programs do not really care). Patch from Alex Miller.
3226 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3227 The latter is generated from $received_to and is only set if the
3228 message has one envelope recipient. SA can use these headers,
3229 obviously out-of-the-box. Patch from Alex Miller.
3231 PH/08 The ${def test on a variable was returning false if the variable's
3232 value was "0", contrary to what the specification has always said!
3233 The result should be true unless the variable is empty.
3235 PH/09 The syntax error of a character other than { following "${if
3236 def:variable_name" (after optional whitespace) was not being diagnosed.
3237 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3238 accidental colon was present, for example, could give incorrect results.
3240 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3241 result is used (not including appendfile, where other changes are about
3244 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3245 This involved changing a lot of size variables from int to off_t. It
3246 should work with maildirs and everything.
3248 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3249 spamd dying while we are connected to it.
3251 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3254 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3255 function for BDB 4.3.
3257 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3259 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3260 the sources for PCRE have been re-arranged and more files are now
3263 PH/15 The code I had for printing potentially long long variables in PH/11
3264 above was not the best (it lost precision). The length of off_t variables
3265 is now inspected at build time, and an appropriate printing format (%ld
3266 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3267 to be "long long int" or "long int". This is needed for the internal
3268 formatting function string_vformat().
3270 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3271 the configuration file to be ":syslog", then the script "guesses" where
3272 the logs files are, rather than using the compiled in default. In our
3273 case the guess is not the same as the compiled default, so the script
3274 suddenly stopped working when I started to use syslog. The patch checks
3275 to see if log_file_path is "". If so, it attempts to read it from exim
3276 with no configuration file to get the compiled in version, before it
3277 falls back to the previous guessing code."
3279 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3280 implementing BATV in an Exim configuration. See NewStuff for the gory
3283 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3284 Makefile that are specific to HP-UX.
3286 PH/18 If the "use_postmaster" option was set for a recipient callout together
3287 with the "random" option, the postmaster address was used as the MAIL
3288 FROM address for the random test, but not for the subsequent recipient
3289 test. It is now used for both.
3291 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3292 patch removes a few documentation additions to RFC 3028, because the
3293 latest draft now contains them. It adds the new en;ascii-case comparator
3294 and a new error check for 8bit text in MIME parts. Comparator and
3295 require names are now matched exactly. I enabled the subaddress
3296 extension, but it is not well tested yet (read: it works for me)."
3298 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3299 rework some of the code of TK/09 above to avoid the hardwired use of
3300 "%lld" and "long long". Replaced the call to snprintf() with a call to
3303 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3304 records point to non-existent hosts", "retry timeout exceeded", and
3305 "retry time not reached for any host after a long failure period".
3307 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3308 experimental DomainKeys support:
3310 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3311 (2) On an error such as an illegally used "control", the wrong name for
3312 the control was given.
3314 These problems did NOT occur unless DomainKeys support was compiled.
3316 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3318 PH/24 Added ${if match_ip condition.
3320 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3321 databases so that it will be absolutely obvious if a crash occurs in the
3322 DB library. This is a regular occurrence (often caused by mis-matched
3325 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3326 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3327 picked up on a user's system that detects such things. There doesn't seem
3328 to be a gcc warning option for this - only an attribute that has to be
3329 put on the function's prototype. It seems that in Fedora Core 4 they have
3330 set this on a number of new functions. No doubt there will be more in due
3333 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3334 specify a transport. However, if an address that was verified by such a
3335 router was the subject of a callout, Exim crashed because it tried to
3336 read the rcpt_include_affixes from the non-existent transport. Now it
3337 just assumes that the setting of that option is false. This bug was
3338 introduced by 4.51/PH/31.
3340 PH/28 Changed -d+all to exclude +memory, because that information is very
3341 rarely of interest, but it makes the output a lot bigger. People tend to
3342 do -d+all out of habit.
3344 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3345 code in os-type was giving problems when libc.so lives in lib64, like on
3348 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3349 aren't the modern standard, and it seems that some systems' include files
3350 don't always have them. Exim was already checking for some of the newer
3351 ones like T_AAAA, and defining it itself. I've added checks for all the
3352 record types that Exim uses.
3354 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3355 not automatically generating a new one, as it is supposed to. This
3356 prevented TLS from working. If the file did exist, but contained invalid
3357 data, a new version was generated, as expected. It was only the case of a
3358 non-existent file that was broken.
3360 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3361 with a change in libdomainkeys > 0.64.
3363 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3364 from DNS. If the selector record carries the flag, it now has
3365 precedence over the domain-wide flag.
3367 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3369 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3370 the use of a port name that isn't defined in /etc/services) occurred, the
3371 message was deferred as in a normal delivery, and thus remained on the
3372 spool, instead of being failed because of the mua_wrapper setting. This
3373 is now fixed, and I tidied up some of the mua_wrapper messages at the
3376 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3377 the messages in a hash of arrays rather than using individual hashes.
3378 This is a bit cleaner and results in dramatic memory savings, albeit
3379 at a slight CPU cost.
3381 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3382 as requested by Marc Sherman.
3384 SC/10 Eximstats: added histograms for user specified patterns as requested
3387 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3389 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3390 fopen() in the content-scanning modules that did not already have it.
3396 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3397 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3399 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3401 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3403 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3404 to test Sieve filters that use "vacation".
3406 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3407 that changes the way the GnuTLS parameters are stored in the cache file.
3408 The new format can be generated externally. For backward compatibility,
3409 if the data in the cache doesn't make sense, Exim assumes it has read an
3410 old-format file, and it generates new data and writes a new file. This
3411 means that you can't go back to an older release without removing the
3414 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3415 work if there are any delivery delays because "one_time" forces the
3416 parent to be marked "delivered", so its unseen clone is never tried
3417 again. For this reason, Exim now forbids the simultaneous setting of
3420 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3421 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3422 entry if you want to know the details. The fix, however, neglected to
3423 consider the case when local delivery batching is involved. The test for
3424 "previously delivered" was not happening when checking to see if an
3425 address could be batched with a previous (undelivered) one; under
3426 certain circumstances this could lead to multiple deliveries to the same
3429 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3430 in its include files, and this causes problems building Exim.
3432 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3433 header_syntax) but Exim was just ignoring anything given after a slash.
3434 In particular, this caused confusion with an attempt to use "verify =
3435 reverse_host_lookup/defer_ok". An error is now given when options are
3436 supplied for verify items that do not have them. (Maybe reverse_host_
3437 lookup should have a defer_ok option, but that's a different point.)
3439 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3440 defined by RFC 821) to 2048, because there were problems with some AUTH
3441 commands, and RFC 1869 says the size should be increased for extended
3442 SMTP commands that take arguments.
3444 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3447 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3448 "unknown" error; now it says that the functionality isn't in the binary.
3450 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3451 an address' error message when a string expansion fails (syntax or
3452 whatever). Otherwise the password may appear in the log. Following change
3453 PH/42 below, there is no longer a chance of it appearing in a bounce
3456 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3458 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3459 of its addresses was ever tried. (Bugzilla bug #2.)
3461 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3462 the result incorrectly in the debug output. (It correctly added a newline
3463 to what was transported.)
3465 TF/01 Added $received_time.
3467 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3468 commented out examples of how to interface to a virus scanner and to
3469 SpamAssassin. Also added commented examples of av_scanner and
3470 spamd_address settings.
3472 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3473 and controls are allowed in which ACLs. There were a couple of minor
3474 errors. Some of the entries in the conditions table (which is a table of
3475 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3476 negation of where the condition IS allowed.
3478 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3480 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3481 header file does not have a version number, so I've had to invent a new
3482 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3483 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3484 radiusclient), but it was contributed by a Radius user.
3486 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3487 files or directories, for OpenSSL.
3489 PH/19 When an Exim process that is running as root has to create an Exim log
3490 file, it does so in a subprocess that runs as exim:exim so as to get the
3491 ownership right at creation (otherwise, other Exim processes might see
3492 the file with the wrong ownership). There was no test for failure of this
3493 fork() call, which would lead to the process getting stuck as it waited
3494 for a non-existent subprocess. Forks do occasionally fail when resources
3495 run out. I reviewed all the other calls to fork(); they all seem to check
3498 PH/20 When checking for unexpected SMTP input at connect time (before writing
3499 the banner), Exim was not dealing correctly with a non-positive return
3500 from the read() function. If the client had disconnected by this time,
3501 the result was a log entry for a synchronization error with an empty
3502 string after "input=" when read() returned zero. If read() returned -1
3503 (an event I could not check), uninitialized data bytes were printed.
3504 There were reports of junk text (parts of files, etc) appearing after
3507 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3509 PH/22 Added support for macro redefinition, and (re)definition in between
3510 driver and ACL definitions.
3512 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3513 forgetting to use the resulting value; it was using the unexpanded value.
3515 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3516 hadn't been configured. The fix is from Juergen Kreileder, who
3517 understands it better than I do:
3519 "Here's what I see happening with three configured cyrus_sasl
3520 authenticators configured (plain, login, cram-md5):
3522 On startup auth_cyrus_sasl_init() gets called for each of these.
3523 This means three calls to sasl_listmech() without a specified mech_list.
3524 => SASL tests which mechs of all available mechs actually work
3525 => three warnings about OTP not working
3526 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3528 With the patch, sasl_listmech() also gets called three times. But now
3529 SASL's mech_list option is set to the server_mech specified in the the
3530 authenticator. Or in other words, the answer from sasl_listmech()
3531 gets limited to just the mech you're testing for (which is different
3533 => the return list contains just 'plain' or 'login', 'cram-md5' or
3534 nothing depending on the value of ob->server_mech.
3536 I've just tested the patch: Authentication still works fine,
3537 unavailable mechs specified in the exim configuration are still
3538 caught, and the auth.log warnings about OTP are gone."
3540 PH/25 When debugging is enabled, the contents of the command line are added
3541 to the debugging output, even when log_selector=+arguments is not
3544 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3545 answer is "GNU", and only if the return is "GNU/something" is the answer
3548 PH/27 $acl_verify_message is now set immediately after the failure of a
3549 verification in an ACL, and so is available in subsequent modifiers. In
3550 particular, the message can be preserved by coding like this:
3552 warn !verify = sender
3553 set acl_m0 = $acl_verify_message
3555 Previously, $acl_verify_message was set only while expanding "message"
3556 and "log_message" when a very denied access.
3558 PH/28 Modified OS/os.c-Linux with
3560 -#ifndef OS_LOAD_AVERAGE
3561 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3563 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3564 nomenclature these days.)
3566 PH/29 Installed patch from the Sieve maintainer that adds the options
3567 sieve_useraddress and sieve_subaddress to the redirect router.
3569 PH/30 In these circumstances:
3570 . Two addresses routed to the same list of hosts;
3571 . First host does not offer TLS;
3572 . First host accepts first address;
3573 . First host gives temporary error to second address;
3574 . Second host offers TLS and a TLS session is established;
3575 . Second host accepts second address.
3576 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3577 and peerdn, if requested) that were in fact used only for the second
3580 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3581 attention to any local part prefix or suffix that was matched by the
3582 router that accepted the address. It now behaves in the same way as it
3583 does for delivery: the affixes are removed from the local part unless
3584 rcpt_include_affixes is set on the transport.
3586 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3587 timeout during the DATA phase of an incoming message.
3589 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3590 applied a patch sanctioned by the Sieve maintainer.
3592 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3593 the uid or gid is negative. A case of a negative gid caused this to be
3594 noticed. The fix allows for either to be negative.
3596 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3597 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3600 PH/36 The change PH/12 above was broken. Fixed it.
3602 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3603 the grounds that routing the same address twice would always produce the
3604 same answer. This might have been true once, but it is certainly no
3605 longer true now. Routing a child address may depend on the previous
3606 routing that produced that child. Some complicated redirection strategies
3607 went wrong when messages had multiple recipients, and made Exim's
3608 behaviour dependent on the order in which the addresses were given.
3610 I have moved the duplicate checking until after the routing is complete.
3611 Exim scans the addresses that are assigned to local and remote
3612 transports, and removes any duplicates. This means that more work will be
3613 done, as duplicates will always all be routed, but duplicates are
3614 presumably rare, so I don't expect this is of any significance.
3616 For deliveries to pipes, files, and autoreplies, the duplicate checking
3617 still happens during the routing process, since they are not going to be
3620 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3621 It corrects a timeout issue with spamd. This is Ian's comment: "The
3622 background is that sometimes spamd either never reads data from a
3623 connection it has accepted, or it never writes response data. The exiscan
3624 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3625 blindly assumes that writes won't block so it may never time out."
3627 PH/39 Allow G after quota size as well as K and M.
3629 PH/40 The value set for $authenticated_id in an authenticator may not contain
3630 binary zeroes or newlines because the value is written to log lines and
3631 to spool files. There was no check on this. Now the value is run through
3632 the string_printing() function so that such characters are converted to
3633 printable escape sequences.
3635 PH/41 $message_linecount is a new variable that contains the total number of
3636 lines in the message. Compare $body_linecount, which is the count for the
3639 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3640 bounce and delay warning messages, except in certain special cases, which
3643 (a) An SMTP error message from a remote host;
3644 (b) A message specified in a :fail: redirection;
3645 (c) A message specified in a "fail" command in a system filter;
3646 (d) A message specified in a FAIL return from the queryprogram router;
3647 (e) A message specified by the cannot_route_message router option.
3649 In these cases only, Exim does include the error details in bounce and
3650 warning messages. There are also a few cases where bland messages such
3651 as "unrouteable address" or "local delivery error" are given.
3653 PH/43 $value is now also set for the "else" part of a ${run expansion.
3655 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3656 being worked on, but at least Exim now implements the latest version to
3659 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3660 process to complete was treated as a delivery failure, a timeout while
3661 writing the message to the pipe was logged, but erroneously treated as a
3662 successful delivery. Such timeouts include transport filter timeouts. For
3663 consistency with the overall process timeout, these timeouts are now
3664 treated as errors, giving rise to delivery failures by default. However,
3665 there is now a new Boolean option for the pipe transport called
3666 timeout_defer, which, if set TRUE, converts the failures into defers for
3667 both kinds of timeout. A transport filter timeout is now identified in
3670 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3671 systems where "make" and "gmake" are different, calling "gmake" at top
3672 level broke things. I've arranged for the value of $(MAKE) to be passed
3673 from the Makefile to this script so that it can call the same version of
3677 A note about Exim versions 4.44 and 4.50
3678 ----------------------------------------
3680 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3681 changes of various kinds. As a consequence, a big documentation update was
3682 needed. This delayed the release for rather longer than seemed good, especially
3683 in the light of a couple of (minor) security issues. Therefore, the changes
3684 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3685 release. So 4.44 and 4.50 are in effect two different branches that both start
3688 I have left the 4.50 change log unchanged; it contains all the changes since
3689 4.43. The change log for 4.44 is below; many of its items are identical to
3690 those for 4.50. This seems to be the most sensible way to preserve the
3691 historical information.
3697 1. Minor wording change to the doc/README.SIEVE file.
3699 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3700 computation of the current number of files was incorrect.
3702 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3703 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3706 4. Give more explanation in the error message when the command for a transport
3707 filter fails to execute.
3709 5. There are several places where Exim runs a non-Exim command in a
3710 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3711 was being done only for the command run by the queryprogram router. It is
3712 now done for all such subprocesses. The other cases are: ${run, transport
3713 filters, and the commands run by the lmtp and pipe transports.
3715 6. Added CONFIGURE_GROUP build-time option.
3717 7. Some older OS have a limit of 256 on the maximum number of file
3718 descriptors. Exim was using setrlimit() to set 1000 as a large value
3719 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3720 systems. I've change it so that if it can't get 1000, it tries for 256.
3722 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3723 was an oversight, and furthermore, ever since the addition of extra
3724 controls (e.g. 4.43/32), the checks on when to allow different forms of
3725 "control" were broken. There should now be diagnostics for all cases when a
3726 control that does not make sense is encountered.
3728 9. Added the /retain_sender option to "control=submission".
3730 10. $recipients is now available in the predata ACL (oversight).
3732 11. Tidy the search cache before the fork to do a delivery from a message
3733 received from the command line. Otherwise the child will trigger a lookup
3734 failure and thereby defer the delivery if it tries to use (for example) a
3735 cached ldap connection that the parent has called unbind on.
3737 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3738 of $address_data from the recipient verification was clobbered by the
3739 sender verification.
3741 13. The value of address_data from a sender verification is now available in
3742 $sender_address_data in subsequent conditions in the ACL statement.
3744 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3746 15. Added a new option "connect=<time>" to callout options, to set a different
3749 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3750 was its contents. (It was OK if the option was not defined at all.)
3752 17. A "Completed" log line is now written for messages that are removed from
3753 the spool by the -Mrm option.
3755 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3756 information about exactly what failed.
3758 19. Added -dd to debug only the daemon process.
3760 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3761 handles timeouts, both on the server side and network timeouts. Renamed the
3762 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3764 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3765 It is now set to "smtps".
3767 22. $host_address is now set to the target address during the checking of
3768 ignore_target_hosts.
3770 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3771 being passed; this would have caused $sender_host_name to have been used if
3772 matching the list had actually called for a host name (not very likely,
3773 since this list is usually IP addresses). A host name is now passed as
3776 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3777 code that specifies a non-restarting handler (typically sigaction() in
3778 modern systems) in an attempt to fix a rare and obscure crash bug.
3780 25. Narrowed the window for a race in the daemon that could cause it to ignore
3781 SIGCHLD signals. This is not a major problem, because they are used only to
3782 wake it up if nothing else does.
3784 26. A malformed maildirsize file could cause Exim to calculate negative values
3785 for the mailbox size or file count. Odd effects could occur as a result.
3786 The maildirsize information is now recalculated if the size or filecount
3789 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3790 support for a long time. Removed HAVE_SYS_VFS_H.
3792 28. Installed the latest version of exipick from John Jetmore.
3794 29. In an address list, if the pattern was not a regular expression, an empty
3795 subject address (from a bounce message) matched only if the pattern was an
3796 empty string. Non-empty patterns were not even tested. This was the wrong
3797 because it is perfectly reasonable to use an empty address as part of a
3798 database query. An empty address is now tested by patterns that are
3799 lookups. However, all the other forms of pattern expect the subject to
3800 contain a local part and a domain, and therefore, for them, an empty
3801 address still always fails if the pattern is not itself empty.
3803 30. Exim went into a mad DNS loop when attempting to do a callout where the
3804 host was specified on an smtp transport, and looking it up yielded more
3805 than one IP address.
3807 31. Re-factored the code for checking spool and log partition space into a
3808 function that finds that data and another that does the check. The former
3809 is then used to implement four new variables: $spool_space, $log_space,
3810 $spool_inodes, and $log_inodes.
3812 32. The RFC2047 encoding function was originally intended for short strings
3813 such as real names; it was not keeping to the 75-character limit for
3814 encoded words that the RFC imposes. It now respects the limit, and
3815 generates multiple encoded words if necessary. To be on the safe side, I
3816 have increased the buffer size for the ${rfc2047: expansion operator from
3819 33. It is now permitted to omit both strings after an "if" condition; if the
3820 condition is true, the result is "true". As before, when the second string
3821 is omitted, a false condition yields an empty string. This makes it less
3822 cumbersome to write custom ACL and router conditions.
3824 34. Failure to deliver a bounce message always caused it to be frozen, even if
3825 there was an errors_to setting on the router. The errors_to setting is now
3828 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3829 canonical form (fully expanded) before being placed in
3830 $sender_host_address.
3832 36. The table in the code that translates DNS record types into text (T_A to
3833 "A" for instance) was missing entries for NS and CNAME. It is just possible
3834 that this could have caused confusion if both these types were looked up
3835 for the same domain, because the text type is used as part of Exim's
3836 per-process caching. But the chance of anyone hitting this buglet seems
3839 37. The dnsdb lookup has been extended in a number of ways.
3841 (1) There is a new type, "zns", which walks up the domain tree until it
3842 finds some nameserver records. It should be used with care.
3844 (2) There is a new type, "mxh", which is like "mx" except that it returns
3845 just the host names, not the priorities.
3847 (3) It is now possible to give a list of domains (or IP addresses) to be
3848 looked up. The behaviour when one of the lookups defers can be
3849 controlled by a keyword.
3851 (4) It is now possible to specify the separator character for use when
3852 multiple records are returned.
3854 38. The dnslists ACL condition has been extended: it is now possible to supply
3855 a list of IP addresses and/or domains to be looked up in a particular DNS
3858 39. Added log_selector=+queue_time_overall.
3860 40. When running the queue in the test harness, wait just a tad after forking a
3861 delivery process, to get repeatability of debugging output.
3863 41. Include certificate and key file names in error message when GnuTLS fails
3864 to set them up, because the GnuTLS error message doesn't include the name
3865 of the failing file when there is a problem reading it.
3867 42. Allow both -bf and -bF in the same test run.
3869 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3871 44. The "Exiscan patch" is now merged into the mainline Exim source.
3873 45. Sometimes the final signoff response after QUIT could fail to get
3874 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3875 < 0 before doing a fflush(). This bug looks as though it goes back to the
3876 introduction of TLS in release 3.20, but "sometimes" must have been rare
3877 because the tests only now provoked it.
3879 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3880 (this can affect the format of dates).
3882 47. exim_tidydb, when checking for the continued existence of a message for
3883 which it has found a message-specific retry record, was not finding
3884 messages that were in split spool directories. Consequently, it was
3885 deleting retry records that should have stayed in existence.
3887 48. Steve fixed some bugs in eximstats.
3889 49. The SPA authentication driver was not abandoning authentication and moving
3890 on to the next authenticator when an expansion was forced to fail,
3891 contradicting the general specification for all authenticators. Instead it
3892 was generating a temporary error. It now behaves as specified.
3894 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3895 (the order specifies the preference for clients). The order is now AES256,
3896 AES128, 3DES, ARCFOUR128.
3898 51. Small patch to Sieve code - explicitly set From: when generating an
3901 52. Exim crashed if a remote delivery caused a very long error message to be
3902 recorded - for instance if somebody sent an entire SpamAssassin report back
3903 as a large number of 550 error lines. This bug was coincidentally fixed by
3904 increasing the size of one of Exim's internal buffers (big_buffer) that
3905 happened as part of the Exiscan merge. However, to be on the safe side, I
3906 have made the code more robust (and fixed the comments that describe what
3909 53. Now that there can be additional text after "Completed" in log lines (if
3910 the queue_time_overall log selector is set), a one-byte patch to exigrep
3911 was needed to allow it to recognize "Completed" as not the last thing in
3914 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3915 patch that reportedly fixes this has been added. I am not expert enough to
3916 create a test for it. This is what the patch creator wrote:
3918 "I found a little strange behaviour of ldap code when working with
3919 Windows 2003 AD Domain, where users was placed in more than one
3920 Organization Units. When I tried to give exim partial DN, the exit code
3921 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3922 But simultaneously result of request was absolutely normal ldap result,
3923 so I produce this patch..."
3925 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3926 REFERENCE, so I have modified the code to exclude the patch when that macro
3929 55. Some experimental protocols are using DNS PTR records for new purposes. The
3930 keys for these records are domain names, not reversed IP addresses. The
3931 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3932 leaves it alone. Component reversal etc. now happens only for IP addresses.
3935 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3937 57. Double the size of the debug message buffer (to 2048) so that more of very
3938 long debug lines gets shown.
3940 58. The exicyclog utility now does better if the number of log files to keep
3941 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3943 59. Two changes related to the smtp_active_hostname option:
3945 (1) $smtp_active_hostname is now available as a variable.
3946 (2) The default for smtp_banner uses $smtp_active_hostname instead
3947 of $primary_hostname.
3949 60. The host_aton() function is supposed to be passed a string that is known
3950 to be a valid IP address. However, in the case of IPv6 addresses, it was
3951 not checking this. This is a hostage to fortune. Exim now panics and dies
3952 if the condition is not met. A case was found where this could be provoked
3953 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3954 components; fortuitously, this particular loophole had already been fixed
3955 by change 4.50/55 above.
3957 If there are any other similar loopholes, the new check in host_aton()
3958 itself should stop them being exploited. The report I received stated that
3959 data on the command line could provoke the exploit when Exim was running as
3960 exim, but did not say which command line option was involved. All I could
3961 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3962 running as the user.
3965 61. There was a buffer overflow vulnerability in the SPA authentication code
3966 (which came originally from the Samba project). I have added a test to the
3967 spa_base64_to_bits() function which I hope fixes it.
3970 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3971 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3973 63. The daemon start-up calls getloadavg() while still root for those OS that
3974 need the first call to be done as root, but it missed one case: when
3975 deliver_queue_load_max is set with deliver_drop_privilege. This is
3976 necessary for the benefit of the queue runner, because there is no re-exec
3977 when deliver_drop_privilege is set.
3979 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3980 This has been fixed.
3982 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3983 was in use, was not putting the data itself into the right store pool;
3984 consequently, it could be overwritten for a subsequent message in the same
3985 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3988 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3990 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3991 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3992 the function were treating the return as a boolean value, which happened to
3993 work because 0=false and not-0=true, but is not correct code.
3995 68. The host_aton() function was not handling scoped IPv6 addresses (those
3996 with, for example, "%eth0" on the end) correctly.
3998 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3999 negated items (that is, ~something) in unsigned ints. Some compilers
4000 apparently mutter when there is no cast.
4002 70. If an address verification called from an ACL failed, and did not produce a
4003 user-specific message (i.e. there was only a "system" message), nothing was
4004 put in $acl_verify_message. In this situation, it now puts the system
4007 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4008 session; change 4.31/43 added the unwanted input to the log line - except
4009 that it did not do this in the start of session case. It now does.
4011 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4012 This is wrong and can cause the other end to generate a synchronization
4013 error if it is another Exim or anything else that does the synchronization
4014 check. A QUIT command is no longer sent after a timeout.
4016 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4017 during host lookups.
4019 74. The defer_ok option of callout verification was not working if it was used
4020 when verifying addresses in header lines, that is, for this case:
4022 verify = header_sender/callout=defer_ok
4024 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4025 those file descriptors could be used for SMTP connections. If anything
4026 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4027 could be sent to the SMTP client, causing chaos. The daemon now opens
4028 stdin, stdout, and stderr to /dev/null when it puts itself into the
4031 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4032 log by default. The user can override this with suitable Perl magic.
4034 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4035 the log message when discard triggers, was not working for the DATA ACL or
4036 for the non-SMTP ACL.
4038 78. Error message wording change in sieve.c.
4040 79. If smtp_accept_max_per_host was set, the number of connections could be
4041 restricted to fewer than expected, because the daemon was trying to set up
4042 a new connection before checking whether the processes handling previous
4043 connections had finished. The check for completed processes is now done
4044 earlier. On busy systems, this bug wouldn't be noticed because something
4045 else would have woken the daemon, and it would have reaped the completed
4048 80. If a message was submitted locally by a user whose login name contained one
4049 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4050 It caused a spool format error. I have fixed the spool reading code. A
4051 related problem was that the "from" clause in the Received: line became
4052 illegal because of the space(s). It is now covered by ${quote_local_part.
4054 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4057 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4059 83. Patch from Sieve maintainer to fix unterminated string problem in
4060 "vacation" handling.
4062 84. Some minor changes to the Linux configuration files to help with other
4063 OS variants using glibc.
4065 85. One more patch for Sieve to update vacation handling to latest spec.
4068 ----------------------------------------------------
4069 See the note above about the 4.44 and 4.50 releases.
4070 ----------------------------------------------------
4076 1. Change 4.43/35 introduced a bug that caused file counts to be
4077 incorrectly computed when quota_filecount was set in an appendfile
4080 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4081 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4084 3. Give more explanation in the error message when the command for a transport
4085 filter fails to execute.
4087 4. There are several places where Exim runs a non-Exim command in a
4088 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4089 was being done only for the command run by the queryprogram router. It is
4090 now done for all such subprocesses. The other cases are: ${run, transport
4091 filters, and the commands run by the lmtp and pipe transports.
4093 5. Some older OS have a limit of 256 on the maximum number of file
4094 descriptors. Exim was using setrlimit() to set 1000 as a large value
4095 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4096 systems. I've change it so that if it can't get 1000, it tries for 256.
4098 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4099 was an oversight, and furthermore, ever since the addition of extra
4100 controls (e.g. 4.43/32), the checks on when to allow different forms of
4101 "control" were broken. There should now be diagnostics for all cases when a
4102 control that does not make sense is encountered.
4104 7. $recipients is now available in the predata ACL (oversight).
4106 8. Tidy the search cache before the fork to do a delivery from a message
4107 received from the command line. Otherwise the child will trigger a lookup
4108 failure and thereby defer the delivery if it tries to use (for example) a
4109 cached ldap connection that the parent has called unbind on.
4111 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4112 of $address_data from the recipient verification was clobbered by the
4113 sender verification.
4115 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4116 was its contents. (It was OK if the option was not defined at all.)
4118 11. A "Completed" log line is now written for messages that are removed from
4119 the spool by the -Mrm option.
4121 12. $host_address is now set to the target address during the checking of
4122 ignore_target_hosts.
4124 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4125 being passed; this would have caused $sender_host_name to have been used if
4126 matching the list had actually called for a host name (not very likely,
4127 since this list is usually IP addresses). A host name is now passed as
4130 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4131 code that specifies a non-restarting handler (typically sigaction() in
4132 modern systems) in an attempt to fix a rare and obscure crash bug.
4134 15. Narrowed the window for a race in the daemon that could cause it to ignore
4135 SIGCHLD signals. This is not a major problem, because they are used only to
4136 wake it up if nothing else does.
4138 16. A malformed maildirsize file could cause Exim to calculate negative values
4139 for the mailbox size or file count. Odd effects could occur as a result.
4140 The maildirsize information is now recalculated if the size or filecount
4143 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4144 support for a long time. Removed HAVE_SYS_VFS_H.
4146 18. Updated exipick to current release from John Jetmore.
4148 19. Allow an empty sender to be matched against a lookup in an address list.
4149 Previously the only cases considered were a regular expression, or an
4152 20. Exim went into a mad DNS lookup loop when doing a callout where the
4153 host was specified on the transport, if the DNS lookup yielded more than
4156 21. The RFC2047 encoding function was originally intended for short strings
4157 such as real names; it was not keeping to the 75-character limit for
4158 encoded words that the RFC imposes. It now respects the limit, and
4159 generates multiple encoded words if necessary. To be on the safe side, I
4160 have increased the buffer size for the ${rfc2047: expansion operator from
4163 22. Failure to deliver a bounce message always caused it to be frozen, even if
4164 there was an errors_to setting on the router. The errors_to setting is now
4167 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4168 canonical form (fully expanded) before being placed in
4169 $sender_host_address.
4171 24. Updated eximstats to version 1.33
4173 25. Include certificate and key file names in error message when GnuTLS fails
4174 to set them up, because the GnuTLS error message doesn't include the name
4175 of the failing file when there is a problem reading it.
4177 26. Expand error message when OpenSSL has problems setting up cert/key files.
4180 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4181 (this can affect the format of dates).
4183 28. exim_tidydb, when checking for the continued existence of a message for
4184 which it has found a message-specific retry record, was not finding
4185 messages that were in split spool directories. Consequently, it was
4186 deleting retry records that should have stayed in existence.
4188 29. eximstats updated to version 1.35
4189 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4190 1.35 - bugfix such that pie charts by volume are generated correctly
4192 30. The SPA authentication driver was not abandoning authentication and moving
4193 on to the next authenticator when an expansion was forced to fail,
4194 contradicting the general specification for all authenticators. Instead it
4195 was generating a temporary error. It now behaves as specified.
4197 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4198 (the order specifies the preference for clients). The order is now AES256,
4199 AES128, 3DES, ARCFOUR128.
4201 31. Small patch to Sieve code - explicitly set From: when generating an
4204 32. Exim crashed if a remote delivery caused a very long error message to be
4205 recorded - for instance if somebody sent an entire SpamAssassin report back
4206 as a large number of 550 error lines. This bug was coincidentally fixed by
4207 increasing the size of one of Exim's internal buffers (big_buffer) that
4208 happened as part of the Exiscan merge. However, to be on the safe side, I
4209 have made the code more robust (and fixed the comments that describe what
4212 33. Some experimental protocols are using DNS PTR records for new purposes. The
4213 keys for these records are domain names, not reversed IP addresses. The
4214 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4215 leaves it alone. Component reversal etc. now happens only for IP addresses.
4218 34. The host_aton() function is supposed to be passed a string that is known
4219 to be a valid IP address. However, in the case of IPv6 addresses, it was
4220 not checking this. This is a hostage to fortune. Exim now panics and dies
4221 if the condition is not met. A case was found where this could be provoked
4222 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4223 components; fortuitously, this particular loophole had already been fixed
4224 by change 4.50/55 or 4.44/33 above.
4226 If there are any other similar loopholes, the new check in host_aton()
4227 itself should stop them being exploited. The report I received stated that
4228 data on the command line could provoke the exploit when Exim was running as
4229 exim, but did not say which command line option was involved. All I could
4230 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4231 running as the user.
4234 35. There was a buffer overflow vulnerability in the SPA authentication code
4235 (which came originally from the Samba project). I have added a test to the
4236 spa_base64_to_bits() function which I hope fixes it.
4239 36. The daemon start-up calls getloadavg() while still root for those OS that
4240 need the first call to be done as root, but it missed one case: when
4241 deliver_queue_load_max is set with deliver_drop_privilege. This is
4242 necessary for the benefit of the queue runner, because there is no re-exec
4243 when deliver_drop_privilege is set.
4245 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4246 was in use, was not putting the data itself into the right store pool;
4247 consequently, it could be overwritten for a subsequent message in the same
4248 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4251 38. Sometimes the final signoff response after QUIT could fail to get
4252 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4253 < 0 before doing a fflush(). This bug looks as though it goes back to the
4254 introduction of TLS in release 3.20, but "sometimes" must have been rare
4255 because the tests only now provoked it.
4261 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4262 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4263 Now it returns an integer. A number of calls were still expecting a T/F
4264 return. Fortuitously, in all cases, the tests worked in OK situations,
4265 which is the norm. However, things would have gone wrong on any write
4266 failures on the smtp file descriptor. This function is used when sending
4267 messages over SMTP and also when doing verify callouts.
4269 2. When Exim is called to do synchronous delivery of a locally submitted
4270 message (the -odf or -odi options), it no longer closes stderr before doing
4273 3. Implemented the mua_wrapper option.
4275 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4277 5. Implemented the functions header_remove(), header_testname(),
4278 header_add_at_position(), and receive_remove_recipient(), and exported them
4281 6. If an ACL "warn" statement specified the addition of headers, Exim already
4282 inserted X-ACL-Warn: at the start if there was no header name. However, it
4283 was not making this test for the second and subsequent header lines if
4284 there were newlines in the string. This meant that an invalid header could
4285 be inserted if Exim was badly configured.
4287 7. Allow an ACL "warn" statement to add header lines at the start or after all
4288 the Received: headers, as well as at the end.
4290 8. Added the rcpt_4xx retry error code.
4292 9. Added postmaster_mailfrom=xxx to callout verification option.
4294 10. Added mailfrom=xxxx to the callout verification option, for verify=
4297 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4298 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4300 12. Inserted some casts to stop certain compilers warning when using pointer
4301 differences as field lengths or precisions in printf-type calls (mostly
4302 affecting debugging statements).
4304 13. Added optional readline() support for -be (dynamically loaded).
4306 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4307 same clock tick as a message's arrival, so that its received time was the
4308 same as the "first fail" time on the retry record, and that message
4309 remained on the queue past the ultimate address timeout, every queue runner
4310 would try a delivery (because it was past the ultimate address timeout) but
4311 after another failure, the ultimate address timeout, which should have then
4312 bounced the address, did not kick in. This was a "< instead of <=" error;
4313 in most cases the first failure would have been in the next clock tick
4314 after the received time, and all would be well.
4316 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4317 being recognized when the domain list was tested by the match_domain
4318 condition in an expansion string.
4320 16. Added the ${str2b64: operator.
4322 17. Exim was always calling setrlimit() to set a large limit for the number of
4323 processes, without checking whether the existing limit was already
4324 adequate. (It did check for the limit on file descriptors.) Furthermore,
4325 errors from getrlimit() and setrlimit() were being ignored. Now they are
4326 logged to the main and panic logs, but Exim does carry on, to try to do its
4327 job under whatever limits there are.
4329 18. Imported PCRE 5.0.
4331 19. Trivial typo in log message " temporarily refused connection" (the leading
4334 20. If the log selector return_path_on_delivery was set and an address was
4335 redirected to /dev/null, the delivery process crashed because it assumed
4336 that a return path would always be set for a "successful" delivery. In this
4337 case, the whole delivery is bypassed as an optimization, and therefore no
4340 21. Internal re-arrangement: the function for sending a challenge and reading
4341 a response while authentication was assuming a zero-terminated challenge
4342 string. It's now changed to take a pointer and a length, to allow for
4343 binary data in such strings.
4345 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4347 23. Exim was not respecting finduser_retries when seeking the login of the
4348 uid under which it was called; it was always trying 10 times. (The default
4349 setting of finduser_retries is zero.) Also, it was sleeping after the final
4350 failure, which is pointless.
4352 24. Implemented tls_on_connect_ports.
4354 25. Implemented acl_smtp_predata.
4356 26. If the domain in control=submission is set empty, Exim assumes that the
4357 authenticated id is a complete email address when it generates From: or
4358 Sender: header lines.
4360 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4361 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4362 chown and chgrp in /bin and hostname in /usr/bin.
4364 28. Exim was keeping the "process log" file open after each use, just as it
4365 does for the main log. This opens the possibility of it remaining open for
4366 long periods when the USR1 signal hits a daemon. Occasional processlog
4367 errors were reported, that could have been caused by this. Anyway, it seems
4368 much more sensible not to leave this file open at all, so that is what now
4371 29. The long-running daemon process does not normally write to the log once it
4372 has entered its main loop, and it closes the log before doing so. This is
4373 so that log files can straightforwardly be renamed and moved. However,
4374 there are a couple of unusual error situations where the daemon does write
4375 log entries, and I had neglected to close the log afterwards.
4377 30. The text of an SMTP error response that was received during a remote
4378 delivery was being truncated at 512 bytes. This is too short for some of
4379 the long messages that one sometimes sees. I've increased the limit to
4382 31. It is now possible to make retry rules that apply only when a message has a
4383 specific sender, in particular, an empty sender.
4385 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4386 it possible to be selective about when SMTP synchronization is enforced.
4388 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4390 32. Implemented hosts_connection_nolog.
4392 33. Added an ACL for QUIT.
4394 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4397 35. Added mailbox_size and mailbox_filecount to appendfile.
4399 36. Added control = no_multiline_responses to ACLs.
4401 37. There was a bug in the logic of the code that waits for the clock to tick
4402 in the case where the clock went backwards by a substantial amount such
4403 that the microsecond fraction of "now" was more than the microsecond
4404 fraction of "then" (but the whole seconds number was less).
4406 38. Added support for the libradius Radius client library this is found on
4407 FreeBSD (previously only the radiusclient library was supported).
4413 1. When certain lookups returned multiple values in the form name=value, the
4414 quoting of the values was not always being done properly. Specifically:
4415 (a) If the value started with a double quote, but contained no whitespace,
4417 (b) If the value contained whitespace other than a space character (i.e.
4418 tabs or newlines or carriage returns) it was not quoted.
4419 This fix has been applied to the mysql and pgsql lookups by writing a
4420 separate quoting function and calling it from the lookup code. The fix
4421 should probably also be applied to nisplus, ibase and oracle lookups, but
4422 since I cannot test any of those, I have not disturbed their existing code.
4424 2. A hit in the callout cache for a specific address caused a log line with no
4425 reason for rejecting RCPT. Now it says "Previous (cached) callout
4426 verification failure".
4428 3. There was an off-by-one bug in the queryprogram router. An over-long
4429 return line was truncated at 256 instead of 255 characters, thereby
4430 overflowing its buffer with the terminating zero. As well as fixing this, I
4431 have increased the buffer size to 1024 (and made a note to document this).
4433 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4434 when Exim is waiting for an SMTP response from a remote server, Exim
4435 restarts its select() call on the socket, thereby resetting its timeout.
4436 This is not a problem when such interrupts are rare. Somebody set up a cron
4437 job to run exiwhat every 2 minutes, which is less than the normal select()
4438 timeout (5 or 10 minutes). This meant that the select() timeout never
4439 kicked in because it was always reset. I have fixed this by comparing the
4440 time when an interrupt arrives with the time at the start of the first call
4441 to select(). If more time than the timeout has elapsed, the interrupt is
4442 treated as a timeout.
4444 5. Some internal re-factoring in preparation for the addition of Sieve
4445 extensions (by MH). In particular, the "personal" test is moved to a
4446 separate function, and given an option for scanning Cc: and Bcc: (which is
4447 not set for Exim filters).
4449 6. When Exim created an email address using the login of the caller as the
4450 local part (e.g. when creating a From: or Sender: header line), it was not
4451 quoting the local part when it contained special characters such as @.
4453 7. Installed new OpenBSD configuration files.
4455 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4456 try to make them clearer.
4458 9. Callout options, other than the timeout value, were being ignored when
4459 verifying sender addresses in header lines. For example, when using
4461 verify = header_sender/callout=no_cache
4463 the cache was (incorrectly) being used.
4465 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4466 only the Cygwin environment.
4468 11. When return_path_on_delivery was set as a log selector, if different remote
4469 addresses in the same message used different return paths and parallel
4470 remote delivery occurred, the wrong values would sometimes be logged.
4471 (Whenever a remote delivery process finished, the return path value from
4472 the most recently started remote delivery process was logged.)
4474 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4475 lines when AUTH and/or TLS are in use. This is the "received protocol"
4476 field. Exim used to use "asmtp" for authenticated SMTP, without any
4477 indication (in the protocol name) for TLS use. Now it follows the RFC and
4478 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4479 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4480 names appear in log lines as well as in Received: header lines.
4482 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4483 extensions, and comparison tests, and to fix some bugs.
4485 14. Changes to the "personal" filter test:
4487 (1) The test was buggy in that it was just doing the equivalent of
4488 "contains" tests on header lines. For example, if a user's address was
4489 anne@some.where, the "personal" test would incorrectly be true for
4491 To: susanne@some.where
4493 This test is now done by extracting each address from the header in turn,
4494 and checking the entire address. Other tests that are part of "personal"
4495 are now done using regular expressions (for example, to check local parts
4496 of addresses in From: header lines).
4498 (2) The list of non-personal local parts in From: addresses has been
4499 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4500 taken from the Sieve specification recommendations.
4502 (3) If the message contains any header line starting with "List-" it is
4503 treated as non-personal.
4505 (4) The test for "circular" in the Subject: header line has been removed
4506 because it now seems ill-conceived.
4508 15. Minor typos in src/EDITME comments corrected.
4510 16. Installed latest exipick from John Jetmore.
4512 17. If headers_add on a router specified a text string that was too long for
4513 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4514 of string_sprintf() is now avoided.
4516 18. $message_body_size was not set (it was always zero) when running the DATA
4517 ACL and the local_scan() function.
4519 19. For the "mail" command in an Exim filter, no default was being set for
4520 the once_repeat time, causing a random time value to be used if "once" was
4521 specified. (If the value happened to be <= 0, no repeat happened.) The
4522 default is now 0s, meaning "never repeat". The "vacation" command was OK
4523 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4524 (I found it when inspecting the code).
4526 20. There is now an overall timeout for performing a callout verification. It
4527 defaults to 4 times the callout timeout, which applies to individual SMTP
4528 commands during the callout. The overall timeout applies when there is more
4529 than one host that can be tried. The timeout is checked before trying the
4530 next host. This prevents very long delays if there are a large number of
4531 hosts and all are timing out (e.g. when the network connections are timing
4532 out). The value of the overall timeout can be changed by specifying an
4533 additional sub-option for "callout", called "maxwait". For example:
4535 verify = sender/callout=5s,maxwait=20s
4537 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4538 to the end before writing, but this should make it even safer).
4540 22. Exim was forgetting that it had advertised PIPELINING for the second and
4541 subsequent messages on an SMTP connection. It was also not resetting its
4542 memory on STARTTLS and an internal HELO.
4544 23. When Exim logs an SMTP synchronization error within a session, it now
4545 records whether PIPELINING has been advertised or not.
4547 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4548 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4549 rather than long int.
4551 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4553 26. Added the never_mail option to autoreply.
4559 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4560 crash if the getsockname() call failed; this can happen if a connection is
4561 closed very soon after it is established. The problem was simply in the
4562 order in which certain operations were done, causing Exim to try to write
4563 to the SMTP stream before it had set up the file descriptor. The bug has
4564 been fixed by making things happen in the correct order.
4570 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4571 before the connection was closed, thus losing the rejection response.
4573 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4574 some early Solaris releases, but causes trouble in current releases where
4575 socklen_t is defined.
4577 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4580 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4583 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4584 that releasing the top of it at the end releases what was used for sub-
4585 expansions (unless the block got too big). However, discard this block if
4586 the first thing is a variable or header, so that we can use its block when
4587 it is dynamic (useful for very large $message_headers, for example).
4589 6. Lookups now cache *every* query, not just the most recent. A new, separate
4590 store pool is used for this. It can be recovered when all lookup caches are
4591 flushed. Lookups now release memory at the end of their result strings.
4592 This has involved some general refactoring of the lookup sources.
4594 7. Some code has been added to the store_xxx() functions to reduce the amount
4595 of flapping under certain conditions.
4597 8. log_incoming_interface used to affect only the <= reception log lines. Now
4598 it causes the local interface and port to be added to several more SMTP log
4599 lines, for example "SMTP connection from", and rejection lines.
4601 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4603 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4605 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4606 could be overwritten at the end of the current message (or the start of a
4607 new message if it was set in a HELO ACL). The value is now preserved for
4608 the duration of the SMTP connection.
4610 12. If a transport had a headers_rewrite setting, and a matching header line
4611 contained an unqualified address, that address was qualified, even if it
4612 did not match any rewriting rules. The underlying bug was that the values
4613 of the flags that permit the existence of unqualified sender and recipient
4614 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4615 non-local messages, and by -bnq for local messages) were not being
4616 preserved with the message after it was received.
4618 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4619 "next input=" as part of the text comprising the host identity instead of
4620 the correct text. The code was using the same buffer for two different
4621 strings. However, depending on which order the printing function evaluated
4622 its arguments, the bug did not always show up. Under Linux, for example, my
4623 test suite worked just fine.
4625 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4626 doesn't work with some older versions of Perl. It has been changed to "my",
4627 which in any case is probably the better facility to use.
4629 15. A really picky compiler found some instances of statements for creating
4630 error messages that either had too many or two few arguments for the format
4633 16. The size of the buffer for calls to the DNS resolver has been increased
4634 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4635 for addresses that have a lot of PTR records. This alleviates a problem; it
4636 does not fully solve it.
4638 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4639 buffer now truncates the list and logs the incident, which is the same
4640 action as happens when Exim is looking up a host name and its aliases.
4641 Previously in this situation something unpredictable would happen;
4642 sometimes it was "internal error: store_reset failed".
4644 18. If a server dropped the connection unexpectedly when an Exim client was
4645 using GnuTLS and trying to read a response, the client delivery process
4646 crashed while trying to generate an error log message.
4648 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4649 string, for example:
4651 warn message = H1: something\nH2: something
4653 the text was added as a single header line from Exim's point of view
4654 though it ended up OK in the delivered message. However, searching for the
4655 second and subsequent header lines using $h_h2: did not work. This has been
4656 fixed. Similarly, if a system filter added multiple headers in this way,
4657 the routers could not see them.
4659 20. Expanded the error message when iplsearch is called with an invalid key to
4660 suggest using net-iplsearch in a host list.
4662 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4663 ACLs are no longer actually imposed (and a message to that effect is
4666 22. If a "gecos" field in a passwd entry contained escaped characters, in
4667 particular, if it contained a \" sequence, Exim got it wrong when building
4668 a From: or a Sender: header from that name. A second bug also caused
4669 incorrect handling when an unquoted " was present following a character
4670 that needed quoting.
4672 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4673 was not being matched caselessly.
4675 24. Arranged for all hyphens in the exim.8 source to be escaped with
4678 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4679 properly. Recipient callout cache records were still being keyed to include
4680 the sender, even when use_sender was set false. This led to far more
4681 callouts that were necessary. The sender is no longer included in the key
4682 when use_sender is false.
4684 26. Added "control = submission" modifier to ACLs.
4686 27. Added the ${base62d: operator to decode base 62 numbers.
4688 28. dnsdb lookups can now access SRV records.
4690 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4691 the configuration file.
4693 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4694 (-v) mode. This makes the output for a verbose queue run more intelligible.
4696 31. Added a use_postmaster feature to recipient callouts.
4698 32. Added the $body_zerocount variable, containing the number of binary zero
4699 bytes in the message body.
4701 33. The time of last modification of the "new" subdirectory is now used as the
4702 "mailbox time last read" when there is a quota error for a maildir
4705 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4707 35. Added +ignore_unknown as a special item in host lists.
4709 36. Code for decoding IPv6 addresses in host lists is now included, even if
4710 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4711 address was recognized as an IP address, but was then not correctly decoded
4712 into binary, causing unexpected and incorrect effects when compared with
4719 1. Very minor rewording of debugging text in manualroute to say "list of
4720 hosts" instead of "hostlist".
4722 2. If verify=header_syntax was set, and a header line with an unqualified
4723 address (no domain) and a large number of spaces between the end of the
4724 name and the colon was received, the reception process suffered a buffer
4725 overflow, and (when I tested it) crashed. This was caused by some obsolete
4726 code that should have been removed. The fix is to remove it!
4728 3. When running in the test harness, delay a bit after writing a bounce
4729 message to get a bit more predictability in the log output.
4731 4. Added a call to search_tidyup() just before forking a reception process. In
4732 theory, someone could use a lookup in the expansion of smtp_accept_max_
4733 per_host which, without the tidyup, could leave open a database connection.
4735 5. Added the variables $recipient_data and $sender_data which get set from a
4736 lookup success in an ACL "recipients" or "senders" condition, or a router
4737 "senders" option, similar to $domain_data and $local_part_data.
4739 6. Moved the writing of debug_print from before to after the "senders" test
4742 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4743 problems for message scanning, either using a data ACL, or using
4744 local_scan() because the Received: header was not generated till after they
4745 were called (in order to set the time as the time of reception completion).
4746 I have revised the way this works. The header is now generated after the
4747 body is received, but before the ACL or local_scan() are called. After they
4748 are run, the timestamp in the header is updated.
4754 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4755 before starting a queue runner without re-exec. This happened only when
4756 deliver_drop_privilege was set or when the Exim user was set to root. The
4757 effect of the bug was that timeouts during subsequent deliveries caused
4758 crashes instead of being properly handled. The handler is now left at its
4759 default (and expected) setting.
4761 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4762 message, again when deliver_drop_privilege is set or Exim is run as root.
4763 The bug described in (1) was not present in this case, but the tidying up
4764 of the other signals was missing. I have made the two cases consistent.
4766 3. The ignore_target_hosts setting on a manualroute router was being ignored
4767 for hosts that were looked up using the /MX notation.
4769 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4772 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4773 operated on the sender address. After changing the $sender_address to <>
4774 for the sender address verify, Exim was re-instated it as the original
4775 (before rewriting) address, but remembering that it had rewritten it, so it
4776 wasn't rewriting it again. This bug also had the effect of breaking the
4777 sender address verification caching when the sender address was rewritten.
4779 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4780 This has been changed so that if the ip literal address matches
4781 ignore_target_hosts, the router declines.
4783 7. Added expansion conditions match_domain, match_address, and match_local_
4784 part (NOT match_host).
4786 8. The placeholder for the Received: header didn't have a length field set.
4788 9. Added code to Exim itself and to exim_lock to test for a specific race
4789 condition that could lead to file corruption when using MBX delivery. The
4790 issue is with the lockfile that is created in /tmp. If this file is removed
4791 after a process has opened it but before that process has acquired a lock,
4792 there is the potential for a second process to recreate the file and also
4793 acquire a lock. This could lead to two Exim processes writing to the file
4794 at the same time. The added code performs the same test as UW imapd; it
4795 checks after acquiring the lock that its file descriptor still refers to
4796 the same named file.
4798 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4799 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4802 11. Added the smtp_active_hostname option. If used, this will typically be made
4803 to depend on the incoming interface address. Because $interface_address is
4804 not set up until the daemon has forked a reception process, error responses
4805 that can happen earlier (such as "too many connections") no longer contain
4808 12. If an expansion in a condition on a "warn" statement fails because a lookup
4809 defers, the "warn" statement is abandoned, and the next ACL statement is
4810 processed. Previously this caused the whole ACL to be aborted.
4812 13. Added the iplsearch lookup type.
4814 14. Added ident_timeout as a log selector.
4816 15. Added tls_certificate_verified as a log selector.
4818 16. Added a global option tls_require_ciphers (compare the smtp transport
4819 option of the same name). This controls incoming TLS connections.
4821 17. I finally figured out how to make tls_require_ciphers do a similar thing
4822 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4823 before starting the TLS session.
4825 18. Tabs are now shown as \t in -bP output.
4827 19. If the log selector return_path_on_delivery was set, Exim crashed when
4828 bouncing a message because it had too many Received: header lines.
4830 20. If two routers both had headers_remove settings, and the first one included
4831 a superfluous trailing colon, the final name in the first list and the
4832 first name in the second list were incorrectly joined into one item (with a
4833 colon in the middle).
4839 1. Added -C and -D options to the exinext utility, mainly to make it easier
4840 to include in the automated testing, but these could be helpful when
4841 multiple configurations are in use.
4843 2. The exinext utility was not formatting the output nicely when there was
4844 an alternate port involved in the retry record key, nor when there was a
4845 message id as well (for retries that were specific to a specific message
4846 and a specific host). It was also confused by IPv6 addresses, because of
4847 the additional colons they contain. I have fixed the IPv4 problem, and
4848 patched it up to do a reasonable job for IPv6.
4850 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4851 delivery, the log line now contains "pipelined" if PIPELINING was used.
4853 4. An SMTP transport process used to panic and die if the bind() call to set
4854 an explicit outgoing interface failed. This has been changed; it is now
4855 treated in the same way as a connect() failure.
4857 5. A reference to $sender_host_name in the part of a conditional expansion
4858 that was being skipped was still causing a DNS lookup. This no longer
4861 6. The def: expansion condition was not recognizing references to header lines
4862 that used bh_ and bheader_.
4864 7. Added the _cache feature to named lists.
4866 8. The code for checking quota_filecount in the appendfile transport was
4867 allowing one more file than it should have been.
4869 9. For compatibility with Sendmail, the command line option
4877 and sets the incoming protocol and host name (for trusted callers). The
4878 host name and its colon can be omitted when only the protocol is to be set.
4879 Note the Exim already has two private options, -pd and -ps, that refer to
4880 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4881 "s", but I don't think that's a major issue.
4883 10. A number of refactoring changes to the code, none of which should affect
4886 (a) The number of logging options was getting close to filling up the
4887 32-bit word that was used as a bit map. I have split them into two classes:
4888 those that are passed in the argument to log_write(), and those that are
4889 only ever tested independently outside of that function. These are now in
4890 separate 32-bit words, so there is plenty of room for expansion again.
4891 There is no change in the user interface or the logging behaviour.
4893 (b) When building, for example, log lines, the code previously used a
4894 macro that called string_cat() twice, in order to add two strings. This is
4895 not really sufficiently general. Furthermore, there was one instance where
4896 it was actually wrong because one of the argument was used twice, and in
4897 one call a function was used. (As it happened, calling the function twice
4898 did not affect the overall behaviour.) The macro has been replaced by a
4899 function that can join an arbitrary number of extra strings onto a growing
4902 (c) The code for expansion conditions now uses a table and a binary chop
4903 instead of a serial search (which was left over from when there were very
4904 few conditions). Also, it now recognizes conditions like "pam" even when
4905 the relevant support is not compiled in: a suitably worded error message is
4906 given if an attempt is made to use such a condition.
4908 11. Added ${time_interval:xxxxx}.
4910 12. A bug was causing one of the ddress fields not to be passed back correctly
4911 from remote delivery subprocesses. The field in question was not being
4912 subsequently used, so this caused to problems in practice.
4914 13. Added new log selectors queue_time and deliver_time.
4916 14. Might have fixed a bug in maildirsizefile handling that threw up
4917 "unexpected character" debug warnings, and recalculated the data
4918 unnecessarily. In any case, I expanded the warning message to give more
4921 15. Added the message "Restricted characters in address" to the statements in
4922 the default ACL that block characters like @ and % in local parts.
4924 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4925 Three changes have been made:
4927 (a) There was a serious bug; a negative response to MAIL caused the whole
4928 recipient domain to be cached as invalid, thereby blocking all messages
4929 to all local parts at the same domain, from all senders. This bug has
4930 been fixed. The domain is no longer cached after a negative response to
4931 MAIL if the sender used is not empty.
4933 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4936 (c) A new callout option, "use_sender" has been added for people who want
4937 the modified behaviour.
4943 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4946 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4947 indeed breaks things for older releases.
4949 3. Added additional logging to the case where there is a problem reading data
4950 from a filter that is running in a subprocess using a pipe, in order to
4951 try to track down a specific problem.
4953 4. Testing facility fudge: when running in the test harness and attempting
4954 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4955 getting "No route to host". Convert this to a timeout.
4957 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4960 6. Some OS don't have socklen_t but use size_t instead. This affects the
4961 fifth argument of getsockopt() amongst other things. This is now
4962 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4963 can be set for individual OS. I have set it for SunOS5, OSF1, and
4964 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4965 some earlier ones do not.
4967 7. Change 4.30/15 was not doing the test caselessly.
4969 8. The standard form for an IPv6 address literal was being rejected by address
4970 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4971 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4972 this, as well as the form without the "IPv6" on the front (but only when
4973 address literals are enabled, of course).
4975 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4977 10. Exim crashed if a message with an empty sender address specified by -f
4978 encountered a router with an errors_to setting. This could be provoked only
4979 by a command such as
4983 where an empty string was supplied; "<>" did not hit this bug.
4985 11. Installed PCRE release 4.5.
4987 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4988 remained set. It is now erased.
4990 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4991 times from message ids (which are base 36 rather than the normal 62).
4993 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4994 were being counted as actual protocol errors, and logged if the log
4995 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4996 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4997 and DATA following a set of rejected RCPTs do not count as protocol errors.
4998 In other words, Exim assumes they were pipelined, though this may not
4999 actually be the case. Of course, in all cases the client gets an
5000 appropriate error code.
5002 15. If a lookup fails in an ACL condition, a message about the failure may
5003 be available; it is used if testing the ACL cannot continue, because most
5004 such messages specify what the cause of the deferral is. However, some
5005 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5006 that caused an old message to be retained and used if a later statement
5007 caused a defer, replacing the real cause of the deferral.
5009 16. If an IP address had so many PTR records that the DNS lookup buffer
5010 was not large enough to hold them, Exim could crash while trying to process
5011 the truncated data. It now detects and logs this case.
5013 17. Further to 4.21/58, another change has been made: if (and only if) the
5014 first line of a message (the first header line) ends with CRLF, a bare LF
5015 in a subsequent header line has a space inserted after it, so as not to
5016 terminate the header.
5018 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5019 unnecessarily, used atoi() instead of strtol(), and didn't check the
5020 termination when getting file sizes from file names by regex.
5022 19. Completely re-implemented the support for maildirsize files, in the light
5023 of a number of problems with the previous contributed implementation
5024 (4.30/29). In particular:
5026 . If the quota is zero, the maildirsize file is maintained, but no quota is
5029 . If the maildir directory does not exist, it is created before any attempt
5030 to write a maildirsize file.
5032 . The quota value in the file is just a cache; if the quota is changed in
5033 the transport, the new value overrides.
5035 . A regular expression is available for excluding directories from the
5038 20. The autoreply transport checks the characters in options that define the
5039 message's headers; it allows continued headers, but it was checking with
5040 isspace() after an embedded newline instead of explicitly looking for a
5043 21. If all the "regular" hosts to which an address was routed had passed their
5044 expiry times, and had not reached their retry times, the address was
5045 bounced, even if fallback hosts were defined. Now Exim should go on to try
5048 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5049 equivalent code in the SMTP transport. Some hosts send humungous responses
5050 to HELO/EHLO, more than 1024 it seems.
5052 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5053 gives compiler warnings in some environments. I've now done it "properly",
5056 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5057 (because of problems with the built-in one) was declared to return uschar *
5058 instead of char *, causing compiler failure.
5060 25. Fixed a file descriptor leak when processing alias/forward files.
5062 26. Fixed a minor format string issue in dbfn.c.
5064 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5066 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5067 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5070 29. When a sender address is verified, it is cached, to save repeating the test
5071 when there is more than one recipient in a message. However, when the
5072 verification involves a callout, it is possible for different callout
5073 options to be set for different recipients. It is too complicated to keep
5074 track of this in the cache, so now Exim always runs a verification when a
5075 callout is required, relying on the callout cache for the optimization.
5076 The overhead is duplication of the address routing, but this should not be
5079 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5080 to be verified with callout=postmaster, and the main callout worked but the
5081 postmaster check failed, the verification correctly failed. However, if a
5082 subsequent RCPT command asked for sender verification *without* the
5083 postmaster check, incorrect caching caused this verification also to fail,
5086 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5087 it was not caching the DNS options (qualify_single, search_parents) that
5088 were used when the lookup failed. A subsequent lookup with different
5089 options therefore always gave the same answer, though there were cases
5090 where it should not have. (Example: a "domains = !$mx_any" option on a
5091 dnslookup router: the "domains" option is always processed without any
5092 widening, but the router might have qualify_single set.) Now Exim uses the
5093 cached value only when the same options are set.
5095 32. Added John Jetmore's "exipick" utility to the distribution.
5097 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5098 than a timeout (e.g. a certificate is required, and is not provided), an
5099 Exim server now closes the connection immediately. Previously it waited for
5100 the client to close - but if the client is SSL, it seems that they each
5101 wait for each other, leading to a delay before one of them times out.
5103 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5104 maintained 0.8.x compatibility because I don't think many are using it, and
5105 it is clearly obsolete.
5107 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5110 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5111 client certificate was expired. A simple patch fixes this, though I don't
5112 understand the full logic of why the verify callback is called multiple
5115 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5116 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5117 which causes problems with some clients (such as the Certicom SSL Plus
5118 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5119 disables the coutermeasure allowing Eudora to connect."
5121 38. Exim was not checking that a write() to a log file succeeded. This could
5122 lead to Bad Things if a log got too big, in particular if it hit a file
5123 size limit. Exim now panics and dies if it cannot write to a log file, just
5124 as it does if it cannot open a log file.
5126 39. Modified OS/Makefile-Linux so that it now contains
5128 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5130 The two -D definitions ensure that Exim is compiled with large file
5131 support, which makes it possible to handle log files that are bigger than
5134 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5135 instance) a domain was checked against a named list that involved a lookup,
5136 causing $domain_data to be set, then another domain was checked against the
5137 same list, then the first domain was re-checked, the value of $domain_data
5138 after the final check could be wrong. In particular, if the second check
5139 failed, it could be set empty. This bug probably also applied to
5142 41. The strip_trailing_dot option was not being applied to the address given
5143 with the -f command-line option.
5145 42. The code for reading a message's header from the spool was incrementing
5146 $received_count, but never initializing it. This meant that the value was
5147 incorrect (doubled) while delivering a message in the same process in which
5148 it was received. In the most common configuration of Exim, this never
5149 happens - a fresh exec is done - but it can happen when
5150 deliver_drop_privilege is set.
5152 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5153 it now includes up to 150 characters of the unexpected data in the log
5156 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5157 and building data strings. The size of both of these buffers was 10 000
5158 bytes - far larger than anybody would *ever* want, thought I. Needless to
5159 say, somebody hit the limit. I have increased the maximum line length to
5160 20 000 and the maximum data length of concatenated lines to 100 000. I have
5161 also fixed two bugs, because there was no checking on these buffers. Tsk,
5162 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5163 buffer is too small.
5165 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5166 lsearch lookups. Now it does.
5168 46. When parsing a route_list item in a manualroute router, a fixed-length
5169 buffer was used for the list of hosts. I made this 1024 bytes long,
5170 thinking that nobody would ever have a list of hosts that long. Wrong.
5171 Somebody had a whole pile of complicated expansion conditions, and the
5172 string was silently truncated, leading to an expansion error. It turns out
5173 that it is easier to change to an unlimited length (owing to other changes
5174 that have happened since this code was originally written) than to build
5175 structure for giving a limitation error. The length of the item that
5176 expands into the list of hosts is now unlimited.
5178 47. The lsearch lookup could not handle data where the length of text line was
5179 more than 4095 characters. Such lines were truncated, leading to shortened
5180 data being returned. It should now handle lines of any length.
5182 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5183 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5186 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5187 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5188 no longer happens in comment lines. A list of replacements is now placed
5189 at the head of all of the source files, except those whose only change is
5190 to replace PERL_COMMAND in the very first #! line.
5192 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5193 messages on the queue, with a bottom-up merge sort, using code contributed
5194 by Michael Haardt. This should make operations like -bp somewhat faster on
5195 large queues. It won't affect queue runners, except when queue_run_in_order
5198 51. Installed eximstats 1.31 in the distribution.
5200 52. Added support for SRV lookups to the dnslookup router.
5202 53. If an ACL referred to $message_body or $message_body_end, the value was not
5203 reset for any messages that followed in the same SMTP session.
5205 54. The store-handling optimization for building very long strings was not
5206 differentiating between the different store pools. I don't think this
5207 actually made any difference in practice, but I've tidied it.
5209 55. While running the routers to verify a sender address, $sender_address
5210 was still set to the sender address. This is wrong, because when routing to
5211 send a bounce to the sender, it would be empty. Therefore, I have changed
5212 it so that, while verifying a sender address, $sender_address is set to <>.
5213 (There is no change to what happens when verifying a recipient address.)
5215 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5216 target A or AAAA records (if not already returned) without resetting the
5217 qualify_single or search_parents options of the DNS resolver. These are
5218 inappropriate in this case because the targets of MX and SRV records must
5219 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5220 target that, when qualified, matched something in the local domain. These
5221 two options are now turned off when doing these lookups.
5223 57. It seems that at least some releases of Reiserfs (which does not have the
5224 concept of a fixed number of inodes) returns zero and not -1 for the
5225 number of available inodes. This interacted badly with check_spool_inodes,
5226 which assumed that -1 was the "no such thing" setting. What I have done is
5227 to check that the total number of inodes is greater than zero before doing
5228 the test of how many are available.
5230 58. When a "warn" ACL statement has a log_message modifier, the message is
5231 remembered, and not repeated. This is to avoid a lot of repetition when a
5232 message has many recipients that cause the same warning to be written.
5233 Howewer, Exim was preserving the list of already written lines for an
5234 entire SMTP session, which doesn't seem right. The memory is now reset if a
5235 new message is started.
5237 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5238 debugging output unless log_rewrite was also set.
5240 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5241 of (int)(handle) when we know that handle contains (void *)(-1).
5243 61. The Exim daemon panic-logs an error return when it closes the incoming
5244 connection. However "connection reset by peer" seems to be common, and
5245 isn't really an error worthy of noting specially, so that particular error
5248 62. When Exim is trying to find all the local interfaces, it used to panic and
5249 die if the ioctl to get the interface flags failed. However, it seems that
5250 on at least one OS (Solaris 9) it is possible to have an interface that is
5251 included in the list of interfaces, but for which you get a failure error
5252 for this call. This happens when the interface is not "plumbed" into a
5253 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5254 failure of the "get flags" call assumes that the interface is down.
5256 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5257 makes life easier for people who are doing arithmetic on fields extracted
5258 from dates, where you often get leading zeros that should not be
5259 interpreted as octal.
5261 64. Added qualify_domain to the redirect router, to override the global
5264 65. If a pathologically long header line contained very many addresses (the
5265 report of this problem mentioned 10 000) and each of them was rewritten,
5266 Exim could use up a very large amount of memory. (It kept on making new
5267 copies of the header line as it rewrote, and never released the old ones.)
5268 At the expense of a bit more processing, the header rewriting function has
5269 been changed so that it no longer eats memory in this way.
5271 66. The generation of the Received: header has been moved from the time that a
5272 message starts to be received, to the time that it finishes. The timestamp
5273 in the Received: header should now be very close to that of the <= log
5274 line. There are two side-effects of this change:
5276 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5277 logged header lines no longer include the local Received: line, because
5278 it has not yet been created. The same applies to a copy of the message
5279 that is returned to a non-SMTP sender when a message is rejected.
5281 (b) When a filter file is tested using -bf, no additional Received: header
5282 is added to the test message. After some thought, I decided that this
5285 This change does not affect the value of $received_for. It is still set
5286 after address rewriting, but before local_scan() is called.
5288 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5290 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5291 gave an unhelpful panic error message, and a defer error. I have managed to
5292 change this behaviour so that it now rejects any supplied certificate,
5293 which seems right, as the list of acceptable certificates is empty.
5295 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5296 gave an unhelpful defer error. I have not managed to make this reject any
5297 supplied certificates, but the error message it gives is "no certificate
5298 supplied", which is not helpful.
5300 70. exigrep's output now also includes lines that are not associated with any
5301 message, but which match the given pattern. Implemented by a patch from
5302 Martin Sluka, which also tidied up the Perl a bit.
5304 71. Recipient callout verification, like sender verification, was using <> in
5305 the MAIL FROM command. This isn't really the right thing, since the actual
5306 sender may affect whether the remote host accepts the recipient or not. I
5307 have changed it to use the actual sender in the callout; this means that
5308 the cache record is now keyed on a recipient/sender pair, not just the
5309 recipient address. There doesn't seem to be a real danger of callout loops,
5310 since a callout by the remote host to check the sender would use <>.
5311 [SEE ABOVE: changed after hitting problems.]
5313 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5314 temporary errors. However, in the case of such a code being given after
5315 the end of a data transmission (i.e. after ".") Exim was failing to write
5316 a retry record for the message. (Yes, there was some broken host that was
5317 actually sending 8xx at this point.)
5319 73. An unknown lookup type in a host list could cause Exim to panic-die when
5320 the list was checked. (An example that provoked this was putting <; in the
5321 middle of a list instead of at the start.) If this happened during a DATA
5322 ACL check, a -D file could be left lying around. This kind of configuration
5323 error no longer causes Exim to die; instead it causes a defer errror. The
5324 incident is still logged to the main and panic logs.
5326 74. Buglet left over from Exim 3 conversion. The message "too many messages
5327 in one connection" was written to the rejectlog but not the mainlog, except
5328 when address rewriting (yes!) was being logged.
5330 75. Added write_rejectlog option.
5332 76. When a system filter was run not as root (that is, when system_filter_user
5333 was set), the values of the $n variables were not being returned to the
5334 main process; thus, they were not subsequently available in the $sn
5337 77. Added +return_path_on_delivery log selector.
5339 78. A connection timeout was being treated differently from recipients deferred
5340 when testing hosts_max_try with a message that was older than the host's
5341 retry timeout. (The host should not be counted, thus allowing all hosts to
5342 be tried at least once before bouncing.) This may have been the cause of an
5343 occasionally reported bug whereby a message would remain on the queue
5344 longer than the retry timeout, but would be bounced if a delivery was
5345 forced. I say "may" because I never totally pinned down the problem;
5346 setting up timeout/retry tests is difficult. See also the next item.
5348 79. The ultimate address timeout was not being applied to errors that involved
5349 a combination of host plus message (for example, a timeout on a MAIL
5350 command). When an address resolved to a number of possible hosts, and they
5351 were not all tried for each delivery (e.g. because of hosts_max_try), a
5352 message could remain on the queue longer than the retry timeout.
5354 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5357 81. Fixed an obscure SMTP outgoing bug which required at least the following
5358 conditions: (a) there was another message waiting for the same server;
5359 (b) the server returned 5xx to all RCPT commands in the first message so
5360 that the message was not completed; (c) the server dropped the connection
5361 or gave a negative response to the RSET that Exim sends to abort the
5362 transaction. The observed case was a dropped connection after DATA that had
5363 been sent in pipelining mode. That is, the server had advertised PIPELINING
5364 but was not implementing it correctly. The effect of the bug was incorrect
5365 behaviour, such as trying another host, and this could lead to a crash.
5371 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5372 and daemon.c were passed as pointers to ints; they should have been
5373 pointers to socklen_t variables (which are typically unsigned ints).
5375 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5378 3. Fixed a really odd bug that affected only the testing scheme; patching a
5379 certain fixed string in the binary changed the value of another string that
5380 happened to be identical to the end of the original first string.
5382 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5383 name", it returns that address as the IP address. On some operating
5384 systems (e.g. Solaris), it also passes back the IP address string as the
5385 "host name". However, on others (e.g. Linux), it passes back an empty
5386 string. Exim wasn't checking for this, and was changing the host name to an
5387 empty string, assuming it had been canonicized.
5389 5. Although rare, it is permitted to have more than one PTR record for a given
5390 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5391 all the names associated with an address, because they do in Solaris.
5392 However, it seems that they do not in Linux for data that comes from the
5393 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5394 I found this out when I moved to a new Linux workstation and tried to run
5395 the Exim test suite.
5397 To get round this problem I have changed the code so that it now does its
5398 own call to the DNS to look up PTR records when searching for a host name.
5399 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5400 addresses that are only in /etc/hosts are still found.
5402 This behaviour is, however, controlled by an option called host_lookup_
5403 order, which defaults to "bydns:byaddr". If people want to use the other
5404 order, or indeed, just use one or the other means of lookup, they can
5405 specify it in this variable.
5407 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5408 some operating systems, this comes back from gethostbyaddr() as an empty
5409 string, and this is what Exim used to test for. However, it seems that in
5410 other systems, "." is yielded. Exim now tests for this case too.
5412 7. The values of check_spool_space and check_log_space are now held internally
5413 as a number of kilobytes instead of an absolute number of bytes. If a
5414 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5415 kilobyte. This means that much larger values can be stored.
5417 8. Exim monitor: an attempt to get the action menu when not actually pointing
5418 at a message produces an empty menu entitled "No message selected". This
5419 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5420 no entries in it ("Shell widget menu has zero width and/or height"). So I
5421 have added a single, blank menu entry in this case.
5423 9. Added ${quote_local_part.
5425 10. MIME decoding is now applied to the contents of Subject: header lines when
5428 11. Now that a reference to $sender_host_address automatically causes a reverse
5429 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5430 host lookup before query-style lookups in lists that might use this
5431 variable. This has therefore been abolished, and the "net-" prefix is no
5432 longer necessary for query-style lookups.
5434 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5435 have been a typo for LFLAGS, so it has been changed.
5437 13. The install script calls Exim with "-C /dev/null" in order to find the
5438 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5439 to be output. Howeve, since Exim outputs its version number before the
5440 error, it didn't break the script. It just looked ugly. I fixed this by
5441 always allowing "-C /dev/null" if the caller is root.
5443 14. Ignore overlarge ACL variable number when reading spool file - insurance
5444 against a later release with more variables having written the file.
5446 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5447 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5448 this, as well as the form without the "IPv6" on the front.
5450 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5451 OS/Makefile-Darwin file.
5453 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5454 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5456 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5457 variable. However, some OS use ints for inodes. I've added cast to long int
5458 to get rid of the compiler warning.
5460 19. I had forgotten to lock out "/../" in configuration file names when
5461 ALT_CONFIG_PREFIX was set.
5463 20. Routers used for verification do not need to specify transports. However,
5464 if such a router generated a host list, and callout was configured, Exim
5465 crashed, because it could not find a port number from the (non-existent)
5466 transport. It now assumes port 25 in this circumstance.
5468 21. Added the -t option to exigrep.
5470 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5471 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5472 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5473 them is set, LOOKUP_LSEARCH is forced.
5475 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5477 24. Added sender and host information to the "rejected by local_scan()" log
5478 line; previously there was no indication of these.
5480 25. Added .include_if_exists.
5482 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5483 Linux. It turns out that not all file systems support this. Apparently some
5484 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5485 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5486 not supported on the file descriptor, is now ignored when Exim is trying to
5487 sync a directory. This applies only to Linux.
5489 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5491 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5492 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5495 29. Added support for maildirsize files from supplied patch (modified a bit).
5497 30. The use of :fail: followed by an empty string could lead Exim to respond to
5498 sender verification failures with (e.g.):
5500 550 Verification failed for <xxx>
5501 550 Sender verify failed
5503 where the first response line was missing the '-' that indicates it is not
5504 the final line of the response.
5506 31. The loop for finding the name of the user that called Exim had a hardwired
5507 limit of 10; it now uses the value of finduser_retries, which is used for
5508 all other user lookups.
5510 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5513 33. Exim was neglecting to zero errno before one call of strtol() when
5514 expanding a string and expecting an integer value. On some systems this
5515 resulted in spurious "integer overflow" errors. Also, it was casting the
5516 result into an int without checking.
5518 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5519 did not work. The code looks as if it has *never* worked, though it appears
5520 to have been documented since at least releast 1.62. I have made it work.
5522 35. The "timeout_DNS" error in retry rules, also documented since at least
5523 1.62, also never worked. As it isn't clear exactly what this means, and
5524 clearly it isn't a major issue, I have abolished the feature by treating it
5525 as "timeout", and writing a warning to the main and panic logs.
5527 36. The display of retry rules for -brt wasn't always showing the error code
5530 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5531 timeout_connect_A, timeout_connect_MX.
5533 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5534 to the empty sender.
5536 39. The daemon was not analysing the content of -oX till after it had closed
5537 stderr and disconnected from the controlling terminal. This meant that any
5538 syntax errors were only noted on the panic log, and the return code from
5539 the command was 0. By re-arranging the code a little, I've made the
5540 decoding happen first, so such errors now appear on stderr, and the return
5541 code is 1. However, the actual setting up of the sockets still happens in
5542 the disconnected process, so errors there are still only recorded on the
5545 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5546 connections (as happens on some IP stacks) was logged at start up time as
5547 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5548 it from "IPv6 and IPv4", which means that two separate sockets are being
5551 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5552 says whether AF_INET or AF_INET6 was passed as an argument.
5554 42. Exiwhat output was messed up when time zones were included in log
5557 43. Exiwhat now gives more information about the daemon's listening ports,
5558 and whether -tls-on-connect was used.
5560 44. The "port" option of the smtp transport is now expanded.
5562 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5563 silently ignored. Now an error message is written to the main and panic
5566 46. There's a new ACL modifier called "logwrite" which writes to a log file
5567 as soon as it is encountered.
5569 47. Added $local_user_uid and $local_user_gid at routing time.
5571 48. Exim crashed when trying to verify a sender address that was being
5574 49. Exim was recognizing only a space character after ".include". It now also
5575 recognizes a tab character.
5577 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5578 extracting the relevant information from the specification. The man page no
5579 longer contains scrambled data for the -d option, and I've added a section
5580 at the front about calling Exim under different names.
5582 51. Added "extra_headers" argument to the "mail" command in filter files.
5584 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5587 53. Installed eximstats 1.29.
5589 54. Added transport_filter_timeout as a generic transport option.
5591 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5592 Cc: header lines. This was required by RFC 822, but it not required by RFC
5595 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5596 incoming messages that did not have them. Now it does so only if the
5597 message originates locally, that is, if there is no associated remote host
5598 address. When Resent- header lines are present, this applies to the Resent-
5599 lines rather than the non-Resent- lines.
5601 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5602 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5604 58. Messages for configuration errors now include the name of the main
5605 configuration file - useful now that there may be more than one file in a
5606 list (.included file names were always shown).
5608 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5609 for those rare installations that do not start the daemon as root or run it
5610 setuid root. I've cut out the call to initgroups() if the daemon is not
5613 60. The Exim user and group can now be bound into the binary as text strings
5614 that are looked up at the start of Exim's processing.
5616 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5618 62. Added $mailstore_basename variable.
5620 63. Installed patch to sieve.c from Michael Haardt.
5622 64. When Exim failed to open the panic log after failing to open the main log,
5623 the original message it was trying to log was written to stderr and debug
5624 output, but if they were not available (the usual case in production), it
5625 was lost. Now it is written to syslog before the two lines that record the
5626 failures to open the logs.
5628 65. Users' Exim filters run in subprocesses under the user's uid. It is
5629 possible for a "deliver" command or an alias in a "personal" command to
5630 provoke an address rewrite. If logging of address rewriting is configured,
5631 this fails because the process is not running as root or exim. There may be
5632 a better way of dealing with this, but for the moment (because 4.30 needs
5633 to be released), I have disabled address rewrite logging when running a
5634 filter in a non-root, non-exim process.
5640 1. The buildconfig auxiliary program wasn't quoting the value set for
5641 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5642 not defined. This bug was masked in 4.22 by the effect that was fixed in
5645 2. Some messages that were rejected after a message id was allocated were
5646 shown as "incomplete" by exigrep. It no longer does this for messages that
5647 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5649 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5650 have allow_domain_literals set, the ID did not get logged in the <= line.
5651 Domain literals are now always recognized in Message-ID: header lines.
5653 4. The first argument for a ${extract expansion item is the key name or field
5654 number. Leading and trailing spaces in this item were not being ignored,
5655 causing some misleading effects.
5657 5. When deliver_drop_privilege was set, single queue runner processes started
5658 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5659 same command in the process it spins off) were not dropping privilege.
5661 6. When the daemon running as "exim" started a queue runner, it always
5662 re-executed Exim in the spun-off process. This is a waste of effort when
5663 deliver_drop_privilege is set. The new process now just calls the
5664 queue-runner function directly.
5670 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5673 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5674 ignored. Though the use of -oP was forcing the writing of a pid file, it
5675 was always written to the default place.
5677 3. If the message "no IP address found for host xxxx" is generated during
5678 incoming verification, it is now followed by identification of the incoming
5679 connection (so you can more easily find what provoked it).
5681 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5683 5. Added some features to "harden" Exim a bit more against certain attacks:
5685 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5686 be put in Local/Makefile. This is like the never_users runtime option,
5687 but it cannot be overridden. The default setting is "root".
5689 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5690 prefix string with which any file named in a -C command line option
5693 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5694 is retained for -C and -D only if the caller of Exim is root. Without
5695 it, the exim user may also use -C and -D and retain privilege.
5697 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5698 command line option is disabled.
5700 6. Macro names set by the -D option must start with an upper case letter, just
5701 like macro names defined in the configuration file.
5703 7. Added "dereference=" facility to LDAP.
5705 8. Two instances of the typo "uknown" in the source files are fixed.
5707 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5708 the Configure-Makefile script screwed up while processing it.
5710 10. Incorporated PCRE 4.4.
5712 11. The SMTP synchronization check was not operating right at the start of an
5713 SMTP session. For example, it could not catch a HELO sent before the client
5714 waited for the greeting. There is now a check for outstanding input at the
5715 point when the greeting is written. Because of the duplex, asynchronous
5716 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5717 way, but not yet received, when the check is performed.
5719 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5720 on TCP/IP sockets, because this apparently causes some broken clients to
5723 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5724 unchanged) from the Cygwin maintainer.
5726 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5727 was supported instead of testing for mbx. Effectively a typo.
5729 15. The spa authenticator server code was not checking that the input it
5730 received was valid base64.
5732 16. The debug output line for the "set" modifier in ACLs was not showing the
5733 name of the variable that was being set.
5735 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5737 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5738 lookup on its own. Something else was needed to trigger the lookup. For
5739 example, a match in host_lookup or the need for a host name in a host list.
5740 Now, if $sender_host_name is referenced and the host name has not yet been
5741 looked up, a lookup is performed. If the lookup fails, the variable remains
5742 empty, and $host_lookup_failed is set to "1".
5744 19. Added "eqi" as a case-independent comparison operator.
5746 20. The saslauthd authentication condition could segfault if neither service
5747 nor realm was specified.
5749 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5750 error message that was logged was misleading, and incoming SMTP
5751 connections were dropped. The message is now more accurate, and temporary
5752 errors are given to SMTP connections.
5754 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5755 (except RSET) with a 421 error, until QUIT is received. However, it was
5756 failing to send a response to QUIT.
5758 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5759 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5760 ACL is now run after the helo_try_verify_hosts code.
5762 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5763 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5764 case-independent, so other case variants are also recognized). Apparently
5765 some systems use these upper case variants.
5767 25. If more than two messages were waiting for the same host, and a transport
5768 filter was specified for the transport, Exim sent two messages over the
5769 same TCP/IP connection, and then failed with "socket operation on non-
5770 socket" when it tried to send the third.
5772 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5774 27. The extern definition of crypt16() in expand.c was not being excluded when
5775 the OS had its own crypt16() function.
5777 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5778 as a preferred synonym for return_size_limit, both as an option and as an
5781 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5783 30. Changed the default configuration ACL to relax the local part checking rule
5784 for addresses that are not in any local domains. For these addresses,
5785 slashes and pipe symbols are allowed within local parts, but the sequence
5786 /../ is explicitly forbidden.
5788 31. SPA server authentication was not clearing the challenge buffer before
5791 32. log_message in a "warn" ACL statement was writing to the reject log as
5792 well as to the main log, which contradicts the documentation and doesn't
5793 seem right (because no rejection is happening). So I have stopped it.
5795 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5796 However, I am unable to do any testing of this.
5798 34. Fixed an infelicity in the appendfile transport. When checking directories
5799 for a mailbox, to see if any needed to be created, it was accidentally
5800 using path names with one or more superfluous leading slashes; tracing
5801 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5803 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5804 contents are added to the log line that is written for every discarded
5805 recipient. (Previously a log_message setting was ignored.)
5807 36. The ${quote: operator now quotes the string if it is empty.
5809 37. The install script runs exim in order to find its version number. If for
5810 some reason other than non-existence or emptiness, which it checks, it
5811 could not run './exim', it was installing it with an empty version number,
5812 i.e. as "exim-". This error state is now caught, and the installation is
5815 38. An argument was missing from the function that creates an error message
5816 when Exim fails to connect to the socket for saslauthd authentication.
5817 This could cause Exim to crash, or give a corrupted message.
5819 39. Added isip, isip4, and isip6 to ${if conditions.
5821 40. The ACL variables $acl_xx are now saved with the message, and can be
5822 accessed later in routers, transports, and filters.
5824 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5825 strings in the file are not string-expanded.
5827 42. If a MAIL command specified a SIZE value that was too large to fit into an
5828 int variable, the check against message_size_limit failed. Such values are
5829 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5830 day this will have to be increased, but I don't think I want to be around
5831 when emails are that large.
5838 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5839 iconv() is not standard in FreeBSD.
5841 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5842 IPv6 enabled. The observed symptom was a segmentation fault on return from
5843 the function os_common_find_running_interfaces() in src/os.c.
5845 3. In the check_special_case() function in daemon.c I had used "errno" as an
5846 argument name, which causes warnings on some systems. This was basically a
5847 typo, since it was named "eno" in the comments!
5849 4. The code that waits for the clock to tick (at a resolution of some fraction
5850 of a second) so as to ensure message-id uniqueness was always waiting for
5851 at least one whole tick, when it could have waited for less. [This is
5852 almost certainly not relevant at current processor speeds, where it is
5853 unlikely to ever wait at all. But we try to future-proof.]
5855 5. The function that sleeps for a time interval that includes fractions of a
5856 second contained a race. It did not block SIGALRM between setting the
5857 timer, and suspending (a couple of lines later). If the interval was short
5858 and the sigsuspend() was delayed until after it had expired, the suspension
5859 never ended. On busy systems this could lead to processes getting stuck for
5862 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5863 process, before it forks any delivery processes. The open lookup caching
5864 mechanism meant that the open file or database connection was passed into
5865 the delivery process. The problem was that delivery processes always tidy
5866 up cached lookup data. This could cause a problem for the next delivery
5867 process started by the queue runner, because the external queue runner
5868 process does not know about the closure. So the next delivery process
5869 still has data in the lookup cache. In the case of a file lookup, there was
5870 no problem because closing a file descriptor in a subprocess doesn't affect
5871 the parent. However, if the lookup was caching a connection to a database,
5872 the connection was closed, and the second delivery process was likely to
5873 see errors such as "PGSQL: query failed: server closed the connection
5874 unexpectedly". The problem has been fixed by closing all cached lookups
5875 in a queue runner before running a delivery process.
5877 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5878 seem to have the "const" qualifier which it has on other OS. I've
5881 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5882 *of the same type* (client or server) with the same public name that an
5883 error should be diagnosed.
5885 9. When Exim looked up a host name for an IP address, but failed to find the
5886 original IP address when looking up the host name (a safety check), it
5887 output the message "<ip address> does not match any IP for NULL", which was
5888 confusing, to say the least. The bug was that the host name should have
5889 appeared instead of "NULL".
5891 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5892 user that is built into the binary, and the -C or -D options is used, root
5893 privilege is dropped before the configuration file is read. In addition,
5894 logging is switched to stderr instead of the normal log files. If the
5895 configuration then re-defines the Exim user, the unprivileged environment
5896 is probably not what is expected, so Exim logs a panic warning message (but
5899 However, if deliver_drop_privilege is set, the unprivileged state may well
5900 be exactly what is intended, so the warning has been cut out in that case,
5901 and Exim is allowed to try to write to its normal log files.
5907 1. smtp_return_error_details was not giving details for temporary sender
5908 or receiver verification errors.
5910 2. Diagnose a configuration error if two authenticators have the same public
5913 3. Exim used not to create the message log file for a message until the first
5914 delivery attempt. This could be confusing when incoming messages were held
5915 for policy or load reasons. The message log file is now created at the time
5916 the message is received, and an initial "Received" line is written to it.
5918 4. The automatically generated man page for command line options had a minor
5919 bug that caused no ill effects; however, a more serious problem was that
5920 the procedure for building the man page automatically didn't always
5921 operate. Consequently, release 4.20 contains an out-of-date version. This
5922 shouldn't happen again.
5924 5. When building Exim with embedded Perl support, the script that builds the
5925 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5926 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5928 6. The freeze_tell option was not being used for messages that were frozen on
5929 arrival, either by an ACL or by local_scan().
5931 7. Added the smtp_incomplete_transaction log selector.
5933 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5934 was accepting AUTH without a new EHLO.
5936 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5937 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5940 10. Logging of TCP/IP connections (when configured) now happens in the main
5941 daemon process instead of the child process, so that the TCP/IP connection
5942 count is more accurate (but it can never be perfect).
5944 11. The use of "drop" in a nested ACL was not being handled correctly in the
5945 outer ACL. Now, if condition failure induced by the nested "drop" causes
5946 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5947 or "require"), the connection is dropped.
5949 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5950 that yield "discard" can now be used with an "accept" or a "discard" verb,
5951 but an error is generated for any others (because I can't see a useful way
5952 to define what should happen).
5954 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5955 are now processed in the same way as lines in the Exim configuration file.
5956 In particular, continuation lines are supported.
5958 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5960 15. Added -ti meaning -t -i.
5962 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5963 domains, and warn by logging if others are found.
5965 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5966 returned from GIFCONF when Exim is trying to find the list of interfaces on
5967 a host. The code in os.c has been modified to copy each ifreq to an aligned
5968 structure in all cases.
5970 Also, in some cases, the returned ifreq's were being copied to a 'struct
5971 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5972 means the last couple of bytes of an IPv6 address could be chopped if the
5973 ifreq contained only a normal sockaddr (14 bytes storage).
5975 18. Named domain lists were not supported in the hosts_treat_as_local option.
5976 An entry such as +xxxx was not recognized, and was treated as a literal
5979 19. Ensure that header lines added by a DATA ACL are included in the reject log
5980 if the ACL subsequently rejects the message.
5982 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5983 MD5 (which is deprecated).
5985 21. When testing a filter file using -bf, Exim was writing a message when it
5986 took the sender from a "From " line in the message, but it was not doing so
5987 when it took $return_path from a Return-Path: header line. It now does.
5989 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5990 with a valid header line field name (a series of printing characters
5991 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5993 23. Changed "disc" in the source to "disk" to conform to the documentation and
5994 the book and for uniformity.
5996 24. Ignore Sendmail's -Ooption=value command line item.
5998 25. When execve() failed while trying to run a command in a pipe transport,
5999 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6000 could be confused with a return value of 69 from the command itself. This
6001 has been changed to 127, the value the shell returns if it is asked to run
6002 a non-existent command. The wording for the related log line suggests a
6003 non-existent command as the problem.
6005 26. If received_header_text expands to an empty string, do not add a Received:
6006 header line to the message. (Well, it adds a token one on the spool, but
6007 marks it "old" so that it doesn't get used or transmitted.)
6009 27. Installed eximstats 1.28 (addition of -nt option).
6011 28. There was no check for failure on the call to getsockname() in the daemon
6012 code. This can fail if there is a shortage of resources on the system, with
6013 ENOMEM, for example. A temporary error is now given on failure.
6015 29. Contrary to the C standard, it seems that in some environments, the
6016 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6017 program. Exim now does this explicitly; it affects the formatting of
6018 timestamps using strftime().
6020 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6021 complaints. I've now initialized all the variables, to avoid this.
6023 32. Header lines added by a system filter were not being "seen" during
6024 transport-time rewrites.
6026 33. The info_callback() function passed to OpenSSL is set up with type void
6027 (*)(SSL *, int, int), as described somewhere. However, when calling the
6028 function (actually a macro) that sets it up, the type void(*)() is
6029 expected. I've put in a cast to prevent warnings from picky compilers.
6031 34. If a DNS black list lookup found a CNAME record, but there were no A
6032 records associated with the domain it pointed at, Exim crashed.
6034 35. If a DNS black list lookup returned more than one A record, Exim ignored
6035 all but the first. It now scans all returned addresses if a particular IP
6036 value is being sought. In this situation, the contents of the
6037 $dnslist_value variable are a list of all the addresses, separated by a
6040 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6041 to accept a host name and all its aliases if the forward lookup for any of
6042 them yielded the IP address of the incoming connection. Now it accepts only
6043 those names whose forward lookup yields the correct IP address. Any other
6044 names are discarded. This closes a loophole whereby a rogue DNS
6045 administrator could create reverse DNS records to break through a
6046 wildcarded host restriction in an ACL.
6048 37. If a user filter or a system filter that ran in a subprocess used any of
6049 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6050 the wrong values were passed to the pipe command ($thisaddress had the
6051 value of $0, $0 had the value of $1, etc). This bug was introduced by
6052 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6054 38. Improved the line breaking for long SMTP error messages from ACLs.
6055 Previously, if there was no break point between 40 and 75 characters, Exim
6056 left the rest of the message alone. Two changes have been made: (a) I've
6057 reduced the minimum length to 35 characters; (b) if it can't find a break
6058 point between 35 and 75 characters, it looks ahead and uses the first one
6059 that it finds. This may give the occasional overlong line, but at least the
6060 remaining text gets split now.
6062 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6063 file descriptors might be low, and that setting 1000 would always raise it.
6064 It turns out that in some environments, the limit is already over 1000 and
6065 that lowering it causes trouble. So now Exim takes care not to decrease it.
6067 40. When delivering a message, the value of $return_path is set to $sender_
6068 address at the start of routing (routers may change the value). By an
6069 oversight, this default was not being set up when an address was tested by
6070 -bt or -bv, which affected the outcome if any router or filter referred to
6073 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6074 log only when "message" or "log_message" are set. However, if one of the
6075 conditions was an address verification, or a call to a nested ACL, the
6076 messages generated by the underlying test were being passed through. This
6077 no longer happens. The underlying message is available in $acl_verify_
6078 message for both "message" and "log_message" expansions, so it can be
6079 passed through if needed.
6081 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6082 new expansion $bh_ to give the encoded byte string without charset
6083 translation. Translation happens only if iconv() is available; HAVE_ICONV
6084 indicates this at build time. HEADERS_CHARSET gives the charset to
6085 translate to; headers_charset can change it in the configuration, and
6086 "headers charset" can change it in an individual filter file.
6088 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6089 that creates RFC 2047 encoded "words" labels them as that charset instead
6090 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6091 expansion operator; (ii) when Exim creates a From: line for a local
6092 message; (iii) when a header line is rewritten to include a "phrase" part.
6094 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6095 buggy, causing it to skip the first lines of messages whose message ID
6096 ended in 'D'. This would not have bitten before Exim release 4.14, because
6097 message IDs were unlikely to end in 'D' before then. The effect was to have
6098 incorrect size information for certain domains.
6100 45. #include "config.h" was missing at the start of the crypt16.c module. This
6101 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6104 46. If there was a timeout during a "random" callout check, Exim treated it as
6105 a failure of the random address, and carried on sending RSET and the real
6106 address. If the delay was just some slowness somewhere, the response to the
6107 original RCPT would be taken as a response to RSET and so on, causing
6108 mayhem of various kinds.
6110 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6111 when I implemented it. It didn't allow for the fact that some option values
6112 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6113 the right test for positive values.
6115 48. Domain names in DNS records are case-independent. Exim always looks them up
6116 in lower case. Some resolvers return domain names in exactly the case they
6117 appear in the zone file, that is, they may contain uppercase letters. Not
6118 all resolvers do this - some return always lower case. Exim was treating a
6119 change of case by a resolver as a change of domain, similar to a widening
6120 of a domain abbreviation. This triggered its re-routing code and so it was
6121 trying to route what was effectively the same domain again. This normally
6122 caused routing to fail (because the router wouldn't handle the domain
6123 twice). Now Exim checks for this case specially, and just changes the
6124 casing of the domain that it ultimately uses when it transmits the message
6127 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6130 50. If a filter generated a file delivery with a non-absolute name (possible if
6131 no home directory exists for the router), the forbid_file option was not
6134 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6135 the existing equality matching.
6137 52. Exim was using ints instead of ino_t variables in some places where it was
6138 dealing with inode numbers.
6140 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6141 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6142 called TMPDIR, and if it finds it is different, it changes its value.
6144 54. The smtp_printf() function is now made available to local_scan() so
6145 additional output lines can be written before returning. There is also an
6146 smtp_fflush() function to enable the detection of a dropped connection.
6147 The variables smtp_input and smtp_batched_input are exported to
6150 55. Changed the default runtime configuration: the message "Unknown user"
6151 has been removed from the ACL, and instead placed on the localuser router,
6152 using the cannot_route_message feature. This means that any verification
6153 failures that generate their own messages won't get overridden. Similarly,
6154 the "Unrouteable address" message that was in the ACL for unverifiable
6155 relay addresses has also been removed.
6157 56. Added hosts_avoid_esmtp to the smtp transport.
6159 57. The exicyclog script was not checking for the esoteric option
6160 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6161 will work only if exicyclog is run under the appropriate euid.
6163 58. Following a discussion on the list, the rules by which Exim recognises line
6164 endings on incoming messages have been changed. The -dropcr and drop_cr
6165 options are now no-ops, retained only for backwards compatibility. The
6166 following line terminators are recognized: LF CRLF CR. However, special
6167 processing applies to CR:
6169 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6170 nor a local message in the state where . is a terminator.
6172 (ii) If a bare CR is encountered in a header line, an extra space is added
6173 after the line terminator so as not to end the header. The reasoning
6174 behind this is that bare CRs in header lines are most likely either
6175 to be mistakes, or people trying to play silly games.
6177 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6178 was being incorrectly given as 18 bytes larger than it should have been.
6179 This is a VOB (very old bug).
6181 60. This may never have affected anything current, but just in case it has:
6182 When the local host is found other than at the start of a list of hosts,
6183 the local host, those with the same MX, and any that follow, are discarded.
6184 When the list in question was part of a longer list of hosts, the following
6185 hosts (not currently being processed) were also being discarded. This no
6186 longer happens. I'm not sure if this situation could ever has previously
6189 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6192 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6193 header. This is something that is recommended in a new Internet Draft, and
6194 is something that is documented as being done by Sendmail. There are two
6195 possible values. For messages generated by the autoreply transport, Exim
6198 Auto-Submitted: auto-replied
6200 whereas for all other generated messages (e.g. bounces) it adds
6202 Auto-Submitted: auto-generated
6204 63. The "personal" condition in filters now includes a test for the
6205 Auto-Submitted: header. If it contains the string "auto-" the message it
6206 not considered personal.
6208 64. Added rcpt_include_affixes as a generic transport option.
6210 65. Added queue_only_override (default true).
6212 66. Added the syslog_duplication option.
6214 67. If what should have been the first header line of a message consisted of
6215 a space followed by a colon, Exim was mis-interpreting it as a header line.
6216 It isn't of course - it is syntactically invalid and should therefore be
6217 treated as the start of the message body. The misbehaviour could have
6218 caused a number of strange effects, including loss of data in subsequent
6219 header lines, and spool format errors.
6221 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6222 client host had authenticated. This control can now be exercised by an ACL
6223 for more flexibility.
6225 69. By default, callouts do not happen when testing with -bh. There is now a
6226 variant, -bhc, which does actually run the callout code, including
6227 consulting and updating the callout cache.
6229 70. Added support for saslauthd authentication, courtesy of Alexander
6232 71. If statvfs() failed on the spool or log directories while checking their
6233 size for availability, Exim confusingly gave the error "space shortage".
6234 Furthermore, in debugging mode it crashed with a floating point exception.
6235 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6236 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6237 serious problem, Exim now writes to the main and panic logs when this
6238 happens, with details of the failure. It then refuses to accept the
6239 incoming message, giving the message "spool directory problem" or "log
6240 directory problem" with a 421 code for SMTP messages.
6242 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6243 0, 1, and 2 exist, because some OS complain for execs without them (see
6244 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6245 descriptors. However, the code omitted to check that the open succeeded,
6246 causing mysterious errors if for some reason the permissions on /dev/null
6247 got screwed. Now Exim writes a message to the main and panic logs, and
6248 bombs out if it can't open /dev/null.
6250 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6251 interact so that it is all more flexible. It is supposed to remain
6252 backwards compatible. Also added extra_local_interfaces.
6254 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6255 to bomb out with an assertion failure - to the client this appears as a
6256 connection drop. This problem occurs in the part of the code that was taken
6257 from the Samba project. Fortunately, the assertion is in a very simple
6258 function, so I have fixed this by reproducing the function inline in the
6259 one place where it is called, and arranging for authentication to fail
6260 instead of killing the process with assert().
6262 75. The SPA client code was not working when the server requested OEM rather
6263 than Unicode encoding.
6265 76. Added code to make require_files with a specific uid setting more usable in
6266 the case where statting the file as root fails - usually a non-root-mounted
6267 NFS file system. When this happens and the failure is EACCES, Exim now
6268 forks a subprocess and does the per-uid checking as the relevant uid.
6270 77. Added process_log_path.
6272 78. If log_file_path was not explicitly set, a setting of check_log_space or
6273 check_log_inodes was ignored.
6275 79. If a space check for the spool or log partitions fails, the incident is now
6276 logged. Of course, in the latter case the data may get lost...
6278 80. Added the %p formatting code to string_format() so that it can be used to
6279 print addresses in debug_print(). Adjusted all the address printing in the
6280 debugging in store.c to use %p rather than %d.
6282 81. There was a concern that a line of code in smtp_in.c could overflow a
6283 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6284 initially expressed, the concern was not well-founded, because trailing
6285 spaces are removed early. However, if the trailing spaces were followed by
6286 a NULL, they did not get removed, so the overflow was possible. Two fixes
6289 (a) I re-wrote the offending code in a cleaner fashion.
6290 (b) If an incoming SMTP command contains a NULL character, it is rejected
6293 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6294 runs initgroups(), so that if the Exim user is in any additional groups,
6295 they will be used during message reception.
6301 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob