1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
8 and 3.4.4 - once the server is enabled to respond to an OCSP request
9 it does even when not requested, resulting in a stapling non-aware
10 client dropping the TLS connection.
12 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
13 support variable-length bit vectors. No functional change.
15 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
16 The I= interface field on outgoing lines is now after the H= remote
17 host field, same as incoming lines. There is a separate
18 outgoing_interface log selector which allows you to disable the
21 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
22 If not running log_selector +smtp_connection the mainlog would be held
23 open indefinitely after a "too many connections" event, including to a
24 deleted file after a log rotate. Leave the per net connection logging
25 leaving it open for efficiency as that will be quickly detected by the
26 check on the next write.
28 HS/01 Bug 1671: Fix post transport crash.
29 Processing the wait-<transport> messages could crash the delivery
30 process if the message IDs didn't exist for some reason. When
31 using 'split_spool_directory=yes' the construction of the spool
32 file name failed already, exposing the same netto behaviour.
34 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
35 mime_regex ACL conditions.
37 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
38 to DSN fail messages (bounces): remote IP, remote greeting, remote response
39 to HELO, local diagnostic string.
44 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
47 JH/02 The smtp transport option "multi_domain" is now expanded.
49 JH/03 The smtp transport now requests PRDR by default, if the server offers
52 JH/04 Certificate name checking on server certificates, when exim is a client,
53 is now done by default. The transport option tls_verify_cert_hostnames
54 can be used to disable this per-host. The build option
55 EXPERIMENTAL_CERTNAMES is withdrawn.
57 JH/05 The value of the tls_verify_certificates smtp transport and main options
58 default to the word "system" to access the system default CA bundle.
59 For GnuTLS, only version 3.0.20 or later.
61 JH/06 Verification of the server certificate for a TLS connection is now tried
62 (but not required) by default. The verification status is now logged by
63 default, for both outbound TLS and client-certificate supplying inbound
66 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
69 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
70 Status Notification (bounce) messages are now MIME format per RFC 3464.
71 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
72 under the control of the dsn_advertise_hosts option, and routers may
73 have a dsn_lasthop option.
75 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
76 default, modifiable by a malware= option. The list separator for
77 the options can now be changed in the usual way. Bug 68.
79 JH/10 The smtp_receive_timeout main option is now expanded before use.
81 JH/11 The incoming_interface log option now also enables logging of the
82 local interface on delivery outgoing connections.
84 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
85 if the interface and destination host and port all match.
87 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
90 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
91 Patch from Andrew Lewis.
93 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
94 now supports optional time-restrictions, weighting, and priority
95 modifiers per server. Patch originally by <rommer@active.by>.
97 JH/16 The spamd_address main option now supports a mixed list of local
98 and remote servers. Remote servers can be IPv6 addresses, and
101 JH/17 Bug 68: The spamd_address main option now supports an optional
102 timeout value per server.
104 JH/18 Bug 1581: Router and transport options headers_add/remove can
105 now have the list separator specified.
107 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
110 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
113 JH/21 Support for the A6 type of dns record is withdrawn.
115 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
116 rather than the verbs used.
118 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
119 from 255 to 1024 chars.
121 JH/24 Verification callouts now attempt to use TLS by default.
123 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
124 are generic router options now. The defaults didn't change.
126 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
127 Original patch from Alexander Shikoff, worked over by JH.
129 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
130 files if ZCAT_COMMAND is not executable.
132 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
134 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
136 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
137 Normally benign, it bites when the pair was led to by a CNAME;
138 modern usage is to not canoicalize the domain to a CNAME target
139 (and we were inconsistent anyway for A-only vs AAAA+A).
141 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
143 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
144 when evaluating $sender_host_dnssec.
146 JH/31 Check the HELO verification lookup for DNSSEC, adding new
147 $sender_helo_dnssec variable.
149 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
151 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
153 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
155 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
156 documented as working, but never had. Support all but $spam_report.
158 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
159 added for tls authenticator.
164 TL/01 When running the test suite, the README says that variables such as
165 no_msglog_check are global and can be placed anywhere in a specific
166 test's script, however it was observed that placement needed to be near
167 the beginning for it to behave that way. Changed the runtest perl
168 script to read through the entire script once to detect and set these
169 variables, reset to the beginning of the script, and then run through
170 the script parsing/test process like normal.
172 TL/02 The BSD's have an arc4random API. One of the functions to induce
173 adding randomness was arc4random_stir(), but it has been removed in
174 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
175 function when detected.
177 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
178 cause callback expansion.
180 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
181 syntax errors in an expansion can be treated as a string instead of
182 logging or causing an error, due to the internal use of bool_lax
183 instead of bool when processing it.
185 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
186 server certificates when making smtp deliveries.
188 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
190 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
192 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
194 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
195 Merged patch from Sebastian Wiedenroth.
197 JH/05 Fix results-pipe from transport process. Several recipients, combined
198 with certificate use, exposed issues where response data items split
199 over buffer boundaries were not parsed properly. This eventually
200 resulted in duplicates being sent. This issue only became common enough
201 to notice due to the introduction of conection certificate information,
202 the item size being so much larger. Found and fixed by Wolfgang Breyha.
204 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
205 size buffer was used, resulting in syntax errors when an expansion
208 JH/07 Add support for directories of certificates when compiled with a GnuTLS
209 version 3.3.6 or later.
211 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
212 is EXPERIMENTAL_EVENT, the main-configuration and transport options
213 both become "event_action", the variables become $event_name, $event_data
214 and $event_defer_errno. There is a new variable $verify_mode, usable in
215 routers, transports and related events. The tls:cert event is now also
216 raised for inbound connections, if the main configuration event_action
219 TL/06 In test suite, disable OCSP for old versions of openssl which contained
220 early OCSP support, but no stapling (appears to be less than 1.0.0).
222 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
223 server certificate names available under the smtp transport option
224 "tls_verify_cert_hostname" now do not permit multi-component wildcard
227 JH/10 Time-related extraction expansions from certificates now use the main
228 option "timezone" setting for output formatting, and are consistent
229 between OpenSSL and GnuTLS compilations. Bug 1541.
231 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
232 encoded parameter in the incoming message. Bug 1558.
234 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
235 include certificate info, eximon was claiming there were spoolfile
238 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
240 JH/14 Log delivery-related information more consistently, using the sequence
241 "H=<name> [<ip>]" wherever possible.
243 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
244 are problematic for Debian distribution, omit them from the release
247 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
249 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
252 JH/17 Fix a null-indirection in certextract expansions when a nondefault
253 output list separator was used.
258 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
259 checkers that were complaining about end of non-void function with no
262 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
263 This was a regression intruduced in 4.83 by another bugfix.
265 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
267 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
268 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
274 TF/01 Correctly close the server side of TLS when forking for delivery.
276 When a message was received over SMTP with TLS, Exim failed to clear up
277 the incoming connection properly after forking off the child process to
278 deliver the message. In some situations the subsequent outgoing
279 delivery connection happened to have the same fd number as the incoming
280 connection previously had. Exim would try to use TLS and fail, logging
281 a "Bad file descriptor" error.
283 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
284 utilities have not been installed.
286 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
287 temporary space as the ACL may create new global variables.
289 TL/01 LDAP support uses per connection or global context settings, depending
290 upon the detected version of the libraries at build time.
292 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
293 to extract and use the src ip:port in logging and expansions as if it
294 were a direct connection from the outside internet. PPv2 support was
295 updated based on HAProxy spec change in May 2014.
297 JH/02 Add ${listextract {number}{list}{success}{fail}}.
299 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
300 Properly escape header and check for NULL return.
302 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
305 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
307 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
308 characters in header names, implemented as a verify condition.
309 Contributed by Michael Fischer v. Mollard.
311 TL/05 Rename SPF condition results err_perm and err_temp to standardized
312 results permerror and temperror. Previous values are deprecated but
313 still accepted. In a future release, err_perm and err_temp will be
314 completely removed, which will be a backward incompatibility if the
315 ACL tests for either of these two old results. Patch contributed by
316 user bes-internal on the mailing list.
318 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
320 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
321 selectors, in both main and reject logs.
323 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
326 JH/07 Add malware type "sock" for talking to simple daemon.
328 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
330 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
331 routers/transports under cutthrough routing.
333 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
334 numbers. Touch up "bool" conditional to keep the same definition.
336 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
338 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
340 JH/12 Expand items in router/transport headers_add or headers_remove lists
341 individually rather than the list as a whole. Bug 1452.
343 Required for reasonable handling of multiple headers_ options when
344 they may be empty; requires that headers_remove items with embedded
345 colons must have them doubled (or the list-separator changed).
347 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
348 view the policy declared in the DMARC record. Currently, $dmarc_status
349 is a combined value of both the record presence and the result of the
352 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
354 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
355 dnslookup router and the smtp transport (applying to the forward
358 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
359 of ldap servers used for a specific lookup. Patch provided by Heiko
362 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
363 New variable $lookup_dnssec_authenticated for observability.
365 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
366 Patch submitted by Lars Timman.
368 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
370 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
371 Requires trusted mode and valid format message id, aborts otherwise.
372 Patch contributed by Heiko Schlichting.
374 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
375 certextract with support for various fields. Bug 1358.
377 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
378 is requested by default, modifiable by smtp transport option
381 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
382 operate on certificate variables to give certificate fingerprints
383 Also new ${sha256:cert_variable}.
385 JH/23 The PRDR feature is moved from being Experimental into the mainline.
387 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
388 Christian Aistleitner.
390 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
392 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
393 file. Patch from Wolfgang Breyha.
395 JH/25 Expand the coverage of the delivery $host and $host_address to
396 client authenticators run in verify callout. Bug 1476.
398 JH/26 Port service names are now accepted for tls_on_connect_ports, to
399 align with daemon_smtp_ports. Bug 72.
401 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
402 support and error reporting did not work properly.
404 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
405 and is readable. Patch from Andrew Colin Kissa.
407 TL/14 Enhance documentation of ${run expansion and how it parses the
408 commandline after expansion, particularly in the case when an
409 unquoted variable expansion results in an empty value.
411 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
413 PP/02 Fix internal collision of T_APL on systems which support RFC3123
414 by renaming away from it. Addresses GH issue 15, reported by
417 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
419 TL/15 SECURITY: prevent double expansion in math comparison functions
420 (can expand unsanitized data). Not remotely exploitable.
427 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
429 PP/02 Make -n do something, by making it not do something.
430 When combined with -bP, the name of an option is not output.
432 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
435 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
436 $sender_host_name and config options to manage this, and basic check
439 PP/05 DSCP support for outbound connections and control modifier for inbound.
441 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
442 (Only plugin which currently uses this is kerberos4, which nobody should
443 be using, but we should make it available and other future plugins might
444 conceivably use it, even though it would break NAT; stuff *should* be
445 using channel bindings instead).
447 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
448 name; added for Sendmail compatibility; requires admin caller.
449 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
450 just ignore it); requires trusted caller.
451 Also parse but ignore: -Ac -Am -X<logfile>
454 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
456 TL/02 Add +smtp_confirmation as a default logging option.
458 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
459 Patch by Magnus Holmgren from 2007-02-20.
461 TL/04 Bugzilla 1281 - Spec typo.
462 Bugzilla 1283 - Spec typo.
463 Bugzilla 1290 - Spec grammar fixes.
465 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
467 TL/06 Add Experimental DMARC support using libopendmarc libraries.
469 TL/07 Fix an out of order global option causing a segfault. Reported to dev
470 mailing list by by Dmitry Isaikin.
472 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
474 JH/02 Support "G" suffix to numbers in ${if comparisons.
476 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
478 NM/01 Bugzilla 1197 - Spec typo
479 Bugzilla 1196 - Spec examples corrections
481 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
483 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
484 gnutls_enable_pkcs11, but renamed to more accurately indicate its
487 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
488 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
490 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
491 "acl {{name}{arg}...}", and optional args on acl condition
494 JH/05 Permit multiple router/transport headers_add/remove lines.
496 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
498 JH/07 Avoid using a waiting database for a single-message-only transport.
499 Performance patch from Paul Fisher. Bugzilla 1262.
501 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
504 JH/09 Add $headers_added variable, with content from use of ACL modifier
505 add_header (but not yet added to the message). Bugzilla 199.
507 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
508 Pulled from Bugzilla 817 by Wolfgang Breyha.
510 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
512 (nb: this is the same fix as in Exim 4.80.1)
514 JH/11 Add A= logging on delivery lines, and a client_set_id option on
517 JH/12 Add optional authenticated_sender logging to A= and a log_selector
520 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
522 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
523 advertise SMTP AUTH mechanism to us, instead of a generic
524 protocol violation error. Also, make Exim more robust to bad
525 data from the Dovecot auth socket.
527 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
529 When a queue runner is handling a message, Exim first routes the
530 recipient addresses, during which it prunes them based on the retry
531 hints database. After that it attempts to deliver the message to
532 any remaining recipients. It then updates the hints database using
535 So if a recipient address works intermittently, it can get repeatedly
536 deferred at routing time. The retry hints record remains fresh so the
537 address never reaches the final cutoff time.
539 This is a fairly common occurrence when a user is bumping up against
540 their storage quota. Exim had some logic in its local delivery code
541 to deal with this. However it did not apply to per-recipient defers
542 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
544 This change adds a proper retry rule check during routing so that the
545 final cutoff time is checked against the message's age. We only do
546 this check if there is an address retry record and there is not a
547 domain retry record; this implies that previous attempts to handle
548 the address had the retry_use_local_parts option turned on. We use
549 this as an approximation for the destination being like a local
550 delivery, as in LMTP.
552 I suspect this new check makes the old local delivery cutoff check
553 redundant, but I have not verified this so I left the code in place.
555 TF/02 Correct gecos expansion when From: is a prefix of the username.
557 Test 0254 submits a message to Exim with the header
561 When I ran the test suite under the user fanf2, Exim expanded
562 the header to contain my full name, whereas it should have added
563 a Resent-Sender: header. It erroneously treats any prefix of the
564 username as equal to the username.
566 This change corrects that bug.
568 GF/01 DCC debug and logging tidyup
569 Error conditions log to paniclog rather than rejectlog.
570 Debug lines prefixed by "DCC: " to remove any ambiguity.
572 TF/03 Avoid unnecessary rebuilds of lookup-related code.
574 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
575 Bug spotted by Jeremy Harris; was flawed since initial commit.
576 Would have resulted in OCSP responses post-SNI triggering an Exim
577 NULL dereference and crash.
579 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
581 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
582 Bug detection, analysis and fix by Samuel Thibault.
583 Bugzilla 1331, Debian bug #698092.
585 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
587 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
588 Server implementation by Todd Lyons, client by JH.
589 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
590 config variable "prdr_enable" controls whether the server
591 advertises the facility. If the client requests PRDR a new
592 acl_data_smtp_prdr ACL is called once for each recipient, after
593 the body content is received and before the acl_smtp_data ACL.
594 The client is controlled by bolth of: a hosts_try_prdr option
595 on the smtp transport, and the server advertisement.
596 Default client logging of deliveries and rejections involving
597 PRDR are flagged with the string "PRDR".
599 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
600 fclose(). Diagnosis by Todd Lyons.
602 PP/17 Update configure.default to handle IPv6 localhost better.
603 Patch by Alain Williams (plus minor tweaks).
606 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
607 This is now consistent with GnuTLS, and is now documented: the
608 previous undocumented portable approach to treating the option as
609 unset was to force an expansion failure. That still works, and
610 an empty string is now equivalent.
612 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
613 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
614 not performing validation itself.
616 PP/20 Added force_command boolean option to pipe transport.
617 Patch from Nick Koston, of cPanel Inc.
619 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
622 TF/04 Added udpsend ACL modifer and hexquote expansion operator
624 PP/21 Fix eximon continuous updating with timestamped log-files.
625 Broken in a format-string cleanup in 4.80, missed when I repaired the
626 other false fix of the same issue.
627 Report and fix from Heiko Schlichting.
630 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
631 Report from Prashanth Katuri.
633 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
634 It's SecureTransport, so affects any MacOS clients which use the
635 system-integrated TLS libraries, including email clients.
637 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
638 using a MIME ACL for non-SMTP local injection.
639 Report and assistance in diagnosis by Warren Baker.
641 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
643 JH/16 Fix comparisons for 64b. Bugzilla 1385.
645 TL/09 Add expansion variable $authenticated_fail_id to keep track of
646 last id that failed so it may be referenced in subsequent ACL's.
648 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
651 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
652 ldap library initialization, allowing self-signed CA's to be
653 used. Also properly sets require_cert option later in code by
654 using NULL (global ldap config) instead of ldap handle (per
655 session). Bug diagnosis and testing by alxgomz.
657 TL/12 Enhanced documentation in the ratelimit.pl script provided in
658 the src/util/ subdirectory.
660 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
661 renamed to Transport Post Delivery Action by Jeremy Harris, as
664 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
665 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
666 redis_servers = needs to be configured which will be used by the redis
667 lookup. Patch from Warren Baker, of The Packet Hub.
669 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
671 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
672 hostname or reverse DNS when processing a host list. Used suggestions
673 from multiple comments on this bug.
675 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
677 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
678 Missed a few lines, added it to make the runtest require no keyboard
681 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
682 contains upper case chars. Make router use caseful_local_part.
684 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
685 support when GnuTLS has been built with p11-kit.
691 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
693 This, or similar/improved, will also be change PP/11 of 4.82.
699 PP/01 Handle short writes when writing local log-files.
700 In practice, only affects FreeBSD (8 onwards).
701 Bugzilla 1053, with thanks to Dmitry Isaikin.
703 NM/01 Bugzilla 949 - Documentation tweak
705 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
708 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
710 PP/02 Implemented gsasl authenticator.
712 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
714 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
715 `pkg-config foo` for cflags/libs.
717 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
718 with rest of GSASL and with heimdal_gssapi.
720 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
721 `pkg-config foo` for cflags/libs for the TLS implementation.
723 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
724 properties get this fed in as external SSF. A number of robustness
725 and debugging improvements to the cyrus_sasl authenticator.
727 PP/08 cyrus_sasl server now expands the server_realm option.
729 PP/09 Bugzilla 1214 - Log authentication information in reject log.
730 Patch by Jeremy Harris.
732 PP/10 Added dbmjz lookup type.
734 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
736 PP/12 MAIL args handles TAB as well as SP, for better interop with
737 non-compliant senders.
738 Analysis and variant patch by Todd Lyons.
740 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
741 Bug report from Lars Müller <lars@samba.org> (via SUSE),
742 Patch from Dirk Mueller <dmueller@suse.com>
744 PP/13 tls_peerdn now print-escaped for spool files.
745 Observed some $tls_peerdn in wild which contained \n, which resulted
746 in spool file corruption.
748 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
749 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
750 or write after TLS renegotiation, which otherwise led to messages
753 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
754 as a tracking header (ie: a signed header comes before the signature).
755 Patch from Wolfgang Breyha.
757 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
758 comma-sep list; embedded commas doubled.
760 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
762 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
764 Report and patch from Dmitry Banschikov.
766 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
767 Removed SSL_clear() after SSL_new() which led to protocol negotiation
768 failures. We appear to now support TLS1.1+ with Exim.
770 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
771 lets Exim select keys and certificates based upon TLS SNI from client.
772 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
773 before an outbound SMTP session. New log_selector, +tls_sni.
775 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
776 NULL dereference. Report and patch from Alun Jones.
778 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
779 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
780 Not seeing resolver debug output on NetBSD, but suspect this is a
781 resolver implementation change.
783 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
784 Left warnings. Added "eximon gdb" invocation mode.
786 PP/21 Defaulting "accept_8bitmime" to true, not false.
788 PP/22 Added -bw for inetd wait mode support.
790 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
791 locate the relevant includes and libraries. Made this the default.
793 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
794 Bugzilla 1246, report and most of solution from Tomasz Kusy.
796 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
797 This may cause build issues on older platforms.
799 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
800 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
801 gnutls_require_mac & gnutls_require_protocols (no longer supported).
802 Added SNI support via GnuTLS too.
803 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
805 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
807 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
808 Applied second patch from Janne, implementing suggestion to default
809 multiple-strings-in-record handling to match SPF spec.
811 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
813 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
814 read-only, out of scope).
815 Patch from Wolfgang Breyha, report from Stuart Northfield.
817 PP/29 Fix three issues highlighted by clang analyser static analysis.
818 Only crash-plausible issue would require the Cambridge-specific
819 iplookup router and a misconfiguration.
820 Report from Marcin Mirosław.
822 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
824 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
825 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
826 As part of this, removing so much warning spew let me fix some minor
827 real issues in debug logging.
829 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
830 assignment on my part. Fixed.
832 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
833 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
834 Janne Snabb (who went above and beyond: thank you).
836 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
837 string otherwise requires a connection and a bunch more work and it's
838 relatively easy to get wrong. Should also expose TLS library linkage
841 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
842 64-bit ${eval} (JH/03).
844 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
845 GNU libc to support some of the 64-bit stuff, should not lead to
846 conflicts. Defined before os.h is pulled in, so if a given platform
847 needs to override this, it can.
849 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
850 protection layer was required, which is not implemented.
851 Bugzilla 1254, patch from Wolfgang Breyha.
853 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
854 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
855 tls_dhparam take prime identifiers. Also unbreak combination of
856 OpenSSL+DH_params+TLSSNI.
858 PP/39 Disable SSLv2 by default in OpenSSL support.
864 PP/01 Solaris build fix for Oracle's LDAP libraries.
865 Bugzilla 1109, patch from Stephen Usher.
867 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
869 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
872 TF/02 Fix a couple more cases where we did not log the error message
873 when unlink() failed. See also change 4.74-TF/03.
875 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
876 lock up or crash if it happened to be inside a call to libc when it
877 got a SIGUSR1 from exiwhat.
879 The SIGUSR1 handler appends the current process status to the process
880 log which is later printed by exiwhat. It used to use the general
881 purpose logging code to do this, but several functions it calls are
882 not safe for signals.
884 The new output code in the SIGUSR1 handler is specific to the process
885 log, and simple enough that it's easy to inspect for signal safety.
886 Removing some special cases also simplifies the general logging code.
887 Removing the spurious timestamps from the process log simplifies
890 TF/04 Improved ratelimit ACL condition.
892 The /noupdate option has been deprecated in favour of /readonly which
893 has clearer semantics. The /leaky, /strict, and /readonly update modes
894 are mutually exclusive. The update mode is no longer included in the
895 database key; it just determines when the database is updated. (This
896 means that when you upgrde Exim will forget old rate measurements.)
898 Exim now checks that the per_* options are used with an update mode that
899 makes sense for the current ACL. For example, when Exim is processing a
900 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
901 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
902 must specify per_mail/readonly. If you omit the update mode it defaults to
903 /leaky where that makes sense (as before) or /readonly where required.
905 The /noupdate option is now undocumented but still supported for
906 backwards compatibility. It is equivalent to /readonly except that in
907 ACLs where /readonly is required you may specify /leaky/noupdate or
908 /strict/noupdate which are treated the same as /readonly.
910 A useful new feature is the /count= option. This is a generalization
911 of the per_byte option, so that you can measure the throughput of other
912 aggregate values. For example, the per_byte option is now equivalent
913 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
915 The per_rcpt option has been generalized using the /count= mechanism
916 (though it's more complicated than the per_byte equivalence). When it is
917 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
918 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
919 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
920 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
921 Note that using per_rcpt with a non-readonly update mode in more than
922 one ACL will cause the recipients to be double-counted. (The per_mail
923 and per_byte options don't have this problem.)
925 The handling of very low rates has changed slightly. If the computed rate
926 is less than the event's count (usually one) then this event is the first
927 after a long gap. In this case the rate is set to the same as this event's
928 count, so that the first message of a spam run is counted properly.
930 The major new feature is a mechanism for counting the rate of unique
931 events. The new per_addr option counts the number of different
932 recipients that someone has sent messages to in the last time period. It
933 behaves like per_rcpt if all the recipient addresses are different, but
934 duplicate recipient addresses do not increase the measured rate. Like
935 the /count= option this is a general mechanism, so the per_addr option
936 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
937 example, measure the rate that a client uses different sender addresses
938 with the options per_mail/unique=$sender_address. There are further
939 details in the main documentation.
941 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
943 TF/06 Removed a few PCRE remnants.
945 TF/07 Automatically extract Exim's version number from tags in the git
946 repository when doing development or release builds.
948 PP/02 Raise smtp_cmd_buffer_size to 16kB.
949 Bugzilla 879. Patch from Paul Fisher.
951 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
952 Heavily based on revision 40f9a89a from Simon Arlott's tree.
955 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
957 PP/05 Variable $av_failed, true if the AV scanner deferred.
958 Bugzilla 1078. Patch from John Horne.
960 PP/06 Stop make process more reliably on build failure.
961 Bugzilla 1087. Patch from Heiko Schlittermann.
963 PP/07 Make maildir_use_size_file an _expandable_ boolean.
964 Bugzilla 1089. Patch from Heiko Schlittermann.
966 PP/08 Handle ${run} returning more data than OS pipe buffer size.
967 Bugzilla 1131. Patch from Holger Weiß.
969 PP/09 Handle IPv6 addresses with SPF.
970 Bugzilla 860. Patch from Wolfgang Breyha.
972 PP/10 GnuTLS: support TLS 1.2 & 1.1.
974 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
977 PP/11 match_* no longer expand right-hand-side by default.
978 New compile-time build option, EXPAND_LISTMATCH_RHS.
979 New expansion conditions, "inlist", "inlisti".
981 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
983 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
985 PP/14 fix log_write() format string regression from TF/03.
986 Bugzilla 1152. Patch from Dmitry Isaikin.
992 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
994 PP/02 Harmonised TLS library version reporting; only show if debugging.
995 Layout now matches that introduced for other libraries in 4.74 PP/03.
997 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
999 PP/04 New "dns_use_edns0" global option.
1001 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1004 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1005 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1007 TK/01 Updated PolarSSL code to 0.14.2.
1008 Bugzilla 1097. Patch from Andreas Metzler.
1010 PP/07 Catch divide-by-zero in ${eval:...}.
1011 Fixes bugzilla 1102.
1013 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1016 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1017 format-string attack -- SECURITY: remote arbitrary code execution.
1019 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1020 time unintentionally subject to list matching rules, letting the header
1021 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1022 arbitrary string expansion). This allowed for information disclosure.
1024 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1025 INT_MIN/-1 -- value coerced to INT_MAX.
1031 NM/01 Workround for PCRE version dependency in version reporting
1034 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1035 This fixes portability to compilers other than gcc, notably
1036 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1038 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1039 makefiles for portability to HP-UX and POSIX correctness.
1041 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1042 Also via indented variable definition in the Makefile.
1043 (Debugging by Oliver Heesakkers).
1045 PP/02 Restore caching of spamd results with expanded spamd_address.
1046 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1048 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1049 Improves build reliability. Fix from: Frank Elsner
1051 NM/02 Fix wide character breakage in the rfc2047 coding
1052 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1054 NM/03 Allow underscore in dnslist lookups
1055 Fixes bug 1026. Patch from Graeme Fowler
1057 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1058 Code patches from Adam Ciarcinski of NetBSD.
1060 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1063 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1064 is logged, to avoid truncation. Patch from John Horne.
1066 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1067 Patch from Jakob Hirsch.
1069 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1070 SQL string expansion failure details.
1071 Patch from Andrey Oktyabrski.
1073 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1074 Patch from Simon Arlott.
1076 PP/09 New lookups functionality failed to compile on old gcc which rejects
1077 extern declarations in function scope.
1078 Patch from Oliver Fleischmann
1080 PP/10 Use sig_atomic_t for flags set from signal handlers.
1081 Check getgroups() return and improve debugging.
1082 Fixed developed for diagnosis in bug 927 (which turned out to be
1085 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1086 Patch from Mark Zealey.
1088 PP/12 Bugzilla 1056: Improved spamd server selection.
1089 Patch from Mark Zealey.
1091 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1092 Based on patch from Heiko Schlittermann.
1094 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1095 Patch from Uwe Doering, sign-off by Michael Haardt.
1097 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1098 variable declaration deep within a block. Bug and patch from
1101 PP/15 lookups-Makefile IRIX compatibilty coercion.
1103 PP/16 Make DISABLE_DKIM build knob functional.
1105 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1106 Patch by Simon Arlott
1108 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1109 variable argument macros. Our copy now differs from upstream.
1115 TF/01 Failure to get a lock on a hints database can have serious
1116 consequences so log it to the panic log.
1118 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1119 controlled using the smtp_confirmation log selector.
1121 TF/03 Include the error message when we fail to unlink a spool file.
1123 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1124 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1125 for maintaining out-of-tree patches for some time.
1127 PP/01 Bugzilla 139: Documentation and portability issues.
1128 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1129 Handle per-OS dynamic-module compilation flags.
1131 PP/02 Let /dev/null have normal permissions.
1132 The 4.73 fixes were a little too stringent and complained about the
1133 permissions on /dev/null. Exempt it from some checks.
1134 Reported by Andreas M. Kirchwitz.
1136 PP/03 Report version information for many libraries, including
1137 Exim version information for dynamically loaded libraries. Created
1138 version.h, now support a version extension string for distributors
1139 who patch heavily. Dynamic module ABI change.
1141 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1142 privilege escalation vulnerability whereby the Exim run-time user
1143 can cause root to append content of the attacker's choosing to
1146 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1149 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1150 If dropping privileges for untrusted macros, we disabled normal logging
1151 on the basis that it would fail; for the Exim run-time user, this is not
1152 the case, and it resulted in successful deliveries going unlogged.
1153 Fixed. Reported by Andreas Metzler.
1159 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1160 only prepend for the Resent-* case. Fixes regression introduced in
1161 Exim 4.70 by NM/22 for Bugzilla 607.
1163 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1164 increasing numbers of administrators be bitten by this.
1166 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1168 PP/03 Bugzilla 994: added openssl_options main configuration option.
1170 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1172 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1174 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1176 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1177 without a peer certificate, leading to a segfault because of an
1178 assumption that peers always have certificates. Be a little more
1179 paranoid. Problem reported by Martin Tscholak.
1181 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1182 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1183 NB: ClamAV planning to remove STREAM in "middle of 2010".
1184 CL also introduces -bmalware, various -d+acl logging additions and
1185 more caution in buffer sizes.
1187 PP/09 Implemented reverse_ip expansion operator.
1189 PP/10 Bugzilla 937: provide a "debug" ACL control.
1191 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1193 PP/12 Bugzilla 973: Implement --version.
1195 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1197 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1199 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1201 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1202 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1203 ignore trailing whitespace.
1205 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1207 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1210 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1211 Notification from Dr Andrew Aitchison.
1213 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1214 ExtendedDetectionInfo response format.
1215 Notification from John Horne.
1217 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1220 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1221 XSL and documented dependency on system catalogs, with examples of how
1224 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1227 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1228 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1229 configuration file which is writeable by the Exim user or group.
1231 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1232 of configuration files to cover files specified with the -C option if
1233 they are going to be used with root privileges, not just the default
1236 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1237 option (effectively making it always true).
1239 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1240 files to be used while preserving root privileges.
1242 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1243 that rogue child processes cannot use them.
1245 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1246 run-time user, instead of root.
1248 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1249 Exim run-time user without dropping privileges.
1251 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1252 result string, instead of calling string_vformat() twice with the same
1255 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1256 for other users. Others should always drop root privileges if they use
1257 -C on the command line, even for a whitelisted configure file.
1259 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1261 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1267 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1268 $data_path, and $header_path variables; fixed documentation bugs and
1271 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1272 exipick to access non-standard spools, including the "frozen" queue
1275 NM/01 Bugzilla 965: Support mysql stored procedures.
1276 Patch from Alain Williams
1278 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1280 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1281 Patch from Andreas Metzler
1283 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1284 Patch from Kirill Miazine
1286 NM/05 Bugzilla 671: Added umask to procmail example.
1288 JJ/03 installed exipick 20100323.0, fixing doc bug
1290 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1291 directory. Notification and patch from Dan Rosenberg.
1293 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1295 TK/02 Improve log output when DKIM signing operation fails.
1297 MH/01 Treat the transport option dkim_domain as a colon separated
1298 list, not as a single string, and sign the message with each element,
1299 omitting multiple occurences of the same signer.
1301 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1302 Bugzilla 985, 986. Patch by Simon Arlott
1304 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1305 Patch by Simon Arlott
1307 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1308 MBX locking. Notification from Dan Rosenberg.
1314 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1316 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1318 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1320 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1322 NM/04 Bugzilla 915: Flag broken perl installation during build.
1328 TK/01 Added patch by Johannes Berg that expands the main option
1329 "spamd_address" if it starts with a dollar sign.
1331 TK/02 Write list of recipients to X-Envelope-Sender header when building
1332 the mbox-format spool file for content scanning (suggested by Jakob
1335 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1336 (http://www.dcc-servers.net/) support via dccifd. Activated by
1337 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1339 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1340 by Mark Daniel Reidel <mr@df.eu>.
1342 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1343 When building exim an external PCRE library is now needed -
1344 PCRE is a system library on the majority of modern systems.
1345 See entry on PCRE_LIBS in EDITME file.
1347 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1348 conversation. Added nologin parameter to request.
1349 Patch contributed by Kirill Miazine.
1351 TF/01 Do not log submission mode rewrites if they do not change the address.
1353 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1355 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1356 log files in place. Contributed by Roberto Lima.
1358 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1360 TF/03 Bugzilla 615: When checking the local_parts router precondition
1361 after a local_part_suffix or local_part_prefix option, Exim now
1362 does not use the address's named list lookup cache, since this
1363 contains cached lookups for the whole local part.
1365 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1366 Robert Millan. Documentation is in experimental-spec.txt.
1368 TF/04 Bugzilla 668: Fix parallel build (make -j).
1370 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1372 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1373 Patch provided by Jan Srzednicki.
1375 TF/05 Leading white space used to be stripped from $spam_report which
1376 wrecked the formatting. Now it is preserved.
1378 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1379 that they are available at delivery time.
1381 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1383 TF/08 TLS error reporting now respects the incoming_interface and
1384 incoming_port log selectors.
1386 TF/09 Produce a more useful error message if an SMTP transport's hosts
1387 setting expands to an empty string.
1389 NM/06 Bugzilla 744: EXPN did not work under TLS.
1390 Patch provided by Phil Pennock.
1392 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1393 Patch provided by Richard Godbee.
1395 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1396 acl_smtp_notquit, added index entry.
1398 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1399 Patch provided by Eugene Bujak.
1401 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1402 accept(). Patch provided by Maxim Dounin.
1404 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1405 Patch provided by Phil Pennock.
1407 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1409 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1410 Patch provided by Brad "anomie" Jorsch.
1412 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1413 Patch provided by Dean Brooks.
1415 TK/05 Add native DKIM support (does not depend on external libraries).
1417 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1418 Patch provided by Graeme Fowler.
1420 NM/16 Bugzilla 851: Documentation example syntax fix.
1422 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1424 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1427 NM/19 Bugzilla 745: TLS version reporting.
1428 Patch provided by Phil Pennock.
1430 NM/20 Bugzilla 167: bool: condition support.
1431 Patch provided by Phil Pennock.
1433 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1434 clients. Patch provided by Phil Pennock.
1436 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1437 Patch provided by Brad "anomie" Jorsch.
1439 NM/23 Bugzilla 687: Fix misparses in eximstats.
1440 Patch provided by Heiko Schlittermann.
1442 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1443 Patch provided by Heiko Schlittermann.
1445 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1446 plus update to original patch.
1448 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1450 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1451 Patch provided by David Brownlee.
1453 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1455 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1457 NM/30 Bugzilla 888: TLS documentation bugfixes.
1459 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1461 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1462 Unlike the original bugzilla I have changed all shell scripts in src tree.
1464 NM/33 Bugzilla 898: Transport filter timeout fix.
1465 Patch by Todd Rinaldo.
1467 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1468 Patch by Serge Demonchaux.
1470 NM/35 Bugzilla 39: Base64 decode bug fixes.
1471 Patch by Jakob Hirsch.
1473 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1475 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1477 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1479 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1485 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1486 ALT-N's libdkim that I have put here:
1487 http://duncanthrax.net/exim-experimental/
1489 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1490 were called 'true' and it seems that C99 defines that as a reserved
1491 keyword to be used with 'bool' variable types. That means you could
1492 not include C99-style headers which use bools without triggering
1493 build errors in sieve.c.
1495 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1496 as mailq or other aliases. Changed the --help handling significantly
1497 to do whats expected. exim_usage() emits usage/help information.
1499 SC/01 Added the -bylocaldomain option to eximstats.
1501 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1503 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1505 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1511 PH/01 Another patch from the Sieve maintainer.
1513 PH/02 When an IPv6 address is converted to a string for single-key lookup
1514 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1515 dots are used instead of colons so that keys in lsearch files need not
1516 contain colons. This was done some time before quoting was made available
1517 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1518 (notated using the quote facility) so as to distinguish them from IPv4
1519 keys. This meant that lookups for IP addresses in host lists did not work
1520 for iplsearch lookups.
1522 This has been fixed by arranging for IPv6 addresses to be expressed with
1523 colons if the lookup type is iplsearch. This is not incompatible, because
1524 previously such lookups could never work.
1526 The situation is now rather anomolous, since one *can* have colons in
1527 ordinary lsearch keys. However, making the change in all cases is
1528 incompatible and would probably break a number of configurations.
1530 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1533 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1534 conversion specification without a maximum field width, thereby enabling
1535 a rogue spamd server to cause a buffer overflow. While nobody in their
1536 right mind would setup Exim to query an untrusted spamd server, an
1537 attacker that gains access to a server running spamd could potentially
1538 exploit this vulnerability to run arbitrary code as the Exim user.
1540 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1541 $primary_hostname instead of what libspf2 thinks the hosts name is.
1543 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1544 a directory entry by the name of the lookup key. Previously, if a
1545 symlink pointed to a non-existing file or a file in a directory that
1546 Exim lacked permissions to read, a lookup for a key matching that
1547 symlink would fail. Now it is enough that a matching directory entry
1548 exists, symlink or not. (Bugzilla 503.)
1550 PH/03 The body_linecount and body_zerocount variables are now exported in the
1553 PH/04 Added the $dnslist_matched variable.
1555 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1556 This means they are set thereafter only if the connection becomes
1559 PH/06 Added the client_condition to authenticators so that some can be skipped
1560 by clients under certain conditions.
1562 PH/07 The error message for a badly-placed control=no_multiline_responses left
1563 "_responses" off the end of the name.
1565 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1567 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1568 (without spaces) instead of just copying the configuration text.
1570 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1572 PH/11 Added $max_received_linelength.
1574 PH/12 Added +ignore_defer and +include_defer to host lists.
1576 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1577 way in which PCRE > 7.0 is built.
1579 PH/14 Implemented queue_only_load_latch.
1581 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1582 MAIL command. The effect was to mangle the value on 64-bit systems.
1584 PH/16 Another patch from the Sieve maintainer.
1586 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1588 PH/18 If a system quota error occurred while trying to create the file for
1589 a maildir delivery, the message "Mailbox is full" was not appended to the
1590 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1591 only to a quota excession during the actual writing of the file.
1593 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1594 characters?) which causes problems in log lines. The DN values are now
1595 passed through string_printing() before being added to log lines.
1597 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1598 and InterBase are left for another time.)
1600 PH/21 Added message_body_newlines option.
1602 PH/22 Guard against possible overflow in moan_check_errorcopy().
1604 PH/23 POSIX allows open() to be a macro; guard against that.
1606 PH/24 If the recipient of an error message contained an @ in the local part
1607 (suitably quoted, of course), incorrect values were put in $domain and
1608 $local_part during the evaluation of errors_copy.
1614 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1615 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1618 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1619 issue a MAIL command.
1621 PH/02 In an ACL statement such as
1623 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1625 if a client was not listed at all, or was listed with a value other than
1626 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1627 the condition was not true (as it should be), so access was not denied.
1628 The bug was that the ! inversion was incorrectly passed on to the second
1629 item. This has been fixed.
1631 PH/03 Added additional dnslists conditions == and =& which are different from
1632 = and & when the dns lookup returns more than one IP address.
1634 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1635 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1637 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1638 FSYNC, which compiles an option called disable_fsync that allows for
1639 bypassing fsync(). The documentation is heavily laced with warnings.
1641 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1643 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1644 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1645 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1646 including adding "make clean"; (3) Added -fPIC when compiling the test
1647 dynamically loaded module, to get rid of a warning.
1649 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1650 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1651 = 0s. The bug is otherwise harmless.
1653 PH/07 There was a bug in the dovecot authenticator such that the value of
1654 $auth1 could be overwritten, and so not correctly preserved, after a
1655 successful authentication. This usually meant that the value preserved by
1656 the server_setid option was incorrect.
1658 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1660 PH/09 Installed PCRE release 7.0.
1662 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1663 run for batched SMTP input. It is now run at the start of every message
1664 in the batch. While fixing this I discovered that the process information
1665 (output by running exiwhat) was not always getting set for -bs and -bS
1666 input. This is fixed, and it now also says "batched" for BSMTP.
1668 PH/11 Added control=no_pipelining.
1670 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1671 patch, slightly modified), and move the expansion of helo_data till after
1672 the connection is made in the smtp transport (so it can use these
1675 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1677 PH/14 Added log_selector = +pid.
1679 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1681 PH/16 Add ${if forany and ${if forall.
1683 PH/17 Added dsn_from option to vary the From: line in DSNs.
1685 PH/18 Flush SMTP output before performing a callout, unless control =
1686 no_callout_flush is set.
1688 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1689 was true (the default) a successful delivery failed to delete the retry
1690 item, thus causing premature timeout of the address. The bug is now
1693 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1695 PH/21 Long custom messages for fakedefer and fakereject are now split up
1696 into multiline reponses in the same way that messages for "deny" and
1697 other ACL rejections are.
1699 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1700 with slight modification.
1702 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1703 draft, changing the syntax and factoring some duplicate code".
1705 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1706 for deliveries of the second and subsequent messages over the same SMTP
1709 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1710 ${reduce, with only minor "tidies".
1712 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1714 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1715 expansion side effects.
1717 PH/27 When a message times out after an over-quota error from an Exim-imposed
1718 quota, the bounce message says "mailbox is full". This message was not
1719 being given when it was a system quota that was exceeded. It now should
1722 MH/03 Made $recipients available in local_scan(). local_scan() already has
1723 better access to the recipient list through recipients_list[], but
1724 $recipients can be useful in postmaster-provided expansion strings.
1726 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1727 in the case of a MAIL command with additional options following the
1728 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1729 were accidentally chopped off.
1731 PH/29 SMTP synchronization checks are implemented when a command is read -
1732 there is a check that no more input is waiting when there shouldn't be
1733 any. However, for some commands, a delay in an ACL can mean that it is
1734 some time before the response is written. In this time, more input might
1735 arrive, invalidly. So now there are extra checks after an ACL has run for
1736 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1737 pipelining has not been advertised.
1739 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1741 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1742 returned in the SMTP response when smtp_return_error_details was false.
1743 This has been fixed.
1745 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1746 instead of the C I/O that was originally supplied, because problems were
1747 reported on Solaris.
1749 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1750 Exim which did not show up earlier: it was assuming that a call to
1751 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1752 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1753 was a macro that became an assignment, so it seemed to work. This has
1754 changed to a proper function call with a void return, hence the compile
1755 error. Exim's code has been fixed.
1757 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1760 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1762 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1764 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1765 because it turns out to be too big or there is a timeout) there is no
1766 "Completed" line in the log. When some messages of this type were
1767 selected by exigrep, they were listed as "not completed". Others were
1768 picked up by some special patterns. I have improved the selection
1769 criteria to be more general.
1771 PH/38 The host_find_failed option in the manualroute router can now be set
1772 to "ignore", to completely ignore a host whose IP address cannot be
1773 found. If all hosts are ignored, the behaviour is controlled by the new
1774 host_all_ignored option.
1776 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1777 homing or because of multiple MX records with /mx) generated more than
1778 one IP address, and the following item turned out to be the local host,
1779 all the secondary addresses of the first item were incorrectly removed
1780 from the list, along with the local host and any following hosts (which
1781 is what is supposed to happen).
1783 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1784 whoever called Exim into the -H file. In the case of the daemon it was
1785 behaving confusingly. When first started, it used values for whoever
1786 started the daemon, but after a SIGHUP it used the Exim user (because it
1787 calls itself on a restart). I have changed the code so that it now always
1790 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1791 message are rejected with the same error (e.g. no authentication or bad
1792 sender address), and a DATA command is nevertheless sent (as can happen
1793 with PIPELINING or a stupid MUA), the error message that was given to the
1794 RCPT commands is included in the rejection of the DATA command. This is
1795 intended to be helpful for MUAs that show only the final error to their
1798 PH/42 Another patch from the Sieve maintainer.
1800 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1801 Eximstats - Fixed some broken HTML links and added missing column headers
1803 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1804 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1806 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1808 PH/43 Yet another patch from the Sieve maintainer.
1810 PH/44 I found a way to check for a TCP/IP connection going away before sending
1811 the response to the final '.' that terminates a message, but only in the
1812 case where the client has not sent further data following the '.'
1813 (unfortunately, this is allowed). However, in many cases there won't be
1814 any further data because there won't be any more messages to send. A call
1815 to select() can be used: if it shows that the input is "ready", there is
1816 either input waiting, or the socket has been closed. An attempt to read
1817 the next input character can distinguish the two cases. Previously, Exim
1818 would have sent an OK response which the client would never have see.
1819 This could lead to message repetition. This fix should cure that, at
1820 least in a lot of common cases.
1822 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1823 advertised in response to EHLO.
1829 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1830 fixed by 4.65/MH/01 (is this a record?) are fixed:
1832 (i) An empty string was always treated as zero by the numeric comparison
1833 operators. This behaviour has been restored.
1835 (ii) It is documented that the numeric comparison operators always treat
1836 their arguments as decimal numbers. This was broken in that numbers
1837 starting with 0 were being interpreted as octal.
1839 While fixing these problems I realized that there was another issue that
1840 hadn't been noticed. Values of message_size_limit (both the global option
1841 and the transport option) were treated as octal if they started with 0.
1842 The documentation was vague. These values are now always treated as
1843 decimal, and I will make that clear in the documentation.
1849 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1850 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1853 MH/01 Don't check that the operands of numeric comparison operators are
1854 integers when their expansion is in "skipping" mode (fixes bug
1855 introduced by 4.64-PH/07).
1857 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1858 child addresses, Exim now panics and dies. Previously, because the count
1859 is held in a short int, deliveries were likely to be lost. As such a
1860 large number of recipients for a single message is ridiculous
1861 (performance will be very, very poor), I have chosen to impose a limit
1862 rather than extend the field.
1868 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1869 leftover -K file (the existence of which was triggered by #402).
1870 While we were at it, introduced process PID as part of the -K
1871 filename. This should rule out race conditions when creating
1874 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1875 processing considerably. Previous code took too long for large mails,
1876 triggering a timeout which in turn triggers #401.
1878 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1879 in the DK code in transports.c. sendfile() is not really portable,
1880 hence the _LINUX specificness.
1882 TF/01 In the add_headers option to the mail command in an Exim filter,
1883 there was a bug that Exim would claim a syntax error in any
1884 header after the first one which had an odd number of characters
1887 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1888 callout verification, Exim cached a "reject" for the entire domain. This
1889 is correct for most verifications, but it is not correct for a recipient
1890 verification with use_sender or use_postmaster set, because in that case
1891 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1892 case of MAIL FROM:<> rejection from other early rejections (e.g.
1893 rejection of HELO). When verifying a recipient using a non-null MAIL
1894 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1895 Whatever the result of the callout, the value of the domain cache is
1896 left unchanged (for any other kind of callout, getting as far as trying
1897 RCPT means that the domain itself is ok).
1899 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1902 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1903 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1904 session, the daemon ignores ECONNECT errors and logs others; it now
1905 ignores EPIPE as well.
1907 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1908 (quoted-printable decoding).
1910 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1911 later the small subsequent patch to fix an introduced bug.
1913 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1915 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1917 PH/08 An error is now given if message_size_limit is specified negative.
1919 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1920 to be given (somewhat) arbitrary names.
1922 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1925 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1926 miscellaneous code fixes
1928 PH/10 Added the log_reject_target ACL modifier to specify where to log
1931 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1932 hostname. This is wrong, because it relates to the incoming message (and
1933 probably the interface on which it is arriving) and not to the outgoing
1934 callout (which could be using a different interface). This has been
1935 changed to use the value of the helo_data option from the smtp transport
1936 instead - this is what is used when a message is actually being sent. If
1937 there is no remote transport (possible with a router that sets up host
1938 addresses), $smtp_active_hostname is used.
1940 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1941 tweaks were necessary in order to get it to work (see also 21 below):
1942 (a) The code assumed that strncpy() returns a negative number on buffer
1943 overflow, which isn't the case. Replaced with Exim's string_format()
1945 (b) There were several signed/unsigned issues. I just did the minimum
1946 hacking in of casts. There is scope for a larger refactoring.
1947 (c) The code used strcasecmp() which is not a standard C function.
1948 Replaced with Exim's strcmpic() function.
1949 (d) The code set only $1; it now sets $auth1 as well.
1950 (e) A simple test gave the error "authentication client didn't specify
1951 service in request". It would seem that Dovecot has changed its
1952 interface. Fortunately there's a specification; I followed it and
1953 changed what the client sends and it appears to be working now.
1955 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1958 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1959 address A is aliased to B and C, where B exists and C does not. Without
1960 -v the output is "A verified" because verification stops after a
1961 successful redirection if more than one address is generated. However,
1962 with -v the child addresses are also verified. Exim was outputting "A
1963 failed to verify" and then showing the successful verification for C,
1964 with its parentage. It now outputs "B failed to verify", showing B's
1965 parentage before showing the successful verification of C.
1967 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1968 look up a TXT record in a specific list after matching in a combined
1971 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1972 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1973 they consult the DNS. I had assumed they would set it the way they
1974 wanted; and indeed my experiments on Linux seem to show that in some
1975 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1976 To be on the safe side, however, I have now made the interface to
1977 host_find_byname() similar to host_find_bydns(), with an argument
1978 containing the DNS resolver options. The host_find_byname() function now
1979 sets these options at its start, just as host_find_bydns() does. The smtp
1980 transport options dns_qualify_single and dns_search_parents are passed to
1981 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1982 of host_find_byname() use the default settings of RES_DEFNAMES
1983 (qualify_single) but not RES_DNSRCH (search_parents).
1985 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1986 spool_read_header() do less string testing, by means of a preliminary
1987 switch on the second character of optional "-foo" lines. (This is
1988 overdue, caused by the large number of possibilities that now exist.
1989 Originally there were few.) While I was there, I also converted the
1990 str(n)cmp tests so they don't re-test the leading "-" and the first
1991 character, in the hope this might squeeze out yet more improvement.
1993 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1994 flag allowing group syntax was set by the header_syntax check but not
1995 turned off, possible causing trouble later; (2) The flag was not being
1996 set at all for the header_verify test, causing "group"-style headers to
1997 be rejected. I have now set it in this case, and also caused header_
1998 verify to ignore an empty address taken from a group. While doing this, I
1999 came across some other cases where the code for allowing group syntax
2000 while scanning a header line wasn't quite right (mostly, not resetting
2001 the flag correctly in the right place). These bugs could have caused
2002 trouble for malformed header lines. I hope it is now all correct.
2004 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2005 with the "reply" argument non-NULL. The code, however (which originally
2006 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2007 but it didn't always do it. This confused somebody who was copying the
2008 code for some other use. I have removed all the tests.
2010 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2011 feature that was used to support insecure browsers during the U.S. crypto
2012 embargo. It requires special client support, and Exim is probably the
2013 only MTA that supported it -- and would never use it because real RSA is
2014 always available. This code has been removed, because it had the bad
2015 effect of slowing Exim down by computing (never used) parameters for the
2016 RSA_EXPORT functionality.
2018 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2019 authenticator to fail if there's a tab character in the incoming data
2020 (there should never be unless someone is messing about, as it's supposed
2021 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2022 if the connection is using TLS or if the remote IP is the same as the
2023 local IP, and the "valid-client-cert option" if a client certificate has
2026 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2027 authenticators. This can be used for authorization after authentication
2028 succeeds. (In the case of plaintext, it servers for both authentication
2031 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2032 if any retry times were supplied.
2034 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2035 connection, where there is no client IP address to check. In this
2036 situation, the verify now always succeeds.
2038 PH/25 Applied John Jetmore's -Mset patch.
2040 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2042 PH/27 In a string expansion for a processed (not raw) header when multiple
2043 headers of the same name were present, leading whitespace was being
2044 removed from all of them, but trailing whitespace was being removed only
2045 from the last one. Now trailing whitespace is removed from each header
2046 before concatenation. Completely empty headers in a concatenation (as
2047 before) are ignored.
2049 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2050 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2052 PH/29 [Removed. This was a change that I later backed out, and forgot to
2053 correct the ChangeLog entry (that I had efficiently created) before
2054 committing the later change.]
2056 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2057 address errors (4xx response to RCPT) over the same connection as other
2058 messages routed to the same hosts. Such deliveries are always "forced",
2059 so retry times are not inspected. This resulted in far too many retries
2060 for the affected addresses. The effect occurred only when there were more
2061 hosts than the hosts_max_try setting in the smtp transport when it had
2062 the 4xx errors. Those hosts that it had tried were not added to the list
2063 of hosts for which the message was waiting, so if all were tried, there
2064 was no problem. Two fixes have been applied:
2066 (i) If there are any address or message errors in an SMTP delivery, none
2067 of the hosts (tried or untried) are now added to the list of hosts
2068 for which the message is waiting, so the message should not be a
2069 candidate for sending over the same connection that was used for a
2070 successful delivery of some other message. This seems entirely
2071 reasonable: after all the message is NOT "waiting for some host".
2072 This is so "obvious" that I'm not sure why it wasn't done
2073 previously. Hope I haven't missed anything, but it can't do any
2074 harm, as the worst effect is to miss an optimization.
2076 (ii) If, despite (i), such a delivery is accidentally attempted, the
2077 routing retry time is respected, so at least it doesn't keep
2078 hammering the server.
2080 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2081 in ${readsocket because some servers need this prod.
2083 PH/32 Added some extra debug output when updating a wait-xxx database.
2085 PH/33 The hint "could be header name not terminated by colon", which has been
2086 given for certain expansion errors for a long time, was not being given
2087 for the ${if def:h_colon_omitted{... case.
2089 PH/34 The spec says: "With one important exception, whenever a domain list is
2090 being scanned, $domain contains the subject domain." There was at least
2091 one case where this was not true.
2093 PH/35 The error "getsockname() failed: connection reset by peer" was being
2094 written to the panic log as well as the main log, but it isn't really
2095 panic-worthy as it just means the connection died rather early on. I have
2096 removed the panic log writing for the ECONNRESET error when getsockname()
2099 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2100 runs only) independently of the message's sender address. This meant
2101 that, if the 4xx error was in fact related to the sender, a different
2102 message to the same recipient with a different sender could confuse
2103 things. In particualar, this can happen when sending to a greylisting
2104 server, but other circumstances could also provoke similar problems.
2105 I have changed the default so that the retry time for these errors is now
2106 based a combination of the sender and recipient addresses. This change
2107 can be overridden by setting address_retry_include_sender=false in the
2110 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2111 remote server are returned as part of bounce messages. This was not
2112 happening for LMTP over a pipe (the lmtp transport), but now it is the
2113 same for both kinds of LMTP.
2115 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2116 in header lines that were in fact CNAMEs. This is no longer the case.
2118 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2119 and queue runs started by the daemon processed all messages. This has
2120 been fixed so that -R and -S can now usefully be given with -q<time>.
2122 PH/40 Import PCRE release 6.7 (fixes some bugs).
2124 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2126 PH/42 Give an error if -q is specified more than once.
2128 PH/43 Renamed the variables $interface_address and $interface_port as
2129 $received_ip_address and $received_port, to make it clear that these
2130 values apply to message reception, and not to the outgoing interface when
2131 a message is delivered. (The old names remain recognized, of course.)
2133 PH/44 There was no timeout on the connect() call when using a Unix domain
2134 socket in the ${readsocket expansion. There now is.
2136 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2137 be meaningful with "accept".
2139 SC/01 Eximstats V1.43
2140 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2142 SC/02 Eximstats V1.44
2143 Use a glob alias rather than an array ref in the generated
2144 parser. This improves both readability and performance.
2146 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2147 Collect SpamAssassin and rejection statistics.
2148 Don't display local sender or destination tables unless
2149 there is data to show.
2150 Added average volumes into the top table text output.
2152 SC/04 Eximstats V1.46
2153 Collect data on the number of addresses (recipients)
2154 as well as the number of messages.
2156 SC/05 Eximstats V1.47
2157 Added 'Message too big' to the list of mail rejection
2158 reasons (thanks to Marco Gaiarin).
2160 SC/06 Eximstats V1.48
2161 Mainlog lines which have GMT offsets and are too short to
2162 have a flag are now skipped.
2164 SC/07 Eximstats V1.49 (Alain Williams)
2165 Added the -emptyok flag.
2167 SC/08 Eximstats V1.50
2168 Fixes for obtaining the IP address from reject messages.
2170 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2171 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2172 whitesspace changes from 4.64-PH/27
2174 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2177 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2178 are found, allow negative numbers in numeric criteria)
2180 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2182 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2183 to match changes made in 4.64-PH/43
2185 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2187 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2188 "domain" is actually an IP address, and force a failure. This locks out
2189 those revolvers/nameservers that support "A-for-A" lookups, in
2190 contravention of the specifications.
2192 PH/48 When a host name was looked up from an IP address, and the subsequent
2193 forward lookup of the name timed out, the host name was left in
2194 $sender_host_name, contrary to the specification.
2196 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2197 restricted to single-key lookups, Exim was not diagnosing an error if
2198 * or *@ was used with a query-style lookup.
2200 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2202 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2203 long ago, but noone interested enough thought of it. Let's just say that
2204 the "1.1" means that there are some new functions that weren't there at
2205 some point in the past.
2207 PH/51 Error processing for expansion failure of helo_data from an smtp
2208 transport during callout processing was broken.
2210 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2211 tested/used via the -bh/-bhc/-bs options.
2213 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2214 bug, fixed in subsequent PCRE releases).
2216 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2217 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2219 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2225 SC/01 Use a glob alias rather than an array ref in eximstats generated
2226 parser. This improves both readability and performance.
2228 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2229 Don't display local sender or destination tables in eximstats unless
2230 there is data to show.
2231 Added average volumes into the eximstats top table text output.
2233 SC/03 Collect data on the number of addresses (recipients) as well
2234 as the number of messages in eximstats.
2236 TF/01 Correct an error in the documentation for the redirect router. Exim
2237 does (usually) call initgroups() when daemonizing.
2239 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2240 with consistent privilege compared to when running as a daemon.
2242 TF/03 Note in the spec that $authenticated_id is not set for local
2243 submissions from trusted users.
2245 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2246 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2248 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2249 by adding some example configuration directives to the default
2250 configuration file. A little bit of work is required to uncomment the
2251 directives and define how usernames and passwords are checked, but
2252 there is now a framework to start from.
2254 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2255 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2256 without this. I don't know how relevant this is to other LDAP libraries.
2258 PH/02 Add the verb name to the "unknown ACL verb" error.
2260 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2262 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2264 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2265 directory not expanded when it should be if an expanded home directory
2266 was set for the address (which is overridden by the transport).
2268 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2271 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2272 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2273 because it is too late at that time, and has no effect.
2275 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2276 security issue with \' (bugzilla #107). I could not use the
2277 PQescapeStringConn() function, because it needs a PGconn value as one of
2280 PH/08 When testing addresses using -bt, indicate those final addresses that
2281 are duplicates that would not cause an additional delivery. At least one
2282 person was confused, thinking that -bt output corresponded to deliveries.
2283 (Suppressing duplicates isn't a good idea as you lose the information
2284 about possibly different redirections that led to the duplicates.)
2286 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2287 systems where poll() doesn't work, in particular OS X.
2289 PH/10 Added more information to debugging output for retry time not reached.
2291 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2292 operations in malware.c.
2294 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2297 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2298 syslog_duplication set false, log lines that would normally be written
2299 both the the main log and to the reject log were not written to syslog at
2302 PH/14 In the default configuration, change the use of "message" in ACL warn
2303 statements to "add_header".
2305 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2306 not followed by a command (e.g. "seen endif").
2308 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2309 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2312 PH/17 Added extra conditions to the default value of delay_warning_condition
2316 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2317 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2318 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2321 The Auto-Submitted: and various List- headers are standardised, whereas I
2322 don't think Precedence: ever was.
2324 PH/18 Refactored debugging code in route_finduser() to show more information,
2325 in particular, the error code if getpwnam() issues one.
2327 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2328 This is apparently needed in addition to the PH/07 change above to avoid
2329 any possible encoding problems.
2331 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2332 but not after initializing Perl.
2334 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2335 output them only if debugging. By default they are written stderr,
2336 apparently, which is not desirable.
2338 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2341 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2344 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2346 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2347 authenticated or an ident call has been made. Suppress the default
2348 values for $authenticated_id and $authenticated_sender (but permit -oMai
2349 and -oMas) when testing with -bh.
2351 PH/24 Re-jigged the order of the tests in the default configuration so that the
2352 tests for valid domains and recipients precede the DNS black list and CSA
2353 tests, on the grounds that those ones are more expensive.
2355 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2356 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2357 This bug exists in every version of Exim that I still have, right back to
2360 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2361 However, an attempt to turn on case-sensitivity in a regex key by
2362 including (?-i) didn't work because the subject string was already
2363 lowercased, and the effects were non-intuitive. It turns out that a
2364 one-line patch can be used to allow (?-i) to work as expected.
2370 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2371 other effects) broke the use of negated acl sub-conditions.
2373 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2376 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2377 "Deny" causes Exim to reject the incoming connection with a 554 error.
2378 Unfortunately, if there is a major crisis, such as a disk failure,
2379 tcp-wrappers gives "deny", whereas what one would like would be some
2380 kind of temporary error. A kludge has been added to help with this.
2381 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2382 554 error is used if errno is still zero or contains ENOENT (which occurs
2383 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2386 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2388 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2389 errors. Otherwise a message that provokes a temporary error (when other
2390 messages do not) can cause a whole host to time out.
2392 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2393 addresses were routed directly to files or pipes from a redirect router.
2394 File deliveries just didn't batch; pipe deliveries might have suffered
2397 PH/06 A failure to get a lock for a hints database would erroneously always say
2398 "Failed to get write lock", even when it was really a read lock.
2400 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2401 of 0600. This has been changed to use the value of the lockfile_mode
2402 option (which defaults to 0600).
2404 PH/08 Applied small patch from the Sieve maintainer.
2406 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2407 folder from quota calculations, a direct delivery into this folder messed
2408 up the contents of the maildirsize file. This was because the regex was
2409 used only to exclude .Trash (or whatever) when the size of the mailbox
2410 was calculated. There was no check that a delivery was happening into an
2411 excluded directory. This bug has been fixed by ignoring all quota
2412 processing for deliveries into excluded directories.
2414 PH/10 Added the maildirfolder_create_regex option to appendfile.
2420 PH/01 The code for finding all the local interface addresses on a FreeBSD
2421 system running IPv6 was broken. This may well have applied to all BSD
2422 systems, as well as to others that have similar system calls. The broken
2423 code found IPv4 interfaces correctly, but gave incorrect values for the
2424 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2425 that it would not match correctly against @[] and not recognize the IPv6
2428 PH/02 The ipliteral router was not recognizing addresses of the form user@
2429 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2431 PH/03 Added disable_ipv6.
2433 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2434 decoded form, because it is most often used to construct To: headers
2435 lines in autoreplies, and the decoded form may well be syntactically
2436 invalid. However, $reply_address has leading white space removed, and all
2437 newlines turned into spaces so that the autoreply transport does not
2440 PH/05 If group was specified without a user on a router, and no group or user
2441 was specified on a transport, the group from the router was ignored.
2443 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2444 for visible compile-time settings that can be used to change these
2445 numbers, for those that want even more. Backwards compatibility with old
2446 spool files has been maintained. However, going back to a previous Exim
2447 release will lost any variables that are in spool files.
2449 PH/07 Two small changes when running in the test harness: increase delay when
2450 passing a TCP/IP connection to a new process, in case the original
2451 process has to generate a bounce, and remove special handling of
2452 127.0.0.2 (sic), which is no longer necessary.
2454 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2455 be the same on different OS.
2457 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2460 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2461 whether --show-vars was specified or not
2463 JJ/02 exipick: Added support for new ACL variable spool format introduced
2466 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2467 syntactically invalid From: or Reply-to: line, and a filter used this to
2468 generate an autoreply, and therefore failed to obtain an address for the
2469 autoreply, Exim could try to deliver to a non-existent relative file
2470 name, causing unrelated and misleading errors. What now happens is that
2471 it logs this as a hard delivery error, but does not attempt to create a
2474 PH/11 The exinext utility has a -C option for testing purposes, but although
2475 the given file was scanned by exinext itself; it wasn't being passed on
2476 when Exim was called.
2478 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2479 an end-of-file indication when reading a command response.
2481 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2482 compiled. In many other places in Exim, IPv6 addresses are always
2483 recognized, so I have changed this. It also means that IPv4 domain
2484 literals of the form [IPV4:n.n.n.n] are now always recognized.
2486 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2487 used if the router is not running as root, for example, when verifying at
2488 ACL time, or when using -bh. The debugging output from this situation was
2489 non-existent - all you got was a failure to exec. I have made two
2492 (a) Failures to set uid/gid, the current directory, or a process leader
2493 in a subprocess such as that created by queryprogram now generate
2494 suitable debugging ouput when -d is set.
2496 (b) The queryprogram router detects when it is not running as root,
2497 outputs suitable debugging information if -d is set, and then runs
2498 the subprocess without attempting to change uid/gid.
2500 PH/15 Minor change to Makefile for building test_host (undocumented testing
2503 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2504 additional section of a DNS packet that returns MX or SRV records.
2505 Instead, it always explicitly searches for A/AAAA records. This avoids
2506 major problems that occur when a DNS server includes only records of one
2507 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2508 fixed another bug: if SRV records were looked up and the corresponding
2509 address records were *not* found in the additional section, the port
2510 values from the SRV records were lost.
2512 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2513 using the correct key (the original address) when searching the retry
2514 rules in order to find which one to use for generating the retry hint.
2516 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2517 adding the default one. Similarly, if it contains a Reply-To: header, the
2518 errors_reply_to option, if set, is not used.
2520 PH/19 When calculating a retry time, Exim used to measure the "time since
2521 failure" by looking at the "first failed" field in the retry record. Now
2522 it does not use this if it is later than than the arrival time of the
2523 message. Instead it uses the arrival time. This makes for better
2524 behaviour in cases where some deliveries succeed, thus re-setting the
2525 "first failed" field. An example is a quota failure for a huge message
2526 when small messages continue to be delivered. Without this change, the
2527 "time since failure" will always be short, possible causing more frequent
2528 delivery attempts for the huge message than are intended.
2529 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2531 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2532 $1, $2, $3) because the numerical variables can be reset during some
2533 expansion items (e.g. "match"), thereby losing the authentication data.
2535 PH/21 Make -bV show the size of off_t variables so that the test suite can
2536 decide whether to run tests for quotas > 2G.
2538 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2539 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2540 filecount value is greater than 2G or if a quota value is greater than 2G
2541 on a system where the size of off_t is not greater than 4, a panic error
2544 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2545 never match. The debug and -bh output now contains an explicit error
2546 message indicating a malformed IPv4 address or mask.
2548 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2549 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2550 PH/23 above applies.
2552 PH/25 Do not write to syslog when running in the test harness. The only
2553 occasion when this arises is a failure to open the main or panic logs
2554 (for which there is an explicit test).
2556 PH/26 Added the /no_tell option to "control=freeze".
2558 PH/27 If a host name lookup failed very early in a connection, for example, if
2559 the IP address matched host_lookup and the reverse lookup yielded a name
2560 that did not have a forward lookup, an error message of the form "no IP
2561 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2562 could be logged. Now it outputs the IP address instead of "NULL".
2564 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2565 allows the sender and the authenticated sender to be set when
2566 submitting a message from within Exim. Since child_open_exim() is
2567 documented for local_scan(), the new function should be too.
2569 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2570 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2571 results in an empty string is now treated as unset.
2573 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2575 PH/31 Added sender_verify_fail logging option.
2577 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2578 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2579 tidied the source and removed it altogether.
2581 PH/33 When a queue run was abandoned because the load average was too high, a
2582 log line was always written; now it is written only if the queue_run log
2583 selector is set. In addition, the log line for abandonment now contains
2584 information about the queue run such as the pid. This is always present
2585 in "start" and "stop" lines but was omitted from the "abandon" line.
2587 PH/34 Omit spaces between a header name and the colon in the error message that
2588 is given when verify = headers_syntax fails (if there are lots of them,
2589 the message gets confusing).
2591 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2592 names, as there are now some PTR records that contain slashes. This check
2593 is only to protect against broken name servers that fall over on strange
2594 characters, so the fact that it applies to all lookups doesn't matter.
2596 PH/36 Now that the new test suite is complete, we can remove some of the
2597 special code in Exim that was needed for the old test suite. For example,
2598 sorting DNS records because real resolvers return them in an arbitrary
2599 order. The new test suite's fake resolver always returns records in the
2602 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2603 bounces) except when queue_only is set, to avoid logging races between
2604 the different processes.
2606 PH/38 Panic-die if .include specifies a non-absolute path.
2608 PH/39 A tweak to the "H" retry rule from its user.
2610 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2611 a label. They prevented compilation on older perls.
2613 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2614 a warning to be raised on newish perls.
2616 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2617 on queue. Changes to match documented behaviour of showing count of
2618 messages matching specified criteria.
2620 PH/40 Changed the default ident timeout from 30s to 5s.
2622 PH/41 Added support for the use of login_cap features, on those BSD systems
2623 that have them, for controlling the resources used by pipe deliveries.
2625 PH/42 The content-scanning code uses fopen() to create files in which to put
2626 message data. Previously it was not paying any attention to the mode of
2627 the files. Exim runs with umask(0) because the rest of the code creates
2628 files with open(), and sets the required mode explicitly. Thus, these
2629 files were ending up world-writeable. This was not a big issue, because,
2630 being within the spool directory, they were not world-accessible. I have
2631 created a function called modefopen, which takes an additional mode
2632 argument. It sets umask(777), creates the file, chmods it to the required
2633 mode, then resets the umask. All the relevant calls to fopen() in the
2634 content scanning code have been changed to use this function.
2636 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2637 to 24 hours. This avoids potential overflow problems when processing G
2638 and H retry rules. I suspect nobody ever tinkers with this value.
2640 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2642 PH/45 When the plaintext authenticator is running as a client, the server's
2643 challenges are checked to ensure they are valid base64 strings. By
2644 default, the authentication attempt is cancelled if an invalid string is
2645 received. Setting client_ignore_invalid_base64 true ignores these errors.
2646 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2647 they are received. Thus, the responses can be made to depend on the
2648 challenges. If an invalid string is ignored, an empty string is placed in
2651 PH/46 Messages that are created by the autoreply transport now contains a
2652 References: header, in accordance with RFCs 2822 and 3834.
2654 PH/47 Added authenticated_sender_force to the smtp transport.
2656 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2658 PH/49 Installed latest patch from the Sieve maintainer.
2660 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2661 was also set, the appendfile transport was unnecessarily scanning a
2662 directory of message files (e.g. for maildir delivery) to find the count
2663 of files (along with the size), even though it did not need this
2664 information. It now does the scan only if it needs to find either the
2665 size of the count of files.
2667 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2669 PH/52 Two bugs concerned with error handling when the smtp transport is
2672 (i) Exim was not creating retry information for temporary errors given
2673 for individual recipients after the DATA command when the smtp transport
2674 was used in LMTP mode. This meant that they could be retried too
2675 frequently, and not timed out correctly.
2677 (ii) Exim was setting the flag that allows error details to be returned
2678 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2679 recipients that were returned after the DATA command.
2681 PH/53 This is related to PH/52, but is more general: for any failing address,
2682 when detailed error information was permitted to be returned to the
2683 sender, but the error was temporary, then after the final timeout, only
2684 "retry timeout exceeded" was returned. Now it returns the full error as
2685 well as "retry timeout exceeded".
2687 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2688 do this, and (what is worse) MTAs that accept it.
2690 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2691 will now be deprecated.
2693 PH/56 New os.c-cygwin from the Cygwin maintainer.
2695 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2696 formats (previously only available in exim formats via -bpr, -bpru,
2697 and -bpra. Now also available in native and exiqgrep formats)
2699 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2700 with very large, slow to parse queues
2702 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2704 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2706 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2707 responses to authentication challenges, though it was showing the
2708 challenges; (ii) I've removed the CR characters from the debug output for
2711 PH/58 Allow for the insertion of a newline as well as a space when a string
2712 is turned into more than one encoded-word during RFC 2047 encoding. The
2713 Sieve code now uses this.
2715 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2716 data_4xx, lost_connection, tls_required.
2718 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2719 message was being sent as an SMTP response.
2721 PH/61 Add -l and -k options to exicyclog.
2723 PH/62 When verifying, if an address was redirected to one new address, so that
2724 verification continued, and the new address failed or deferred after
2725 having set something in $address_data, the value of $address_data was not
2726 passed back to the ACL. This was different to the case when no
2727 redirection occurred. The value is now passed back in both cases.
2729 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2730 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2731 use login_cap.h, so on its own it isn't the distinguishing feature. The
2732 new name refers directly to the setclassresources() function.
2734 PH/65 Added configuration files for NetBSD3.
2736 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2738 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2739 is preferred over IPv4.
2741 PH/68 The bounce_return_message and bounce_return_body options were not being
2742 honoured for bounces generated during the reception of non-SMTP messages.
2743 In particular, this applied to messages rejected by the ACL. This bug has
2744 been fixed. However, if bounce_return_message is true and bounce_return_
2745 body is false, the headers that are returned for a non-SMTP message
2746 include only those that have been read before the error was detected.
2747 (In the case of an ACL rejection, they have all been read.)
2749 PH/69 The HTML version of the specification is now built in a directory called
2750 spec_html instead of spec.html, because the latter looks like a path with
2751 a MIME-type, and this confuses some software.
2753 PH/70 Catch two compiler warnings in sieve.c.
2755 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2756 function verify_get_ident() calls ip_connect() to connect a socket, but
2757 if the "connect()" function timed out, ip_connect() used to close the
2758 socket. However, verify_get_ident() also closes the socket later, and in
2759 between Exim writes to the log, which may get opened at this point. When
2760 the socket was closed in ip_connect(), the log could get the same file
2761 descriptor number as the socket. This naturally causes chaos. The fix is
2762 not to close the socket in ip_connect(); the socket should be closed by
2763 the function that creates it. There was only one place in the code where
2764 this was missing, in the iplookup router, which I don't think anybody now
2765 uses, but I've fixed it anyway.
2767 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2768 well as to direct DNS lookups. Otherwise the handling of names in host
2769 lists is inconsistent and therefore confusing.
2775 PH/01 Two changes to the default runtime configuration:
2777 (1) Move the checks for relay_from_hosts and authenticated clients from
2778 after to before the (commented out) DNS black list checks.
2780 (2) Add control=submission to the relay_from_hosts and authenticated
2781 clients checks, on the grounds that messages accepted by these
2782 statements are most likely to be submissions.
2784 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2786 (1) Generate an error if the third argument for the ${prvs expansion is
2789 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2792 (3) Reset the variables that are obtained from the first argument of
2793 ${prvscheck and used in the second argument before leaving the code,
2794 because their memory is reclaimed, so using them afterwards may do
2797 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2798 one (it's much easier than Tom thought :-).
2800 (5) Because of (4), we can now allow for the use of $prvscheck_result
2801 inside the third argument.
2803 PH/03 For some reason, the default setting of PATH when running a command from
2804 a pipe transport was just "/usr/bin". I have changed it to
2807 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2808 anything to be listed in the output from -bV.
2810 PH/05 When a filter generated an autoreply, the entire To: header line was
2811 quoted in the delivery log line, like this:
2813 => >A.N.Other <ano@some.domain> <original@ddress> ...
2815 This has been changed so that it extracts the operative address. There
2816 may be more than one such address. If so, they are comma-separated, like
2819 => >ano@some.domain,ona@other.domain <original@ddress> ...
2821 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2822 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2823 not being looked up in the rDNS to get a host name, Exim was showing the
2824 IP address twice in Received: lines, even though the IP addresses were
2825 identical. For example:
2827 Received: from [1.2.3.4] (helo=[1.2.3.4])
2829 However, if the real host name was known, it was omitting the HELO data
2830 if it matched the actual IP address. This has been tidied up so that it
2831 doesn't show the same IP address twice.
2833 PH/07 When both +timestamp and +memory debugging was on, the value given by
2834 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2835 called by the debug printing, thereby overwriting the timestamp buffer.
2836 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2838 PH/08 When the original message was included in an autoreply transport, it
2839 always said "this is a copy of the message, including all the headers",
2840 even if body_only or headers_only was set. It now gives an appropriate
2843 PH/09 Applied a patch from the Sieve maintainer which:
2845 o fixes some comments
2846 o adds the (disabled) notify extension core
2847 o adds some debug output for the result of if/elsif tests
2848 o points to the current vacation draft in the documentation
2849 and documents the missing references header update
2853 o fixes a bug in processing the envelope test (when testing
2854 multiple envelope elements, the last element determinted the
2857 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2858 Electronic Mail") by including:
2860 Auto-submitted: auto-generated
2862 in the messages that it generates (bounce messages and others, such as
2863 warnings). In the case of bounce messages for non-SMTP mesages, there was
2864 also a typo: it was using "Auto_submitted" (underscore instead of
2865 hyphen). Since every message generated by Exim is necessarily in response
2866 to another message, thes have all been changed to:
2868 Auto-Submitted: auto-replied
2870 in accordance with these statements in the RFC:
2872 The auto-replied keyword:
2874 - SHOULD be used on messages sent in direct response to another
2875 message by an automatic process,
2877 - MUST NOT be used on manually-generated messages,
2879 - MAY be used on Delivery Status Notifications (DSNs) and Message
2880 Disposition Notifications (MDNs),
2882 - MUST NOT be used on messages generated by automatic or periodic
2883 processes, except for messages which are automatic responses to
2886 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2887 to the default Received: header definition.
2889 PH/12 Added log selector acl_warn_skipped (default on).
2891 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2892 variables because (a) they are in the wrong storage pool and (b) even if
2893 they were copied, it wouldn't work properly because of the caching.
2895 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2896 checking when decoding. Apparently there are clients that generate
2897 overlong encoded strings. Why am I not surprised?
2899 PH/15 If the first argument of "${if match_address" was not empty, but did not
2900 contain an "@" character, Exim crashed. Now it writes a panic log message
2901 and treats the condition as false.
2903 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2905 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2906 extension "envelope-auth". The code is finished and in agreement with
2907 other implementations, but there is no documentation so far and in fact,
2908 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2909 not changing the active code.
2911 Print executed "if" and "elsif" statements when debugging is used. This
2912 helps a great deal to understand what a filter does.
2914 Document more things not specified clearly in RFC3028. I had all this
2915 sorted out, when out of a sudden new issues came to my mind. Oops."
2917 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2920 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2921 canonical form (as documented). However, after a host name lookup from
2922 the IP address, check_host() was doing a simple string comparison with
2923 addresses acquired from the DNS when checking that the found name did
2924 have the original IP as one of its addresses. Since any found IPv6
2925 addresses are likely to be in abbreviated form, the comparison could
2926 fail. Luckily, there already exists a function for doing the comparison
2927 by converting both addresses to binary, so now that is used instead of
2928 the text comparison.
2930 PH/20 There was another similar case to PH/19, when a complete host name was
2931 given in a host list; looking up its IP address could give an abbreviated
2932 form, whereas the current host's name might or might not be abbreviated.
2933 The same fix has been applied.
2939 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2940 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2943 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2944 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2946 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2948 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2949 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2950 Also, for "Auto-Submitted", treat anything other than "no" as
2951 non-personal, in accordance with RFC 3834. (Previously it treated
2952 anything starting "auto-" as non-personal.)
2954 TF/01 The control=submission/name=... option had a problem with syntax
2955 errors if the name included a slash character. The /name= option
2956 now slurps the rest of the string, so it can include any characters
2957 but it must come last in the list of options (after /sender_retain
2960 PH/05 Some modifications to the interface to the fake nameserver for the new
2968 TK/01 Added the "success_on_redirect" address verification option. See
2969 NewStuff for rationale and an example.
2971 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2973 PH/02 Patch to exigrep to allow it to work on syslog lines.
2975 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2976 fread() to skip over the body file's header line, because in Cygwin the
2977 header line is locked and is inaccessible.
2979 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2980 co-exist for some time) to make it clear that it is the Exim ID that is
2981 referenced, not the Message-ID: header line.
2983 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2984 string_format() function, because snprintf() does not exist on all
2987 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2988 ${stat: expansion item.
2990 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2991 protocol synchronization error", to keep the pedants happy.
2993 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2994 well as for IRIX systems, when gcc is being used. See the host.c source
2997 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2999 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3002 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3003 transport and to the smtp transport in LMTP mode.
3005 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3007 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3008 run a filter in a subprocess. This could lead to confusion in subsequent
3009 lookups in the parent process. There should also be a search_tidyup() at
3010 the end of the subprocess.
3012 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3013 only if the host matched helo_try_verify_hosts, which caused the
3014 verification to occur when the EHLO/HELO command was issued. The ACL just
3015 tested the remembered result. Now, if a previous verification attempt has
3016 not happened, "verify = helo" does it there and then.
3018 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3020 TK/03 Fix log output including CR from clamd.
3022 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3023 exist provoked a memory error which could cause a segfault.
3025 PH/15 Installed PCRE 6.2
3027 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3029 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3030 of the problem. Specifically, suggested +O2 rather than +O1 for the
3033 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3035 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3036 record, the port was not correctly logged when the outgoing_port log
3037 selector was set (it logged the transort's default port).
3039 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3040 fallback_hosts, and "hosts" in the smtp transport.
3042 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3043 host errors such as "Connection refused".
3045 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3046 authentication with radiusclient 0.4.9:
3048 - Error returned from rc_read_config was caught wrongly
3049 - Username/password not passed on to radius server due to wrong length.
3051 The presumption is that some radiusclient API changes for 4.51/PH/17
3052 were not taken care of correctly. The code is still untested by me (my
3053 Linux distribution still has 0.3.2 of radiusclient), but it was
3054 contributed by a Radius user.
3056 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3057 expanding the "port" option of the smtp transport.
3059 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3060 while reading a MIME header. Thanks to Tom Hughes for a patch.
3062 PH/24 Include config.h inside local_scan.h so that configuration settings are
3065 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3066 that in an ACL for RCPT (for example), you can examine exactly what was
3069 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3070 commands, but it was not correctly comparing the address with the actual
3071 client host address. Thus, it would show the EHLO address in Received:
3072 header lines when this was not necessary.
3074 PH/27 Added the % operator to ${eval:}.
3076 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3077 it should be ignoring failures (because with -C, for example, it has lost
3078 privilege). It wasn't ignoring creation failures other than "already
3081 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3084 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3085 because an input line was too long, either on its own, or by virtue of
3086 too many continuations, the temporary file was not being removed, and the
3087 return code was incorrect.
3089 PH/31 Missing "BOOL" in function definition in filtertest.c.
3091 PH/32 Applied Sieve patches from the maintainer.
3093 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3095 PH/33 Added "verify = not_blind".
3097 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3098 Local/Makefile (with some defaults set). These are used in built scripts
3099 such as exicyclog, but they have never been used in the exim_install
3100 script (though there are many overriding facilities there). I have
3101 arranged that the exim_install script now takes note of these two
3104 PH/35 Installed configuration files for Dragonfly.
3106 PH/36 When a locally submitted message by a trusted user did not contain a
3107 From: header, and the sender address was obtained from -f or from an SMTP
3108 MAIL command, and the trusted user did not use -F to supply a sender
3109 name, $originator_name was incorrectly used when constructing a From:
3110 header. Furthermore, $originator_name was used for submission mode
3111 messages from external hosts without From: headers in a similar way,
3112 which is clearly wrong.
3114 PH/37 Added control=suppress_local_fixups.
3116 PH/38 When log_selector = +received_sender was set, and the addition of the
3117 sender made the log line's construction buffer exactly full, or one byte
3118 less than full, an overflow happened when the terminating "\n" was
3121 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3122 when the result of a list match is failure because a DNS lookup failed.
3124 PH/40 RM_COMMAND is now used in the building process.
3126 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3127 the "build-* directories that it finds.
3129 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3130 address in a domain literal was a prefix of an interface address.
3132 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3133 when verifying a sender address, unless rewrite_headers is false.
3135 PH/44 Wrote a long comment about why errors_to addresses are verified as
3136 recipients, not senders.
3138 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3139 the ratelimit ACL was added.
3141 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3143 PH/46 Added extra information about PostgreSQL errors to the error string.
3145 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3146 avoiding the need to install special zones in a real server. This is
3147 backwards compatible; if it can't find the fake resolver, it drops back.
3148 Thus, both old and new test suites can be run.
3150 TF/02 Added util/ratelimit.pl
3152 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3153 clock is set back in time.
3155 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3156 Candler <B.Candler@pobox.com>.
3158 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3159 OK for addresses that are the result of redirections.
3161 PH/48 A number of further additions for the benefit of the new test suite,
3162 including a fake gethostbyname() that interfaces to the fake DNS resolver
3165 TF/06 The fix for widen_domains has also been applied to qualify_single and
3166 search_parents which are the other dnslookup options that can cause
3169 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3172 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3174 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3175 service" if it encountered an MX record with an empty target hostname.
3176 The message is now "an MX or SRV record indicated no SMTP service".
3178 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3179 if the DNS of the sending site is misconfigured. This is quite a
3180 common situation. This change restores the behaviour of treating a
3181 helo verification defer as a failure.
3183 PH/51 If self=fail was set on a router, the bounce message did not include the
3184 actual error message.
3190 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3192 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3193 command itself ended in error, the underlying message about the transport
3194 filter timeout was being overwritten with the pipe command error. Now the
3195 underlying error message should be appended to the second error message.
3197 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3199 PH/02 Reduce the amount of output that "make" produces by default. Full output
3200 can still be requested.
3202 PH/03 The warning log line about a condition test deferring for a "warn" verb
3203 was being output only once per connection, rather than after each
3204 occurrence (because it was using the same function as for successful
3205 "warn" verbs). This seems wrong, so I have changed it.
3207 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3208 it should not have, which might have caused a crash in the right
3209 circumstances, but probably never did.
3211 PH/04 Installed a modified version of Tony Finch's patch to make submission
3212 mode fix the return path as well as the Sender: header line, and to
3213 add a /name= option so that you can make the user's friendly name appear
3216 TF/03 Added the control = fakedefer ACL modifier.
3218 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3219 Mark Lowes for thorough testing.
3221 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3223 TK/03 Merged latest SRS patch from Miles Wilton.
3225 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3226 with the definition in sysexits.h (which is #included earlier).
3227 Fortunately, Exim does not actually use EX_OK. The code used to try to
3228 preserve the sysexits.h value, by assumimg that macro definitions were
3229 scanned for macro replacements. I have been disabused of this notion,
3230 so now the code just undefines EX_OK before #including unistd.h.
3232 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3233 in the smtp transport. When a block could not be written in a single
3234 write() function, the timeout was being re-applied to each part-write.
3235 This seems wrong - if the receiver was accepting one byte at a time it
3236 would take for ever. The timeout is now adjusted when this happens. It
3237 doesn't have to be particularly precise.
3239 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3240 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3242 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3243 without a domain if the check to <postmaster@domain> fails.
3245 SC/01 Eximstats: added -xls and the ability to specify output files
3246 (patch written by Frank Heydlauf).
3248 SC/02 Eximstats: use FileHandles for outputing results.
3250 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3252 SC/04 Eximstats: fixed display of large numbers with -nvr option
3254 SC/05 Eximstats: fixed merging of reports with empty tables.
3256 SC/06 Eximstats: added the -include_original_destination flag
3258 SC/07 Eximstats: removed tabs and trailing whitespace.
3260 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3262 TK/06 MBOX spool code: Add real "From " MBOX separator line
3263 so the .eml file is really in mbox format (even though
3264 most programs do not really care). Patch from Alex Miller.
3266 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3267 The latter is generated from $received_to and is only set if the
3268 message has one envelope recipient. SA can use these headers,
3269 obviously out-of-the-box. Patch from Alex Miller.
3271 PH/08 The ${def test on a variable was returning false if the variable's
3272 value was "0", contrary to what the specification has always said!
3273 The result should be true unless the variable is empty.
3275 PH/09 The syntax error of a character other than { following "${if
3276 def:variable_name" (after optional whitespace) was not being diagnosed.
3277 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3278 accidental colon was present, for example, could give incorrect results.
3280 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3281 result is used (not including appendfile, where other changes are about
3284 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3285 This involved changing a lot of size variables from int to off_t. It
3286 should work with maildirs and everything.
3288 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3289 spamd dying while we are connected to it.
3291 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3294 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3295 function for BDB 4.3.
3297 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3299 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3300 the sources for PCRE have been re-arranged and more files are now
3303 PH/15 The code I had for printing potentially long long variables in PH/11
3304 above was not the best (it lost precision). The length of off_t variables
3305 is now inspected at build time, and an appropriate printing format (%ld
3306 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3307 to be "long long int" or "long int". This is needed for the internal
3308 formatting function string_vformat().
3310 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3311 the configuration file to be ":syslog", then the script "guesses" where
3312 the logs files are, rather than using the compiled in default. In our
3313 case the guess is not the same as the compiled default, so the script
3314 suddenly stopped working when I started to use syslog. The patch checks
3315 to see if log_file_path is "". If so, it attempts to read it from exim
3316 with no configuration file to get the compiled in version, before it
3317 falls back to the previous guessing code."
3319 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3320 implementing BATV in an Exim configuration. See NewStuff for the gory
3323 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3324 Makefile that are specific to HP-UX.
3326 PH/18 If the "use_postmaster" option was set for a recipient callout together
3327 with the "random" option, the postmaster address was used as the MAIL
3328 FROM address for the random test, but not for the subsequent recipient
3329 test. It is now used for both.
3331 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3332 patch removes a few documentation additions to RFC 3028, because the
3333 latest draft now contains them. It adds the new en;ascii-case comparator
3334 and a new error check for 8bit text in MIME parts. Comparator and
3335 require names are now matched exactly. I enabled the subaddress
3336 extension, but it is not well tested yet (read: it works for me)."
3338 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3339 rework some of the code of TK/09 above to avoid the hardwired use of
3340 "%lld" and "long long". Replaced the call to snprintf() with a call to
3343 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3344 records point to non-existent hosts", "retry timeout exceeded", and
3345 "retry time not reached for any host after a long failure period".
3347 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3348 experimental DomainKeys support:
3350 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3351 (2) On an error such as an illegally used "control", the wrong name for
3352 the control was given.
3354 These problems did NOT occur unless DomainKeys support was compiled.
3356 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3358 PH/24 Added ${if match_ip condition.
3360 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3361 databases so that it will be absolutely obvious if a crash occurs in the
3362 DB library. This is a regular occurrence (often caused by mis-matched
3365 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3366 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3367 picked up on a user's system that detects such things. There doesn't seem
3368 to be a gcc warning option for this - only an attribute that has to be
3369 put on the function's prototype. It seems that in Fedora Core 4 they have
3370 set this on a number of new functions. No doubt there will be more in due
3373 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3374 specify a transport. However, if an address that was verified by such a
3375 router was the subject of a callout, Exim crashed because it tried to
3376 read the rcpt_include_affixes from the non-existent transport. Now it
3377 just assumes that the setting of that option is false. This bug was
3378 introduced by 4.51/PH/31.
3380 PH/28 Changed -d+all to exclude +memory, because that information is very
3381 rarely of interest, but it makes the output a lot bigger. People tend to
3382 do -d+all out of habit.
3384 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3385 code in os-type was giving problems when libc.so lives in lib64, like on
3388 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3389 aren't the modern standard, and it seems that some systems' include files
3390 don't always have them. Exim was already checking for some of the newer
3391 ones like T_AAAA, and defining it itself. I've added checks for all the
3392 record types that Exim uses.
3394 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3395 not automatically generating a new one, as it is supposed to. This
3396 prevented TLS from working. If the file did exist, but contained invalid
3397 data, a new version was generated, as expected. It was only the case of a
3398 non-existent file that was broken.
3400 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3401 with a change in libdomainkeys > 0.64.
3403 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3404 from DNS. If the selector record carries the flag, it now has
3405 precedence over the domain-wide flag.
3407 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3409 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3410 the use of a port name that isn't defined in /etc/services) occurred, the
3411 message was deferred as in a normal delivery, and thus remained on the
3412 spool, instead of being failed because of the mua_wrapper setting. This
3413 is now fixed, and I tidied up some of the mua_wrapper messages at the
3416 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3417 the messages in a hash of arrays rather than using individual hashes.
3418 This is a bit cleaner and results in dramatic memory savings, albeit
3419 at a slight CPU cost.
3421 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3422 as requested by Marc Sherman.
3424 SC/10 Eximstats: added histograms for user specified patterns as requested
3427 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3429 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3430 fopen() in the content-scanning modules that did not already have it.
3436 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3437 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3439 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3441 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3443 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3444 to test Sieve filters that use "vacation".
3446 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3447 that changes the way the GnuTLS parameters are stored in the cache file.
3448 The new format can be generated externally. For backward compatibility,
3449 if the data in the cache doesn't make sense, Exim assumes it has read an
3450 old-format file, and it generates new data and writes a new file. This
3451 means that you can't go back to an older release without removing the
3454 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3455 work if there are any delivery delays because "one_time" forces the
3456 parent to be marked "delivered", so its unseen clone is never tried
3457 again. For this reason, Exim now forbids the simultaneous setting of
3460 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3461 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3462 entry if you want to know the details. The fix, however, neglected to
3463 consider the case when local delivery batching is involved. The test for
3464 "previously delivered" was not happening when checking to see if an
3465 address could be batched with a previous (undelivered) one; under
3466 certain circumstances this could lead to multiple deliveries to the same
3469 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3470 in its include files, and this causes problems building Exim.
3472 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3473 header_syntax) but Exim was just ignoring anything given after a slash.
3474 In particular, this caused confusion with an attempt to use "verify =
3475 reverse_host_lookup/defer_ok". An error is now given when options are
3476 supplied for verify items that do not have them. (Maybe reverse_host_
3477 lookup should have a defer_ok option, but that's a different point.)
3479 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3480 defined by RFC 821) to 2048, because there were problems with some AUTH
3481 commands, and RFC 1869 says the size should be increased for extended
3482 SMTP commands that take arguments.
3484 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3487 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3488 "unknown" error; now it says that the functionality isn't in the binary.
3490 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3491 an address' error message when a string expansion fails (syntax or
3492 whatever). Otherwise the password may appear in the log. Following change
3493 PH/42 below, there is no longer a chance of it appearing in a bounce
3496 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3498 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3499 of its addresses was ever tried. (Bugzilla bug #2.)
3501 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3502 the result incorrectly in the debug output. (It correctly added a newline
3503 to what was transported.)
3505 TF/01 Added $received_time.
3507 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3508 commented out examples of how to interface to a virus scanner and to
3509 SpamAssassin. Also added commented examples of av_scanner and
3510 spamd_address settings.
3512 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3513 and controls are allowed in which ACLs. There were a couple of minor
3514 errors. Some of the entries in the conditions table (which is a table of
3515 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3516 negation of where the condition IS allowed.
3518 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3520 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3521 header file does not have a version number, so I've had to invent a new
3522 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3523 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3524 radiusclient), but it was contributed by a Radius user.
3526 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3527 files or directories, for OpenSSL.
3529 PH/19 When an Exim process that is running as root has to create an Exim log
3530 file, it does so in a subprocess that runs as exim:exim so as to get the
3531 ownership right at creation (otherwise, other Exim processes might see
3532 the file with the wrong ownership). There was no test for failure of this
3533 fork() call, which would lead to the process getting stuck as it waited
3534 for a non-existent subprocess. Forks do occasionally fail when resources
3535 run out. I reviewed all the other calls to fork(); they all seem to check
3538 PH/20 When checking for unexpected SMTP input at connect time (before writing
3539 the banner), Exim was not dealing correctly with a non-positive return
3540 from the read() function. If the client had disconnected by this time,
3541 the result was a log entry for a synchronization error with an empty
3542 string after "input=" when read() returned zero. If read() returned -1
3543 (an event I could not check), uninitialized data bytes were printed.
3544 There were reports of junk text (parts of files, etc) appearing after
3547 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3549 PH/22 Added support for macro redefinition, and (re)definition in between
3550 driver and ACL definitions.
3552 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3553 forgetting to use the resulting value; it was using the unexpanded value.
3555 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3556 hadn't been configured. The fix is from Juergen Kreileder, who
3557 understands it better than I do:
3559 "Here's what I see happening with three configured cyrus_sasl
3560 authenticators configured (plain, login, cram-md5):
3562 On startup auth_cyrus_sasl_init() gets called for each of these.
3563 This means three calls to sasl_listmech() without a specified mech_list.
3564 => SASL tests which mechs of all available mechs actually work
3565 => three warnings about OTP not working
3566 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3568 With the patch, sasl_listmech() also gets called three times. But now
3569 SASL's mech_list option is set to the server_mech specified in the the
3570 authenticator. Or in other words, the answer from sasl_listmech()
3571 gets limited to just the mech you're testing for (which is different
3573 => the return list contains just 'plain' or 'login', 'cram-md5' or
3574 nothing depending on the value of ob->server_mech.
3576 I've just tested the patch: Authentication still works fine,
3577 unavailable mechs specified in the exim configuration are still
3578 caught, and the auth.log warnings about OTP are gone."
3580 PH/25 When debugging is enabled, the contents of the command line are added
3581 to the debugging output, even when log_selector=+arguments is not
3584 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3585 answer is "GNU", and only if the return is "GNU/something" is the answer
3588 PH/27 $acl_verify_message is now set immediately after the failure of a
3589 verification in an ACL, and so is available in subsequent modifiers. In
3590 particular, the message can be preserved by coding like this:
3592 warn !verify = sender
3593 set acl_m0 = $acl_verify_message
3595 Previously, $acl_verify_message was set only while expanding "message"
3596 and "log_message" when a very denied access.
3598 PH/28 Modified OS/os.c-Linux with
3600 -#ifndef OS_LOAD_AVERAGE
3601 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3603 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3604 nomenclature these days.)
3606 PH/29 Installed patch from the Sieve maintainer that adds the options
3607 sieve_useraddress and sieve_subaddress to the redirect router.
3609 PH/30 In these circumstances:
3610 . Two addresses routed to the same list of hosts;
3611 . First host does not offer TLS;
3612 . First host accepts first address;
3613 . First host gives temporary error to second address;
3614 . Second host offers TLS and a TLS session is established;
3615 . Second host accepts second address.
3616 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3617 and peerdn, if requested) that were in fact used only for the second
3620 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3621 attention to any local part prefix or suffix that was matched by the
3622 router that accepted the address. It now behaves in the same way as it
3623 does for delivery: the affixes are removed from the local part unless
3624 rcpt_include_affixes is set on the transport.
3626 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3627 timeout during the DATA phase of an incoming message.
3629 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3630 applied a patch sanctioned by the Sieve maintainer.
3632 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3633 the uid or gid is negative. A case of a negative gid caused this to be
3634 noticed. The fix allows for either to be negative.
3636 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3637 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3640 PH/36 The change PH/12 above was broken. Fixed it.
3642 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3643 the grounds that routing the same address twice would always produce the
3644 same answer. This might have been true once, but it is certainly no
3645 longer true now. Routing a child address may depend on the previous
3646 routing that produced that child. Some complicated redirection strategies
3647 went wrong when messages had multiple recipients, and made Exim's
3648 behaviour dependent on the order in which the addresses were given.
3650 I have moved the duplicate checking until after the routing is complete.
3651 Exim scans the addresses that are assigned to local and remote
3652 transports, and removes any duplicates. This means that more work will be
3653 done, as duplicates will always all be routed, but duplicates are
3654 presumably rare, so I don't expect this is of any significance.
3656 For deliveries to pipes, files, and autoreplies, the duplicate checking
3657 still happens during the routing process, since they are not going to be
3660 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3661 It corrects a timeout issue with spamd. This is Ian's comment: "The
3662 background is that sometimes spamd either never reads data from a
3663 connection it has accepted, or it never writes response data. The exiscan
3664 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3665 blindly assumes that writes won't block so it may never time out."
3667 PH/39 Allow G after quota size as well as K and M.
3669 PH/40 The value set for $authenticated_id in an authenticator may not contain
3670 binary zeroes or newlines because the value is written to log lines and
3671 to spool files. There was no check on this. Now the value is run through
3672 the string_printing() function so that such characters are converted to
3673 printable escape sequences.
3675 PH/41 $message_linecount is a new variable that contains the total number of
3676 lines in the message. Compare $body_linecount, which is the count for the
3679 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3680 bounce and delay warning messages, except in certain special cases, which
3683 (a) An SMTP error message from a remote host;
3684 (b) A message specified in a :fail: redirection;
3685 (c) A message specified in a "fail" command in a system filter;
3686 (d) A message specified in a FAIL return from the queryprogram router;
3687 (e) A message specified by the cannot_route_message router option.
3689 In these cases only, Exim does include the error details in bounce and
3690 warning messages. There are also a few cases where bland messages such
3691 as "unrouteable address" or "local delivery error" are given.
3693 PH/43 $value is now also set for the "else" part of a ${run expansion.
3695 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3696 being worked on, but at least Exim now implements the latest version to
3699 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3700 process to complete was treated as a delivery failure, a timeout while
3701 writing the message to the pipe was logged, but erroneously treated as a
3702 successful delivery. Such timeouts include transport filter timeouts. For
3703 consistency with the overall process timeout, these timeouts are now
3704 treated as errors, giving rise to delivery failures by default. However,
3705 there is now a new Boolean option for the pipe transport called
3706 timeout_defer, which, if set TRUE, converts the failures into defers for
3707 both kinds of timeout. A transport filter timeout is now identified in
3710 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3711 systems where "make" and "gmake" are different, calling "gmake" at top
3712 level broke things. I've arranged for the value of $(MAKE) to be passed
3713 from the Makefile to this script so that it can call the same version of
3717 A note about Exim versions 4.44 and 4.50
3718 ----------------------------------------
3720 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3721 changes of various kinds. As a consequence, a big documentation update was
3722 needed. This delayed the release for rather longer than seemed good, especially
3723 in the light of a couple of (minor) security issues. Therefore, the changes
3724 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3725 release. So 4.44 and 4.50 are in effect two different branches that both start
3728 I have left the 4.50 change log unchanged; it contains all the changes since
3729 4.43. The change log for 4.44 is below; many of its items are identical to
3730 those for 4.50. This seems to be the most sensible way to preserve the
3731 historical information.
3737 1. Minor wording change to the doc/README.SIEVE file.
3739 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3740 computation of the current number of files was incorrect.
3742 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3743 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3746 4. Give more explanation in the error message when the command for a transport
3747 filter fails to execute.
3749 5. There are several places where Exim runs a non-Exim command in a
3750 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3751 was being done only for the command run by the queryprogram router. It is
3752 now done for all such subprocesses. The other cases are: ${run, transport
3753 filters, and the commands run by the lmtp and pipe transports.
3755 6. Added CONFIGURE_GROUP build-time option.
3757 7. Some older OS have a limit of 256 on the maximum number of file
3758 descriptors. Exim was using setrlimit() to set 1000 as a large value
3759 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3760 systems. I've change it so that if it can't get 1000, it tries for 256.
3762 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3763 was an oversight, and furthermore, ever since the addition of extra
3764 controls (e.g. 4.43/32), the checks on when to allow different forms of
3765 "control" were broken. There should now be diagnostics for all cases when a
3766 control that does not make sense is encountered.
3768 9. Added the /retain_sender option to "control=submission".
3770 10. $recipients is now available in the predata ACL (oversight).
3772 11. Tidy the search cache before the fork to do a delivery from a message
3773 received from the command line. Otherwise the child will trigger a lookup
3774 failure and thereby defer the delivery if it tries to use (for example) a
3775 cached ldap connection that the parent has called unbind on.
3777 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3778 of $address_data from the recipient verification was clobbered by the
3779 sender verification.
3781 13. The value of address_data from a sender verification is now available in
3782 $sender_address_data in subsequent conditions in the ACL statement.
3784 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3786 15. Added a new option "connect=<time>" to callout options, to set a different
3789 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3790 was its contents. (It was OK if the option was not defined at all.)
3792 17. A "Completed" log line is now written for messages that are removed from
3793 the spool by the -Mrm option.
3795 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3796 information about exactly what failed.
3798 19. Added -dd to debug only the daemon process.
3800 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3801 handles timeouts, both on the server side and network timeouts. Renamed the
3802 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3804 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3805 It is now set to "smtps".
3807 22. $host_address is now set to the target address during the checking of
3808 ignore_target_hosts.
3810 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3811 being passed; this would have caused $sender_host_name to have been used if
3812 matching the list had actually called for a host name (not very likely,
3813 since this list is usually IP addresses). A host name is now passed as
3816 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3817 code that specifies a non-restarting handler (typically sigaction() in
3818 modern systems) in an attempt to fix a rare and obscure crash bug.
3820 25. Narrowed the window for a race in the daemon that could cause it to ignore
3821 SIGCHLD signals. This is not a major problem, because they are used only to
3822 wake it up if nothing else does.
3824 26. A malformed maildirsize file could cause Exim to calculate negative values
3825 for the mailbox size or file count. Odd effects could occur as a result.
3826 The maildirsize information is now recalculated if the size or filecount
3829 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3830 support for a long time. Removed HAVE_SYS_VFS_H.
3832 28. Installed the latest version of exipick from John Jetmore.
3834 29. In an address list, if the pattern was not a regular expression, an empty
3835 subject address (from a bounce message) matched only if the pattern was an
3836 empty string. Non-empty patterns were not even tested. This was the wrong
3837 because it is perfectly reasonable to use an empty address as part of a
3838 database query. An empty address is now tested by patterns that are
3839 lookups. However, all the other forms of pattern expect the subject to
3840 contain a local part and a domain, and therefore, for them, an empty
3841 address still always fails if the pattern is not itself empty.
3843 30. Exim went into a mad DNS loop when attempting to do a callout where the
3844 host was specified on an smtp transport, and looking it up yielded more
3845 than one IP address.
3847 31. Re-factored the code for checking spool and log partition space into a
3848 function that finds that data and another that does the check. The former
3849 is then used to implement four new variables: $spool_space, $log_space,
3850 $spool_inodes, and $log_inodes.
3852 32. The RFC2047 encoding function was originally intended for short strings
3853 such as real names; it was not keeping to the 75-character limit for
3854 encoded words that the RFC imposes. It now respects the limit, and
3855 generates multiple encoded words if necessary. To be on the safe side, I
3856 have increased the buffer size for the ${rfc2047: expansion operator from
3859 33. It is now permitted to omit both strings after an "if" condition; if the
3860 condition is true, the result is "true". As before, when the second string
3861 is omitted, a false condition yields an empty string. This makes it less
3862 cumbersome to write custom ACL and router conditions.
3864 34. Failure to deliver a bounce message always caused it to be frozen, even if
3865 there was an errors_to setting on the router. The errors_to setting is now
3868 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3869 canonical form (fully expanded) before being placed in
3870 $sender_host_address.
3872 36. The table in the code that translates DNS record types into text (T_A to
3873 "A" for instance) was missing entries for NS and CNAME. It is just possible
3874 that this could have caused confusion if both these types were looked up
3875 for the same domain, because the text type is used as part of Exim's
3876 per-process caching. But the chance of anyone hitting this buglet seems
3879 37. The dnsdb lookup has been extended in a number of ways.
3881 (1) There is a new type, "zns", which walks up the domain tree until it
3882 finds some nameserver records. It should be used with care.
3884 (2) There is a new type, "mxh", which is like "mx" except that it returns
3885 just the host names, not the priorities.
3887 (3) It is now possible to give a list of domains (or IP addresses) to be
3888 looked up. The behaviour when one of the lookups defers can be
3889 controlled by a keyword.
3891 (4) It is now possible to specify the separator character for use when
3892 multiple records are returned.
3894 38. The dnslists ACL condition has been extended: it is now possible to supply
3895 a list of IP addresses and/or domains to be looked up in a particular DNS
3898 39. Added log_selector=+queue_time_overall.
3900 40. When running the queue in the test harness, wait just a tad after forking a
3901 delivery process, to get repeatability of debugging output.
3903 41. Include certificate and key file names in error message when GnuTLS fails
3904 to set them up, because the GnuTLS error message doesn't include the name
3905 of the failing file when there is a problem reading it.
3907 42. Allow both -bf and -bF in the same test run.
3909 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3911 44. The "Exiscan patch" is now merged into the mainline Exim source.
3913 45. Sometimes the final signoff response after QUIT could fail to get
3914 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3915 < 0 before doing a fflush(). This bug looks as though it goes back to the
3916 introduction of TLS in release 3.20, but "sometimes" must have been rare
3917 because the tests only now provoked it.
3919 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3920 (this can affect the format of dates).
3922 47. exim_tidydb, when checking for the continued existence of a message for
3923 which it has found a message-specific retry record, was not finding
3924 messages that were in split spool directories. Consequently, it was
3925 deleting retry records that should have stayed in existence.
3927 48. Steve fixed some bugs in eximstats.
3929 49. The SPA authentication driver was not abandoning authentication and moving
3930 on to the next authenticator when an expansion was forced to fail,
3931 contradicting the general specification for all authenticators. Instead it
3932 was generating a temporary error. It now behaves as specified.
3934 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3935 (the order specifies the preference for clients). The order is now AES256,
3936 AES128, 3DES, ARCFOUR128.
3938 51. Small patch to Sieve code - explicitly set From: when generating an
3941 52. Exim crashed if a remote delivery caused a very long error message to be
3942 recorded - for instance if somebody sent an entire SpamAssassin report back
3943 as a large number of 550 error lines. This bug was coincidentally fixed by
3944 increasing the size of one of Exim's internal buffers (big_buffer) that
3945 happened as part of the Exiscan merge. However, to be on the safe side, I
3946 have made the code more robust (and fixed the comments that describe what
3949 53. Now that there can be additional text after "Completed" in log lines (if
3950 the queue_time_overall log selector is set), a one-byte patch to exigrep
3951 was needed to allow it to recognize "Completed" as not the last thing in
3954 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3955 patch that reportedly fixes this has been added. I am not expert enough to
3956 create a test for it. This is what the patch creator wrote:
3958 "I found a little strange behaviour of ldap code when working with
3959 Windows 2003 AD Domain, where users was placed in more than one
3960 Organization Units. When I tried to give exim partial DN, the exit code
3961 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3962 But simultaneously result of request was absolutely normal ldap result,
3963 so I produce this patch..."
3965 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3966 REFERENCE, so I have modified the code to exclude the patch when that macro
3969 55. Some experimental protocols are using DNS PTR records for new purposes. The
3970 keys for these records are domain names, not reversed IP addresses. The
3971 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3972 leaves it alone. Component reversal etc. now happens only for IP addresses.
3975 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3977 57. Double the size of the debug message buffer (to 2048) so that more of very
3978 long debug lines gets shown.
3980 58. The exicyclog utility now does better if the number of log files to keep
3981 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3983 59. Two changes related to the smtp_active_hostname option:
3985 (1) $smtp_active_hostname is now available as a variable.
3986 (2) The default for smtp_banner uses $smtp_active_hostname instead
3987 of $primary_hostname.
3989 60. The host_aton() function is supposed to be passed a string that is known
3990 to be a valid IP address. However, in the case of IPv6 addresses, it was
3991 not checking this. This is a hostage to fortune. Exim now panics and dies
3992 if the condition is not met. A case was found where this could be provoked
3993 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3994 components; fortuitously, this particular loophole had already been fixed
3995 by change 4.50/55 above.
3997 If there are any other similar loopholes, the new check in host_aton()
3998 itself should stop them being exploited. The report I received stated that
3999 data on the command line could provoke the exploit when Exim was running as
4000 exim, but did not say which command line option was involved. All I could
4001 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4002 running as the user.
4005 61. There was a buffer overflow vulnerability in the SPA authentication code
4006 (which came originally from the Samba project). I have added a test to the
4007 spa_base64_to_bits() function which I hope fixes it.
4010 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4011 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4013 63. The daemon start-up calls getloadavg() while still root for those OS that
4014 need the first call to be done as root, but it missed one case: when
4015 deliver_queue_load_max is set with deliver_drop_privilege. This is
4016 necessary for the benefit of the queue runner, because there is no re-exec
4017 when deliver_drop_privilege is set.
4019 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4020 This has been fixed.
4022 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4023 was in use, was not putting the data itself into the right store pool;
4024 consequently, it could be overwritten for a subsequent message in the same
4025 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4028 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4030 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4031 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4032 the function were treating the return as a boolean value, which happened to
4033 work because 0=false and not-0=true, but is not correct code.
4035 68. The host_aton() function was not handling scoped IPv6 addresses (those
4036 with, for example, "%eth0" on the end) correctly.
4038 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4039 negated items (that is, ~something) in unsigned ints. Some compilers
4040 apparently mutter when there is no cast.
4042 70. If an address verification called from an ACL failed, and did not produce a
4043 user-specific message (i.e. there was only a "system" message), nothing was
4044 put in $acl_verify_message. In this situation, it now puts the system
4047 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4048 session; change 4.31/43 added the unwanted input to the log line - except
4049 that it did not do this in the start of session case. It now does.
4051 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4052 This is wrong and can cause the other end to generate a synchronization
4053 error if it is another Exim or anything else that does the synchronization
4054 check. A QUIT command is no longer sent after a timeout.
4056 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4057 during host lookups.
4059 74. The defer_ok option of callout verification was not working if it was used
4060 when verifying addresses in header lines, that is, for this case:
4062 verify = header_sender/callout=defer_ok
4064 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4065 those file descriptors could be used for SMTP connections. If anything
4066 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4067 could be sent to the SMTP client, causing chaos. The daemon now opens
4068 stdin, stdout, and stderr to /dev/null when it puts itself into the
4071 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4072 log by default. The user can override this with suitable Perl magic.
4074 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4075 the log message when discard triggers, was not working for the DATA ACL or
4076 for the non-SMTP ACL.
4078 78. Error message wording change in sieve.c.
4080 79. If smtp_accept_max_per_host was set, the number of connections could be
4081 restricted to fewer than expected, because the daemon was trying to set up
4082 a new connection before checking whether the processes handling previous
4083 connections had finished. The check for completed processes is now done
4084 earlier. On busy systems, this bug wouldn't be noticed because something
4085 else would have woken the daemon, and it would have reaped the completed
4088 80. If a message was submitted locally by a user whose login name contained one
4089 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4090 It caused a spool format error. I have fixed the spool reading code. A
4091 related problem was that the "from" clause in the Received: line became
4092 illegal because of the space(s). It is now covered by ${quote_local_part.
4094 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4097 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4099 83. Patch from Sieve maintainer to fix unterminated string problem in
4100 "vacation" handling.
4102 84. Some minor changes to the Linux configuration files to help with other
4103 OS variants using glibc.
4105 85. One more patch for Sieve to update vacation handling to latest spec.
4108 ----------------------------------------------------
4109 See the note above about the 4.44 and 4.50 releases.
4110 ----------------------------------------------------
4116 1. Change 4.43/35 introduced a bug that caused file counts to be
4117 incorrectly computed when quota_filecount was set in an appendfile
4120 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4121 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4124 3. Give more explanation in the error message when the command for a transport
4125 filter fails to execute.
4127 4. There are several places where Exim runs a non-Exim command in a
4128 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4129 was being done only for the command run by the queryprogram router. It is
4130 now done for all such subprocesses. The other cases are: ${run, transport
4131 filters, and the commands run by the lmtp and pipe transports.
4133 5. Some older OS have a limit of 256 on the maximum number of file
4134 descriptors. Exim was using setrlimit() to set 1000 as a large value
4135 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4136 systems. I've change it so that if it can't get 1000, it tries for 256.
4138 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4139 was an oversight, and furthermore, ever since the addition of extra
4140 controls (e.g. 4.43/32), the checks on when to allow different forms of
4141 "control" were broken. There should now be diagnostics for all cases when a
4142 control that does not make sense is encountered.
4144 7. $recipients is now available in the predata ACL (oversight).
4146 8. Tidy the search cache before the fork to do a delivery from a message
4147 received from the command line. Otherwise the child will trigger a lookup
4148 failure and thereby defer the delivery if it tries to use (for example) a
4149 cached ldap connection that the parent has called unbind on.
4151 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4152 of $address_data from the recipient verification was clobbered by the
4153 sender verification.
4155 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4156 was its contents. (It was OK if the option was not defined at all.)
4158 11. A "Completed" log line is now written for messages that are removed from
4159 the spool by the -Mrm option.
4161 12. $host_address is now set to the target address during the checking of
4162 ignore_target_hosts.
4164 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4165 being passed; this would have caused $sender_host_name to have been used if
4166 matching the list had actually called for a host name (not very likely,
4167 since this list is usually IP addresses). A host name is now passed as
4170 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4171 code that specifies a non-restarting handler (typically sigaction() in
4172 modern systems) in an attempt to fix a rare and obscure crash bug.
4174 15. Narrowed the window for a race in the daemon that could cause it to ignore
4175 SIGCHLD signals. This is not a major problem, because they are used only to
4176 wake it up if nothing else does.
4178 16. A malformed maildirsize file could cause Exim to calculate negative values
4179 for the mailbox size or file count. Odd effects could occur as a result.
4180 The maildirsize information is now recalculated if the size or filecount
4183 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4184 support for a long time. Removed HAVE_SYS_VFS_H.
4186 18. Updated exipick to current release from John Jetmore.
4188 19. Allow an empty sender to be matched against a lookup in an address list.
4189 Previously the only cases considered were a regular expression, or an
4192 20. Exim went into a mad DNS lookup loop when doing a callout where the
4193 host was specified on the transport, if the DNS lookup yielded more than
4196 21. The RFC2047 encoding function was originally intended for short strings
4197 such as real names; it was not keeping to the 75-character limit for
4198 encoded words that the RFC imposes. It now respects the limit, and
4199 generates multiple encoded words if necessary. To be on the safe side, I
4200 have increased the buffer size for the ${rfc2047: expansion operator from
4203 22. Failure to deliver a bounce message always caused it to be frozen, even if
4204 there was an errors_to setting on the router. The errors_to setting is now
4207 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4208 canonical form (fully expanded) before being placed in
4209 $sender_host_address.
4211 24. Updated eximstats to version 1.33
4213 25. Include certificate and key file names in error message when GnuTLS fails
4214 to set them up, because the GnuTLS error message doesn't include the name
4215 of the failing file when there is a problem reading it.
4217 26. Expand error message when OpenSSL has problems setting up cert/key files.
4220 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4221 (this can affect the format of dates).
4223 28. exim_tidydb, when checking for the continued existence of a message for
4224 which it has found a message-specific retry record, was not finding
4225 messages that were in split spool directories. Consequently, it was
4226 deleting retry records that should have stayed in existence.
4228 29. eximstats updated to version 1.35
4229 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4230 1.35 - bugfix such that pie charts by volume are generated correctly
4232 30. The SPA authentication driver was not abandoning authentication and moving
4233 on to the next authenticator when an expansion was forced to fail,
4234 contradicting the general specification for all authenticators. Instead it
4235 was generating a temporary error. It now behaves as specified.
4237 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4238 (the order specifies the preference for clients). The order is now AES256,
4239 AES128, 3DES, ARCFOUR128.
4241 31. Small patch to Sieve code - explicitly set From: when generating an
4244 32. Exim crashed if a remote delivery caused a very long error message to be
4245 recorded - for instance if somebody sent an entire SpamAssassin report back
4246 as a large number of 550 error lines. This bug was coincidentally fixed by
4247 increasing the size of one of Exim's internal buffers (big_buffer) that
4248 happened as part of the Exiscan merge. However, to be on the safe side, I
4249 have made the code more robust (and fixed the comments that describe what
4252 33. Some experimental protocols are using DNS PTR records for new purposes. The
4253 keys for these records are domain names, not reversed IP addresses. The
4254 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4255 leaves it alone. Component reversal etc. now happens only for IP addresses.
4258 34. The host_aton() function is supposed to be passed a string that is known
4259 to be a valid IP address. However, in the case of IPv6 addresses, it was
4260 not checking this. This is a hostage to fortune. Exim now panics and dies
4261 if the condition is not met. A case was found where this could be provoked
4262 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4263 components; fortuitously, this particular loophole had already been fixed
4264 by change 4.50/55 or 4.44/33 above.
4266 If there are any other similar loopholes, the new check in host_aton()
4267 itself should stop them being exploited. The report I received stated that
4268 data on the command line could provoke the exploit when Exim was running as
4269 exim, but did not say which command line option was involved. All I could
4270 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4271 running as the user.
4274 35. There was a buffer overflow vulnerability in the SPA authentication code
4275 (which came originally from the Samba project). I have added a test to the
4276 spa_base64_to_bits() function which I hope fixes it.
4279 36. The daemon start-up calls getloadavg() while still root for those OS that
4280 need the first call to be done as root, but it missed one case: when
4281 deliver_queue_load_max is set with deliver_drop_privilege. This is
4282 necessary for the benefit of the queue runner, because there is no re-exec
4283 when deliver_drop_privilege is set.
4285 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4286 was in use, was not putting the data itself into the right store pool;
4287 consequently, it could be overwritten for a subsequent message in the same
4288 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4291 38. Sometimes the final signoff response after QUIT could fail to get
4292 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4293 < 0 before doing a fflush(). This bug looks as though it goes back to the
4294 introduction of TLS in release 3.20, but "sometimes" must have been rare
4295 because the tests only now provoked it.
4301 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4302 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4303 Now it returns an integer. A number of calls were still expecting a T/F
4304 return. Fortuitously, in all cases, the tests worked in OK situations,
4305 which is the norm. However, things would have gone wrong on any write
4306 failures on the smtp file descriptor. This function is used when sending
4307 messages over SMTP and also when doing verify callouts.
4309 2. When Exim is called to do synchronous delivery of a locally submitted
4310 message (the -odf or -odi options), it no longer closes stderr before doing
4313 3. Implemented the mua_wrapper option.
4315 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4317 5. Implemented the functions header_remove(), header_testname(),
4318 header_add_at_position(), and receive_remove_recipient(), and exported them
4321 6. If an ACL "warn" statement specified the addition of headers, Exim already
4322 inserted X-ACL-Warn: at the start if there was no header name. However, it
4323 was not making this test for the second and subsequent header lines if
4324 there were newlines in the string. This meant that an invalid header could
4325 be inserted if Exim was badly configured.
4327 7. Allow an ACL "warn" statement to add header lines at the start or after all
4328 the Received: headers, as well as at the end.
4330 8. Added the rcpt_4xx retry error code.
4332 9. Added postmaster_mailfrom=xxx to callout verification option.
4334 10. Added mailfrom=xxxx to the callout verification option, for verify=
4337 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4338 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4340 12. Inserted some casts to stop certain compilers warning when using pointer
4341 differences as field lengths or precisions in printf-type calls (mostly
4342 affecting debugging statements).
4344 13. Added optional readline() support for -be (dynamically loaded).
4346 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4347 same clock tick as a message's arrival, so that its received time was the
4348 same as the "first fail" time on the retry record, and that message
4349 remained on the queue past the ultimate address timeout, every queue runner
4350 would try a delivery (because it was past the ultimate address timeout) but
4351 after another failure, the ultimate address timeout, which should have then
4352 bounced the address, did not kick in. This was a "< instead of <=" error;
4353 in most cases the first failure would have been in the next clock tick
4354 after the received time, and all would be well.
4356 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4357 being recognized when the domain list was tested by the match_domain
4358 condition in an expansion string.
4360 16. Added the ${str2b64: operator.
4362 17. Exim was always calling setrlimit() to set a large limit for the number of
4363 processes, without checking whether the existing limit was already
4364 adequate. (It did check for the limit on file descriptors.) Furthermore,
4365 errors from getrlimit() and setrlimit() were being ignored. Now they are
4366 logged to the main and panic logs, but Exim does carry on, to try to do its
4367 job under whatever limits there are.
4369 18. Imported PCRE 5.0.
4371 19. Trivial typo in log message " temporarily refused connection" (the leading
4374 20. If the log selector return_path_on_delivery was set and an address was
4375 redirected to /dev/null, the delivery process crashed because it assumed
4376 that a return path would always be set for a "successful" delivery. In this
4377 case, the whole delivery is bypassed as an optimization, and therefore no
4380 21. Internal re-arrangement: the function for sending a challenge and reading
4381 a response while authentication was assuming a zero-terminated challenge
4382 string. It's now changed to take a pointer and a length, to allow for
4383 binary data in such strings.
4385 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4387 23. Exim was not respecting finduser_retries when seeking the login of the
4388 uid under which it was called; it was always trying 10 times. (The default
4389 setting of finduser_retries is zero.) Also, it was sleeping after the final
4390 failure, which is pointless.
4392 24. Implemented tls_on_connect_ports.
4394 25. Implemented acl_smtp_predata.
4396 26. If the domain in control=submission is set empty, Exim assumes that the
4397 authenticated id is a complete email address when it generates From: or
4398 Sender: header lines.
4400 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4401 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4402 chown and chgrp in /bin and hostname in /usr/bin.
4404 28. Exim was keeping the "process log" file open after each use, just as it
4405 does for the main log. This opens the possibility of it remaining open for
4406 long periods when the USR1 signal hits a daemon. Occasional processlog
4407 errors were reported, that could have been caused by this. Anyway, it seems
4408 much more sensible not to leave this file open at all, so that is what now
4411 29. The long-running daemon process does not normally write to the log once it
4412 has entered its main loop, and it closes the log before doing so. This is
4413 so that log files can straightforwardly be renamed and moved. However,
4414 there are a couple of unusual error situations where the daemon does write
4415 log entries, and I had neglected to close the log afterwards.
4417 30. The text of an SMTP error response that was received during a remote
4418 delivery was being truncated at 512 bytes. This is too short for some of
4419 the long messages that one sometimes sees. I've increased the limit to
4422 31. It is now possible to make retry rules that apply only when a message has a
4423 specific sender, in particular, an empty sender.
4425 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4426 it possible to be selective about when SMTP synchronization is enforced.
4428 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4430 32. Implemented hosts_connection_nolog.
4432 33. Added an ACL for QUIT.
4434 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4437 35. Added mailbox_size and mailbox_filecount to appendfile.
4439 36. Added control = no_multiline_responses to ACLs.
4441 37. There was a bug in the logic of the code that waits for the clock to tick
4442 in the case where the clock went backwards by a substantial amount such
4443 that the microsecond fraction of "now" was more than the microsecond
4444 fraction of "then" (but the whole seconds number was less).
4446 38. Added support for the libradius Radius client library this is found on
4447 FreeBSD (previously only the radiusclient library was supported).
4453 1. When certain lookups returned multiple values in the form name=value, the
4454 quoting of the values was not always being done properly. Specifically:
4455 (a) If the value started with a double quote, but contained no whitespace,
4457 (b) If the value contained whitespace other than a space character (i.e.
4458 tabs or newlines or carriage returns) it was not quoted.
4459 This fix has been applied to the mysql and pgsql lookups by writing a
4460 separate quoting function and calling it from the lookup code. The fix
4461 should probably also be applied to nisplus, ibase and oracle lookups, but
4462 since I cannot test any of those, I have not disturbed their existing code.
4464 2. A hit in the callout cache for a specific address caused a log line with no
4465 reason for rejecting RCPT. Now it says "Previous (cached) callout
4466 verification failure".
4468 3. There was an off-by-one bug in the queryprogram router. An over-long
4469 return line was truncated at 256 instead of 255 characters, thereby
4470 overflowing its buffer with the terminating zero. As well as fixing this, I
4471 have increased the buffer size to 1024 (and made a note to document this).
4473 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4474 when Exim is waiting for an SMTP response from a remote server, Exim
4475 restarts its select() call on the socket, thereby resetting its timeout.
4476 This is not a problem when such interrupts are rare. Somebody set up a cron
4477 job to run exiwhat every 2 minutes, which is less than the normal select()
4478 timeout (5 or 10 minutes). This meant that the select() timeout never
4479 kicked in because it was always reset. I have fixed this by comparing the
4480 time when an interrupt arrives with the time at the start of the first call
4481 to select(). If more time than the timeout has elapsed, the interrupt is
4482 treated as a timeout.
4484 5. Some internal re-factoring in preparation for the addition of Sieve
4485 extensions (by MH). In particular, the "personal" test is moved to a
4486 separate function, and given an option for scanning Cc: and Bcc: (which is
4487 not set for Exim filters).
4489 6. When Exim created an email address using the login of the caller as the
4490 local part (e.g. when creating a From: or Sender: header line), it was not
4491 quoting the local part when it contained special characters such as @.
4493 7. Installed new OpenBSD configuration files.
4495 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4496 try to make them clearer.
4498 9. Callout options, other than the timeout value, were being ignored when
4499 verifying sender addresses in header lines. For example, when using
4501 verify = header_sender/callout=no_cache
4503 the cache was (incorrectly) being used.
4505 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4506 only the Cygwin environment.
4508 11. When return_path_on_delivery was set as a log selector, if different remote
4509 addresses in the same message used different return paths and parallel
4510 remote delivery occurred, the wrong values would sometimes be logged.
4511 (Whenever a remote delivery process finished, the return path value from
4512 the most recently started remote delivery process was logged.)
4514 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4515 lines when AUTH and/or TLS are in use. This is the "received protocol"
4516 field. Exim used to use "asmtp" for authenticated SMTP, without any
4517 indication (in the protocol name) for TLS use. Now it follows the RFC and
4518 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4519 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4520 names appear in log lines as well as in Received: header lines.
4522 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4523 extensions, and comparison tests, and to fix some bugs.
4525 14. Changes to the "personal" filter test:
4527 (1) The test was buggy in that it was just doing the equivalent of
4528 "contains" tests on header lines. For example, if a user's address was
4529 anne@some.where, the "personal" test would incorrectly be true for
4531 To: susanne@some.where
4533 This test is now done by extracting each address from the header in turn,
4534 and checking the entire address. Other tests that are part of "personal"
4535 are now done using regular expressions (for example, to check local parts
4536 of addresses in From: header lines).
4538 (2) The list of non-personal local parts in From: addresses has been
4539 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4540 taken from the Sieve specification recommendations.
4542 (3) If the message contains any header line starting with "List-" it is
4543 treated as non-personal.
4545 (4) The test for "circular" in the Subject: header line has been removed
4546 because it now seems ill-conceived.
4548 15. Minor typos in src/EDITME comments corrected.
4550 16. Installed latest exipick from John Jetmore.
4552 17. If headers_add on a router specified a text string that was too long for
4553 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4554 of string_sprintf() is now avoided.
4556 18. $message_body_size was not set (it was always zero) when running the DATA
4557 ACL and the local_scan() function.
4559 19. For the "mail" command in an Exim filter, no default was being set for
4560 the once_repeat time, causing a random time value to be used if "once" was
4561 specified. (If the value happened to be <= 0, no repeat happened.) The
4562 default is now 0s, meaning "never repeat". The "vacation" command was OK
4563 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4564 (I found it when inspecting the code).
4566 20. There is now an overall timeout for performing a callout verification. It
4567 defaults to 4 times the callout timeout, which applies to individual SMTP
4568 commands during the callout. The overall timeout applies when there is more
4569 than one host that can be tried. The timeout is checked before trying the
4570 next host. This prevents very long delays if there are a large number of
4571 hosts and all are timing out (e.g. when the network connections are timing
4572 out). The value of the overall timeout can be changed by specifying an
4573 additional sub-option for "callout", called "maxwait". For example:
4575 verify = sender/callout=5s,maxwait=20s
4577 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4578 to the end before writing, but this should make it even safer).
4580 22. Exim was forgetting that it had advertised PIPELINING for the second and
4581 subsequent messages on an SMTP connection. It was also not resetting its
4582 memory on STARTTLS and an internal HELO.
4584 23. When Exim logs an SMTP synchronization error within a session, it now
4585 records whether PIPELINING has been advertised or not.
4587 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4588 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4589 rather than long int.
4591 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4593 26. Added the never_mail option to autoreply.
4599 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4600 crash if the getsockname() call failed; this can happen if a connection is
4601 closed very soon after it is established. The problem was simply in the
4602 order in which certain operations were done, causing Exim to try to write
4603 to the SMTP stream before it had set up the file descriptor. The bug has
4604 been fixed by making things happen in the correct order.
4610 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4611 before the connection was closed, thus losing the rejection response.
4613 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4614 some early Solaris releases, but causes trouble in current releases where
4615 socklen_t is defined.
4617 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4620 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4623 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4624 that releasing the top of it at the end releases what was used for sub-
4625 expansions (unless the block got too big). However, discard this block if
4626 the first thing is a variable or header, so that we can use its block when
4627 it is dynamic (useful for very large $message_headers, for example).
4629 6. Lookups now cache *every* query, not just the most recent. A new, separate
4630 store pool is used for this. It can be recovered when all lookup caches are
4631 flushed. Lookups now release memory at the end of their result strings.
4632 This has involved some general refactoring of the lookup sources.
4634 7. Some code has been added to the store_xxx() functions to reduce the amount
4635 of flapping under certain conditions.
4637 8. log_incoming_interface used to affect only the <= reception log lines. Now
4638 it causes the local interface and port to be added to several more SMTP log
4639 lines, for example "SMTP connection from", and rejection lines.
4641 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4643 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4645 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4646 could be overwritten at the end of the current message (or the start of a
4647 new message if it was set in a HELO ACL). The value is now preserved for
4648 the duration of the SMTP connection.
4650 12. If a transport had a headers_rewrite setting, and a matching header line
4651 contained an unqualified address, that address was qualified, even if it
4652 did not match any rewriting rules. The underlying bug was that the values
4653 of the flags that permit the existence of unqualified sender and recipient
4654 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4655 non-local messages, and by -bnq for local messages) were not being
4656 preserved with the message after it was received.
4658 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4659 "next input=" as part of the text comprising the host identity instead of
4660 the correct text. The code was using the same buffer for two different
4661 strings. However, depending on which order the printing function evaluated
4662 its arguments, the bug did not always show up. Under Linux, for example, my
4663 test suite worked just fine.
4665 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4666 doesn't work with some older versions of Perl. It has been changed to "my",
4667 which in any case is probably the better facility to use.
4669 15. A really picky compiler found some instances of statements for creating
4670 error messages that either had too many or two few arguments for the format
4673 16. The size of the buffer for calls to the DNS resolver has been increased
4674 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4675 for addresses that have a lot of PTR records. This alleviates a problem; it
4676 does not fully solve it.
4678 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4679 buffer now truncates the list and logs the incident, which is the same
4680 action as happens when Exim is looking up a host name and its aliases.
4681 Previously in this situation something unpredictable would happen;
4682 sometimes it was "internal error: store_reset failed".
4684 18. If a server dropped the connection unexpectedly when an Exim client was
4685 using GnuTLS and trying to read a response, the client delivery process
4686 crashed while trying to generate an error log message.
4688 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4689 string, for example:
4691 warn message = H1: something\nH2: something
4693 the text was added as a single header line from Exim's point of view
4694 though it ended up OK in the delivered message. However, searching for the
4695 second and subsequent header lines using $h_h2: did not work. This has been
4696 fixed. Similarly, if a system filter added multiple headers in this way,
4697 the routers could not see them.
4699 20. Expanded the error message when iplsearch is called with an invalid key to
4700 suggest using net-iplsearch in a host list.
4702 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4703 ACLs are no longer actually imposed (and a message to that effect is
4706 22. If a "gecos" field in a passwd entry contained escaped characters, in
4707 particular, if it contained a \" sequence, Exim got it wrong when building
4708 a From: or a Sender: header from that name. A second bug also caused
4709 incorrect handling when an unquoted " was present following a character
4710 that needed quoting.
4712 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4713 was not being matched caselessly.
4715 24. Arranged for all hyphens in the exim.8 source to be escaped with
4718 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4719 properly. Recipient callout cache records were still being keyed to include
4720 the sender, even when use_sender was set false. This led to far more
4721 callouts that were necessary. The sender is no longer included in the key
4722 when use_sender is false.
4724 26. Added "control = submission" modifier to ACLs.
4726 27. Added the ${base62d: operator to decode base 62 numbers.
4728 28. dnsdb lookups can now access SRV records.
4730 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4731 the configuration file.
4733 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4734 (-v) mode. This makes the output for a verbose queue run more intelligible.
4736 31. Added a use_postmaster feature to recipient callouts.
4738 32. Added the $body_zerocount variable, containing the number of binary zero
4739 bytes in the message body.
4741 33. The time of last modification of the "new" subdirectory is now used as the
4742 "mailbox time last read" when there is a quota error for a maildir
4745 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4747 35. Added +ignore_unknown as a special item in host lists.
4749 36. Code for decoding IPv6 addresses in host lists is now included, even if
4750 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4751 address was recognized as an IP address, but was then not correctly decoded
4752 into binary, causing unexpected and incorrect effects when compared with
4759 1. Very minor rewording of debugging text in manualroute to say "list of
4760 hosts" instead of "hostlist".
4762 2. If verify=header_syntax was set, and a header line with an unqualified
4763 address (no domain) and a large number of spaces between the end of the
4764 name and the colon was received, the reception process suffered a buffer
4765 overflow, and (when I tested it) crashed. This was caused by some obsolete
4766 code that should have been removed. The fix is to remove it!
4768 3. When running in the test harness, delay a bit after writing a bounce
4769 message to get a bit more predictability in the log output.
4771 4. Added a call to search_tidyup() just before forking a reception process. In
4772 theory, someone could use a lookup in the expansion of smtp_accept_max_
4773 per_host which, without the tidyup, could leave open a database connection.
4775 5. Added the variables $recipient_data and $sender_data which get set from a
4776 lookup success in an ACL "recipients" or "senders" condition, or a router
4777 "senders" option, similar to $domain_data and $local_part_data.
4779 6. Moved the writing of debug_print from before to after the "senders" test
4782 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4783 problems for message scanning, either using a data ACL, or using
4784 local_scan() because the Received: header was not generated till after they
4785 were called (in order to set the time as the time of reception completion).
4786 I have revised the way this works. The header is now generated after the
4787 body is received, but before the ACL or local_scan() are called. After they
4788 are run, the timestamp in the header is updated.
4794 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4795 before starting a queue runner without re-exec. This happened only when
4796 deliver_drop_privilege was set or when the Exim user was set to root. The
4797 effect of the bug was that timeouts during subsequent deliveries caused
4798 crashes instead of being properly handled. The handler is now left at its
4799 default (and expected) setting.
4801 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4802 message, again when deliver_drop_privilege is set or Exim is run as root.
4803 The bug described in (1) was not present in this case, but the tidying up
4804 of the other signals was missing. I have made the two cases consistent.
4806 3. The ignore_target_hosts setting on a manualroute router was being ignored
4807 for hosts that were looked up using the /MX notation.
4809 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4812 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4813 operated on the sender address. After changing the $sender_address to <>
4814 for the sender address verify, Exim was re-instated it as the original
4815 (before rewriting) address, but remembering that it had rewritten it, so it
4816 wasn't rewriting it again. This bug also had the effect of breaking the
4817 sender address verification caching when the sender address was rewritten.
4819 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4820 This has been changed so that if the ip literal address matches
4821 ignore_target_hosts, the router declines.
4823 7. Added expansion conditions match_domain, match_address, and match_local_
4824 part (NOT match_host).
4826 8. The placeholder for the Received: header didn't have a length field set.
4828 9. Added code to Exim itself and to exim_lock to test for a specific race
4829 condition that could lead to file corruption when using MBX delivery. The
4830 issue is with the lockfile that is created in /tmp. If this file is removed
4831 after a process has opened it but before that process has acquired a lock,
4832 there is the potential for a second process to recreate the file and also
4833 acquire a lock. This could lead to two Exim processes writing to the file
4834 at the same time. The added code performs the same test as UW imapd; it
4835 checks after acquiring the lock that its file descriptor still refers to
4836 the same named file.
4838 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4839 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4842 11. Added the smtp_active_hostname option. If used, this will typically be made
4843 to depend on the incoming interface address. Because $interface_address is
4844 not set up until the daemon has forked a reception process, error responses
4845 that can happen earlier (such as "too many connections") no longer contain
4848 12. If an expansion in a condition on a "warn" statement fails because a lookup
4849 defers, the "warn" statement is abandoned, and the next ACL statement is
4850 processed. Previously this caused the whole ACL to be aborted.
4852 13. Added the iplsearch lookup type.
4854 14. Added ident_timeout as a log selector.
4856 15. Added tls_certificate_verified as a log selector.
4858 16. Added a global option tls_require_ciphers (compare the smtp transport
4859 option of the same name). This controls incoming TLS connections.
4861 17. I finally figured out how to make tls_require_ciphers do a similar thing
4862 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4863 before starting the TLS session.
4865 18. Tabs are now shown as \t in -bP output.
4867 19. If the log selector return_path_on_delivery was set, Exim crashed when
4868 bouncing a message because it had too many Received: header lines.
4870 20. If two routers both had headers_remove settings, and the first one included
4871 a superfluous trailing colon, the final name in the first list and the
4872 first name in the second list were incorrectly joined into one item (with a
4873 colon in the middle).
4879 1. Added -C and -D options to the exinext utility, mainly to make it easier
4880 to include in the automated testing, but these could be helpful when
4881 multiple configurations are in use.
4883 2. The exinext utility was not formatting the output nicely when there was
4884 an alternate port involved in the retry record key, nor when there was a
4885 message id as well (for retries that were specific to a specific message
4886 and a specific host). It was also confused by IPv6 addresses, because of
4887 the additional colons they contain. I have fixed the IPv4 problem, and
4888 patched it up to do a reasonable job for IPv6.
4890 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4891 delivery, the log line now contains "pipelined" if PIPELINING was used.
4893 4. An SMTP transport process used to panic and die if the bind() call to set
4894 an explicit outgoing interface failed. This has been changed; it is now
4895 treated in the same way as a connect() failure.
4897 5. A reference to $sender_host_name in the part of a conditional expansion
4898 that was being skipped was still causing a DNS lookup. This no longer
4901 6. The def: expansion condition was not recognizing references to header lines
4902 that used bh_ and bheader_.
4904 7. Added the _cache feature to named lists.
4906 8. The code for checking quota_filecount in the appendfile transport was
4907 allowing one more file than it should have been.
4909 9. For compatibility with Sendmail, the command line option
4917 and sets the incoming protocol and host name (for trusted callers). The
4918 host name and its colon can be omitted when only the protocol is to be set.
4919 Note the Exim already has two private options, -pd and -ps, that refer to
4920 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4921 "s", but I don't think that's a major issue.
4923 10. A number of refactoring changes to the code, none of which should affect
4926 (a) The number of logging options was getting close to filling up the
4927 32-bit word that was used as a bit map. I have split them into two classes:
4928 those that are passed in the argument to log_write(), and those that are
4929 only ever tested independently outside of that function. These are now in
4930 separate 32-bit words, so there is plenty of room for expansion again.
4931 There is no change in the user interface or the logging behaviour.
4933 (b) When building, for example, log lines, the code previously used a
4934 macro that called string_cat() twice, in order to add two strings. This is
4935 not really sufficiently general. Furthermore, there was one instance where
4936 it was actually wrong because one of the argument was used twice, and in
4937 one call a function was used. (As it happened, calling the function twice
4938 did not affect the overall behaviour.) The macro has been replaced by a
4939 function that can join an arbitrary number of extra strings onto a growing
4942 (c) The code for expansion conditions now uses a table and a binary chop
4943 instead of a serial search (which was left over from when there were very
4944 few conditions). Also, it now recognizes conditions like "pam" even when
4945 the relevant support is not compiled in: a suitably worded error message is
4946 given if an attempt is made to use such a condition.
4948 11. Added ${time_interval:xxxxx}.
4950 12. A bug was causing one of the ddress fields not to be passed back correctly
4951 from remote delivery subprocesses. The field in question was not being
4952 subsequently used, so this caused to problems in practice.
4954 13. Added new log selectors queue_time and deliver_time.
4956 14. Might have fixed a bug in maildirsizefile handling that threw up
4957 "unexpected character" debug warnings, and recalculated the data
4958 unnecessarily. In any case, I expanded the warning message to give more
4961 15. Added the message "Restricted characters in address" to the statements in
4962 the default ACL that block characters like @ and % in local parts.
4964 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4965 Three changes have been made:
4967 (a) There was a serious bug; a negative response to MAIL caused the whole
4968 recipient domain to be cached as invalid, thereby blocking all messages
4969 to all local parts at the same domain, from all senders. This bug has
4970 been fixed. The domain is no longer cached after a negative response to
4971 MAIL if the sender used is not empty.
4973 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4976 (c) A new callout option, "use_sender" has been added for people who want
4977 the modified behaviour.
4983 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4986 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4987 indeed breaks things for older releases.
4989 3. Added additional logging to the case where there is a problem reading data
4990 from a filter that is running in a subprocess using a pipe, in order to
4991 try to track down a specific problem.
4993 4. Testing facility fudge: when running in the test harness and attempting
4994 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4995 getting "No route to host". Convert this to a timeout.
4997 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5000 6. Some OS don't have socklen_t but use size_t instead. This affects the
5001 fifth argument of getsockopt() amongst other things. This is now
5002 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5003 can be set for individual OS. I have set it for SunOS5, OSF1, and
5004 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5005 some earlier ones do not.
5007 7. Change 4.30/15 was not doing the test caselessly.
5009 8. The standard form for an IPv6 address literal was being rejected by address
5010 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5011 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5012 this, as well as the form without the "IPv6" on the front (but only when
5013 address literals are enabled, of course).
5015 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5017 10. Exim crashed if a message with an empty sender address specified by -f
5018 encountered a router with an errors_to setting. This could be provoked only
5019 by a command such as
5023 where an empty string was supplied; "<>" did not hit this bug.
5025 11. Installed PCRE release 4.5.
5027 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5028 remained set. It is now erased.
5030 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5031 times from message ids (which are base 36 rather than the normal 62).
5033 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5034 were being counted as actual protocol errors, and logged if the log
5035 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5036 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5037 and DATA following a set of rejected RCPTs do not count as protocol errors.
5038 In other words, Exim assumes they were pipelined, though this may not
5039 actually be the case. Of course, in all cases the client gets an
5040 appropriate error code.
5042 15. If a lookup fails in an ACL condition, a message about the failure may
5043 be available; it is used if testing the ACL cannot continue, because most
5044 such messages specify what the cause of the deferral is. However, some
5045 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5046 that caused an old message to be retained and used if a later statement
5047 caused a defer, replacing the real cause of the deferral.
5049 16. If an IP address had so many PTR records that the DNS lookup buffer
5050 was not large enough to hold them, Exim could crash while trying to process
5051 the truncated data. It now detects and logs this case.
5053 17. Further to 4.21/58, another change has been made: if (and only if) the
5054 first line of a message (the first header line) ends with CRLF, a bare LF
5055 in a subsequent header line has a space inserted after it, so as not to
5056 terminate the header.
5058 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5059 unnecessarily, used atoi() instead of strtol(), and didn't check the
5060 termination when getting file sizes from file names by regex.
5062 19. Completely re-implemented the support for maildirsize files, in the light
5063 of a number of problems with the previous contributed implementation
5064 (4.30/29). In particular:
5066 . If the quota is zero, the maildirsize file is maintained, but no quota is
5069 . If the maildir directory does not exist, it is created before any attempt
5070 to write a maildirsize file.
5072 . The quota value in the file is just a cache; if the quota is changed in
5073 the transport, the new value overrides.
5075 . A regular expression is available for excluding directories from the
5078 20. The autoreply transport checks the characters in options that define the
5079 message's headers; it allows continued headers, but it was checking with
5080 isspace() after an embedded newline instead of explicitly looking for a
5083 21. If all the "regular" hosts to which an address was routed had passed their
5084 expiry times, and had not reached their retry times, the address was
5085 bounced, even if fallback hosts were defined. Now Exim should go on to try
5088 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5089 equivalent code in the SMTP transport. Some hosts send humungous responses
5090 to HELO/EHLO, more than 1024 it seems.
5092 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5093 gives compiler warnings in some environments. I've now done it "properly",
5096 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5097 (because of problems with the built-in one) was declared to return uschar *
5098 instead of char *, causing compiler failure.
5100 25. Fixed a file descriptor leak when processing alias/forward files.
5102 26. Fixed a minor format string issue in dbfn.c.
5104 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5106 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5107 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5110 29. When a sender address is verified, it is cached, to save repeating the test
5111 when there is more than one recipient in a message. However, when the
5112 verification involves a callout, it is possible for different callout
5113 options to be set for different recipients. It is too complicated to keep
5114 track of this in the cache, so now Exim always runs a verification when a
5115 callout is required, relying on the callout cache for the optimization.
5116 The overhead is duplication of the address routing, but this should not be
5119 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5120 to be verified with callout=postmaster, and the main callout worked but the
5121 postmaster check failed, the verification correctly failed. However, if a
5122 subsequent RCPT command asked for sender verification *without* the
5123 postmaster check, incorrect caching caused this verification also to fail,
5126 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5127 it was not caching the DNS options (qualify_single, search_parents) that
5128 were used when the lookup failed. A subsequent lookup with different
5129 options therefore always gave the same answer, though there were cases
5130 where it should not have. (Example: a "domains = !$mx_any" option on a
5131 dnslookup router: the "domains" option is always processed without any
5132 widening, but the router might have qualify_single set.) Now Exim uses the
5133 cached value only when the same options are set.
5135 32. Added John Jetmore's "exipick" utility to the distribution.
5137 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5138 than a timeout (e.g. a certificate is required, and is not provided), an
5139 Exim server now closes the connection immediately. Previously it waited for
5140 the client to close - but if the client is SSL, it seems that they each
5141 wait for each other, leading to a delay before one of them times out.
5143 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5144 maintained 0.8.x compatibility because I don't think many are using it, and
5145 it is clearly obsolete.
5147 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5150 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5151 client certificate was expired. A simple patch fixes this, though I don't
5152 understand the full logic of why the verify callback is called multiple
5155 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5156 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5157 which causes problems with some clients (such as the Certicom SSL Plus
5158 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5159 disables the coutermeasure allowing Eudora to connect."
5161 38. Exim was not checking that a write() to a log file succeeded. This could
5162 lead to Bad Things if a log got too big, in particular if it hit a file
5163 size limit. Exim now panics and dies if it cannot write to a log file, just
5164 as it does if it cannot open a log file.
5166 39. Modified OS/Makefile-Linux so that it now contains
5168 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5170 The two -D definitions ensure that Exim is compiled with large file
5171 support, which makes it possible to handle log files that are bigger than
5174 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5175 instance) a domain was checked against a named list that involved a lookup,
5176 causing $domain_data to be set, then another domain was checked against the
5177 same list, then the first domain was re-checked, the value of $domain_data
5178 after the final check could be wrong. In particular, if the second check
5179 failed, it could be set empty. This bug probably also applied to
5182 41. The strip_trailing_dot option was not being applied to the address given
5183 with the -f command-line option.
5185 42. The code for reading a message's header from the spool was incrementing
5186 $received_count, but never initializing it. This meant that the value was
5187 incorrect (doubled) while delivering a message in the same process in which
5188 it was received. In the most common configuration of Exim, this never
5189 happens - a fresh exec is done - but it can happen when
5190 deliver_drop_privilege is set.
5192 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5193 it now includes up to 150 characters of the unexpected data in the log
5196 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5197 and building data strings. The size of both of these buffers was 10 000
5198 bytes - far larger than anybody would *ever* want, thought I. Needless to
5199 say, somebody hit the limit. I have increased the maximum line length to
5200 20 000 and the maximum data length of concatenated lines to 100 000. I have
5201 also fixed two bugs, because there was no checking on these buffers. Tsk,
5202 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5203 buffer is too small.
5205 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5206 lsearch lookups. Now it does.
5208 46. When parsing a route_list item in a manualroute router, a fixed-length
5209 buffer was used for the list of hosts. I made this 1024 bytes long,
5210 thinking that nobody would ever have a list of hosts that long. Wrong.
5211 Somebody had a whole pile of complicated expansion conditions, and the
5212 string was silently truncated, leading to an expansion error. It turns out
5213 that it is easier to change to an unlimited length (owing to other changes
5214 that have happened since this code was originally written) than to build
5215 structure for giving a limitation error. The length of the item that
5216 expands into the list of hosts is now unlimited.
5218 47. The lsearch lookup could not handle data where the length of text line was
5219 more than 4095 characters. Such lines were truncated, leading to shortened
5220 data being returned. It should now handle lines of any length.
5222 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5223 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5226 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5227 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5228 no longer happens in comment lines. A list of replacements is now placed
5229 at the head of all of the source files, except those whose only change is
5230 to replace PERL_COMMAND in the very first #! line.
5232 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5233 messages on the queue, with a bottom-up merge sort, using code contributed
5234 by Michael Haardt. This should make operations like -bp somewhat faster on
5235 large queues. It won't affect queue runners, except when queue_run_in_order
5238 51. Installed eximstats 1.31 in the distribution.
5240 52. Added support for SRV lookups to the dnslookup router.
5242 53. If an ACL referred to $message_body or $message_body_end, the value was not
5243 reset for any messages that followed in the same SMTP session.
5245 54. The store-handling optimization for building very long strings was not
5246 differentiating between the different store pools. I don't think this
5247 actually made any difference in practice, but I've tidied it.
5249 55. While running the routers to verify a sender address, $sender_address
5250 was still set to the sender address. This is wrong, because when routing to
5251 send a bounce to the sender, it would be empty. Therefore, I have changed
5252 it so that, while verifying a sender address, $sender_address is set to <>.
5253 (There is no change to what happens when verifying a recipient address.)
5255 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5256 target A or AAAA records (if not already returned) without resetting the
5257 qualify_single or search_parents options of the DNS resolver. These are
5258 inappropriate in this case because the targets of MX and SRV records must
5259 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5260 target that, when qualified, matched something in the local domain. These
5261 two options are now turned off when doing these lookups.
5263 57. It seems that at least some releases of Reiserfs (which does not have the
5264 concept of a fixed number of inodes) returns zero and not -1 for the
5265 number of available inodes. This interacted badly with check_spool_inodes,
5266 which assumed that -1 was the "no such thing" setting. What I have done is
5267 to check that the total number of inodes is greater than zero before doing
5268 the test of how many are available.
5270 58. When a "warn" ACL statement has a log_message modifier, the message is
5271 remembered, and not repeated. This is to avoid a lot of repetition when a
5272 message has many recipients that cause the same warning to be written.
5273 Howewer, Exim was preserving the list of already written lines for an
5274 entire SMTP session, which doesn't seem right. The memory is now reset if a
5275 new message is started.
5277 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5278 debugging output unless log_rewrite was also set.
5280 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5281 of (int)(handle) when we know that handle contains (void *)(-1).
5283 61. The Exim daemon panic-logs an error return when it closes the incoming
5284 connection. However "connection reset by peer" seems to be common, and
5285 isn't really an error worthy of noting specially, so that particular error
5288 62. When Exim is trying to find all the local interfaces, it used to panic and
5289 die if the ioctl to get the interface flags failed. However, it seems that
5290 on at least one OS (Solaris 9) it is possible to have an interface that is
5291 included in the list of interfaces, but for which you get a failure error
5292 for this call. This happens when the interface is not "plumbed" into a
5293 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5294 failure of the "get flags" call assumes that the interface is down.
5296 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5297 makes life easier for people who are doing arithmetic on fields extracted
5298 from dates, where you often get leading zeros that should not be
5299 interpreted as octal.
5301 64. Added qualify_domain to the redirect router, to override the global
5304 65. If a pathologically long header line contained very many addresses (the
5305 report of this problem mentioned 10 000) and each of them was rewritten,
5306 Exim could use up a very large amount of memory. (It kept on making new
5307 copies of the header line as it rewrote, and never released the old ones.)
5308 At the expense of a bit more processing, the header rewriting function has
5309 been changed so that it no longer eats memory in this way.
5311 66. The generation of the Received: header has been moved from the time that a
5312 message starts to be received, to the time that it finishes. The timestamp
5313 in the Received: header should now be very close to that of the <= log
5314 line. There are two side-effects of this change:
5316 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5317 logged header lines no longer include the local Received: line, because
5318 it has not yet been created. The same applies to a copy of the message
5319 that is returned to a non-SMTP sender when a message is rejected.
5321 (b) When a filter file is tested using -bf, no additional Received: header
5322 is added to the test message. After some thought, I decided that this
5325 This change does not affect the value of $received_for. It is still set
5326 after address rewriting, but before local_scan() is called.
5328 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5330 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5331 gave an unhelpful panic error message, and a defer error. I have managed to
5332 change this behaviour so that it now rejects any supplied certificate,
5333 which seems right, as the list of acceptable certificates is empty.
5335 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5336 gave an unhelpful defer error. I have not managed to make this reject any
5337 supplied certificates, but the error message it gives is "no certificate
5338 supplied", which is not helpful.
5340 70. exigrep's output now also includes lines that are not associated with any
5341 message, but which match the given pattern. Implemented by a patch from
5342 Martin Sluka, which also tidied up the Perl a bit.
5344 71. Recipient callout verification, like sender verification, was using <> in
5345 the MAIL FROM command. This isn't really the right thing, since the actual
5346 sender may affect whether the remote host accepts the recipient or not. I
5347 have changed it to use the actual sender in the callout; this means that
5348 the cache record is now keyed on a recipient/sender pair, not just the
5349 recipient address. There doesn't seem to be a real danger of callout loops,
5350 since a callout by the remote host to check the sender would use <>.
5351 [SEE ABOVE: changed after hitting problems.]
5353 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5354 temporary errors. However, in the case of such a code being given after
5355 the end of a data transmission (i.e. after ".") Exim was failing to write
5356 a retry record for the message. (Yes, there was some broken host that was
5357 actually sending 8xx at this point.)
5359 73. An unknown lookup type in a host list could cause Exim to panic-die when
5360 the list was checked. (An example that provoked this was putting <; in the
5361 middle of a list instead of at the start.) If this happened during a DATA
5362 ACL check, a -D file could be left lying around. This kind of configuration
5363 error no longer causes Exim to die; instead it causes a defer errror. The
5364 incident is still logged to the main and panic logs.
5366 74. Buglet left over from Exim 3 conversion. The message "too many messages
5367 in one connection" was written to the rejectlog but not the mainlog, except
5368 when address rewriting (yes!) was being logged.
5370 75. Added write_rejectlog option.
5372 76. When a system filter was run not as root (that is, when system_filter_user
5373 was set), the values of the $n variables were not being returned to the
5374 main process; thus, they were not subsequently available in the $sn
5377 77. Added +return_path_on_delivery log selector.
5379 78. A connection timeout was being treated differently from recipients deferred
5380 when testing hosts_max_try with a message that was older than the host's
5381 retry timeout. (The host should not be counted, thus allowing all hosts to
5382 be tried at least once before bouncing.) This may have been the cause of an
5383 occasionally reported bug whereby a message would remain on the queue
5384 longer than the retry timeout, but would be bounced if a delivery was
5385 forced. I say "may" because I never totally pinned down the problem;
5386 setting up timeout/retry tests is difficult. See also the next item.
5388 79. The ultimate address timeout was not being applied to errors that involved
5389 a combination of host plus message (for example, a timeout on a MAIL
5390 command). When an address resolved to a number of possible hosts, and they
5391 were not all tried for each delivery (e.g. because of hosts_max_try), a
5392 message could remain on the queue longer than the retry timeout.
5394 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5397 81. Fixed an obscure SMTP outgoing bug which required at least the following
5398 conditions: (a) there was another message waiting for the same server;
5399 (b) the server returned 5xx to all RCPT commands in the first message so
5400 that the message was not completed; (c) the server dropped the connection
5401 or gave a negative response to the RSET that Exim sends to abort the
5402 transaction. The observed case was a dropped connection after DATA that had
5403 been sent in pipelining mode. That is, the server had advertised PIPELINING
5404 but was not implementing it correctly. The effect of the bug was incorrect
5405 behaviour, such as trying another host, and this could lead to a crash.
5411 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5412 and daemon.c were passed as pointers to ints; they should have been
5413 pointers to socklen_t variables (which are typically unsigned ints).
5415 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5418 3. Fixed a really odd bug that affected only the testing scheme; patching a
5419 certain fixed string in the binary changed the value of another string that
5420 happened to be identical to the end of the original first string.
5422 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5423 name", it returns that address as the IP address. On some operating
5424 systems (e.g. Solaris), it also passes back the IP address string as the
5425 "host name". However, on others (e.g. Linux), it passes back an empty
5426 string. Exim wasn't checking for this, and was changing the host name to an
5427 empty string, assuming it had been canonicized.
5429 5. Although rare, it is permitted to have more than one PTR record for a given
5430 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5431 all the names associated with an address, because they do in Solaris.
5432 However, it seems that they do not in Linux for data that comes from the
5433 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5434 I found this out when I moved to a new Linux workstation and tried to run
5435 the Exim test suite.
5437 To get round this problem I have changed the code so that it now does its
5438 own call to the DNS to look up PTR records when searching for a host name.
5439 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5440 addresses that are only in /etc/hosts are still found.
5442 This behaviour is, however, controlled by an option called host_lookup_
5443 order, which defaults to "bydns:byaddr". If people want to use the other
5444 order, or indeed, just use one or the other means of lookup, they can
5445 specify it in this variable.
5447 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5448 some operating systems, this comes back from gethostbyaddr() as an empty
5449 string, and this is what Exim used to test for. However, it seems that in
5450 other systems, "." is yielded. Exim now tests for this case too.
5452 7. The values of check_spool_space and check_log_space are now held internally
5453 as a number of kilobytes instead of an absolute number of bytes. If a
5454 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5455 kilobyte. This means that much larger values can be stored.
5457 8. Exim monitor: an attempt to get the action menu when not actually pointing
5458 at a message produces an empty menu entitled "No message selected". This
5459 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5460 no entries in it ("Shell widget menu has zero width and/or height"). So I
5461 have added a single, blank menu entry in this case.
5463 9. Added ${quote_local_part.
5465 10. MIME decoding is now applied to the contents of Subject: header lines when
5468 11. Now that a reference to $sender_host_address automatically causes a reverse
5469 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5470 host lookup before query-style lookups in lists that might use this
5471 variable. This has therefore been abolished, and the "net-" prefix is no
5472 longer necessary for query-style lookups.
5474 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5475 have been a typo for LFLAGS, so it has been changed.
5477 13. The install script calls Exim with "-C /dev/null" in order to find the
5478 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5479 to be output. Howeve, since Exim outputs its version number before the
5480 error, it didn't break the script. It just looked ugly. I fixed this by
5481 always allowing "-C /dev/null" if the caller is root.
5483 14. Ignore overlarge ACL variable number when reading spool file - insurance
5484 against a later release with more variables having written the file.
5486 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5487 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5488 this, as well as the form without the "IPv6" on the front.
5490 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5491 OS/Makefile-Darwin file.
5493 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5494 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5496 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5497 variable. However, some OS use ints for inodes. I've added cast to long int
5498 to get rid of the compiler warning.
5500 19. I had forgotten to lock out "/../" in configuration file names when
5501 ALT_CONFIG_PREFIX was set.
5503 20. Routers used for verification do not need to specify transports. However,
5504 if such a router generated a host list, and callout was configured, Exim
5505 crashed, because it could not find a port number from the (non-existent)
5506 transport. It now assumes port 25 in this circumstance.
5508 21. Added the -t option to exigrep.
5510 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5511 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5512 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5513 them is set, LOOKUP_LSEARCH is forced.
5515 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5517 24. Added sender and host information to the "rejected by local_scan()" log
5518 line; previously there was no indication of these.
5520 25. Added .include_if_exists.
5522 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5523 Linux. It turns out that not all file systems support this. Apparently some
5524 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5525 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5526 not supported on the file descriptor, is now ignored when Exim is trying to
5527 sync a directory. This applies only to Linux.
5529 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5531 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5532 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5535 29. Added support for maildirsize files from supplied patch (modified a bit).
5537 30. The use of :fail: followed by an empty string could lead Exim to respond to
5538 sender verification failures with (e.g.):
5540 550 Verification failed for <xxx>
5541 550 Sender verify failed
5543 where the first response line was missing the '-' that indicates it is not
5544 the final line of the response.
5546 31. The loop for finding the name of the user that called Exim had a hardwired
5547 limit of 10; it now uses the value of finduser_retries, which is used for
5548 all other user lookups.
5550 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5553 33. Exim was neglecting to zero errno before one call of strtol() when
5554 expanding a string and expecting an integer value. On some systems this
5555 resulted in spurious "integer overflow" errors. Also, it was casting the
5556 result into an int without checking.
5558 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5559 did not work. The code looks as if it has *never* worked, though it appears
5560 to have been documented since at least releast 1.62. I have made it work.
5562 35. The "timeout_DNS" error in retry rules, also documented since at least
5563 1.62, also never worked. As it isn't clear exactly what this means, and
5564 clearly it isn't a major issue, I have abolished the feature by treating it
5565 as "timeout", and writing a warning to the main and panic logs.
5567 36. The display of retry rules for -brt wasn't always showing the error code
5570 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5571 timeout_connect_A, timeout_connect_MX.
5573 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5574 to the empty sender.
5576 39. The daemon was not analysing the content of -oX till after it had closed
5577 stderr and disconnected from the controlling terminal. This meant that any
5578 syntax errors were only noted on the panic log, and the return code from
5579 the command was 0. By re-arranging the code a little, I've made the
5580 decoding happen first, so such errors now appear on stderr, and the return
5581 code is 1. However, the actual setting up of the sockets still happens in
5582 the disconnected process, so errors there are still only recorded on the
5585 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5586 connections (as happens on some IP stacks) was logged at start up time as
5587 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5588 it from "IPv6 and IPv4", which means that two separate sockets are being
5591 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5592 says whether AF_INET or AF_INET6 was passed as an argument.
5594 42. Exiwhat output was messed up when time zones were included in log
5597 43. Exiwhat now gives more information about the daemon's listening ports,
5598 and whether -tls-on-connect was used.
5600 44. The "port" option of the smtp transport is now expanded.
5602 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5603 silently ignored. Now an error message is written to the main and panic
5606 46. There's a new ACL modifier called "logwrite" which writes to a log file
5607 as soon as it is encountered.
5609 47. Added $local_user_uid and $local_user_gid at routing time.
5611 48. Exim crashed when trying to verify a sender address that was being
5614 49. Exim was recognizing only a space character after ".include". It now also
5615 recognizes a tab character.
5617 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5618 extracting the relevant information from the specification. The man page no
5619 longer contains scrambled data for the -d option, and I've added a section
5620 at the front about calling Exim under different names.
5622 51. Added "extra_headers" argument to the "mail" command in filter files.
5624 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5627 53. Installed eximstats 1.29.
5629 54. Added transport_filter_timeout as a generic transport option.
5631 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5632 Cc: header lines. This was required by RFC 822, but it not required by RFC
5635 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5636 incoming messages that did not have them. Now it does so only if the
5637 message originates locally, that is, if there is no associated remote host
5638 address. When Resent- header lines are present, this applies to the Resent-
5639 lines rather than the non-Resent- lines.
5641 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5642 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5644 58. Messages for configuration errors now include the name of the main
5645 configuration file - useful now that there may be more than one file in a
5646 list (.included file names were always shown).
5648 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5649 for those rare installations that do not start the daemon as root or run it
5650 setuid root. I've cut out the call to initgroups() if the daemon is not
5653 60. The Exim user and group can now be bound into the binary as text strings
5654 that are looked up at the start of Exim's processing.
5656 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5658 62. Added $mailstore_basename variable.
5660 63. Installed patch to sieve.c from Michael Haardt.
5662 64. When Exim failed to open the panic log after failing to open the main log,
5663 the original message it was trying to log was written to stderr and debug
5664 output, but if they were not available (the usual case in production), it
5665 was lost. Now it is written to syslog before the two lines that record the
5666 failures to open the logs.
5668 65. Users' Exim filters run in subprocesses under the user's uid. It is
5669 possible for a "deliver" command or an alias in a "personal" command to
5670 provoke an address rewrite. If logging of address rewriting is configured,
5671 this fails because the process is not running as root or exim. There may be
5672 a better way of dealing with this, but for the moment (because 4.30 needs
5673 to be released), I have disabled address rewrite logging when running a
5674 filter in a non-root, non-exim process.
5680 1. The buildconfig auxiliary program wasn't quoting the value set for
5681 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5682 not defined. This bug was masked in 4.22 by the effect that was fixed in
5685 2. Some messages that were rejected after a message id was allocated were
5686 shown as "incomplete" by exigrep. It no longer does this for messages that
5687 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5689 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5690 have allow_domain_literals set, the ID did not get logged in the <= line.
5691 Domain literals are now always recognized in Message-ID: header lines.
5693 4. The first argument for a ${extract expansion item is the key name or field
5694 number. Leading and trailing spaces in this item were not being ignored,
5695 causing some misleading effects.
5697 5. When deliver_drop_privilege was set, single queue runner processes started
5698 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5699 same command in the process it spins off) were not dropping privilege.
5701 6. When the daemon running as "exim" started a queue runner, it always
5702 re-executed Exim in the spun-off process. This is a waste of effort when
5703 deliver_drop_privilege is set. The new process now just calls the
5704 queue-runner function directly.
5710 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5713 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5714 ignored. Though the use of -oP was forcing the writing of a pid file, it
5715 was always written to the default place.
5717 3. If the message "no IP address found for host xxxx" is generated during
5718 incoming verification, it is now followed by identification of the incoming
5719 connection (so you can more easily find what provoked it).
5721 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5723 5. Added some features to "harden" Exim a bit more against certain attacks:
5725 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5726 be put in Local/Makefile. This is like the never_users runtime option,
5727 but it cannot be overridden. The default setting is "root".
5729 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5730 prefix string with which any file named in a -C command line option
5733 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5734 is retained for -C and -D only if the caller of Exim is root. Without
5735 it, the exim user may also use -C and -D and retain privilege.
5737 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5738 command line option is disabled.
5740 6. Macro names set by the -D option must start with an upper case letter, just
5741 like macro names defined in the configuration file.
5743 7. Added "dereference=" facility to LDAP.
5745 8. Two instances of the typo "uknown" in the source files are fixed.
5747 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5748 the Configure-Makefile script screwed up while processing it.
5750 10. Incorporated PCRE 4.4.
5752 11. The SMTP synchronization check was not operating right at the start of an
5753 SMTP session. For example, it could not catch a HELO sent before the client
5754 waited for the greeting. There is now a check for outstanding input at the
5755 point when the greeting is written. Because of the duplex, asynchronous
5756 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5757 way, but not yet received, when the check is performed.
5759 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5760 on TCP/IP sockets, because this apparently causes some broken clients to
5763 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5764 unchanged) from the Cygwin maintainer.
5766 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5767 was supported instead of testing for mbx. Effectively a typo.
5769 15. The spa authenticator server code was not checking that the input it
5770 received was valid base64.
5772 16. The debug output line for the "set" modifier in ACLs was not showing the
5773 name of the variable that was being set.
5775 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5777 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5778 lookup on its own. Something else was needed to trigger the lookup. For
5779 example, a match in host_lookup or the need for a host name in a host list.
5780 Now, if $sender_host_name is referenced and the host name has not yet been
5781 looked up, a lookup is performed. If the lookup fails, the variable remains
5782 empty, and $host_lookup_failed is set to "1".
5784 19. Added "eqi" as a case-independent comparison operator.
5786 20. The saslauthd authentication condition could segfault if neither service
5787 nor realm was specified.
5789 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5790 error message that was logged was misleading, and incoming SMTP
5791 connections were dropped. The message is now more accurate, and temporary
5792 errors are given to SMTP connections.
5794 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5795 (except RSET) with a 421 error, until QUIT is received. However, it was
5796 failing to send a response to QUIT.
5798 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5799 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5800 ACL is now run after the helo_try_verify_hosts code.
5802 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5803 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5804 case-independent, so other case variants are also recognized). Apparently
5805 some systems use these upper case variants.
5807 25. If more than two messages were waiting for the same host, and a transport
5808 filter was specified for the transport, Exim sent two messages over the
5809 same TCP/IP connection, and then failed with "socket operation on non-
5810 socket" when it tried to send the third.
5812 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5814 27. The extern definition of crypt16() in expand.c was not being excluded when
5815 the OS had its own crypt16() function.
5817 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5818 as a preferred synonym for return_size_limit, both as an option and as an
5821 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5823 30. Changed the default configuration ACL to relax the local part checking rule
5824 for addresses that are not in any local domains. For these addresses,
5825 slashes and pipe symbols are allowed within local parts, but the sequence
5826 /../ is explicitly forbidden.
5828 31. SPA server authentication was not clearing the challenge buffer before
5831 32. log_message in a "warn" ACL statement was writing to the reject log as
5832 well as to the main log, which contradicts the documentation and doesn't
5833 seem right (because no rejection is happening). So I have stopped it.
5835 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5836 However, I am unable to do any testing of this.
5838 34. Fixed an infelicity in the appendfile transport. When checking directories
5839 for a mailbox, to see if any needed to be created, it was accidentally
5840 using path names with one or more superfluous leading slashes; tracing
5841 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5843 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5844 contents are added to the log line that is written for every discarded
5845 recipient. (Previously a log_message setting was ignored.)
5847 36. The ${quote: operator now quotes the string if it is empty.
5849 37. The install script runs exim in order to find its version number. If for
5850 some reason other than non-existence or emptiness, which it checks, it
5851 could not run './exim', it was installing it with an empty version number,
5852 i.e. as "exim-". This error state is now caught, and the installation is
5855 38. An argument was missing from the function that creates an error message
5856 when Exim fails to connect to the socket for saslauthd authentication.
5857 This could cause Exim to crash, or give a corrupted message.
5859 39. Added isip, isip4, and isip6 to ${if conditions.
5861 40. The ACL variables $acl_xx are now saved with the message, and can be
5862 accessed later in routers, transports, and filters.
5864 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5865 strings in the file are not string-expanded.
5867 42. If a MAIL command specified a SIZE value that was too large to fit into an
5868 int variable, the check against message_size_limit failed. Such values are
5869 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5870 day this will have to be increased, but I don't think I want to be around
5871 when emails are that large.
5878 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5879 iconv() is not standard in FreeBSD.
5881 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5882 IPv6 enabled. The observed symptom was a segmentation fault on return from
5883 the function os_common_find_running_interfaces() in src/os.c.
5885 3. In the check_special_case() function in daemon.c I had used "errno" as an
5886 argument name, which causes warnings on some systems. This was basically a
5887 typo, since it was named "eno" in the comments!
5889 4. The code that waits for the clock to tick (at a resolution of some fraction
5890 of a second) so as to ensure message-id uniqueness was always waiting for
5891 at least one whole tick, when it could have waited for less. [This is
5892 almost certainly not relevant at current processor speeds, where it is
5893 unlikely to ever wait at all. But we try to future-proof.]
5895 5. The function that sleeps for a time interval that includes fractions of a
5896 second contained a race. It did not block SIGALRM between setting the
5897 timer, and suspending (a couple of lines later). If the interval was short
5898 and the sigsuspend() was delayed until after it had expired, the suspension
5899 never ended. On busy systems this could lead to processes getting stuck for
5902 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5903 process, before it forks any delivery processes. The open lookup caching
5904 mechanism meant that the open file or database connection was passed into
5905 the delivery process. The problem was that delivery processes always tidy
5906 up cached lookup data. This could cause a problem for the next delivery
5907 process started by the queue runner, because the external queue runner
5908 process does not know about the closure. So the next delivery process
5909 still has data in the lookup cache. In the case of a file lookup, there was
5910 no problem because closing a file descriptor in a subprocess doesn't affect
5911 the parent. However, if the lookup was caching a connection to a database,
5912 the connection was closed, and the second delivery process was likely to
5913 see errors such as "PGSQL: query failed: server closed the connection
5914 unexpectedly". The problem has been fixed by closing all cached lookups
5915 in a queue runner before running a delivery process.
5917 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5918 seem to have the "const" qualifier which it has on other OS. I've
5921 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5922 *of the same type* (client or server) with the same public name that an
5923 error should be diagnosed.
5925 9. When Exim looked up a host name for an IP address, but failed to find the
5926 original IP address when looking up the host name (a safety check), it
5927 output the message "<ip address> does not match any IP for NULL", which was
5928 confusing, to say the least. The bug was that the host name should have
5929 appeared instead of "NULL".
5931 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5932 user that is built into the binary, and the -C or -D options is used, root
5933 privilege is dropped before the configuration file is read. In addition,
5934 logging is switched to stderr instead of the normal log files. If the
5935 configuration then re-defines the Exim user, the unprivileged environment
5936 is probably not what is expected, so Exim logs a panic warning message (but
5939 However, if deliver_drop_privilege is set, the unprivileged state may well
5940 be exactly what is intended, so the warning has been cut out in that case,
5941 and Exim is allowed to try to write to its normal log files.
5947 1. smtp_return_error_details was not giving details for temporary sender
5948 or receiver verification errors.
5950 2. Diagnose a configuration error if two authenticators have the same public
5953 3. Exim used not to create the message log file for a message until the first
5954 delivery attempt. This could be confusing when incoming messages were held
5955 for policy or load reasons. The message log file is now created at the time
5956 the message is received, and an initial "Received" line is written to it.
5958 4. The automatically generated man page for command line options had a minor
5959 bug that caused no ill effects; however, a more serious problem was that
5960 the procedure for building the man page automatically didn't always
5961 operate. Consequently, release 4.20 contains an out-of-date version. This
5962 shouldn't happen again.
5964 5. When building Exim with embedded Perl support, the script that builds the
5965 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5966 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5968 6. The freeze_tell option was not being used for messages that were frozen on
5969 arrival, either by an ACL or by local_scan().
5971 7. Added the smtp_incomplete_transaction log selector.
5973 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5974 was accepting AUTH without a new EHLO.
5976 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5977 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5980 10. Logging of TCP/IP connections (when configured) now happens in the main
5981 daemon process instead of the child process, so that the TCP/IP connection
5982 count is more accurate (but it can never be perfect).
5984 11. The use of "drop" in a nested ACL was not being handled correctly in the
5985 outer ACL. Now, if condition failure induced by the nested "drop" causes
5986 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5987 or "require"), the connection is dropped.
5989 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5990 that yield "discard" can now be used with an "accept" or a "discard" verb,
5991 but an error is generated for any others (because I can't see a useful way
5992 to define what should happen).
5994 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5995 are now processed in the same way as lines in the Exim configuration file.
5996 In particular, continuation lines are supported.
5998 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6000 15. Added -ti meaning -t -i.
6002 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6003 domains, and warn by logging if others are found.
6005 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6006 returned from GIFCONF when Exim is trying to find the list of interfaces on
6007 a host. The code in os.c has been modified to copy each ifreq to an aligned
6008 structure in all cases.
6010 Also, in some cases, the returned ifreq's were being copied to a 'struct
6011 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6012 means the last couple of bytes of an IPv6 address could be chopped if the
6013 ifreq contained only a normal sockaddr (14 bytes storage).
6015 18. Named domain lists were not supported in the hosts_treat_as_local option.
6016 An entry such as +xxxx was not recognized, and was treated as a literal
6019 19. Ensure that header lines added by a DATA ACL are included in the reject log
6020 if the ACL subsequently rejects the message.
6022 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6023 MD5 (which is deprecated).
6025 21. When testing a filter file using -bf, Exim was writing a message when it
6026 took the sender from a "From " line in the message, but it was not doing so
6027 when it took $return_path from a Return-Path: header line. It now does.
6029 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6030 with a valid header line field name (a series of printing characters
6031 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6033 23. Changed "disc" in the source to "disk" to conform to the documentation and
6034 the book and for uniformity.
6036 24. Ignore Sendmail's -Ooption=value command line item.
6038 25. When execve() failed while trying to run a command in a pipe transport,
6039 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6040 could be confused with a return value of 69 from the command itself. This
6041 has been changed to 127, the value the shell returns if it is asked to run
6042 a non-existent command. The wording for the related log line suggests a
6043 non-existent command as the problem.
6045 26. If received_header_text expands to an empty string, do not add a Received:
6046 header line to the message. (Well, it adds a token one on the spool, but
6047 marks it "old" so that it doesn't get used or transmitted.)
6049 27. Installed eximstats 1.28 (addition of -nt option).
6051 28. There was no check for failure on the call to getsockname() in the daemon
6052 code. This can fail if there is a shortage of resources on the system, with
6053 ENOMEM, for example. A temporary error is now given on failure.
6055 29. Contrary to the C standard, it seems that in some environments, the
6056 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6057 program. Exim now does this explicitly; it affects the formatting of
6058 timestamps using strftime().
6060 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6061 complaints. I've now initialized all the variables, to avoid this.
6063 32. Header lines added by a system filter were not being "seen" during
6064 transport-time rewrites.
6066 33. The info_callback() function passed to OpenSSL is set up with type void
6067 (*)(SSL *, int, int), as described somewhere. However, when calling the
6068 function (actually a macro) that sets it up, the type void(*)() is
6069 expected. I've put in a cast to prevent warnings from picky compilers.
6071 34. If a DNS black list lookup found a CNAME record, but there were no A
6072 records associated with the domain it pointed at, Exim crashed.
6074 35. If a DNS black list lookup returned more than one A record, Exim ignored
6075 all but the first. It now scans all returned addresses if a particular IP
6076 value is being sought. In this situation, the contents of the
6077 $dnslist_value variable are a list of all the addresses, separated by a
6080 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6081 to accept a host name and all its aliases if the forward lookup for any of
6082 them yielded the IP address of the incoming connection. Now it accepts only
6083 those names whose forward lookup yields the correct IP address. Any other
6084 names are discarded. This closes a loophole whereby a rogue DNS
6085 administrator could create reverse DNS records to break through a
6086 wildcarded host restriction in an ACL.
6088 37. If a user filter or a system filter that ran in a subprocess used any of
6089 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6090 the wrong values were passed to the pipe command ($thisaddress had the
6091 value of $0, $0 had the value of $1, etc). This bug was introduced by
6092 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6094 38. Improved the line breaking for long SMTP error messages from ACLs.
6095 Previously, if there was no break point between 40 and 75 characters, Exim
6096 left the rest of the message alone. Two changes have been made: (a) I've
6097 reduced the minimum length to 35 characters; (b) if it can't find a break
6098 point between 35 and 75 characters, it looks ahead and uses the first one
6099 that it finds. This may give the occasional overlong line, but at least the
6100 remaining text gets split now.
6102 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6103 file descriptors might be low, and that setting 1000 would always raise it.
6104 It turns out that in some environments, the limit is already over 1000 and
6105 that lowering it causes trouble. So now Exim takes care not to decrease it.
6107 40. When delivering a message, the value of $return_path is set to $sender_
6108 address at the start of routing (routers may change the value). By an
6109 oversight, this default was not being set up when an address was tested by
6110 -bt or -bv, which affected the outcome if any router or filter referred to
6113 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6114 log only when "message" or "log_message" are set. However, if one of the
6115 conditions was an address verification, or a call to a nested ACL, the
6116 messages generated by the underlying test were being passed through. This
6117 no longer happens. The underlying message is available in $acl_verify_
6118 message for both "message" and "log_message" expansions, so it can be
6119 passed through if needed.
6121 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6122 new expansion $bh_ to give the encoded byte string without charset
6123 translation. Translation happens only if iconv() is available; HAVE_ICONV
6124 indicates this at build time. HEADERS_CHARSET gives the charset to
6125 translate to; headers_charset can change it in the configuration, and
6126 "headers charset" can change it in an individual filter file.
6128 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6129 that creates RFC 2047 encoded "words" labels them as that charset instead
6130 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6131 expansion operator; (ii) when Exim creates a From: line for a local
6132 message; (iii) when a header line is rewritten to include a "phrase" part.
6134 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6135 buggy, causing it to skip the first lines of messages whose message ID
6136 ended in 'D'. This would not have bitten before Exim release 4.14, because
6137 message IDs were unlikely to end in 'D' before then. The effect was to have
6138 incorrect size information for certain domains.
6140 45. #include "config.h" was missing at the start of the crypt16.c module. This
6141 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6144 46. If there was a timeout during a "random" callout check, Exim treated it as
6145 a failure of the random address, and carried on sending RSET and the real
6146 address. If the delay was just some slowness somewhere, the response to the
6147 original RCPT would be taken as a response to RSET and so on, causing
6148 mayhem of various kinds.
6150 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6151 when I implemented it. It didn't allow for the fact that some option values
6152 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6153 the right test for positive values.
6155 48. Domain names in DNS records are case-independent. Exim always looks them up
6156 in lower case. Some resolvers return domain names in exactly the case they
6157 appear in the zone file, that is, they may contain uppercase letters. Not
6158 all resolvers do this - some return always lower case. Exim was treating a
6159 change of case by a resolver as a change of domain, similar to a widening
6160 of a domain abbreviation. This triggered its re-routing code and so it was
6161 trying to route what was effectively the same domain again. This normally
6162 caused routing to fail (because the router wouldn't handle the domain
6163 twice). Now Exim checks for this case specially, and just changes the
6164 casing of the domain that it ultimately uses when it transmits the message
6167 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6170 50. If a filter generated a file delivery with a non-absolute name (possible if
6171 no home directory exists for the router), the forbid_file option was not
6174 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6175 the existing equality matching.
6177 52. Exim was using ints instead of ino_t variables in some places where it was
6178 dealing with inode numbers.
6180 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6181 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6182 called TMPDIR, and if it finds it is different, it changes its value.
6184 54. The smtp_printf() function is now made available to local_scan() so
6185 additional output lines can be written before returning. There is also an
6186 smtp_fflush() function to enable the detection of a dropped connection.
6187 The variables smtp_input and smtp_batched_input are exported to
6190 55. Changed the default runtime configuration: the message "Unknown user"
6191 has been removed from the ACL, and instead placed on the localuser router,
6192 using the cannot_route_message feature. This means that any verification
6193 failures that generate their own messages won't get overridden. Similarly,
6194 the "Unrouteable address" message that was in the ACL for unverifiable
6195 relay addresses has also been removed.
6197 56. Added hosts_avoid_esmtp to the smtp transport.
6199 57. The exicyclog script was not checking for the esoteric option
6200 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6201 will work only if exicyclog is run under the appropriate euid.
6203 58. Following a discussion on the list, the rules by which Exim recognises line
6204 endings on incoming messages have been changed. The -dropcr and drop_cr
6205 options are now no-ops, retained only for backwards compatibility. The
6206 following line terminators are recognized: LF CRLF CR. However, special
6207 processing applies to CR:
6209 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6210 nor a local message in the state where . is a terminator.
6212 (ii) If a bare CR is encountered in a header line, an extra space is added
6213 after the line terminator so as not to end the header. The reasoning
6214 behind this is that bare CRs in header lines are most likely either
6215 to be mistakes, or people trying to play silly games.
6217 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6218 was being incorrectly given as 18 bytes larger than it should have been.
6219 This is a VOB (very old bug).
6221 60. This may never have affected anything current, but just in case it has:
6222 When the local host is found other than at the start of a list of hosts,
6223 the local host, those with the same MX, and any that follow, are discarded.
6224 When the list in question was part of a longer list of hosts, the following
6225 hosts (not currently being processed) were also being discarded. This no
6226 longer happens. I'm not sure if this situation could ever has previously
6229 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6232 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6233 header. This is something that is recommended in a new Internet Draft, and
6234 is something that is documented as being done by Sendmail. There are two
6235 possible values. For messages generated by the autoreply transport, Exim
6238 Auto-Submitted: auto-replied
6240 whereas for all other generated messages (e.g. bounces) it adds
6242 Auto-Submitted: auto-generated
6244 63. The "personal" condition in filters now includes a test for the
6245 Auto-Submitted: header. If it contains the string "auto-" the message it
6246 not considered personal.
6248 64. Added rcpt_include_affixes as a generic transport option.
6250 65. Added queue_only_override (default true).
6252 66. Added the syslog_duplication option.
6254 67. If what should have been the first header line of a message consisted of
6255 a space followed by a colon, Exim was mis-interpreting it as a header line.
6256 It isn't of course - it is syntactically invalid and should therefore be
6257 treated as the start of the message body. The misbehaviour could have
6258 caused a number of strange effects, including loss of data in subsequent
6259 header lines, and spool format errors.
6261 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6262 client host had authenticated. This control can now be exercised by an ACL
6263 for more flexibility.
6265 69. By default, callouts do not happen when testing with -bh. There is now a
6266 variant, -bhc, which does actually run the callout code, including
6267 consulting and updating the callout cache.
6269 70. Added support for saslauthd authentication, courtesy of Alexander
6272 71. If statvfs() failed on the spool or log directories while checking their
6273 size for availability, Exim confusingly gave the error "space shortage".
6274 Furthermore, in debugging mode it crashed with a floating point exception.
6275 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6276 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6277 serious problem, Exim now writes to the main and panic logs when this
6278 happens, with details of the failure. It then refuses to accept the
6279 incoming message, giving the message "spool directory problem" or "log
6280 directory problem" with a 421 code for SMTP messages.
6282 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6283 0, 1, and 2 exist, because some OS complain for execs without them (see
6284 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6285 descriptors. However, the code omitted to check that the open succeeded,
6286 causing mysterious errors if for some reason the permissions on /dev/null
6287 got screwed. Now Exim writes a message to the main and panic logs, and
6288 bombs out if it can't open /dev/null.
6290 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6291 interact so that it is all more flexible. It is supposed to remain
6292 backwards compatible. Also added extra_local_interfaces.
6294 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6295 to bomb out with an assertion failure - to the client this appears as a
6296 connection drop. This problem occurs in the part of the code that was taken
6297 from the Samba project. Fortunately, the assertion is in a very simple
6298 function, so I have fixed this by reproducing the function inline in the
6299 one place where it is called, and arranging for authentication to fail
6300 instead of killing the process with assert().
6302 75. The SPA client code was not working when the server requested OEM rather
6303 than Unicode encoding.
6305 76. Added code to make require_files with a specific uid setting more usable in
6306 the case where statting the file as root fails - usually a non-root-mounted
6307 NFS file system. When this happens and the failure is EACCES, Exim now
6308 forks a subprocess and does the per-uid checking as the relevant uid.
6310 77. Added process_log_path.
6312 78. If log_file_path was not explicitly set, a setting of check_log_space or
6313 check_log_inodes was ignored.
6315 79. If a space check for the spool or log partitions fails, the incident is now
6316 logged. Of course, in the latter case the data may get lost...
6318 80. Added the %p formatting code to string_format() so that it can be used to
6319 print addresses in debug_print(). Adjusted all the address printing in the
6320 debugging in store.c to use %p rather than %d.
6322 81. There was a concern that a line of code in smtp_in.c could overflow a
6323 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6324 initially expressed, the concern was not well-founded, because trailing
6325 spaces are removed early. However, if the trailing spaces were followed by
6326 a NULL, they did not get removed, so the overflow was possible. Two fixes
6329 (a) I re-wrote the offending code in a cleaner fashion.
6330 (b) If an incoming SMTP command contains a NULL character, it is rejected
6333 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6334 runs initgroups(), so that if the Exim user is in any additional groups,
6335 they will be used during message reception.
6341 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob