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.
31 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
34 JH/02 The smtp transport option "multi_domain" is now expanded.
36 JH/03 The smtp transport now requests PRDR by default, if the server offers
39 JH/04 Certificate name checking on server certificates, when exim is a client,
40 is now done by default. The transport option tls_verify_cert_hostnames
41 can be used to disable this per-host. The build option
42 EXPERIMENTAL_CERTNAMES is withdrawn.
44 JH/05 The value of the tls_verify_certificates smtp transport and main options
45 default to the word "system" to access the system default CA bundle.
46 For GnuTLS, only version 3.0.20 or later.
48 JH/06 Verification of the server certificate for a TLS connection is now tried
49 (but not required) by default. The verification status is now logged by
50 default, for both outbound TLS and client-certificate supplying inbound
53 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
56 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
57 Status Notification (bounce) messages are now MIME format per RFC 3464.
58 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
59 under the control of the dsn_advertise_hosts option, and routers may
60 have a dsn_lasthop option.
62 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
63 default, modifiable by a malware= option. The list separator for
64 the options can now be changed in the usual way. Bug 68.
66 JH/10 The smtp_receive_timeout main option is now expanded before use.
68 JH/11 The incoming_interface log option now also enables logging of the
69 local interface on delivery outgoing connections.
71 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
72 if the interface and destination host and port all match.
74 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
77 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
78 Patch from Andrew Lewis.
80 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
81 now supports optional time-restrictions, weighting, and priority
82 modifiers per server. Patch originally by <rommer@active.by>.
84 JH/16 The spamd_address main option now supports a mixed list of local
85 and remote servers. Remote servers can be IPv6 addresses, and
88 JH/17 Bug 68: The spamd_address main option now supports an optional
89 timeout value per server.
91 JH/18 Bug 1581: Router and transport options headers_add/remove can
92 now have the list separator specified.
94 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
97 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
100 JH/21 Support for the A6 type of dns record is withdrawn.
102 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
103 rather than the verbs used.
105 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
106 from 255 to 1024 chars.
108 JH/24 Verification callouts now attempt to use TLS by default.
110 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
111 are generic router options now. The defaults didn't change.
113 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
114 Original patch from Alexander Shikoff, worked over by JH.
116 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
117 files if ZCAT_COMMAND is not executable.
119 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
121 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
123 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
124 Normally benign, it bites when the pair was led to by a CNAME;
125 modern usage is to not canoicalize the domain to a CNAME target
126 (and we were inconsistent anyway for A-only vs AAAA+A).
128 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
130 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
131 when evaluating $sender_host_dnssec.
133 JH/31 Check the HELO verification lookup for DNSSEC, adding new
134 $sender_helo_dnssec variable.
136 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
138 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
140 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
142 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
143 documented as working, but never had. Support all but $spam_report.
145 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
146 added for tls authenticator.
151 TL/01 When running the test suite, the README says that variables such as
152 no_msglog_check are global and can be placed anywhere in a specific
153 test's script, however it was observed that placement needed to be near
154 the beginning for it to behave that way. Changed the runtest perl
155 script to read through the entire script once to detect and set these
156 variables, reset to the beginning of the script, and then run through
157 the script parsing/test process like normal.
159 TL/02 The BSD's have an arc4random API. One of the functions to induce
160 adding randomness was arc4random_stir(), but it has been removed in
161 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
162 function when detected.
164 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
165 cause callback expansion.
167 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
168 syntax errors in an expansion can be treated as a string instead of
169 logging or causing an error, due to the internal use of bool_lax
170 instead of bool when processing it.
172 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
173 server certificates when making smtp deliveries.
175 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
177 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
179 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
181 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
182 Merged patch from Sebastian Wiedenroth.
184 JH/05 Fix results-pipe from transport process. Several recipients, combined
185 with certificate use, exposed issues where response data items split
186 over buffer boundaries were not parsed properly. This eventually
187 resulted in duplicates being sent. This issue only became common enough
188 to notice due to the introduction of conection certificate information,
189 the item size being so much larger. Found and fixed by Wolfgang Breyha.
191 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
192 size buffer was used, resulting in syntax errors when an expansion
195 JH/07 Add support for directories of certificates when compiled with a GnuTLS
196 version 3.3.6 or later.
198 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
199 is EXPERIMENTAL_EVENT, the main-configuration and transport options
200 both become "event_action", the variables become $event_name, $event_data
201 and $event_defer_errno. There is a new variable $verify_mode, usable in
202 routers, transports and related events. The tls:cert event is now also
203 raised for inbound connections, if the main configuration event_action
206 TL/06 In test suite, disable OCSP for old versions of openssl which contained
207 early OCSP support, but no stapling (appears to be less than 1.0.0).
209 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
210 server certificate names available under the smtp transport option
211 "tls_verify_cert_hostname" now do not permit multi-component wildcard
214 JH/10 Time-related extraction expansions from certificates now use the main
215 option "timezone" setting for output formatting, and are consistent
216 between OpenSSL and GnuTLS compilations. Bug 1541.
218 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
219 encoded parameter in the incoming message. Bug 1558.
221 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
222 include certificate info, eximon was claiming there were spoolfile
225 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
227 JH/14 Log delivery-related information more consistently, using the sequence
228 "H=<name> [<ip>]" wherever possible.
230 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
231 are problematic for Debian distribution, omit them from the release
234 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
236 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
239 JH/17 Fix a null-indirection in certextract expansions when a nondefault
240 output list separator was used.
245 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
246 checkers that were complaining about end of non-void function with no
249 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
250 This was a regression intruduced in 4.83 by another bugfix.
252 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
254 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
255 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
261 TF/01 Correctly close the server side of TLS when forking for delivery.
263 When a message was received over SMTP with TLS, Exim failed to clear up
264 the incoming connection properly after forking off the child process to
265 deliver the message. In some situations the subsequent outgoing
266 delivery connection happened to have the same fd number as the incoming
267 connection previously had. Exim would try to use TLS and fail, logging
268 a "Bad file descriptor" error.
270 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
271 utilities have not been installed.
273 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
274 temporary space as the ACL may create new global variables.
276 TL/01 LDAP support uses per connection or global context settings, depending
277 upon the detected version of the libraries at build time.
279 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
280 to extract and use the src ip:port in logging and expansions as if it
281 were a direct connection from the outside internet. PPv2 support was
282 updated based on HAProxy spec change in May 2014.
284 JH/02 Add ${listextract {number}{list}{success}{fail}}.
286 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
287 Properly escape header and check for NULL return.
289 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
292 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
294 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
295 characters in header names, implemented as a verify condition.
296 Contributed by Michael Fischer v. Mollard.
298 TL/05 Rename SPF condition results err_perm and err_temp to standardized
299 results permerror and temperror. Previous values are deprecated but
300 still accepted. In a future release, err_perm and err_temp will be
301 completely removed, which will be a backward incompatibility if the
302 ACL tests for either of these two old results. Patch contributed by
303 user bes-internal on the mailing list.
305 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
307 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
308 selectors, in both main and reject logs.
310 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
313 JH/07 Add malware type "sock" for talking to simple daemon.
315 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
317 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
318 routers/transports under cutthrough routing.
320 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
321 numbers. Touch up "bool" conditional to keep the same definition.
323 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
325 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
327 JH/12 Expand items in router/transport headers_add or headers_remove lists
328 individually rather than the list as a whole. Bug 1452.
330 Required for reasonable handling of multiple headers_ options when
331 they may be empty; requires that headers_remove items with embedded
332 colons must have them doubled (or the list-separator changed).
334 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
335 view the policy declared in the DMARC record. Currently, $dmarc_status
336 is a combined value of both the record presence and the result of the
339 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
341 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
342 dnslookup router and the smtp transport (applying to the forward
345 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
346 of ldap servers used for a specific lookup. Patch provided by Heiko
349 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
350 New variable $lookup_dnssec_authenticated for observability.
352 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
353 Patch submitted by Lars Timman.
355 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
357 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
358 Requires trusted mode and valid format message id, aborts otherwise.
359 Patch contributed by Heiko Schlichting.
361 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
362 certextract with support for various fields. Bug 1358.
364 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
365 is requested by default, modifiable by smtp transport option
368 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
369 operate on certificate variables to give certificate fingerprints
370 Also new ${sha256:cert_variable}.
372 JH/23 The PRDR feature is moved from being Experimental into the mainline.
374 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
375 Christian Aistleitner.
377 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
379 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
380 file. Patch from Wolfgang Breyha.
382 JH/25 Expand the coverage of the delivery $host and $host_address to
383 client authenticators run in verify callout. Bug 1476.
385 JH/26 Port service names are now accepted for tls_on_connect_ports, to
386 align with daemon_smtp_ports. Bug 72.
388 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
389 support and error reporting did not work properly.
391 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
392 and is readable. Patch from Andrew Colin Kissa.
394 TL/14 Enhance documentation of ${run expansion and how it parses the
395 commandline after expansion, particularly in the case when an
396 unquoted variable expansion results in an empty value.
398 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
400 PP/02 Fix internal collision of T_APL on systems which support RFC3123
401 by renaming away from it. Addresses GH issue 15, reported by
404 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
406 TL/15 SECURITY: prevent double expansion in math comparison functions
407 (can expand unsanitized data). Not remotely exploitable.
414 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
416 PP/02 Make -n do something, by making it not do something.
417 When combined with -bP, the name of an option is not output.
419 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
422 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
423 $sender_host_name and config options to manage this, and basic check
426 PP/05 DSCP support for outbound connections and control modifier for inbound.
428 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
429 (Only plugin which currently uses this is kerberos4, which nobody should
430 be using, but we should make it available and other future plugins might
431 conceivably use it, even though it would break NAT; stuff *should* be
432 using channel bindings instead).
434 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
435 name; added for Sendmail compatibility; requires admin caller.
436 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
437 just ignore it); requires trusted caller.
438 Also parse but ignore: -Ac -Am -X<logfile>
441 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
443 TL/02 Add +smtp_confirmation as a default logging option.
445 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
446 Patch by Magnus Holmgren from 2007-02-20.
448 TL/04 Bugzilla 1281 - Spec typo.
449 Bugzilla 1283 - Spec typo.
450 Bugzilla 1290 - Spec grammar fixes.
452 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
454 TL/06 Add Experimental DMARC support using libopendmarc libraries.
456 TL/07 Fix an out of order global option causing a segfault. Reported to dev
457 mailing list by by Dmitry Isaikin.
459 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
461 JH/02 Support "G" suffix to numbers in ${if comparisons.
463 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
465 NM/01 Bugzilla 1197 - Spec typo
466 Bugzilla 1196 - Spec examples corrections
468 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
470 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
471 gnutls_enable_pkcs11, but renamed to more accurately indicate its
474 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
475 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
477 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
478 "acl {{name}{arg}...}", and optional args on acl condition
481 JH/05 Permit multiple router/transport headers_add/remove lines.
483 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
485 JH/07 Avoid using a waiting database for a single-message-only transport.
486 Performance patch from Paul Fisher. Bugzilla 1262.
488 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
491 JH/09 Add $headers_added variable, with content from use of ACL modifier
492 add_header (but not yet added to the message). Bugzilla 199.
494 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
495 Pulled from Bugzilla 817 by Wolfgang Breyha.
497 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
499 (nb: this is the same fix as in Exim 4.80.1)
501 JH/11 Add A= logging on delivery lines, and a client_set_id option on
504 JH/12 Add optional authenticated_sender logging to A= and a log_selector
507 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
509 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
510 advertise SMTP AUTH mechanism to us, instead of a generic
511 protocol violation error. Also, make Exim more robust to bad
512 data from the Dovecot auth socket.
514 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
516 When a queue runner is handling a message, Exim first routes the
517 recipient addresses, during which it prunes them based on the retry
518 hints database. After that it attempts to deliver the message to
519 any remaining recipients. It then updates the hints database using
522 So if a recipient address works intermittently, it can get repeatedly
523 deferred at routing time. The retry hints record remains fresh so the
524 address never reaches the final cutoff time.
526 This is a fairly common occurrence when a user is bumping up against
527 their storage quota. Exim had some logic in its local delivery code
528 to deal with this. However it did not apply to per-recipient defers
529 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
531 This change adds a proper retry rule check during routing so that the
532 final cutoff time is checked against the message's age. We only do
533 this check if there is an address retry record and there is not a
534 domain retry record; this implies that previous attempts to handle
535 the address had the retry_use_local_parts option turned on. We use
536 this as an approximation for the destination being like a local
537 delivery, as in LMTP.
539 I suspect this new check makes the old local delivery cutoff check
540 redundant, but I have not verified this so I left the code in place.
542 TF/02 Correct gecos expansion when From: is a prefix of the username.
544 Test 0254 submits a message to Exim with the header
548 When I ran the test suite under the user fanf2, Exim expanded
549 the header to contain my full name, whereas it should have added
550 a Resent-Sender: header. It erroneously treats any prefix of the
551 username as equal to the username.
553 This change corrects that bug.
555 GF/01 DCC debug and logging tidyup
556 Error conditions log to paniclog rather than rejectlog.
557 Debug lines prefixed by "DCC: " to remove any ambiguity.
559 TF/03 Avoid unnecessary rebuilds of lookup-related code.
561 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
562 Bug spotted by Jeremy Harris; was flawed since initial commit.
563 Would have resulted in OCSP responses post-SNI triggering an Exim
564 NULL dereference and crash.
566 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
568 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
569 Bug detection, analysis and fix by Samuel Thibault.
570 Bugzilla 1331, Debian bug #698092.
572 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
574 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
575 Server implementation by Todd Lyons, client by JH.
576 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
577 config variable "prdr_enable" controls whether the server
578 advertises the facility. If the client requests PRDR a new
579 acl_data_smtp_prdr ACL is called once for each recipient, after
580 the body content is received and before the acl_smtp_data ACL.
581 The client is controlled by bolth of: a hosts_try_prdr option
582 on the smtp transport, and the server advertisement.
583 Default client logging of deliveries and rejections involving
584 PRDR are flagged with the string "PRDR".
586 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
587 fclose(). Diagnosis by Todd Lyons.
589 PP/17 Update configure.default to handle IPv6 localhost better.
590 Patch by Alain Williams (plus minor tweaks).
593 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
594 This is now consistent with GnuTLS, and is now documented: the
595 previous undocumented portable approach to treating the option as
596 unset was to force an expansion failure. That still works, and
597 an empty string is now equivalent.
599 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
600 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
601 not performing validation itself.
603 PP/20 Added force_command boolean option to pipe transport.
604 Patch from Nick Koston, of cPanel Inc.
606 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
609 TF/04 Added udpsend ACL modifer and hexquote expansion operator
611 PP/21 Fix eximon continuous updating with timestamped log-files.
612 Broken in a format-string cleanup in 4.80, missed when I repaired the
613 other false fix of the same issue.
614 Report and fix from Heiko Schlichting.
617 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
618 Report from Prashanth Katuri.
620 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
621 It's SecureTransport, so affects any MacOS clients which use the
622 system-integrated TLS libraries, including email clients.
624 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
625 using a MIME ACL for non-SMTP local injection.
626 Report and assistance in diagnosis by Warren Baker.
628 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
630 JH/16 Fix comparisons for 64b. Bugzilla 1385.
632 TL/09 Add expansion variable $authenticated_fail_id to keep track of
633 last id that failed so it may be referenced in subsequent ACL's.
635 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
638 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
639 ldap library initialization, allowing self-signed CA's to be
640 used. Also properly sets require_cert option later in code by
641 using NULL (global ldap config) instead of ldap handle (per
642 session). Bug diagnosis and testing by alxgomz.
644 TL/12 Enhanced documentation in the ratelimit.pl script provided in
645 the src/util/ subdirectory.
647 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
648 renamed to Transport Post Delivery Action by Jeremy Harris, as
651 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
652 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
653 redis_servers = needs to be configured which will be used by the redis
654 lookup. Patch from Warren Baker, of The Packet Hub.
656 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
658 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
659 hostname or reverse DNS when processing a host list. Used suggestions
660 from multiple comments on this bug.
662 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
664 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
665 Missed a few lines, added it to make the runtest require no keyboard
668 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
669 contains upper case chars. Make router use caseful_local_part.
671 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
672 support when GnuTLS has been built with p11-kit.
678 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
680 This, or similar/improved, will also be change PP/11 of 4.82.
686 PP/01 Handle short writes when writing local log-files.
687 In practice, only affects FreeBSD (8 onwards).
688 Bugzilla 1053, with thanks to Dmitry Isaikin.
690 NM/01 Bugzilla 949 - Documentation tweak
692 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
695 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
697 PP/02 Implemented gsasl authenticator.
699 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
701 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
702 `pkg-config foo` for cflags/libs.
704 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
705 with rest of GSASL and with heimdal_gssapi.
707 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
708 `pkg-config foo` for cflags/libs for the TLS implementation.
710 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
711 properties get this fed in as external SSF. A number of robustness
712 and debugging improvements to the cyrus_sasl authenticator.
714 PP/08 cyrus_sasl server now expands the server_realm option.
716 PP/09 Bugzilla 1214 - Log authentication information in reject log.
717 Patch by Jeremy Harris.
719 PP/10 Added dbmjz lookup type.
721 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
723 PP/12 MAIL args handles TAB as well as SP, for better interop with
724 non-compliant senders.
725 Analysis and variant patch by Todd Lyons.
727 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
728 Bug report from Lars Müller <lars@samba.org> (via SUSE),
729 Patch from Dirk Mueller <dmueller@suse.com>
731 PP/13 tls_peerdn now print-escaped for spool files.
732 Observed some $tls_peerdn in wild which contained \n, which resulted
733 in spool file corruption.
735 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
736 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
737 or write after TLS renegotiation, which otherwise led to messages
740 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
741 as a tracking header (ie: a signed header comes before the signature).
742 Patch from Wolfgang Breyha.
744 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
745 comma-sep list; embedded commas doubled.
747 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
749 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
751 Report and patch from Dmitry Banschikov.
753 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
754 Removed SSL_clear() after SSL_new() which led to protocol negotiation
755 failures. We appear to now support TLS1.1+ with Exim.
757 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
758 lets Exim select keys and certificates based upon TLS SNI from client.
759 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
760 before an outbound SMTP session. New log_selector, +tls_sni.
762 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
763 NULL dereference. Report and patch from Alun Jones.
765 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
766 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
767 Not seeing resolver debug output on NetBSD, but suspect this is a
768 resolver implementation change.
770 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
771 Left warnings. Added "eximon gdb" invocation mode.
773 PP/21 Defaulting "accept_8bitmime" to true, not false.
775 PP/22 Added -bw for inetd wait mode support.
777 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
778 locate the relevant includes and libraries. Made this the default.
780 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
781 Bugzilla 1246, report and most of solution from Tomasz Kusy.
783 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
784 This may cause build issues on older platforms.
786 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
787 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
788 gnutls_require_mac & gnutls_require_protocols (no longer supported).
789 Added SNI support via GnuTLS too.
790 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
792 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
794 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
795 Applied second patch from Janne, implementing suggestion to default
796 multiple-strings-in-record handling to match SPF spec.
798 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
800 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
801 read-only, out of scope).
802 Patch from Wolfgang Breyha, report from Stuart Northfield.
804 PP/29 Fix three issues highlighted by clang analyser static analysis.
805 Only crash-plausible issue would require the Cambridge-specific
806 iplookup router and a misconfiguration.
807 Report from Marcin Mirosław.
809 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
811 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
812 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
813 As part of this, removing so much warning spew let me fix some minor
814 real issues in debug logging.
816 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
817 assignment on my part. Fixed.
819 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
820 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
821 Janne Snabb (who went above and beyond: thank you).
823 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
824 string otherwise requires a connection and a bunch more work and it's
825 relatively easy to get wrong. Should also expose TLS library linkage
828 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
829 64-bit ${eval} (JH/03).
831 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
832 GNU libc to support some of the 64-bit stuff, should not lead to
833 conflicts. Defined before os.h is pulled in, so if a given platform
834 needs to override this, it can.
836 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
837 protection layer was required, which is not implemented.
838 Bugzilla 1254, patch from Wolfgang Breyha.
840 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
841 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
842 tls_dhparam take prime identifiers. Also unbreak combination of
843 OpenSSL+DH_params+TLSSNI.
845 PP/39 Disable SSLv2 by default in OpenSSL support.
851 PP/01 Solaris build fix for Oracle's LDAP libraries.
852 Bugzilla 1109, patch from Stephen Usher.
854 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
856 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
859 TF/02 Fix a couple more cases where we did not log the error message
860 when unlink() failed. See also change 4.74-TF/03.
862 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
863 lock up or crash if it happened to be inside a call to libc when it
864 got a SIGUSR1 from exiwhat.
866 The SIGUSR1 handler appends the current process status to the process
867 log which is later printed by exiwhat. It used to use the general
868 purpose logging code to do this, but several functions it calls are
869 not safe for signals.
871 The new output code in the SIGUSR1 handler is specific to the process
872 log, and simple enough that it's easy to inspect for signal safety.
873 Removing some special cases also simplifies the general logging code.
874 Removing the spurious timestamps from the process log simplifies
877 TF/04 Improved ratelimit ACL condition.
879 The /noupdate option has been deprecated in favour of /readonly which
880 has clearer semantics. The /leaky, /strict, and /readonly update modes
881 are mutually exclusive. The update mode is no longer included in the
882 database key; it just determines when the database is updated. (This
883 means that when you upgrde Exim will forget old rate measurements.)
885 Exim now checks that the per_* options are used with an update mode that
886 makes sense for the current ACL. For example, when Exim is processing a
887 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
888 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
889 must specify per_mail/readonly. If you omit the update mode it defaults to
890 /leaky where that makes sense (as before) or /readonly where required.
892 The /noupdate option is now undocumented but still supported for
893 backwards compatibility. It is equivalent to /readonly except that in
894 ACLs where /readonly is required you may specify /leaky/noupdate or
895 /strict/noupdate which are treated the same as /readonly.
897 A useful new feature is the /count= option. This is a generalization
898 of the per_byte option, so that you can measure the throughput of other
899 aggregate values. For example, the per_byte option is now equivalent
900 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
902 The per_rcpt option has been generalized using the /count= mechanism
903 (though it's more complicated than the per_byte equivalence). When it is
904 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
905 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
906 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
907 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
908 Note that using per_rcpt with a non-readonly update mode in more than
909 one ACL will cause the recipients to be double-counted. (The per_mail
910 and per_byte options don't have this problem.)
912 The handling of very low rates has changed slightly. If the computed rate
913 is less than the event's count (usually one) then this event is the first
914 after a long gap. In this case the rate is set to the same as this event's
915 count, so that the first message of a spam run is counted properly.
917 The major new feature is a mechanism for counting the rate of unique
918 events. The new per_addr option counts the number of different
919 recipients that someone has sent messages to in the last time period. It
920 behaves like per_rcpt if all the recipient addresses are different, but
921 duplicate recipient addresses do not increase the measured rate. Like
922 the /count= option this is a general mechanism, so the per_addr option
923 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
924 example, measure the rate that a client uses different sender addresses
925 with the options per_mail/unique=$sender_address. There are further
926 details in the main documentation.
928 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
930 TF/06 Removed a few PCRE remnants.
932 TF/07 Automatically extract Exim's version number from tags in the git
933 repository when doing development or release builds.
935 PP/02 Raise smtp_cmd_buffer_size to 16kB.
936 Bugzilla 879. Patch from Paul Fisher.
938 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
939 Heavily based on revision 40f9a89a from Simon Arlott's tree.
942 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
944 PP/05 Variable $av_failed, true if the AV scanner deferred.
945 Bugzilla 1078. Patch from John Horne.
947 PP/06 Stop make process more reliably on build failure.
948 Bugzilla 1087. Patch from Heiko Schlittermann.
950 PP/07 Make maildir_use_size_file an _expandable_ boolean.
951 Bugzilla 1089. Patch from Heiko Schlittermann.
953 PP/08 Handle ${run} returning more data than OS pipe buffer size.
954 Bugzilla 1131. Patch from Holger Weiß.
956 PP/09 Handle IPv6 addresses with SPF.
957 Bugzilla 860. Patch from Wolfgang Breyha.
959 PP/10 GnuTLS: support TLS 1.2 & 1.1.
961 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
964 PP/11 match_* no longer expand right-hand-side by default.
965 New compile-time build option, EXPAND_LISTMATCH_RHS.
966 New expansion conditions, "inlist", "inlisti".
968 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
970 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
972 PP/14 fix log_write() format string regression from TF/03.
973 Bugzilla 1152. Patch from Dmitry Isaikin.
979 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
981 PP/02 Harmonised TLS library version reporting; only show if debugging.
982 Layout now matches that introduced for other libraries in 4.74 PP/03.
984 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
986 PP/04 New "dns_use_edns0" global option.
988 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
991 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
992 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
994 TK/01 Updated PolarSSL code to 0.14.2.
995 Bugzilla 1097. Patch from Andreas Metzler.
997 PP/07 Catch divide-by-zero in ${eval:...}.
1000 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1003 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1004 format-string attack -- SECURITY: remote arbitrary code execution.
1006 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1007 time unintentionally subject to list matching rules, letting the header
1008 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1009 arbitrary string expansion). This allowed for information disclosure.
1011 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1012 INT_MIN/-1 -- value coerced to INT_MAX.
1018 NM/01 Workround for PCRE version dependency in version reporting
1021 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1022 This fixes portability to compilers other than gcc, notably
1023 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1025 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1026 makefiles for portability to HP-UX and POSIX correctness.
1028 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1029 Also via indented variable definition in the Makefile.
1030 (Debugging by Oliver Heesakkers).
1032 PP/02 Restore caching of spamd results with expanded spamd_address.
1033 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1035 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1036 Improves build reliability. Fix from: Frank Elsner
1038 NM/02 Fix wide character breakage in the rfc2047 coding
1039 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1041 NM/03 Allow underscore in dnslist lookups
1042 Fixes bug 1026. Patch from Graeme Fowler
1044 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1045 Code patches from Adam Ciarcinski of NetBSD.
1047 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1050 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1051 is logged, to avoid truncation. Patch from John Horne.
1053 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1054 Patch from Jakob Hirsch.
1056 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1057 SQL string expansion failure details.
1058 Patch from Andrey Oktyabrski.
1060 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1061 Patch from Simon Arlott.
1063 PP/09 New lookups functionality failed to compile on old gcc which rejects
1064 extern declarations in function scope.
1065 Patch from Oliver Fleischmann
1067 PP/10 Use sig_atomic_t for flags set from signal handlers.
1068 Check getgroups() return and improve debugging.
1069 Fixed developed for diagnosis in bug 927 (which turned out to be
1072 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1073 Patch from Mark Zealey.
1075 PP/12 Bugzilla 1056: Improved spamd server selection.
1076 Patch from Mark Zealey.
1078 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1079 Based on patch from Heiko Schlittermann.
1081 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1082 Patch from Uwe Doering, sign-off by Michael Haardt.
1084 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1085 variable declaration deep within a block. Bug and patch from
1088 PP/15 lookups-Makefile IRIX compatibilty coercion.
1090 PP/16 Make DISABLE_DKIM build knob functional.
1092 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1093 Patch by Simon Arlott
1095 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1096 variable argument macros. Our copy now differs from upstream.
1102 TF/01 Failure to get a lock on a hints database can have serious
1103 consequences so log it to the panic log.
1105 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1106 controlled using the smtp_confirmation log selector.
1108 TF/03 Include the error message when we fail to unlink a spool file.
1110 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1111 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1112 for maintaining out-of-tree patches for some time.
1114 PP/01 Bugzilla 139: Documentation and portability issues.
1115 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1116 Handle per-OS dynamic-module compilation flags.
1118 PP/02 Let /dev/null have normal permissions.
1119 The 4.73 fixes were a little too stringent and complained about the
1120 permissions on /dev/null. Exempt it from some checks.
1121 Reported by Andreas M. Kirchwitz.
1123 PP/03 Report version information for many libraries, including
1124 Exim version information for dynamically loaded libraries. Created
1125 version.h, now support a version extension string for distributors
1126 who patch heavily. Dynamic module ABI change.
1128 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1129 privilege escalation vulnerability whereby the Exim run-time user
1130 can cause root to append content of the attacker's choosing to
1133 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1136 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1137 If dropping privileges for untrusted macros, we disabled normal logging
1138 on the basis that it would fail; for the Exim run-time user, this is not
1139 the case, and it resulted in successful deliveries going unlogged.
1140 Fixed. Reported by Andreas Metzler.
1146 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1147 only prepend for the Resent-* case. Fixes regression introduced in
1148 Exim 4.70 by NM/22 for Bugzilla 607.
1150 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1151 increasing numbers of administrators be bitten by this.
1153 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1155 PP/03 Bugzilla 994: added openssl_options main configuration option.
1157 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1159 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1161 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1163 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1164 without a peer certificate, leading to a segfault because of an
1165 assumption that peers always have certificates. Be a little more
1166 paranoid. Problem reported by Martin Tscholak.
1168 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1169 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1170 NB: ClamAV planning to remove STREAM in "middle of 2010".
1171 CL also introduces -bmalware, various -d+acl logging additions and
1172 more caution in buffer sizes.
1174 PP/09 Implemented reverse_ip expansion operator.
1176 PP/10 Bugzilla 937: provide a "debug" ACL control.
1178 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1180 PP/12 Bugzilla 973: Implement --version.
1182 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1184 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1186 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1188 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1189 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1190 ignore trailing whitespace.
1192 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1194 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1197 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1198 Notification from Dr Andrew Aitchison.
1200 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1201 ExtendedDetectionInfo response format.
1202 Notification from John Horne.
1204 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1207 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1208 XSL and documented dependency on system catalogs, with examples of how
1211 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1214 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1215 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1216 configuration file which is writeable by the Exim user or group.
1218 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1219 of configuration files to cover files specified with the -C option if
1220 they are going to be used with root privileges, not just the default
1223 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1224 option (effectively making it always true).
1226 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1227 files to be used while preserving root privileges.
1229 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1230 that rogue child processes cannot use them.
1232 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1233 run-time user, instead of root.
1235 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1236 Exim run-time user without dropping privileges.
1238 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1239 result string, instead of calling string_vformat() twice with the same
1242 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1243 for other users. Others should always drop root privileges if they use
1244 -C on the command line, even for a whitelisted configure file.
1246 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1248 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1254 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1255 $data_path, and $header_path variables; fixed documentation bugs and
1258 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1259 exipick to access non-standard spools, including the "frozen" queue
1262 NM/01 Bugzilla 965: Support mysql stored procedures.
1263 Patch from Alain Williams
1265 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1267 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1268 Patch from Andreas Metzler
1270 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1271 Patch from Kirill Miazine
1273 NM/05 Bugzilla 671: Added umask to procmail example.
1275 JJ/03 installed exipick 20100323.0, fixing doc bug
1277 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1278 directory. Notification and patch from Dan Rosenberg.
1280 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1282 TK/02 Improve log output when DKIM signing operation fails.
1284 MH/01 Treat the transport option dkim_domain as a colon separated
1285 list, not as a single string, and sign the message with each element,
1286 omitting multiple occurences of the same signer.
1288 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1289 Bugzilla 985, 986. Patch by Simon Arlott
1291 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1292 Patch by Simon Arlott
1294 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1295 MBX locking. Notification from Dan Rosenberg.
1301 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1303 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1305 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1307 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1309 NM/04 Bugzilla 915: Flag broken perl installation during build.
1315 TK/01 Added patch by Johannes Berg that expands the main option
1316 "spamd_address" if it starts with a dollar sign.
1318 TK/02 Write list of recipients to X-Envelope-Sender header when building
1319 the mbox-format spool file for content scanning (suggested by Jakob
1322 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1323 (http://www.dcc-servers.net/) support via dccifd. Activated by
1324 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1326 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1327 by Mark Daniel Reidel <mr@df.eu>.
1329 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1330 When building exim an external PCRE library is now needed -
1331 PCRE is a system library on the majority of modern systems.
1332 See entry on PCRE_LIBS in EDITME file.
1334 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1335 conversation. Added nologin parameter to request.
1336 Patch contributed by Kirill Miazine.
1338 TF/01 Do not log submission mode rewrites if they do not change the address.
1340 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1342 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1343 log files in place. Contributed by Roberto Lima.
1345 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1347 TF/03 Bugzilla 615: When checking the local_parts router precondition
1348 after a local_part_suffix or local_part_prefix option, Exim now
1349 does not use the address's named list lookup cache, since this
1350 contains cached lookups for the whole local part.
1352 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1353 Robert Millan. Documentation is in experimental-spec.txt.
1355 TF/04 Bugzilla 668: Fix parallel build (make -j).
1357 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1359 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1360 Patch provided by Jan Srzednicki.
1362 TF/05 Leading white space used to be stripped from $spam_report which
1363 wrecked the formatting. Now it is preserved.
1365 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1366 that they are available at delivery time.
1368 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1370 TF/08 TLS error reporting now respects the incoming_interface and
1371 incoming_port log selectors.
1373 TF/09 Produce a more useful error message if an SMTP transport's hosts
1374 setting expands to an empty string.
1376 NM/06 Bugzilla 744: EXPN did not work under TLS.
1377 Patch provided by Phil Pennock.
1379 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1380 Patch provided by Richard Godbee.
1382 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1383 acl_smtp_notquit, added index entry.
1385 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1386 Patch provided by Eugene Bujak.
1388 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1389 accept(). Patch provided by Maxim Dounin.
1391 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1392 Patch provided by Phil Pennock.
1394 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1396 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1397 Patch provided by Brad "anomie" Jorsch.
1399 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1400 Patch provided by Dean Brooks.
1402 TK/05 Add native DKIM support (does not depend on external libraries).
1404 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1405 Patch provided by Graeme Fowler.
1407 NM/16 Bugzilla 851: Documentation example syntax fix.
1409 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1411 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1414 NM/19 Bugzilla 745: TLS version reporting.
1415 Patch provided by Phil Pennock.
1417 NM/20 Bugzilla 167: bool: condition support.
1418 Patch provided by Phil Pennock.
1420 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1421 clients. Patch provided by Phil Pennock.
1423 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1424 Patch provided by Brad "anomie" Jorsch.
1426 NM/23 Bugzilla 687: Fix misparses in eximstats.
1427 Patch provided by Heiko Schlittermann.
1429 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1430 Patch provided by Heiko Schlittermann.
1432 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1433 plus update to original patch.
1435 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1437 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1438 Patch provided by David Brownlee.
1440 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1442 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1444 NM/30 Bugzilla 888: TLS documentation bugfixes.
1446 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1448 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1449 Unlike the original bugzilla I have changed all shell scripts in src tree.
1451 NM/33 Bugzilla 898: Transport filter timeout fix.
1452 Patch by Todd Rinaldo.
1454 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1455 Patch by Serge Demonchaux.
1457 NM/35 Bugzilla 39: Base64 decode bug fixes.
1458 Patch by Jakob Hirsch.
1460 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1462 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1464 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1466 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1472 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1473 ALT-N's libdkim that I have put here:
1474 http://duncanthrax.net/exim-experimental/
1476 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1477 were called 'true' and it seems that C99 defines that as a reserved
1478 keyword to be used with 'bool' variable types. That means you could
1479 not include C99-style headers which use bools without triggering
1480 build errors in sieve.c.
1482 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1483 as mailq or other aliases. Changed the --help handling significantly
1484 to do whats expected. exim_usage() emits usage/help information.
1486 SC/01 Added the -bylocaldomain option to eximstats.
1488 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1490 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1492 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1498 PH/01 Another patch from the Sieve maintainer.
1500 PH/02 When an IPv6 address is converted to a string for single-key lookup
1501 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1502 dots are used instead of colons so that keys in lsearch files need not
1503 contain colons. This was done some time before quoting was made available
1504 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1505 (notated using the quote facility) so as to distinguish them from IPv4
1506 keys. This meant that lookups for IP addresses in host lists did not work
1507 for iplsearch lookups.
1509 This has been fixed by arranging for IPv6 addresses to be expressed with
1510 colons if the lookup type is iplsearch. This is not incompatible, because
1511 previously such lookups could never work.
1513 The situation is now rather anomolous, since one *can* have colons in
1514 ordinary lsearch keys. However, making the change in all cases is
1515 incompatible and would probably break a number of configurations.
1517 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1520 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1521 conversion specification without a maximum field width, thereby enabling
1522 a rogue spamd server to cause a buffer overflow. While nobody in their
1523 right mind would setup Exim to query an untrusted spamd server, an
1524 attacker that gains access to a server running spamd could potentially
1525 exploit this vulnerability to run arbitrary code as the Exim user.
1527 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1528 $primary_hostname instead of what libspf2 thinks the hosts name is.
1530 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1531 a directory entry by the name of the lookup key. Previously, if a
1532 symlink pointed to a non-existing file or a file in a directory that
1533 Exim lacked permissions to read, a lookup for a key matching that
1534 symlink would fail. Now it is enough that a matching directory entry
1535 exists, symlink or not. (Bugzilla 503.)
1537 PH/03 The body_linecount and body_zerocount variables are now exported in the
1540 PH/04 Added the $dnslist_matched variable.
1542 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1543 This means they are set thereafter only if the connection becomes
1546 PH/06 Added the client_condition to authenticators so that some can be skipped
1547 by clients under certain conditions.
1549 PH/07 The error message for a badly-placed control=no_multiline_responses left
1550 "_responses" off the end of the name.
1552 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1554 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1555 (without spaces) instead of just copying the configuration text.
1557 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1559 PH/11 Added $max_received_linelength.
1561 PH/12 Added +ignore_defer and +include_defer to host lists.
1563 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1564 way in which PCRE > 7.0 is built.
1566 PH/14 Implemented queue_only_load_latch.
1568 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1569 MAIL command. The effect was to mangle the value on 64-bit systems.
1571 PH/16 Another patch from the Sieve maintainer.
1573 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1575 PH/18 If a system quota error occurred while trying to create the file for
1576 a maildir delivery, the message "Mailbox is full" was not appended to the
1577 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1578 only to a quota excession during the actual writing of the file.
1580 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1581 characters?) which causes problems in log lines. The DN values are now
1582 passed through string_printing() before being added to log lines.
1584 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1585 and InterBase are left for another time.)
1587 PH/21 Added message_body_newlines option.
1589 PH/22 Guard against possible overflow in moan_check_errorcopy().
1591 PH/23 POSIX allows open() to be a macro; guard against that.
1593 PH/24 If the recipient of an error message contained an @ in the local part
1594 (suitably quoted, of course), incorrect values were put in $domain and
1595 $local_part during the evaluation of errors_copy.
1601 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1602 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1605 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1606 issue a MAIL command.
1608 PH/02 In an ACL statement such as
1610 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1612 if a client was not listed at all, or was listed with a value other than
1613 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1614 the condition was not true (as it should be), so access was not denied.
1615 The bug was that the ! inversion was incorrectly passed on to the second
1616 item. This has been fixed.
1618 PH/03 Added additional dnslists conditions == and =& which are different from
1619 = and & when the dns lookup returns more than one IP address.
1621 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1622 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1624 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1625 FSYNC, which compiles an option called disable_fsync that allows for
1626 bypassing fsync(). The documentation is heavily laced with warnings.
1628 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1630 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1631 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1632 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1633 including adding "make clean"; (3) Added -fPIC when compiling the test
1634 dynamically loaded module, to get rid of a warning.
1636 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1637 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1638 = 0s. The bug is otherwise harmless.
1640 PH/07 There was a bug in the dovecot authenticator such that the value of
1641 $auth1 could be overwritten, and so not correctly preserved, after a
1642 successful authentication. This usually meant that the value preserved by
1643 the server_setid option was incorrect.
1645 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1647 PH/09 Installed PCRE release 7.0.
1649 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1650 run for batched SMTP input. It is now run at the start of every message
1651 in the batch. While fixing this I discovered that the process information
1652 (output by running exiwhat) was not always getting set for -bs and -bS
1653 input. This is fixed, and it now also says "batched" for BSMTP.
1655 PH/11 Added control=no_pipelining.
1657 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1658 patch, slightly modified), and move the expansion of helo_data till after
1659 the connection is made in the smtp transport (so it can use these
1662 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1664 PH/14 Added log_selector = +pid.
1666 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1668 PH/16 Add ${if forany and ${if forall.
1670 PH/17 Added dsn_from option to vary the From: line in DSNs.
1672 PH/18 Flush SMTP output before performing a callout, unless control =
1673 no_callout_flush is set.
1675 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1676 was true (the default) a successful delivery failed to delete the retry
1677 item, thus causing premature timeout of the address. The bug is now
1680 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1682 PH/21 Long custom messages for fakedefer and fakereject are now split up
1683 into multiline reponses in the same way that messages for "deny" and
1684 other ACL rejections are.
1686 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1687 with slight modification.
1689 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1690 draft, changing the syntax and factoring some duplicate code".
1692 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1693 for deliveries of the second and subsequent messages over the same SMTP
1696 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1697 ${reduce, with only minor "tidies".
1699 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1701 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1702 expansion side effects.
1704 PH/27 When a message times out after an over-quota error from an Exim-imposed
1705 quota, the bounce message says "mailbox is full". This message was not
1706 being given when it was a system quota that was exceeded. It now should
1709 MH/03 Made $recipients available in local_scan(). local_scan() already has
1710 better access to the recipient list through recipients_list[], but
1711 $recipients can be useful in postmaster-provided expansion strings.
1713 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1714 in the case of a MAIL command with additional options following the
1715 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1716 were accidentally chopped off.
1718 PH/29 SMTP synchronization checks are implemented when a command is read -
1719 there is a check that no more input is waiting when there shouldn't be
1720 any. However, for some commands, a delay in an ACL can mean that it is
1721 some time before the response is written. In this time, more input might
1722 arrive, invalidly. So now there are extra checks after an ACL has run for
1723 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1724 pipelining has not been advertised.
1726 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1728 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1729 returned in the SMTP response when smtp_return_error_details was false.
1730 This has been fixed.
1732 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1733 instead of the C I/O that was originally supplied, because problems were
1734 reported on Solaris.
1736 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1737 Exim which did not show up earlier: it was assuming that a call to
1738 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1739 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1740 was a macro that became an assignment, so it seemed to work. This has
1741 changed to a proper function call with a void return, hence the compile
1742 error. Exim's code has been fixed.
1744 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1747 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1749 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1751 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1752 because it turns out to be too big or there is a timeout) there is no
1753 "Completed" line in the log. When some messages of this type were
1754 selected by exigrep, they were listed as "not completed". Others were
1755 picked up by some special patterns. I have improved the selection
1756 criteria to be more general.
1758 PH/38 The host_find_failed option in the manualroute router can now be set
1759 to "ignore", to completely ignore a host whose IP address cannot be
1760 found. If all hosts are ignored, the behaviour is controlled by the new
1761 host_all_ignored option.
1763 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1764 homing or because of multiple MX records with /mx) generated more than
1765 one IP address, and the following item turned out to be the local host,
1766 all the secondary addresses of the first item were incorrectly removed
1767 from the list, along with the local host and any following hosts (which
1768 is what is supposed to happen).
1770 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1771 whoever called Exim into the -H file. In the case of the daemon it was
1772 behaving confusingly. When first started, it used values for whoever
1773 started the daemon, but after a SIGHUP it used the Exim user (because it
1774 calls itself on a restart). I have changed the code so that it now always
1777 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1778 message are rejected with the same error (e.g. no authentication or bad
1779 sender address), and a DATA command is nevertheless sent (as can happen
1780 with PIPELINING or a stupid MUA), the error message that was given to the
1781 RCPT commands is included in the rejection of the DATA command. This is
1782 intended to be helpful for MUAs that show only the final error to their
1785 PH/42 Another patch from the Sieve maintainer.
1787 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1788 Eximstats - Fixed some broken HTML links and added missing column headers
1790 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1791 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1793 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1795 PH/43 Yet another patch from the Sieve maintainer.
1797 PH/44 I found a way to check for a TCP/IP connection going away before sending
1798 the response to the final '.' that terminates a message, but only in the
1799 case where the client has not sent further data following the '.'
1800 (unfortunately, this is allowed). However, in many cases there won't be
1801 any further data because there won't be any more messages to send. A call
1802 to select() can be used: if it shows that the input is "ready", there is
1803 either input waiting, or the socket has been closed. An attempt to read
1804 the next input character can distinguish the two cases. Previously, Exim
1805 would have sent an OK response which the client would never have see.
1806 This could lead to message repetition. This fix should cure that, at
1807 least in a lot of common cases.
1809 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1810 advertised in response to EHLO.
1816 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1817 fixed by 4.65/MH/01 (is this a record?) are fixed:
1819 (i) An empty string was always treated as zero by the numeric comparison
1820 operators. This behaviour has been restored.
1822 (ii) It is documented that the numeric comparison operators always treat
1823 their arguments as decimal numbers. This was broken in that numbers
1824 starting with 0 were being interpreted as octal.
1826 While fixing these problems I realized that there was another issue that
1827 hadn't been noticed. Values of message_size_limit (both the global option
1828 and the transport option) were treated as octal if they started with 0.
1829 The documentation was vague. These values are now always treated as
1830 decimal, and I will make that clear in the documentation.
1836 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1837 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1840 MH/01 Don't check that the operands of numeric comparison operators are
1841 integers when their expansion is in "skipping" mode (fixes bug
1842 introduced by 4.64-PH/07).
1844 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1845 child addresses, Exim now panics and dies. Previously, because the count
1846 is held in a short int, deliveries were likely to be lost. As such a
1847 large number of recipients for a single message is ridiculous
1848 (performance will be very, very poor), I have chosen to impose a limit
1849 rather than extend the field.
1855 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1856 leftover -K file (the existence of which was triggered by #402).
1857 While we were at it, introduced process PID as part of the -K
1858 filename. This should rule out race conditions when creating
1861 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1862 processing considerably. Previous code took too long for large mails,
1863 triggering a timeout which in turn triggers #401.
1865 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1866 in the DK code in transports.c. sendfile() is not really portable,
1867 hence the _LINUX specificness.
1869 TF/01 In the add_headers option to the mail command in an Exim filter,
1870 there was a bug that Exim would claim a syntax error in any
1871 header after the first one which had an odd number of characters
1874 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1875 callout verification, Exim cached a "reject" for the entire domain. This
1876 is correct for most verifications, but it is not correct for a recipient
1877 verification with use_sender or use_postmaster set, because in that case
1878 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1879 case of MAIL FROM:<> rejection from other early rejections (e.g.
1880 rejection of HELO). When verifying a recipient using a non-null MAIL
1881 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1882 Whatever the result of the callout, the value of the domain cache is
1883 left unchanged (for any other kind of callout, getting as far as trying
1884 RCPT means that the domain itself is ok).
1886 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1889 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1890 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1891 session, the daemon ignores ECONNECT errors and logs others; it now
1892 ignores EPIPE as well.
1894 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1895 (quoted-printable decoding).
1897 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1898 later the small subsequent patch to fix an introduced bug.
1900 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1902 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1904 PH/08 An error is now given if message_size_limit is specified negative.
1906 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1907 to be given (somewhat) arbitrary names.
1909 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1912 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1913 miscellaneous code fixes
1915 PH/10 Added the log_reject_target ACL modifier to specify where to log
1918 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1919 hostname. This is wrong, because it relates to the incoming message (and
1920 probably the interface on which it is arriving) and not to the outgoing
1921 callout (which could be using a different interface). This has been
1922 changed to use the value of the helo_data option from the smtp transport
1923 instead - this is what is used when a message is actually being sent. If
1924 there is no remote transport (possible with a router that sets up host
1925 addresses), $smtp_active_hostname is used.
1927 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1928 tweaks were necessary in order to get it to work (see also 21 below):
1929 (a) The code assumed that strncpy() returns a negative number on buffer
1930 overflow, which isn't the case. Replaced with Exim's string_format()
1932 (b) There were several signed/unsigned issues. I just did the minimum
1933 hacking in of casts. There is scope for a larger refactoring.
1934 (c) The code used strcasecmp() which is not a standard C function.
1935 Replaced with Exim's strcmpic() function.
1936 (d) The code set only $1; it now sets $auth1 as well.
1937 (e) A simple test gave the error "authentication client didn't specify
1938 service in request". It would seem that Dovecot has changed its
1939 interface. Fortunately there's a specification; I followed it and
1940 changed what the client sends and it appears to be working now.
1942 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1945 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1946 address A is aliased to B and C, where B exists and C does not. Without
1947 -v the output is "A verified" because verification stops after a
1948 successful redirection if more than one address is generated. However,
1949 with -v the child addresses are also verified. Exim was outputting "A
1950 failed to verify" and then showing the successful verification for C,
1951 with its parentage. It now outputs "B failed to verify", showing B's
1952 parentage before showing the successful verification of C.
1954 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1955 look up a TXT record in a specific list after matching in a combined
1958 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1959 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1960 they consult the DNS. I had assumed they would set it the way they
1961 wanted; and indeed my experiments on Linux seem to show that in some
1962 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1963 To be on the safe side, however, I have now made the interface to
1964 host_find_byname() similar to host_find_bydns(), with an argument
1965 containing the DNS resolver options. The host_find_byname() function now
1966 sets these options at its start, just as host_find_bydns() does. The smtp
1967 transport options dns_qualify_single and dns_search_parents are passed to
1968 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1969 of host_find_byname() use the default settings of RES_DEFNAMES
1970 (qualify_single) but not RES_DNSRCH (search_parents).
1972 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1973 spool_read_header() do less string testing, by means of a preliminary
1974 switch on the second character of optional "-foo" lines. (This is
1975 overdue, caused by the large number of possibilities that now exist.
1976 Originally there were few.) While I was there, I also converted the
1977 str(n)cmp tests so they don't re-test the leading "-" and the first
1978 character, in the hope this might squeeze out yet more improvement.
1980 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1981 flag allowing group syntax was set by the header_syntax check but not
1982 turned off, possible causing trouble later; (2) The flag was not being
1983 set at all for the header_verify test, causing "group"-style headers to
1984 be rejected. I have now set it in this case, and also caused header_
1985 verify to ignore an empty address taken from a group. While doing this, I
1986 came across some other cases where the code for allowing group syntax
1987 while scanning a header line wasn't quite right (mostly, not resetting
1988 the flag correctly in the right place). These bugs could have caused
1989 trouble for malformed header lines. I hope it is now all correct.
1991 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1992 with the "reply" argument non-NULL. The code, however (which originally
1993 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1994 but it didn't always do it. This confused somebody who was copying the
1995 code for some other use. I have removed all the tests.
1997 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1998 feature that was used to support insecure browsers during the U.S. crypto
1999 embargo. It requires special client support, and Exim is probably the
2000 only MTA that supported it -- and would never use it because real RSA is
2001 always available. This code has been removed, because it had the bad
2002 effect of slowing Exim down by computing (never used) parameters for the
2003 RSA_EXPORT functionality.
2005 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2006 authenticator to fail if there's a tab character in the incoming data
2007 (there should never be unless someone is messing about, as it's supposed
2008 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2009 if the connection is using TLS or if the remote IP is the same as the
2010 local IP, and the "valid-client-cert option" if a client certificate has
2013 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2014 authenticators. This can be used for authorization after authentication
2015 succeeds. (In the case of plaintext, it servers for both authentication
2018 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2019 if any retry times were supplied.
2021 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2022 connection, where there is no client IP address to check. In this
2023 situation, the verify now always succeeds.
2025 PH/25 Applied John Jetmore's -Mset patch.
2027 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2029 PH/27 In a string expansion for a processed (not raw) header when multiple
2030 headers of the same name were present, leading whitespace was being
2031 removed from all of them, but trailing whitespace was being removed only
2032 from the last one. Now trailing whitespace is removed from each header
2033 before concatenation. Completely empty headers in a concatenation (as
2034 before) are ignored.
2036 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2037 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2039 PH/29 [Removed. This was a change that I later backed out, and forgot to
2040 correct the ChangeLog entry (that I had efficiently created) before
2041 committing the later change.]
2043 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2044 address errors (4xx response to RCPT) over the same connection as other
2045 messages routed to the same hosts. Such deliveries are always "forced",
2046 so retry times are not inspected. This resulted in far too many retries
2047 for the affected addresses. The effect occurred only when there were more
2048 hosts than the hosts_max_try setting in the smtp transport when it had
2049 the 4xx errors. Those hosts that it had tried were not added to the list
2050 of hosts for which the message was waiting, so if all were tried, there
2051 was no problem. Two fixes have been applied:
2053 (i) If there are any address or message errors in an SMTP delivery, none
2054 of the hosts (tried or untried) are now added to the list of hosts
2055 for which the message is waiting, so the message should not be a
2056 candidate for sending over the same connection that was used for a
2057 successful delivery of some other message. This seems entirely
2058 reasonable: after all the message is NOT "waiting for some host".
2059 This is so "obvious" that I'm not sure why it wasn't done
2060 previously. Hope I haven't missed anything, but it can't do any
2061 harm, as the worst effect is to miss an optimization.
2063 (ii) If, despite (i), such a delivery is accidentally attempted, the
2064 routing retry time is respected, so at least it doesn't keep
2065 hammering the server.
2067 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2068 in ${readsocket because some servers need this prod.
2070 PH/32 Added some extra debug output when updating a wait-xxx database.
2072 PH/33 The hint "could be header name not terminated by colon", which has been
2073 given for certain expansion errors for a long time, was not being given
2074 for the ${if def:h_colon_omitted{... case.
2076 PH/34 The spec says: "With one important exception, whenever a domain list is
2077 being scanned, $domain contains the subject domain." There was at least
2078 one case where this was not true.
2080 PH/35 The error "getsockname() failed: connection reset by peer" was being
2081 written to the panic log as well as the main log, but it isn't really
2082 panic-worthy as it just means the connection died rather early on. I have
2083 removed the panic log writing for the ECONNRESET error when getsockname()
2086 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2087 runs only) independently of the message's sender address. This meant
2088 that, if the 4xx error was in fact related to the sender, a different
2089 message to the same recipient with a different sender could confuse
2090 things. In particualar, this can happen when sending to a greylisting
2091 server, but other circumstances could also provoke similar problems.
2092 I have changed the default so that the retry time for these errors is now
2093 based a combination of the sender and recipient addresses. This change
2094 can be overridden by setting address_retry_include_sender=false in the
2097 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2098 remote server are returned as part of bounce messages. This was not
2099 happening for LMTP over a pipe (the lmtp transport), but now it is the
2100 same for both kinds of LMTP.
2102 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2103 in header lines that were in fact CNAMEs. This is no longer the case.
2105 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2106 and queue runs started by the daemon processed all messages. This has
2107 been fixed so that -R and -S can now usefully be given with -q<time>.
2109 PH/40 Import PCRE release 6.7 (fixes some bugs).
2111 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2113 PH/42 Give an error if -q is specified more than once.
2115 PH/43 Renamed the variables $interface_address and $interface_port as
2116 $received_ip_address and $received_port, to make it clear that these
2117 values apply to message reception, and not to the outgoing interface when
2118 a message is delivered. (The old names remain recognized, of course.)
2120 PH/44 There was no timeout on the connect() call when using a Unix domain
2121 socket in the ${readsocket expansion. There now is.
2123 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2124 be meaningful with "accept".
2126 SC/01 Eximstats V1.43
2127 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2129 SC/02 Eximstats V1.44
2130 Use a glob alias rather than an array ref in the generated
2131 parser. This improves both readability and performance.
2133 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2134 Collect SpamAssassin and rejection statistics.
2135 Don't display local sender or destination tables unless
2136 there is data to show.
2137 Added average volumes into the top table text output.
2139 SC/04 Eximstats V1.46
2140 Collect data on the number of addresses (recipients)
2141 as well as the number of messages.
2143 SC/05 Eximstats V1.47
2144 Added 'Message too big' to the list of mail rejection
2145 reasons (thanks to Marco Gaiarin).
2147 SC/06 Eximstats V1.48
2148 Mainlog lines which have GMT offsets and are too short to
2149 have a flag are now skipped.
2151 SC/07 Eximstats V1.49 (Alain Williams)
2152 Added the -emptyok flag.
2154 SC/08 Eximstats V1.50
2155 Fixes for obtaining the IP address from reject messages.
2157 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2158 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2159 whitesspace changes from 4.64-PH/27
2161 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2164 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2165 are found, allow negative numbers in numeric criteria)
2167 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2169 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2170 to match changes made in 4.64-PH/43
2172 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2174 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2175 "domain" is actually an IP address, and force a failure. This locks out
2176 those revolvers/nameservers that support "A-for-A" lookups, in
2177 contravention of the specifications.
2179 PH/48 When a host name was looked up from an IP address, and the subsequent
2180 forward lookup of the name timed out, the host name was left in
2181 $sender_host_name, contrary to the specification.
2183 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2184 restricted to single-key lookups, Exim was not diagnosing an error if
2185 * or *@ was used with a query-style lookup.
2187 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2189 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2190 long ago, but noone interested enough thought of it. Let's just say that
2191 the "1.1" means that there are some new functions that weren't there at
2192 some point in the past.
2194 PH/51 Error processing for expansion failure of helo_data from an smtp
2195 transport during callout processing was broken.
2197 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2198 tested/used via the -bh/-bhc/-bs options.
2200 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2201 bug, fixed in subsequent PCRE releases).
2203 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2204 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2206 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2212 SC/01 Use a glob alias rather than an array ref in eximstats generated
2213 parser. This improves both readability and performance.
2215 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2216 Don't display local sender or destination tables in eximstats unless
2217 there is data to show.
2218 Added average volumes into the eximstats top table text output.
2220 SC/03 Collect data on the number of addresses (recipients) as well
2221 as the number of messages in eximstats.
2223 TF/01 Correct an error in the documentation for the redirect router. Exim
2224 does (usually) call initgroups() when daemonizing.
2226 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2227 with consistent privilege compared to when running as a daemon.
2229 TF/03 Note in the spec that $authenticated_id is not set for local
2230 submissions from trusted users.
2232 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2233 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2235 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2236 by adding some example configuration directives to the default
2237 configuration file. A little bit of work is required to uncomment the
2238 directives and define how usernames and passwords are checked, but
2239 there is now a framework to start from.
2241 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2242 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2243 without this. I don't know how relevant this is to other LDAP libraries.
2245 PH/02 Add the verb name to the "unknown ACL verb" error.
2247 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2249 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2251 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2252 directory not expanded when it should be if an expanded home directory
2253 was set for the address (which is overridden by the transport).
2255 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2258 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2259 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2260 because it is too late at that time, and has no effect.
2262 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2263 security issue with \' (bugzilla #107). I could not use the
2264 PQescapeStringConn() function, because it needs a PGconn value as one of
2267 PH/08 When testing addresses using -bt, indicate those final addresses that
2268 are duplicates that would not cause an additional delivery. At least one
2269 person was confused, thinking that -bt output corresponded to deliveries.
2270 (Suppressing duplicates isn't a good idea as you lose the information
2271 about possibly different redirections that led to the duplicates.)
2273 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2274 systems where poll() doesn't work, in particular OS X.
2276 PH/10 Added more information to debugging output for retry time not reached.
2278 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2279 operations in malware.c.
2281 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2284 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2285 syslog_duplication set false, log lines that would normally be written
2286 both the the main log and to the reject log were not written to syslog at
2289 PH/14 In the default configuration, change the use of "message" in ACL warn
2290 statements to "add_header".
2292 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2293 not followed by a command (e.g. "seen endif").
2295 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2296 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2299 PH/17 Added extra conditions to the default value of delay_warning_condition
2303 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2304 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2305 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2308 The Auto-Submitted: and various List- headers are standardised, whereas I
2309 don't think Precedence: ever was.
2311 PH/18 Refactored debugging code in route_finduser() to show more information,
2312 in particular, the error code if getpwnam() issues one.
2314 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2315 This is apparently needed in addition to the PH/07 change above to avoid
2316 any possible encoding problems.
2318 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2319 but not after initializing Perl.
2321 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2322 output them only if debugging. By default they are written stderr,
2323 apparently, which is not desirable.
2325 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2328 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2331 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2333 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2334 authenticated or an ident call has been made. Suppress the default
2335 values for $authenticated_id and $authenticated_sender (but permit -oMai
2336 and -oMas) when testing with -bh.
2338 PH/24 Re-jigged the order of the tests in the default configuration so that the
2339 tests for valid domains and recipients precede the DNS black list and CSA
2340 tests, on the grounds that those ones are more expensive.
2342 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2343 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2344 This bug exists in every version of Exim that I still have, right back to
2347 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2348 However, an attempt to turn on case-sensitivity in a regex key by
2349 including (?-i) didn't work because the subject string was already
2350 lowercased, and the effects were non-intuitive. It turns out that a
2351 one-line patch can be used to allow (?-i) to work as expected.
2357 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2358 other effects) broke the use of negated acl sub-conditions.
2360 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2363 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2364 "Deny" causes Exim to reject the incoming connection with a 554 error.
2365 Unfortunately, if there is a major crisis, such as a disk failure,
2366 tcp-wrappers gives "deny", whereas what one would like would be some
2367 kind of temporary error. A kludge has been added to help with this.
2368 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2369 554 error is used if errno is still zero or contains ENOENT (which occurs
2370 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2373 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2375 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2376 errors. Otherwise a message that provokes a temporary error (when other
2377 messages do not) can cause a whole host to time out.
2379 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2380 addresses were routed directly to files or pipes from a redirect router.
2381 File deliveries just didn't batch; pipe deliveries might have suffered
2384 PH/06 A failure to get a lock for a hints database would erroneously always say
2385 "Failed to get write lock", even when it was really a read lock.
2387 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2388 of 0600. This has been changed to use the value of the lockfile_mode
2389 option (which defaults to 0600).
2391 PH/08 Applied small patch from the Sieve maintainer.
2393 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2394 folder from quota calculations, a direct delivery into this folder messed
2395 up the contents of the maildirsize file. This was because the regex was
2396 used only to exclude .Trash (or whatever) when the size of the mailbox
2397 was calculated. There was no check that a delivery was happening into an
2398 excluded directory. This bug has been fixed by ignoring all quota
2399 processing for deliveries into excluded directories.
2401 PH/10 Added the maildirfolder_create_regex option to appendfile.
2407 PH/01 The code for finding all the local interface addresses on a FreeBSD
2408 system running IPv6 was broken. This may well have applied to all BSD
2409 systems, as well as to others that have similar system calls. The broken
2410 code found IPv4 interfaces correctly, but gave incorrect values for the
2411 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2412 that it would not match correctly against @[] and not recognize the IPv6
2415 PH/02 The ipliteral router was not recognizing addresses of the form user@
2416 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2418 PH/03 Added disable_ipv6.
2420 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2421 decoded form, because it is most often used to construct To: headers
2422 lines in autoreplies, and the decoded form may well be syntactically
2423 invalid. However, $reply_address has leading white space removed, and all
2424 newlines turned into spaces so that the autoreply transport does not
2427 PH/05 If group was specified without a user on a router, and no group or user
2428 was specified on a transport, the group from the router was ignored.
2430 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2431 for visible compile-time settings that can be used to change these
2432 numbers, for those that want even more. Backwards compatibility with old
2433 spool files has been maintained. However, going back to a previous Exim
2434 release will lost any variables that are in spool files.
2436 PH/07 Two small changes when running in the test harness: increase delay when
2437 passing a TCP/IP connection to a new process, in case the original
2438 process has to generate a bounce, and remove special handling of
2439 127.0.0.2 (sic), which is no longer necessary.
2441 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2442 be the same on different OS.
2444 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2447 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2448 whether --show-vars was specified or not
2450 JJ/02 exipick: Added support for new ACL variable spool format introduced
2453 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2454 syntactically invalid From: or Reply-to: line, and a filter used this to
2455 generate an autoreply, and therefore failed to obtain an address for the
2456 autoreply, Exim could try to deliver to a non-existent relative file
2457 name, causing unrelated and misleading errors. What now happens is that
2458 it logs this as a hard delivery error, but does not attempt to create a
2461 PH/11 The exinext utility has a -C option for testing purposes, but although
2462 the given file was scanned by exinext itself; it wasn't being passed on
2463 when Exim was called.
2465 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2466 an end-of-file indication when reading a command response.
2468 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2469 compiled. In many other places in Exim, IPv6 addresses are always
2470 recognized, so I have changed this. It also means that IPv4 domain
2471 literals of the form [IPV4:n.n.n.n] are now always recognized.
2473 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2474 used if the router is not running as root, for example, when verifying at
2475 ACL time, or when using -bh. The debugging output from this situation was
2476 non-existent - all you got was a failure to exec. I have made two
2479 (a) Failures to set uid/gid, the current directory, or a process leader
2480 in a subprocess such as that created by queryprogram now generate
2481 suitable debugging ouput when -d is set.
2483 (b) The queryprogram router detects when it is not running as root,
2484 outputs suitable debugging information if -d is set, and then runs
2485 the subprocess without attempting to change uid/gid.
2487 PH/15 Minor change to Makefile for building test_host (undocumented testing
2490 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2491 additional section of a DNS packet that returns MX or SRV records.
2492 Instead, it always explicitly searches for A/AAAA records. This avoids
2493 major problems that occur when a DNS server includes only records of one
2494 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2495 fixed another bug: if SRV records were looked up and the corresponding
2496 address records were *not* found in the additional section, the port
2497 values from the SRV records were lost.
2499 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2500 using the correct key (the original address) when searching the retry
2501 rules in order to find which one to use for generating the retry hint.
2503 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2504 adding the default one. Similarly, if it contains a Reply-To: header, the
2505 errors_reply_to option, if set, is not used.
2507 PH/19 When calculating a retry time, Exim used to measure the "time since
2508 failure" by looking at the "first failed" field in the retry record. Now
2509 it does not use this if it is later than than the arrival time of the
2510 message. Instead it uses the arrival time. This makes for better
2511 behaviour in cases where some deliveries succeed, thus re-setting the
2512 "first failed" field. An example is a quota failure for a huge message
2513 when small messages continue to be delivered. Without this change, the
2514 "time since failure" will always be short, possible causing more frequent
2515 delivery attempts for the huge message than are intended.
2516 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2518 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2519 $1, $2, $3) because the numerical variables can be reset during some
2520 expansion items (e.g. "match"), thereby losing the authentication data.
2522 PH/21 Make -bV show the size of off_t variables so that the test suite can
2523 decide whether to run tests for quotas > 2G.
2525 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2526 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2527 filecount value is greater than 2G or if a quota value is greater than 2G
2528 on a system where the size of off_t is not greater than 4, a panic error
2531 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2532 never match. The debug and -bh output now contains an explicit error
2533 message indicating a malformed IPv4 address or mask.
2535 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2536 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2537 PH/23 above applies.
2539 PH/25 Do not write to syslog when running in the test harness. The only
2540 occasion when this arises is a failure to open the main or panic logs
2541 (for which there is an explicit test).
2543 PH/26 Added the /no_tell option to "control=freeze".
2545 PH/27 If a host name lookup failed very early in a connection, for example, if
2546 the IP address matched host_lookup and the reverse lookup yielded a name
2547 that did not have a forward lookup, an error message of the form "no IP
2548 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2549 could be logged. Now it outputs the IP address instead of "NULL".
2551 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2552 allows the sender and the authenticated sender to be set when
2553 submitting a message from within Exim. Since child_open_exim() is
2554 documented for local_scan(), the new function should be too.
2556 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2557 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2558 results in an empty string is now treated as unset.
2560 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2562 PH/31 Added sender_verify_fail logging option.
2564 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2565 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2566 tidied the source and removed it altogether.
2568 PH/33 When a queue run was abandoned because the load average was too high, a
2569 log line was always written; now it is written only if the queue_run log
2570 selector is set. In addition, the log line for abandonment now contains
2571 information about the queue run such as the pid. This is always present
2572 in "start" and "stop" lines but was omitted from the "abandon" line.
2574 PH/34 Omit spaces between a header name and the colon in the error message that
2575 is given when verify = headers_syntax fails (if there are lots of them,
2576 the message gets confusing).
2578 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2579 names, as there are now some PTR records that contain slashes. This check
2580 is only to protect against broken name servers that fall over on strange
2581 characters, so the fact that it applies to all lookups doesn't matter.
2583 PH/36 Now that the new test suite is complete, we can remove some of the
2584 special code in Exim that was needed for the old test suite. For example,
2585 sorting DNS records because real resolvers return them in an arbitrary
2586 order. The new test suite's fake resolver always returns records in the
2589 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2590 bounces) except when queue_only is set, to avoid logging races between
2591 the different processes.
2593 PH/38 Panic-die if .include specifies a non-absolute path.
2595 PH/39 A tweak to the "H" retry rule from its user.
2597 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2598 a label. They prevented compilation on older perls.
2600 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2601 a warning to be raised on newish perls.
2603 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2604 on queue. Changes to match documented behaviour of showing count of
2605 messages matching specified criteria.
2607 PH/40 Changed the default ident timeout from 30s to 5s.
2609 PH/41 Added support for the use of login_cap features, on those BSD systems
2610 that have them, for controlling the resources used by pipe deliveries.
2612 PH/42 The content-scanning code uses fopen() to create files in which to put
2613 message data. Previously it was not paying any attention to the mode of
2614 the files. Exim runs with umask(0) because the rest of the code creates
2615 files with open(), and sets the required mode explicitly. Thus, these
2616 files were ending up world-writeable. This was not a big issue, because,
2617 being within the spool directory, they were not world-accessible. I have
2618 created a function called modefopen, which takes an additional mode
2619 argument. It sets umask(777), creates the file, chmods it to the required
2620 mode, then resets the umask. All the relevant calls to fopen() in the
2621 content scanning code have been changed to use this function.
2623 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2624 to 24 hours. This avoids potential overflow problems when processing G
2625 and H retry rules. I suspect nobody ever tinkers with this value.
2627 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2629 PH/45 When the plaintext authenticator is running as a client, the server's
2630 challenges are checked to ensure they are valid base64 strings. By
2631 default, the authentication attempt is cancelled if an invalid string is
2632 received. Setting client_ignore_invalid_base64 true ignores these errors.
2633 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2634 they are received. Thus, the responses can be made to depend on the
2635 challenges. If an invalid string is ignored, an empty string is placed in
2638 PH/46 Messages that are created by the autoreply transport now contains a
2639 References: header, in accordance with RFCs 2822 and 3834.
2641 PH/47 Added authenticated_sender_force to the smtp transport.
2643 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2645 PH/49 Installed latest patch from the Sieve maintainer.
2647 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2648 was also set, the appendfile transport was unnecessarily scanning a
2649 directory of message files (e.g. for maildir delivery) to find the count
2650 of files (along with the size), even though it did not need this
2651 information. It now does the scan only if it needs to find either the
2652 size of the count of files.
2654 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2656 PH/52 Two bugs concerned with error handling when the smtp transport is
2659 (i) Exim was not creating retry information for temporary errors given
2660 for individual recipients after the DATA command when the smtp transport
2661 was used in LMTP mode. This meant that they could be retried too
2662 frequently, and not timed out correctly.
2664 (ii) Exim was setting the flag that allows error details to be returned
2665 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2666 recipients that were returned after the DATA command.
2668 PH/53 This is related to PH/52, but is more general: for any failing address,
2669 when detailed error information was permitted to be returned to the
2670 sender, but the error was temporary, then after the final timeout, only
2671 "retry timeout exceeded" was returned. Now it returns the full error as
2672 well as "retry timeout exceeded".
2674 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2675 do this, and (what is worse) MTAs that accept it.
2677 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2678 will now be deprecated.
2680 PH/56 New os.c-cygwin from the Cygwin maintainer.
2682 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2683 formats (previously only available in exim formats via -bpr, -bpru,
2684 and -bpra. Now also available in native and exiqgrep formats)
2686 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2687 with very large, slow to parse queues
2689 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2691 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2693 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2694 responses to authentication challenges, though it was showing the
2695 challenges; (ii) I've removed the CR characters from the debug output for
2698 PH/58 Allow for the insertion of a newline as well as a space when a string
2699 is turned into more than one encoded-word during RFC 2047 encoding. The
2700 Sieve code now uses this.
2702 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2703 data_4xx, lost_connection, tls_required.
2705 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2706 message was being sent as an SMTP response.
2708 PH/61 Add -l and -k options to exicyclog.
2710 PH/62 When verifying, if an address was redirected to one new address, so that
2711 verification continued, and the new address failed or deferred after
2712 having set something in $address_data, the value of $address_data was not
2713 passed back to the ACL. This was different to the case when no
2714 redirection occurred. The value is now passed back in both cases.
2716 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2717 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2718 use login_cap.h, so on its own it isn't the distinguishing feature. The
2719 new name refers directly to the setclassresources() function.
2721 PH/65 Added configuration files for NetBSD3.
2723 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2725 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2726 is preferred over IPv4.
2728 PH/68 The bounce_return_message and bounce_return_body options were not being
2729 honoured for bounces generated during the reception of non-SMTP messages.
2730 In particular, this applied to messages rejected by the ACL. This bug has
2731 been fixed. However, if bounce_return_message is true and bounce_return_
2732 body is false, the headers that are returned for a non-SMTP message
2733 include only those that have been read before the error was detected.
2734 (In the case of an ACL rejection, they have all been read.)
2736 PH/69 The HTML version of the specification is now built in a directory called
2737 spec_html instead of spec.html, because the latter looks like a path with
2738 a MIME-type, and this confuses some software.
2740 PH/70 Catch two compiler warnings in sieve.c.
2742 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2743 function verify_get_ident() calls ip_connect() to connect a socket, but
2744 if the "connect()" function timed out, ip_connect() used to close the
2745 socket. However, verify_get_ident() also closes the socket later, and in
2746 between Exim writes to the log, which may get opened at this point. When
2747 the socket was closed in ip_connect(), the log could get the same file
2748 descriptor number as the socket. This naturally causes chaos. The fix is
2749 not to close the socket in ip_connect(); the socket should be closed by
2750 the function that creates it. There was only one place in the code where
2751 this was missing, in the iplookup router, which I don't think anybody now
2752 uses, but I've fixed it anyway.
2754 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2755 well as to direct DNS lookups. Otherwise the handling of names in host
2756 lists is inconsistent and therefore confusing.
2762 PH/01 Two changes to the default runtime configuration:
2764 (1) Move the checks for relay_from_hosts and authenticated clients from
2765 after to before the (commented out) DNS black list checks.
2767 (2) Add control=submission to the relay_from_hosts and authenticated
2768 clients checks, on the grounds that messages accepted by these
2769 statements are most likely to be submissions.
2771 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2773 (1) Generate an error if the third argument for the ${prvs expansion is
2776 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2779 (3) Reset the variables that are obtained from the first argument of
2780 ${prvscheck and used in the second argument before leaving the code,
2781 because their memory is reclaimed, so using them afterwards may do
2784 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2785 one (it's much easier than Tom thought :-).
2787 (5) Because of (4), we can now allow for the use of $prvscheck_result
2788 inside the third argument.
2790 PH/03 For some reason, the default setting of PATH when running a command from
2791 a pipe transport was just "/usr/bin". I have changed it to
2794 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2795 anything to be listed in the output from -bV.
2797 PH/05 When a filter generated an autoreply, the entire To: header line was
2798 quoted in the delivery log line, like this:
2800 => >A.N.Other <ano@some.domain> <original@ddress> ...
2802 This has been changed so that it extracts the operative address. There
2803 may be more than one such address. If so, they are comma-separated, like
2806 => >ano@some.domain,ona@other.domain <original@ddress> ...
2808 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2809 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2810 not being looked up in the rDNS to get a host name, Exim was showing the
2811 IP address twice in Received: lines, even though the IP addresses were
2812 identical. For example:
2814 Received: from [1.2.3.4] (helo=[1.2.3.4])
2816 However, if the real host name was known, it was omitting the HELO data
2817 if it matched the actual IP address. This has been tidied up so that it
2818 doesn't show the same IP address twice.
2820 PH/07 When both +timestamp and +memory debugging was on, the value given by
2821 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2822 called by the debug printing, thereby overwriting the timestamp buffer.
2823 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2825 PH/08 When the original message was included in an autoreply transport, it
2826 always said "this is a copy of the message, including all the headers",
2827 even if body_only or headers_only was set. It now gives an appropriate
2830 PH/09 Applied a patch from the Sieve maintainer which:
2832 o fixes some comments
2833 o adds the (disabled) notify extension core
2834 o adds some debug output for the result of if/elsif tests
2835 o points to the current vacation draft in the documentation
2836 and documents the missing references header update
2840 o fixes a bug in processing the envelope test (when testing
2841 multiple envelope elements, the last element determinted the
2844 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2845 Electronic Mail") by including:
2847 Auto-submitted: auto-generated
2849 in the messages that it generates (bounce messages and others, such as
2850 warnings). In the case of bounce messages for non-SMTP mesages, there was
2851 also a typo: it was using "Auto_submitted" (underscore instead of
2852 hyphen). Since every message generated by Exim is necessarily in response
2853 to another message, thes have all been changed to:
2855 Auto-Submitted: auto-replied
2857 in accordance with these statements in the RFC:
2859 The auto-replied keyword:
2861 - SHOULD be used on messages sent in direct response to another
2862 message by an automatic process,
2864 - MUST NOT be used on manually-generated messages,
2866 - MAY be used on Delivery Status Notifications (DSNs) and Message
2867 Disposition Notifications (MDNs),
2869 - MUST NOT be used on messages generated by automatic or periodic
2870 processes, except for messages which are automatic responses to
2873 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2874 to the default Received: header definition.
2876 PH/12 Added log selector acl_warn_skipped (default on).
2878 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2879 variables because (a) they are in the wrong storage pool and (b) even if
2880 they were copied, it wouldn't work properly because of the caching.
2882 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2883 checking when decoding. Apparently there are clients that generate
2884 overlong encoded strings. Why am I not surprised?
2886 PH/15 If the first argument of "${if match_address" was not empty, but did not
2887 contain an "@" character, Exim crashed. Now it writes a panic log message
2888 and treats the condition as false.
2890 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2892 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2893 extension "envelope-auth". The code is finished and in agreement with
2894 other implementations, but there is no documentation so far and in fact,
2895 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2896 not changing the active code.
2898 Print executed "if" and "elsif" statements when debugging is used. This
2899 helps a great deal to understand what a filter does.
2901 Document more things not specified clearly in RFC3028. I had all this
2902 sorted out, when out of a sudden new issues came to my mind. Oops."
2904 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2907 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2908 canonical form (as documented). However, after a host name lookup from
2909 the IP address, check_host() was doing a simple string comparison with
2910 addresses acquired from the DNS when checking that the found name did
2911 have the original IP as one of its addresses. Since any found IPv6
2912 addresses are likely to be in abbreviated form, the comparison could
2913 fail. Luckily, there already exists a function for doing the comparison
2914 by converting both addresses to binary, so now that is used instead of
2915 the text comparison.
2917 PH/20 There was another similar case to PH/19, when a complete host name was
2918 given in a host list; looking up its IP address could give an abbreviated
2919 form, whereas the current host's name might or might not be abbreviated.
2920 The same fix has been applied.
2926 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2927 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2930 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2931 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2933 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2935 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2936 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2937 Also, for "Auto-Submitted", treat anything other than "no" as
2938 non-personal, in accordance with RFC 3834. (Previously it treated
2939 anything starting "auto-" as non-personal.)
2941 TF/01 The control=submission/name=... option had a problem with syntax
2942 errors if the name included a slash character. The /name= option
2943 now slurps the rest of the string, so it can include any characters
2944 but it must come last in the list of options (after /sender_retain
2947 PH/05 Some modifications to the interface to the fake nameserver for the new
2955 TK/01 Added the "success_on_redirect" address verification option. See
2956 NewStuff for rationale and an example.
2958 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2960 PH/02 Patch to exigrep to allow it to work on syslog lines.
2962 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2963 fread() to skip over the body file's header line, because in Cygwin the
2964 header line is locked and is inaccessible.
2966 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2967 co-exist for some time) to make it clear that it is the Exim ID that is
2968 referenced, not the Message-ID: header line.
2970 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2971 string_format() function, because snprintf() does not exist on all
2974 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2975 ${stat: expansion item.
2977 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2978 protocol synchronization error", to keep the pedants happy.
2980 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2981 well as for IRIX systems, when gcc is being used. See the host.c source
2984 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2986 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2989 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2990 transport and to the smtp transport in LMTP mode.
2992 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2994 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2995 run a filter in a subprocess. This could lead to confusion in subsequent
2996 lookups in the parent process. There should also be a search_tidyup() at
2997 the end of the subprocess.
2999 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3000 only if the host matched helo_try_verify_hosts, which caused the
3001 verification to occur when the EHLO/HELO command was issued. The ACL just
3002 tested the remembered result. Now, if a previous verification attempt has
3003 not happened, "verify = helo" does it there and then.
3005 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3007 TK/03 Fix log output including CR from clamd.
3009 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3010 exist provoked a memory error which could cause a segfault.
3012 PH/15 Installed PCRE 6.2
3014 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3016 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3017 of the problem. Specifically, suggested +O2 rather than +O1 for the
3020 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3022 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3023 record, the port was not correctly logged when the outgoing_port log
3024 selector was set (it logged the transort's default port).
3026 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3027 fallback_hosts, and "hosts" in the smtp transport.
3029 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3030 host errors such as "Connection refused".
3032 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3033 authentication with radiusclient 0.4.9:
3035 - Error returned from rc_read_config was caught wrongly
3036 - Username/password not passed on to radius server due to wrong length.
3038 The presumption is that some radiusclient API changes for 4.51/PH/17
3039 were not taken care of correctly. The code is still untested by me (my
3040 Linux distribution still has 0.3.2 of radiusclient), but it was
3041 contributed by a Radius user.
3043 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3044 expanding the "port" option of the smtp transport.
3046 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3047 while reading a MIME header. Thanks to Tom Hughes for a patch.
3049 PH/24 Include config.h inside local_scan.h so that configuration settings are
3052 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3053 that in an ACL for RCPT (for example), you can examine exactly what was
3056 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3057 commands, but it was not correctly comparing the address with the actual
3058 client host address. Thus, it would show the EHLO address in Received:
3059 header lines when this was not necessary.
3061 PH/27 Added the % operator to ${eval:}.
3063 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3064 it should be ignoring failures (because with -C, for example, it has lost
3065 privilege). It wasn't ignoring creation failures other than "already
3068 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3071 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3072 because an input line was too long, either on its own, or by virtue of
3073 too many continuations, the temporary file was not being removed, and the
3074 return code was incorrect.
3076 PH/31 Missing "BOOL" in function definition in filtertest.c.
3078 PH/32 Applied Sieve patches from the maintainer.
3080 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3082 PH/33 Added "verify = not_blind".
3084 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3085 Local/Makefile (with some defaults set). These are used in built scripts
3086 such as exicyclog, but they have never been used in the exim_install
3087 script (though there are many overriding facilities there). I have
3088 arranged that the exim_install script now takes note of these two
3091 PH/35 Installed configuration files for Dragonfly.
3093 PH/36 When a locally submitted message by a trusted user did not contain a
3094 From: header, and the sender address was obtained from -f or from an SMTP
3095 MAIL command, and the trusted user did not use -F to supply a sender
3096 name, $originator_name was incorrectly used when constructing a From:
3097 header. Furthermore, $originator_name was used for submission mode
3098 messages from external hosts without From: headers in a similar way,
3099 which is clearly wrong.
3101 PH/37 Added control=suppress_local_fixups.
3103 PH/38 When log_selector = +received_sender was set, and the addition of the
3104 sender made the log line's construction buffer exactly full, or one byte
3105 less than full, an overflow happened when the terminating "\n" was
3108 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3109 when the result of a list match is failure because a DNS lookup failed.
3111 PH/40 RM_COMMAND is now used in the building process.
3113 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3114 the "build-* directories that it finds.
3116 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3117 address in a domain literal was a prefix of an interface address.
3119 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3120 when verifying a sender address, unless rewrite_headers is false.
3122 PH/44 Wrote a long comment about why errors_to addresses are verified as
3123 recipients, not senders.
3125 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3126 the ratelimit ACL was added.
3128 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3130 PH/46 Added extra information about PostgreSQL errors to the error string.
3132 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3133 avoiding the need to install special zones in a real server. This is
3134 backwards compatible; if it can't find the fake resolver, it drops back.
3135 Thus, both old and new test suites can be run.
3137 TF/02 Added util/ratelimit.pl
3139 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3140 clock is set back in time.
3142 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3143 Candler <B.Candler@pobox.com>.
3145 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3146 OK for addresses that are the result of redirections.
3148 PH/48 A number of further additions for the benefit of the new test suite,
3149 including a fake gethostbyname() that interfaces to the fake DNS resolver
3152 TF/06 The fix for widen_domains has also been applied to qualify_single and
3153 search_parents which are the other dnslookup options that can cause
3156 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3159 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3161 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3162 service" if it encountered an MX record with an empty target hostname.
3163 The message is now "an MX or SRV record indicated no SMTP service".
3165 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3166 if the DNS of the sending site is misconfigured. This is quite a
3167 common situation. This change restores the behaviour of treating a
3168 helo verification defer as a failure.
3170 PH/51 If self=fail was set on a router, the bounce message did not include the
3171 actual error message.
3177 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3179 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3180 command itself ended in error, the underlying message about the transport
3181 filter timeout was being overwritten with the pipe command error. Now the
3182 underlying error message should be appended to the second error message.
3184 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3186 PH/02 Reduce the amount of output that "make" produces by default. Full output
3187 can still be requested.
3189 PH/03 The warning log line about a condition test deferring for a "warn" verb
3190 was being output only once per connection, rather than after each
3191 occurrence (because it was using the same function as for successful
3192 "warn" verbs). This seems wrong, so I have changed it.
3194 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3195 it should not have, which might have caused a crash in the right
3196 circumstances, but probably never did.
3198 PH/04 Installed a modified version of Tony Finch's patch to make submission
3199 mode fix the return path as well as the Sender: header line, and to
3200 add a /name= option so that you can make the user's friendly name appear
3203 TF/03 Added the control = fakedefer ACL modifier.
3205 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3206 Mark Lowes for thorough testing.
3208 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3210 TK/03 Merged latest SRS patch from Miles Wilton.
3212 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3213 with the definition in sysexits.h (which is #included earlier).
3214 Fortunately, Exim does not actually use EX_OK. The code used to try to
3215 preserve the sysexits.h value, by assumimg that macro definitions were
3216 scanned for macro replacements. I have been disabused of this notion,
3217 so now the code just undefines EX_OK before #including unistd.h.
3219 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3220 in the smtp transport. When a block could not be written in a single
3221 write() function, the timeout was being re-applied to each part-write.
3222 This seems wrong - if the receiver was accepting one byte at a time it
3223 would take for ever. The timeout is now adjusted when this happens. It
3224 doesn't have to be particularly precise.
3226 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3227 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3229 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3230 without a domain if the check to <postmaster@domain> fails.
3232 SC/01 Eximstats: added -xls and the ability to specify output files
3233 (patch written by Frank Heydlauf).
3235 SC/02 Eximstats: use FileHandles for outputing results.
3237 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3239 SC/04 Eximstats: fixed display of large numbers with -nvr option
3241 SC/05 Eximstats: fixed merging of reports with empty tables.
3243 SC/06 Eximstats: added the -include_original_destination flag
3245 SC/07 Eximstats: removed tabs and trailing whitespace.
3247 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3249 TK/06 MBOX spool code: Add real "From " MBOX separator line
3250 so the .eml file is really in mbox format (even though
3251 most programs do not really care). Patch from Alex Miller.
3253 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3254 The latter is generated from $received_to and is only set if the
3255 message has one envelope recipient. SA can use these headers,
3256 obviously out-of-the-box. Patch from Alex Miller.
3258 PH/08 The ${def test on a variable was returning false if the variable's
3259 value was "0", contrary to what the specification has always said!
3260 The result should be true unless the variable is empty.
3262 PH/09 The syntax error of a character other than { following "${if
3263 def:variable_name" (after optional whitespace) was not being diagnosed.
3264 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3265 accidental colon was present, for example, could give incorrect results.
3267 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3268 result is used (not including appendfile, where other changes are about
3271 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3272 This involved changing a lot of size variables from int to off_t. It
3273 should work with maildirs and everything.
3275 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3276 spamd dying while we are connected to it.
3278 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3281 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3282 function for BDB 4.3.
3284 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3286 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3287 the sources for PCRE have been re-arranged and more files are now
3290 PH/15 The code I had for printing potentially long long variables in PH/11
3291 above was not the best (it lost precision). The length of off_t variables
3292 is now inspected at build time, and an appropriate printing format (%ld
3293 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3294 to be "long long int" or "long int". This is needed for the internal
3295 formatting function string_vformat().
3297 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3298 the configuration file to be ":syslog", then the script "guesses" where
3299 the logs files are, rather than using the compiled in default. In our
3300 case the guess is not the same as the compiled default, so the script
3301 suddenly stopped working when I started to use syslog. The patch checks
3302 to see if log_file_path is "". If so, it attempts to read it from exim
3303 with no configuration file to get the compiled in version, before it
3304 falls back to the previous guessing code."
3306 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3307 implementing BATV in an Exim configuration. See NewStuff for the gory
3310 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3311 Makefile that are specific to HP-UX.
3313 PH/18 If the "use_postmaster" option was set for a recipient callout together
3314 with the "random" option, the postmaster address was used as the MAIL
3315 FROM address for the random test, but not for the subsequent recipient
3316 test. It is now used for both.
3318 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3319 patch removes a few documentation additions to RFC 3028, because the
3320 latest draft now contains them. It adds the new en;ascii-case comparator
3321 and a new error check for 8bit text in MIME parts. Comparator and
3322 require names are now matched exactly. I enabled the subaddress
3323 extension, but it is not well tested yet (read: it works for me)."
3325 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3326 rework some of the code of TK/09 above to avoid the hardwired use of
3327 "%lld" and "long long". Replaced the call to snprintf() with a call to
3330 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3331 records point to non-existent hosts", "retry timeout exceeded", and
3332 "retry time not reached for any host after a long failure period".
3334 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3335 experimental DomainKeys support:
3337 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3338 (2) On an error such as an illegally used "control", the wrong name for
3339 the control was given.
3341 These problems did NOT occur unless DomainKeys support was compiled.
3343 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3345 PH/24 Added ${if match_ip condition.
3347 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3348 databases so that it will be absolutely obvious if a crash occurs in the
3349 DB library. This is a regular occurrence (often caused by mis-matched
3352 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3353 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3354 picked up on a user's system that detects such things. There doesn't seem
3355 to be a gcc warning option for this - only an attribute that has to be
3356 put on the function's prototype. It seems that in Fedora Core 4 they have
3357 set this on a number of new functions. No doubt there will be more in due
3360 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3361 specify a transport. However, if an address that was verified by such a
3362 router was the subject of a callout, Exim crashed because it tried to
3363 read the rcpt_include_affixes from the non-existent transport. Now it
3364 just assumes that the setting of that option is false. This bug was
3365 introduced by 4.51/PH/31.
3367 PH/28 Changed -d+all to exclude +memory, because that information is very
3368 rarely of interest, but it makes the output a lot bigger. People tend to
3369 do -d+all out of habit.
3371 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3372 code in os-type was giving problems when libc.so lives in lib64, like on
3375 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3376 aren't the modern standard, and it seems that some systems' include files
3377 don't always have them. Exim was already checking for some of the newer
3378 ones like T_AAAA, and defining it itself. I've added checks for all the
3379 record types that Exim uses.
3381 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3382 not automatically generating a new one, as it is supposed to. This
3383 prevented TLS from working. If the file did exist, but contained invalid
3384 data, a new version was generated, as expected. It was only the case of a
3385 non-existent file that was broken.
3387 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3388 with a change in libdomainkeys > 0.64.
3390 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3391 from DNS. If the selector record carries the flag, it now has
3392 precedence over the domain-wide flag.
3394 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3396 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3397 the use of a port name that isn't defined in /etc/services) occurred, the
3398 message was deferred as in a normal delivery, and thus remained on the
3399 spool, instead of being failed because of the mua_wrapper setting. This
3400 is now fixed, and I tidied up some of the mua_wrapper messages at the
3403 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3404 the messages in a hash of arrays rather than using individual hashes.
3405 This is a bit cleaner and results in dramatic memory savings, albeit
3406 at a slight CPU cost.
3408 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3409 as requested by Marc Sherman.
3411 SC/10 Eximstats: added histograms for user specified patterns as requested
3414 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3416 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3417 fopen() in the content-scanning modules that did not already have it.
3423 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3424 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3426 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3428 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3430 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3431 to test Sieve filters that use "vacation".
3433 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3434 that changes the way the GnuTLS parameters are stored in the cache file.
3435 The new format can be generated externally. For backward compatibility,
3436 if the data in the cache doesn't make sense, Exim assumes it has read an
3437 old-format file, and it generates new data and writes a new file. This
3438 means that you can't go back to an older release without removing the
3441 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3442 work if there are any delivery delays because "one_time" forces the
3443 parent to be marked "delivered", so its unseen clone is never tried
3444 again. For this reason, Exim now forbids the simultaneous setting of
3447 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3448 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3449 entry if you want to know the details. The fix, however, neglected to
3450 consider the case when local delivery batching is involved. The test for
3451 "previously delivered" was not happening when checking to see if an
3452 address could be batched with a previous (undelivered) one; under
3453 certain circumstances this could lead to multiple deliveries to the same
3456 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3457 in its include files, and this causes problems building Exim.
3459 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3460 header_syntax) but Exim was just ignoring anything given after a slash.
3461 In particular, this caused confusion with an attempt to use "verify =
3462 reverse_host_lookup/defer_ok". An error is now given when options are
3463 supplied for verify items that do not have them. (Maybe reverse_host_
3464 lookup should have a defer_ok option, but that's a different point.)
3466 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3467 defined by RFC 821) to 2048, because there were problems with some AUTH
3468 commands, and RFC 1869 says the size should be increased for extended
3469 SMTP commands that take arguments.
3471 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3474 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3475 "unknown" error; now it says that the functionality isn't in the binary.
3477 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3478 an address' error message when a string expansion fails (syntax or
3479 whatever). Otherwise the password may appear in the log. Following change
3480 PH/42 below, there is no longer a chance of it appearing in a bounce
3483 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3485 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3486 of its addresses was ever tried. (Bugzilla bug #2.)
3488 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3489 the result incorrectly in the debug output. (It correctly added a newline
3490 to what was transported.)
3492 TF/01 Added $received_time.
3494 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3495 commented out examples of how to interface to a virus scanner and to
3496 SpamAssassin. Also added commented examples of av_scanner and
3497 spamd_address settings.
3499 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3500 and controls are allowed in which ACLs. There were a couple of minor
3501 errors. Some of the entries in the conditions table (which is a table of
3502 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3503 negation of where the condition IS allowed.
3505 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3507 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3508 header file does not have a version number, so I've had to invent a new
3509 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3510 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3511 radiusclient), but it was contributed by a Radius user.
3513 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3514 files or directories, for OpenSSL.
3516 PH/19 When an Exim process that is running as root has to create an Exim log
3517 file, it does so in a subprocess that runs as exim:exim so as to get the
3518 ownership right at creation (otherwise, other Exim processes might see
3519 the file with the wrong ownership). There was no test for failure of this
3520 fork() call, which would lead to the process getting stuck as it waited
3521 for a non-existent subprocess. Forks do occasionally fail when resources
3522 run out. I reviewed all the other calls to fork(); they all seem to check
3525 PH/20 When checking for unexpected SMTP input at connect time (before writing
3526 the banner), Exim was not dealing correctly with a non-positive return
3527 from the read() function. If the client had disconnected by this time,
3528 the result was a log entry for a synchronization error with an empty
3529 string after "input=" when read() returned zero. If read() returned -1
3530 (an event I could not check), uninitialized data bytes were printed.
3531 There were reports of junk text (parts of files, etc) appearing after
3534 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3536 PH/22 Added support for macro redefinition, and (re)definition in between
3537 driver and ACL definitions.
3539 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3540 forgetting to use the resulting value; it was using the unexpanded value.
3542 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3543 hadn't been configured. The fix is from Juergen Kreileder, who
3544 understands it better than I do:
3546 "Here's what I see happening with three configured cyrus_sasl
3547 authenticators configured (plain, login, cram-md5):
3549 On startup auth_cyrus_sasl_init() gets called for each of these.
3550 This means three calls to sasl_listmech() without a specified mech_list.
3551 => SASL tests which mechs of all available mechs actually work
3552 => three warnings about OTP not working
3553 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3555 With the patch, sasl_listmech() also gets called three times. But now
3556 SASL's mech_list option is set to the server_mech specified in the the
3557 authenticator. Or in other words, the answer from sasl_listmech()
3558 gets limited to just the mech you're testing for (which is different
3560 => the return list contains just 'plain' or 'login', 'cram-md5' or
3561 nothing depending on the value of ob->server_mech.
3563 I've just tested the patch: Authentication still works fine,
3564 unavailable mechs specified in the exim configuration are still
3565 caught, and the auth.log warnings about OTP are gone."
3567 PH/25 When debugging is enabled, the contents of the command line are added
3568 to the debugging output, even when log_selector=+arguments is not
3571 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3572 answer is "GNU", and only if the return is "GNU/something" is the answer
3575 PH/27 $acl_verify_message is now set immediately after the failure of a
3576 verification in an ACL, and so is available in subsequent modifiers. In
3577 particular, the message can be preserved by coding like this:
3579 warn !verify = sender
3580 set acl_m0 = $acl_verify_message
3582 Previously, $acl_verify_message was set only while expanding "message"
3583 and "log_message" when a very denied access.
3585 PH/28 Modified OS/os.c-Linux with
3587 -#ifndef OS_LOAD_AVERAGE
3588 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3590 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3591 nomenclature these days.)
3593 PH/29 Installed patch from the Sieve maintainer that adds the options
3594 sieve_useraddress and sieve_subaddress to the redirect router.
3596 PH/30 In these circumstances:
3597 . Two addresses routed to the same list of hosts;
3598 . First host does not offer TLS;
3599 . First host accepts first address;
3600 . First host gives temporary error to second address;
3601 . Second host offers TLS and a TLS session is established;
3602 . Second host accepts second address.
3603 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3604 and peerdn, if requested) that were in fact used only for the second
3607 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3608 attention to any local part prefix or suffix that was matched by the
3609 router that accepted the address. It now behaves in the same way as it
3610 does for delivery: the affixes are removed from the local part unless
3611 rcpt_include_affixes is set on the transport.
3613 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3614 timeout during the DATA phase of an incoming message.
3616 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3617 applied a patch sanctioned by the Sieve maintainer.
3619 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3620 the uid or gid is negative. A case of a negative gid caused this to be
3621 noticed. The fix allows for either to be negative.
3623 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3624 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3627 PH/36 The change PH/12 above was broken. Fixed it.
3629 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3630 the grounds that routing the same address twice would always produce the
3631 same answer. This might have been true once, but it is certainly no
3632 longer true now. Routing a child address may depend on the previous
3633 routing that produced that child. Some complicated redirection strategies
3634 went wrong when messages had multiple recipients, and made Exim's
3635 behaviour dependent on the order in which the addresses were given.
3637 I have moved the duplicate checking until after the routing is complete.
3638 Exim scans the addresses that are assigned to local and remote
3639 transports, and removes any duplicates. This means that more work will be
3640 done, as duplicates will always all be routed, but duplicates are
3641 presumably rare, so I don't expect this is of any significance.
3643 For deliveries to pipes, files, and autoreplies, the duplicate checking
3644 still happens during the routing process, since they are not going to be
3647 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3648 It corrects a timeout issue with spamd. This is Ian's comment: "The
3649 background is that sometimes spamd either never reads data from a
3650 connection it has accepted, or it never writes response data. The exiscan
3651 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3652 blindly assumes that writes won't block so it may never time out."
3654 PH/39 Allow G after quota size as well as K and M.
3656 PH/40 The value set for $authenticated_id in an authenticator may not contain
3657 binary zeroes or newlines because the value is written to log lines and
3658 to spool files. There was no check on this. Now the value is run through
3659 the string_printing() function so that such characters are converted to
3660 printable escape sequences.
3662 PH/41 $message_linecount is a new variable that contains the total number of
3663 lines in the message. Compare $body_linecount, which is the count for the
3666 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3667 bounce and delay warning messages, except in certain special cases, which
3670 (a) An SMTP error message from a remote host;
3671 (b) A message specified in a :fail: redirection;
3672 (c) A message specified in a "fail" command in a system filter;
3673 (d) A message specified in a FAIL return from the queryprogram router;
3674 (e) A message specified by the cannot_route_message router option.
3676 In these cases only, Exim does include the error details in bounce and
3677 warning messages. There are also a few cases where bland messages such
3678 as "unrouteable address" or "local delivery error" are given.
3680 PH/43 $value is now also set for the "else" part of a ${run expansion.
3682 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3683 being worked on, but at least Exim now implements the latest version to
3686 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3687 process to complete was treated as a delivery failure, a timeout while
3688 writing the message to the pipe was logged, but erroneously treated as a
3689 successful delivery. Such timeouts include transport filter timeouts. For
3690 consistency with the overall process timeout, these timeouts are now
3691 treated as errors, giving rise to delivery failures by default. However,
3692 there is now a new Boolean option for the pipe transport called
3693 timeout_defer, which, if set TRUE, converts the failures into defers for
3694 both kinds of timeout. A transport filter timeout is now identified in
3697 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3698 systems where "make" and "gmake" are different, calling "gmake" at top
3699 level broke things. I've arranged for the value of $(MAKE) to be passed
3700 from the Makefile to this script so that it can call the same version of
3704 A note about Exim versions 4.44 and 4.50
3705 ----------------------------------------
3707 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3708 changes of various kinds. As a consequence, a big documentation update was
3709 needed. This delayed the release for rather longer than seemed good, especially
3710 in the light of a couple of (minor) security issues. Therefore, the changes
3711 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3712 release. So 4.44 and 4.50 are in effect two different branches that both start
3715 I have left the 4.50 change log unchanged; it contains all the changes since
3716 4.43. The change log for 4.44 is below; many of its items are identical to
3717 those for 4.50. This seems to be the most sensible way to preserve the
3718 historical information.
3724 1. Minor wording change to the doc/README.SIEVE file.
3726 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3727 computation of the current number of files was incorrect.
3729 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3730 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3733 4. Give more explanation in the error message when the command for a transport
3734 filter fails to execute.
3736 5. There are several places where Exim runs a non-Exim command in a
3737 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3738 was being done only for the command run by the queryprogram router. It is
3739 now done for all such subprocesses. The other cases are: ${run, transport
3740 filters, and the commands run by the lmtp and pipe transports.
3742 6. Added CONFIGURE_GROUP build-time option.
3744 7. Some older OS have a limit of 256 on the maximum number of file
3745 descriptors. Exim was using setrlimit() to set 1000 as a large value
3746 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3747 systems. I've change it so that if it can't get 1000, it tries for 256.
3749 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3750 was an oversight, and furthermore, ever since the addition of extra
3751 controls (e.g. 4.43/32), the checks on when to allow different forms of
3752 "control" were broken. There should now be diagnostics for all cases when a
3753 control that does not make sense is encountered.
3755 9. Added the /retain_sender option to "control=submission".
3757 10. $recipients is now available in the predata ACL (oversight).
3759 11. Tidy the search cache before the fork to do a delivery from a message
3760 received from the command line. Otherwise the child will trigger a lookup
3761 failure and thereby defer the delivery if it tries to use (for example) a
3762 cached ldap connection that the parent has called unbind on.
3764 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3765 of $address_data from the recipient verification was clobbered by the
3766 sender verification.
3768 13. The value of address_data from a sender verification is now available in
3769 $sender_address_data in subsequent conditions in the ACL statement.
3771 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3773 15. Added a new option "connect=<time>" to callout options, to set a different
3776 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3777 was its contents. (It was OK if the option was not defined at all.)
3779 17. A "Completed" log line is now written for messages that are removed from
3780 the spool by the -Mrm option.
3782 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3783 information about exactly what failed.
3785 19. Added -dd to debug only the daemon process.
3787 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3788 handles timeouts, both on the server side and network timeouts. Renamed the
3789 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3791 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3792 It is now set to "smtps".
3794 22. $host_address is now set to the target address during the checking of
3795 ignore_target_hosts.
3797 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3798 being passed; this would have caused $sender_host_name to have been used if
3799 matching the list had actually called for a host name (not very likely,
3800 since this list is usually IP addresses). A host name is now passed as
3803 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3804 code that specifies a non-restarting handler (typically sigaction() in
3805 modern systems) in an attempt to fix a rare and obscure crash bug.
3807 25. Narrowed the window for a race in the daemon that could cause it to ignore
3808 SIGCHLD signals. This is not a major problem, because they are used only to
3809 wake it up if nothing else does.
3811 26. A malformed maildirsize file could cause Exim to calculate negative values
3812 for the mailbox size or file count. Odd effects could occur as a result.
3813 The maildirsize information is now recalculated if the size or filecount
3816 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3817 support for a long time. Removed HAVE_SYS_VFS_H.
3819 28. Installed the latest version of exipick from John Jetmore.
3821 29. In an address list, if the pattern was not a regular expression, an empty
3822 subject address (from a bounce message) matched only if the pattern was an
3823 empty string. Non-empty patterns were not even tested. This was the wrong
3824 because it is perfectly reasonable to use an empty address as part of a
3825 database query. An empty address is now tested by patterns that are
3826 lookups. However, all the other forms of pattern expect the subject to
3827 contain a local part and a domain, and therefore, for them, an empty
3828 address still always fails if the pattern is not itself empty.
3830 30. Exim went into a mad DNS loop when attempting to do a callout where the
3831 host was specified on an smtp transport, and looking it up yielded more
3832 than one IP address.
3834 31. Re-factored the code for checking spool and log partition space into a
3835 function that finds that data and another that does the check. The former
3836 is then used to implement four new variables: $spool_space, $log_space,
3837 $spool_inodes, and $log_inodes.
3839 32. The RFC2047 encoding function was originally intended for short strings
3840 such as real names; it was not keeping to the 75-character limit for
3841 encoded words that the RFC imposes. It now respects the limit, and
3842 generates multiple encoded words if necessary. To be on the safe side, I
3843 have increased the buffer size for the ${rfc2047: expansion operator from
3846 33. It is now permitted to omit both strings after an "if" condition; if the
3847 condition is true, the result is "true". As before, when the second string
3848 is omitted, a false condition yields an empty string. This makes it less
3849 cumbersome to write custom ACL and router conditions.
3851 34. Failure to deliver a bounce message always caused it to be frozen, even if
3852 there was an errors_to setting on the router. The errors_to setting is now
3855 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3856 canonical form (fully expanded) before being placed in
3857 $sender_host_address.
3859 36. The table in the code that translates DNS record types into text (T_A to
3860 "A" for instance) was missing entries for NS and CNAME. It is just possible
3861 that this could have caused confusion if both these types were looked up
3862 for the same domain, because the text type is used as part of Exim's
3863 per-process caching. But the chance of anyone hitting this buglet seems
3866 37. The dnsdb lookup has been extended in a number of ways.
3868 (1) There is a new type, "zns", which walks up the domain tree until it
3869 finds some nameserver records. It should be used with care.
3871 (2) There is a new type, "mxh", which is like "mx" except that it returns
3872 just the host names, not the priorities.
3874 (3) It is now possible to give a list of domains (or IP addresses) to be
3875 looked up. The behaviour when one of the lookups defers can be
3876 controlled by a keyword.
3878 (4) It is now possible to specify the separator character for use when
3879 multiple records are returned.
3881 38. The dnslists ACL condition has been extended: it is now possible to supply
3882 a list of IP addresses and/or domains to be looked up in a particular DNS
3885 39. Added log_selector=+queue_time_overall.
3887 40. When running the queue in the test harness, wait just a tad after forking a
3888 delivery process, to get repeatability of debugging output.
3890 41. Include certificate and key file names in error message when GnuTLS fails
3891 to set them up, because the GnuTLS error message doesn't include the name
3892 of the failing file when there is a problem reading it.
3894 42. Allow both -bf and -bF in the same test run.
3896 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3898 44. The "Exiscan patch" is now merged into the mainline Exim source.
3900 45. Sometimes the final signoff response after QUIT could fail to get
3901 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3902 < 0 before doing a fflush(). This bug looks as though it goes back to the
3903 introduction of TLS in release 3.20, but "sometimes" must have been rare
3904 because the tests only now provoked it.
3906 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3907 (this can affect the format of dates).
3909 47. exim_tidydb, when checking for the continued existence of a message for
3910 which it has found a message-specific retry record, was not finding
3911 messages that were in split spool directories. Consequently, it was
3912 deleting retry records that should have stayed in existence.
3914 48. Steve fixed some bugs in eximstats.
3916 49. The SPA authentication driver was not abandoning authentication and moving
3917 on to the next authenticator when an expansion was forced to fail,
3918 contradicting the general specification for all authenticators. Instead it
3919 was generating a temporary error. It now behaves as specified.
3921 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3922 (the order specifies the preference for clients). The order is now AES256,
3923 AES128, 3DES, ARCFOUR128.
3925 51. Small patch to Sieve code - explicitly set From: when generating an
3928 52. Exim crashed if a remote delivery caused a very long error message to be
3929 recorded - for instance if somebody sent an entire SpamAssassin report back
3930 as a large number of 550 error lines. This bug was coincidentally fixed by
3931 increasing the size of one of Exim's internal buffers (big_buffer) that
3932 happened as part of the Exiscan merge. However, to be on the safe side, I
3933 have made the code more robust (and fixed the comments that describe what
3936 53. Now that there can be additional text after "Completed" in log lines (if
3937 the queue_time_overall log selector is set), a one-byte patch to exigrep
3938 was needed to allow it to recognize "Completed" as not the last thing in
3941 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3942 patch that reportedly fixes this has been added. I am not expert enough to
3943 create a test for it. This is what the patch creator wrote:
3945 "I found a little strange behaviour of ldap code when working with
3946 Windows 2003 AD Domain, where users was placed in more than one
3947 Organization Units. When I tried to give exim partial DN, the exit code
3948 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3949 But simultaneously result of request was absolutely normal ldap result,
3950 so I produce this patch..."
3952 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3953 REFERENCE, so I have modified the code to exclude the patch when that macro
3956 55. Some experimental protocols are using DNS PTR records for new purposes. The
3957 keys for these records are domain names, not reversed IP addresses. The
3958 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3959 leaves it alone. Component reversal etc. now happens only for IP addresses.
3962 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3964 57. Double the size of the debug message buffer (to 2048) so that more of very
3965 long debug lines gets shown.
3967 58. The exicyclog utility now does better if the number of log files to keep
3968 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3970 59. Two changes related to the smtp_active_hostname option:
3972 (1) $smtp_active_hostname is now available as a variable.
3973 (2) The default for smtp_banner uses $smtp_active_hostname instead
3974 of $primary_hostname.
3976 60. The host_aton() function is supposed to be passed a string that is known
3977 to be a valid IP address. However, in the case of IPv6 addresses, it was
3978 not checking this. This is a hostage to fortune. Exim now panics and dies
3979 if the condition is not met. A case was found where this could be provoked
3980 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3981 components; fortuitously, this particular loophole had already been fixed
3982 by change 4.50/55 above.
3984 If there are any other similar loopholes, the new check in host_aton()
3985 itself should stop them being exploited. The report I received stated that
3986 data on the command line could provoke the exploit when Exim was running as
3987 exim, but did not say which command line option was involved. All I could
3988 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3989 running as the user.
3992 61. There was a buffer overflow vulnerability in the SPA authentication code
3993 (which came originally from the Samba project). I have added a test to the
3994 spa_base64_to_bits() function which I hope fixes it.
3997 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3998 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4000 63. The daemon start-up calls getloadavg() while still root for those OS that
4001 need the first call to be done as root, but it missed one case: when
4002 deliver_queue_load_max is set with deliver_drop_privilege. This is
4003 necessary for the benefit of the queue runner, because there is no re-exec
4004 when deliver_drop_privilege is set.
4006 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4007 This has been fixed.
4009 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4010 was in use, was not putting the data itself into the right store pool;
4011 consequently, it could be overwritten for a subsequent message in the same
4012 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4015 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4017 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4018 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4019 the function were treating the return as a boolean value, which happened to
4020 work because 0=false and not-0=true, but is not correct code.
4022 68. The host_aton() function was not handling scoped IPv6 addresses (those
4023 with, for example, "%eth0" on the end) correctly.
4025 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4026 negated items (that is, ~something) in unsigned ints. Some compilers
4027 apparently mutter when there is no cast.
4029 70. If an address verification called from an ACL failed, and did not produce a
4030 user-specific message (i.e. there was only a "system" message), nothing was
4031 put in $acl_verify_message. In this situation, it now puts the system
4034 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4035 session; change 4.31/43 added the unwanted input to the log line - except
4036 that it did not do this in the start of session case. It now does.
4038 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4039 This is wrong and can cause the other end to generate a synchronization
4040 error if it is another Exim or anything else that does the synchronization
4041 check. A QUIT command is no longer sent after a timeout.
4043 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4044 during host lookups.
4046 74. The defer_ok option of callout verification was not working if it was used
4047 when verifying addresses in header lines, that is, for this case:
4049 verify = header_sender/callout=defer_ok
4051 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4052 those file descriptors could be used for SMTP connections. If anything
4053 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4054 could be sent to the SMTP client, causing chaos. The daemon now opens
4055 stdin, stdout, and stderr to /dev/null when it puts itself into the
4058 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4059 log by default. The user can override this with suitable Perl magic.
4061 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4062 the log message when discard triggers, was not working for the DATA ACL or
4063 for the non-SMTP ACL.
4065 78. Error message wording change in sieve.c.
4067 79. If smtp_accept_max_per_host was set, the number of connections could be
4068 restricted to fewer than expected, because the daemon was trying to set up
4069 a new connection before checking whether the processes handling previous
4070 connections had finished. The check for completed processes is now done
4071 earlier. On busy systems, this bug wouldn't be noticed because something
4072 else would have woken the daemon, and it would have reaped the completed
4075 80. If a message was submitted locally by a user whose login name contained one
4076 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4077 It caused a spool format error. I have fixed the spool reading code. A
4078 related problem was that the "from" clause in the Received: line became
4079 illegal because of the space(s). It is now covered by ${quote_local_part.
4081 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4084 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4086 83. Patch from Sieve maintainer to fix unterminated string problem in
4087 "vacation" handling.
4089 84. Some minor changes to the Linux configuration files to help with other
4090 OS variants using glibc.
4092 85. One more patch for Sieve to update vacation handling to latest spec.
4095 ----------------------------------------------------
4096 See the note above about the 4.44 and 4.50 releases.
4097 ----------------------------------------------------
4103 1. Change 4.43/35 introduced a bug that caused file counts to be
4104 incorrectly computed when quota_filecount was set in an appendfile
4107 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4108 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4111 3. Give more explanation in the error message when the command for a transport
4112 filter fails to execute.
4114 4. There are several places where Exim runs a non-Exim command in a
4115 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4116 was being done only for the command run by the queryprogram router. It is
4117 now done for all such subprocesses. The other cases are: ${run, transport
4118 filters, and the commands run by the lmtp and pipe transports.
4120 5. Some older OS have a limit of 256 on the maximum number of file
4121 descriptors. Exim was using setrlimit() to set 1000 as a large value
4122 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4123 systems. I've change it so that if it can't get 1000, it tries for 256.
4125 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4126 was an oversight, and furthermore, ever since the addition of extra
4127 controls (e.g. 4.43/32), the checks on when to allow different forms of
4128 "control" were broken. There should now be diagnostics for all cases when a
4129 control that does not make sense is encountered.
4131 7. $recipients is now available in the predata ACL (oversight).
4133 8. Tidy the search cache before the fork to do a delivery from a message
4134 received from the command line. Otherwise the child will trigger a lookup
4135 failure and thereby defer the delivery if it tries to use (for example) a
4136 cached ldap connection that the parent has called unbind on.
4138 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4139 of $address_data from the recipient verification was clobbered by the
4140 sender verification.
4142 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4143 was its contents. (It was OK if the option was not defined at all.)
4145 11. A "Completed" log line is now written for messages that are removed from
4146 the spool by the -Mrm option.
4148 12. $host_address is now set to the target address during the checking of
4149 ignore_target_hosts.
4151 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4152 being passed; this would have caused $sender_host_name to have been used if
4153 matching the list had actually called for a host name (not very likely,
4154 since this list is usually IP addresses). A host name is now passed as
4157 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4158 code that specifies a non-restarting handler (typically sigaction() in
4159 modern systems) in an attempt to fix a rare and obscure crash bug.
4161 15. Narrowed the window for a race in the daemon that could cause it to ignore
4162 SIGCHLD signals. This is not a major problem, because they are used only to
4163 wake it up if nothing else does.
4165 16. A malformed maildirsize file could cause Exim to calculate negative values
4166 for the mailbox size or file count. Odd effects could occur as a result.
4167 The maildirsize information is now recalculated if the size or filecount
4170 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4171 support for a long time. Removed HAVE_SYS_VFS_H.
4173 18. Updated exipick to current release from John Jetmore.
4175 19. Allow an empty sender to be matched against a lookup in an address list.
4176 Previously the only cases considered were a regular expression, or an
4179 20. Exim went into a mad DNS lookup loop when doing a callout where the
4180 host was specified on the transport, if the DNS lookup yielded more than
4183 21. The RFC2047 encoding function was originally intended for short strings
4184 such as real names; it was not keeping to the 75-character limit for
4185 encoded words that the RFC imposes. It now respects the limit, and
4186 generates multiple encoded words if necessary. To be on the safe side, I
4187 have increased the buffer size for the ${rfc2047: expansion operator from
4190 22. Failure to deliver a bounce message always caused it to be frozen, even if
4191 there was an errors_to setting on the router. The errors_to setting is now
4194 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4195 canonical form (fully expanded) before being placed in
4196 $sender_host_address.
4198 24. Updated eximstats to version 1.33
4200 25. Include certificate and key file names in error message when GnuTLS fails
4201 to set them up, because the GnuTLS error message doesn't include the name
4202 of the failing file when there is a problem reading it.
4204 26. Expand error message when OpenSSL has problems setting up cert/key files.
4207 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4208 (this can affect the format of dates).
4210 28. exim_tidydb, when checking for the continued existence of a message for
4211 which it has found a message-specific retry record, was not finding
4212 messages that were in split spool directories. Consequently, it was
4213 deleting retry records that should have stayed in existence.
4215 29. eximstats updated to version 1.35
4216 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4217 1.35 - bugfix such that pie charts by volume are generated correctly
4219 30. The SPA authentication driver was not abandoning authentication and moving
4220 on to the next authenticator when an expansion was forced to fail,
4221 contradicting the general specification for all authenticators. Instead it
4222 was generating a temporary error. It now behaves as specified.
4224 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4225 (the order specifies the preference for clients). The order is now AES256,
4226 AES128, 3DES, ARCFOUR128.
4228 31. Small patch to Sieve code - explicitly set From: when generating an
4231 32. Exim crashed if a remote delivery caused a very long error message to be
4232 recorded - for instance if somebody sent an entire SpamAssassin report back
4233 as a large number of 550 error lines. This bug was coincidentally fixed by
4234 increasing the size of one of Exim's internal buffers (big_buffer) that
4235 happened as part of the Exiscan merge. However, to be on the safe side, I
4236 have made the code more robust (and fixed the comments that describe what
4239 33. Some experimental protocols are using DNS PTR records for new purposes. The
4240 keys for these records are domain names, not reversed IP addresses. The
4241 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4242 leaves it alone. Component reversal etc. now happens only for IP addresses.
4245 34. The host_aton() function is supposed to be passed a string that is known
4246 to be a valid IP address. However, in the case of IPv6 addresses, it was
4247 not checking this. This is a hostage to fortune. Exim now panics and dies
4248 if the condition is not met. A case was found where this could be provoked
4249 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4250 components; fortuitously, this particular loophole had already been fixed
4251 by change 4.50/55 or 4.44/33 above.
4253 If there are any other similar loopholes, the new check in host_aton()
4254 itself should stop them being exploited. The report I received stated that
4255 data on the command line could provoke the exploit when Exim was running as
4256 exim, but did not say which command line option was involved. All I could
4257 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4258 running as the user.
4261 35. There was a buffer overflow vulnerability in the SPA authentication code
4262 (which came originally from the Samba project). I have added a test to the
4263 spa_base64_to_bits() function which I hope fixes it.
4266 36. The daemon start-up calls getloadavg() while still root for those OS that
4267 need the first call to be done as root, but it missed one case: when
4268 deliver_queue_load_max is set with deliver_drop_privilege. This is
4269 necessary for the benefit of the queue runner, because there is no re-exec
4270 when deliver_drop_privilege is set.
4272 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4273 was in use, was not putting the data itself into the right store pool;
4274 consequently, it could be overwritten for a subsequent message in the same
4275 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4278 38. Sometimes the final signoff response after QUIT could fail to get
4279 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4280 < 0 before doing a fflush(). This bug looks as though it goes back to the
4281 introduction of TLS in release 3.20, but "sometimes" must have been rare
4282 because the tests only now provoked it.
4288 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4289 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4290 Now it returns an integer. A number of calls were still expecting a T/F
4291 return. Fortuitously, in all cases, the tests worked in OK situations,
4292 which is the norm. However, things would have gone wrong on any write
4293 failures on the smtp file descriptor. This function is used when sending
4294 messages over SMTP and also when doing verify callouts.
4296 2. When Exim is called to do synchronous delivery of a locally submitted
4297 message (the -odf or -odi options), it no longer closes stderr before doing
4300 3. Implemented the mua_wrapper option.
4302 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4304 5. Implemented the functions header_remove(), header_testname(),
4305 header_add_at_position(), and receive_remove_recipient(), and exported them
4308 6. If an ACL "warn" statement specified the addition of headers, Exim already
4309 inserted X-ACL-Warn: at the start if there was no header name. However, it
4310 was not making this test for the second and subsequent header lines if
4311 there were newlines in the string. This meant that an invalid header could
4312 be inserted if Exim was badly configured.
4314 7. Allow an ACL "warn" statement to add header lines at the start or after all
4315 the Received: headers, as well as at the end.
4317 8. Added the rcpt_4xx retry error code.
4319 9. Added postmaster_mailfrom=xxx to callout verification option.
4321 10. Added mailfrom=xxxx to the callout verification option, for verify=
4324 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4325 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4327 12. Inserted some casts to stop certain compilers warning when using pointer
4328 differences as field lengths or precisions in printf-type calls (mostly
4329 affecting debugging statements).
4331 13. Added optional readline() support for -be (dynamically loaded).
4333 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4334 same clock tick as a message's arrival, so that its received time was the
4335 same as the "first fail" time on the retry record, and that message
4336 remained on the queue past the ultimate address timeout, every queue runner
4337 would try a delivery (because it was past the ultimate address timeout) but
4338 after another failure, the ultimate address timeout, which should have then
4339 bounced the address, did not kick in. This was a "< instead of <=" error;
4340 in most cases the first failure would have been in the next clock tick
4341 after the received time, and all would be well.
4343 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4344 being recognized when the domain list was tested by the match_domain
4345 condition in an expansion string.
4347 16. Added the ${str2b64: operator.
4349 17. Exim was always calling setrlimit() to set a large limit for the number of
4350 processes, without checking whether the existing limit was already
4351 adequate. (It did check for the limit on file descriptors.) Furthermore,
4352 errors from getrlimit() and setrlimit() were being ignored. Now they are
4353 logged to the main and panic logs, but Exim does carry on, to try to do its
4354 job under whatever limits there are.
4356 18. Imported PCRE 5.0.
4358 19. Trivial typo in log message " temporarily refused connection" (the leading
4361 20. If the log selector return_path_on_delivery was set and an address was
4362 redirected to /dev/null, the delivery process crashed because it assumed
4363 that a return path would always be set for a "successful" delivery. In this
4364 case, the whole delivery is bypassed as an optimization, and therefore no
4367 21. Internal re-arrangement: the function for sending a challenge and reading
4368 a response while authentication was assuming a zero-terminated challenge
4369 string. It's now changed to take a pointer and a length, to allow for
4370 binary data in such strings.
4372 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4374 23. Exim was not respecting finduser_retries when seeking the login of the
4375 uid under which it was called; it was always trying 10 times. (The default
4376 setting of finduser_retries is zero.) Also, it was sleeping after the final
4377 failure, which is pointless.
4379 24. Implemented tls_on_connect_ports.
4381 25. Implemented acl_smtp_predata.
4383 26. If the domain in control=submission is set empty, Exim assumes that the
4384 authenticated id is a complete email address when it generates From: or
4385 Sender: header lines.
4387 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4388 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4389 chown and chgrp in /bin and hostname in /usr/bin.
4391 28. Exim was keeping the "process log" file open after each use, just as it
4392 does for the main log. This opens the possibility of it remaining open for
4393 long periods when the USR1 signal hits a daemon. Occasional processlog
4394 errors were reported, that could have been caused by this. Anyway, it seems
4395 much more sensible not to leave this file open at all, so that is what now
4398 29. The long-running daemon process does not normally write to the log once it
4399 has entered its main loop, and it closes the log before doing so. This is
4400 so that log files can straightforwardly be renamed and moved. However,
4401 there are a couple of unusual error situations where the daemon does write
4402 log entries, and I had neglected to close the log afterwards.
4404 30. The text of an SMTP error response that was received during a remote
4405 delivery was being truncated at 512 bytes. This is too short for some of
4406 the long messages that one sometimes sees. I've increased the limit to
4409 31. It is now possible to make retry rules that apply only when a message has a
4410 specific sender, in particular, an empty sender.
4412 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4413 it possible to be selective about when SMTP synchronization is enforced.
4415 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4417 32. Implemented hosts_connection_nolog.
4419 33. Added an ACL for QUIT.
4421 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4424 35. Added mailbox_size and mailbox_filecount to appendfile.
4426 36. Added control = no_multiline_responses to ACLs.
4428 37. There was a bug in the logic of the code that waits for the clock to tick
4429 in the case where the clock went backwards by a substantial amount such
4430 that the microsecond fraction of "now" was more than the microsecond
4431 fraction of "then" (but the whole seconds number was less).
4433 38. Added support for the libradius Radius client library this is found on
4434 FreeBSD (previously only the radiusclient library was supported).
4440 1. When certain lookups returned multiple values in the form name=value, the
4441 quoting of the values was not always being done properly. Specifically:
4442 (a) If the value started with a double quote, but contained no whitespace,
4444 (b) If the value contained whitespace other than a space character (i.e.
4445 tabs or newlines or carriage returns) it was not quoted.
4446 This fix has been applied to the mysql and pgsql lookups by writing a
4447 separate quoting function and calling it from the lookup code. The fix
4448 should probably also be applied to nisplus, ibase and oracle lookups, but
4449 since I cannot test any of those, I have not disturbed their existing code.
4451 2. A hit in the callout cache for a specific address caused a log line with no
4452 reason for rejecting RCPT. Now it says "Previous (cached) callout
4453 verification failure".
4455 3. There was an off-by-one bug in the queryprogram router. An over-long
4456 return line was truncated at 256 instead of 255 characters, thereby
4457 overflowing its buffer with the terminating zero. As well as fixing this, I
4458 have increased the buffer size to 1024 (and made a note to document this).
4460 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4461 when Exim is waiting for an SMTP response from a remote server, Exim
4462 restarts its select() call on the socket, thereby resetting its timeout.
4463 This is not a problem when such interrupts are rare. Somebody set up a cron
4464 job to run exiwhat every 2 minutes, which is less than the normal select()
4465 timeout (5 or 10 minutes). This meant that the select() timeout never
4466 kicked in because it was always reset. I have fixed this by comparing the
4467 time when an interrupt arrives with the time at the start of the first call
4468 to select(). If more time than the timeout has elapsed, the interrupt is
4469 treated as a timeout.
4471 5. Some internal re-factoring in preparation for the addition of Sieve
4472 extensions (by MH). In particular, the "personal" test is moved to a
4473 separate function, and given an option for scanning Cc: and Bcc: (which is
4474 not set for Exim filters).
4476 6. When Exim created an email address using the login of the caller as the
4477 local part (e.g. when creating a From: or Sender: header line), it was not
4478 quoting the local part when it contained special characters such as @.
4480 7. Installed new OpenBSD configuration files.
4482 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4483 try to make them clearer.
4485 9. Callout options, other than the timeout value, were being ignored when
4486 verifying sender addresses in header lines. For example, when using
4488 verify = header_sender/callout=no_cache
4490 the cache was (incorrectly) being used.
4492 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4493 only the Cygwin environment.
4495 11. When return_path_on_delivery was set as a log selector, if different remote
4496 addresses in the same message used different return paths and parallel
4497 remote delivery occurred, the wrong values would sometimes be logged.
4498 (Whenever a remote delivery process finished, the return path value from
4499 the most recently started remote delivery process was logged.)
4501 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4502 lines when AUTH and/or TLS are in use. This is the "received protocol"
4503 field. Exim used to use "asmtp" for authenticated SMTP, without any
4504 indication (in the protocol name) for TLS use. Now it follows the RFC and
4505 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4506 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4507 names appear in log lines as well as in Received: header lines.
4509 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4510 extensions, and comparison tests, and to fix some bugs.
4512 14. Changes to the "personal" filter test:
4514 (1) The test was buggy in that it was just doing the equivalent of
4515 "contains" tests on header lines. For example, if a user's address was
4516 anne@some.where, the "personal" test would incorrectly be true for
4518 To: susanne@some.where
4520 This test is now done by extracting each address from the header in turn,
4521 and checking the entire address. Other tests that are part of "personal"
4522 are now done using regular expressions (for example, to check local parts
4523 of addresses in From: header lines).
4525 (2) The list of non-personal local parts in From: addresses has been
4526 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4527 taken from the Sieve specification recommendations.
4529 (3) If the message contains any header line starting with "List-" it is
4530 treated as non-personal.
4532 (4) The test for "circular" in the Subject: header line has been removed
4533 because it now seems ill-conceived.
4535 15. Minor typos in src/EDITME comments corrected.
4537 16. Installed latest exipick from John Jetmore.
4539 17. If headers_add on a router specified a text string that was too long for
4540 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4541 of string_sprintf() is now avoided.
4543 18. $message_body_size was not set (it was always zero) when running the DATA
4544 ACL and the local_scan() function.
4546 19. For the "mail" command in an Exim filter, no default was being set for
4547 the once_repeat time, causing a random time value to be used if "once" was
4548 specified. (If the value happened to be <= 0, no repeat happened.) The
4549 default is now 0s, meaning "never repeat". The "vacation" command was OK
4550 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4551 (I found it when inspecting the code).
4553 20. There is now an overall timeout for performing a callout verification. It
4554 defaults to 4 times the callout timeout, which applies to individual SMTP
4555 commands during the callout. The overall timeout applies when there is more
4556 than one host that can be tried. The timeout is checked before trying the
4557 next host. This prevents very long delays if there are a large number of
4558 hosts and all are timing out (e.g. when the network connections are timing
4559 out). The value of the overall timeout can be changed by specifying an
4560 additional sub-option for "callout", called "maxwait". For example:
4562 verify = sender/callout=5s,maxwait=20s
4564 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4565 to the end before writing, but this should make it even safer).
4567 22. Exim was forgetting that it had advertised PIPELINING for the second and
4568 subsequent messages on an SMTP connection. It was also not resetting its
4569 memory on STARTTLS and an internal HELO.
4571 23. When Exim logs an SMTP synchronization error within a session, it now
4572 records whether PIPELINING has been advertised or not.
4574 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4575 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4576 rather than long int.
4578 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4580 26. Added the never_mail option to autoreply.
4586 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4587 crash if the getsockname() call failed; this can happen if a connection is
4588 closed very soon after it is established. The problem was simply in the
4589 order in which certain operations were done, causing Exim to try to write
4590 to the SMTP stream before it had set up the file descriptor. The bug has
4591 been fixed by making things happen in the correct order.
4597 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4598 before the connection was closed, thus losing the rejection response.
4600 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4601 some early Solaris releases, but causes trouble in current releases where
4602 socklen_t is defined.
4604 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4607 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4610 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4611 that releasing the top of it at the end releases what was used for sub-
4612 expansions (unless the block got too big). However, discard this block if
4613 the first thing is a variable or header, so that we can use its block when
4614 it is dynamic (useful for very large $message_headers, for example).
4616 6. Lookups now cache *every* query, not just the most recent. A new, separate
4617 store pool is used for this. It can be recovered when all lookup caches are
4618 flushed. Lookups now release memory at the end of their result strings.
4619 This has involved some general refactoring of the lookup sources.
4621 7. Some code has been added to the store_xxx() functions to reduce the amount
4622 of flapping under certain conditions.
4624 8. log_incoming_interface used to affect only the <= reception log lines. Now
4625 it causes the local interface and port to be added to several more SMTP log
4626 lines, for example "SMTP connection from", and rejection lines.
4628 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4630 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4632 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4633 could be overwritten at the end of the current message (or the start of a
4634 new message if it was set in a HELO ACL). The value is now preserved for
4635 the duration of the SMTP connection.
4637 12. If a transport had a headers_rewrite setting, and a matching header line
4638 contained an unqualified address, that address was qualified, even if it
4639 did not match any rewriting rules. The underlying bug was that the values
4640 of the flags that permit the existence of unqualified sender and recipient
4641 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4642 non-local messages, and by -bnq for local messages) were not being
4643 preserved with the message after it was received.
4645 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4646 "next input=" as part of the text comprising the host identity instead of
4647 the correct text. The code was using the same buffer for two different
4648 strings. However, depending on which order the printing function evaluated
4649 its arguments, the bug did not always show up. Under Linux, for example, my
4650 test suite worked just fine.
4652 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4653 doesn't work with some older versions of Perl. It has been changed to "my",
4654 which in any case is probably the better facility to use.
4656 15. A really picky compiler found some instances of statements for creating
4657 error messages that either had too many or two few arguments for the format
4660 16. The size of the buffer for calls to the DNS resolver has been increased
4661 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4662 for addresses that have a lot of PTR records. This alleviates a problem; it
4663 does not fully solve it.
4665 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4666 buffer now truncates the list and logs the incident, which is the same
4667 action as happens when Exim is looking up a host name and its aliases.
4668 Previously in this situation something unpredictable would happen;
4669 sometimes it was "internal error: store_reset failed".
4671 18. If a server dropped the connection unexpectedly when an Exim client was
4672 using GnuTLS and trying to read a response, the client delivery process
4673 crashed while trying to generate an error log message.
4675 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4676 string, for example:
4678 warn message = H1: something\nH2: something
4680 the text was added as a single header line from Exim's point of view
4681 though it ended up OK in the delivered message. However, searching for the
4682 second and subsequent header lines using $h_h2: did not work. This has been
4683 fixed. Similarly, if a system filter added multiple headers in this way,
4684 the routers could not see them.
4686 20. Expanded the error message when iplsearch is called with an invalid key to
4687 suggest using net-iplsearch in a host list.
4689 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4690 ACLs are no longer actually imposed (and a message to that effect is
4693 22. If a "gecos" field in a passwd entry contained escaped characters, in
4694 particular, if it contained a \" sequence, Exim got it wrong when building
4695 a From: or a Sender: header from that name. A second bug also caused
4696 incorrect handling when an unquoted " was present following a character
4697 that needed quoting.
4699 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4700 was not being matched caselessly.
4702 24. Arranged for all hyphens in the exim.8 source to be escaped with
4705 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4706 properly. Recipient callout cache records were still being keyed to include
4707 the sender, even when use_sender was set false. This led to far more
4708 callouts that were necessary. The sender is no longer included in the key
4709 when use_sender is false.
4711 26. Added "control = submission" modifier to ACLs.
4713 27. Added the ${base62d: operator to decode base 62 numbers.
4715 28. dnsdb lookups can now access SRV records.
4717 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4718 the configuration file.
4720 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4721 (-v) mode. This makes the output for a verbose queue run more intelligible.
4723 31. Added a use_postmaster feature to recipient callouts.
4725 32. Added the $body_zerocount variable, containing the number of binary zero
4726 bytes in the message body.
4728 33. The time of last modification of the "new" subdirectory is now used as the
4729 "mailbox time last read" when there is a quota error for a maildir
4732 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4734 35. Added +ignore_unknown as a special item in host lists.
4736 36. Code for decoding IPv6 addresses in host lists is now included, even if
4737 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4738 address was recognized as an IP address, but was then not correctly decoded
4739 into binary, causing unexpected and incorrect effects when compared with
4746 1. Very minor rewording of debugging text in manualroute to say "list of
4747 hosts" instead of "hostlist".
4749 2. If verify=header_syntax was set, and a header line with an unqualified
4750 address (no domain) and a large number of spaces between the end of the
4751 name and the colon was received, the reception process suffered a buffer
4752 overflow, and (when I tested it) crashed. This was caused by some obsolete
4753 code that should have been removed. The fix is to remove it!
4755 3. When running in the test harness, delay a bit after writing a bounce
4756 message to get a bit more predictability in the log output.
4758 4. Added a call to search_tidyup() just before forking a reception process. In
4759 theory, someone could use a lookup in the expansion of smtp_accept_max_
4760 per_host which, without the tidyup, could leave open a database connection.
4762 5. Added the variables $recipient_data and $sender_data which get set from a
4763 lookup success in an ACL "recipients" or "senders" condition, or a router
4764 "senders" option, similar to $domain_data and $local_part_data.
4766 6. Moved the writing of debug_print from before to after the "senders" test
4769 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4770 problems for message scanning, either using a data ACL, or using
4771 local_scan() because the Received: header was not generated till after they
4772 were called (in order to set the time as the time of reception completion).
4773 I have revised the way this works. The header is now generated after the
4774 body is received, but before the ACL or local_scan() are called. After they
4775 are run, the timestamp in the header is updated.
4781 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4782 before starting a queue runner without re-exec. This happened only when
4783 deliver_drop_privilege was set or when the Exim user was set to root. The
4784 effect of the bug was that timeouts during subsequent deliveries caused
4785 crashes instead of being properly handled. The handler is now left at its
4786 default (and expected) setting.
4788 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4789 message, again when deliver_drop_privilege is set or Exim is run as root.
4790 The bug described in (1) was not present in this case, but the tidying up
4791 of the other signals was missing. I have made the two cases consistent.
4793 3. The ignore_target_hosts setting on a manualroute router was being ignored
4794 for hosts that were looked up using the /MX notation.
4796 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4799 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4800 operated on the sender address. After changing the $sender_address to <>
4801 for the sender address verify, Exim was re-instated it as the original
4802 (before rewriting) address, but remembering that it had rewritten it, so it
4803 wasn't rewriting it again. This bug also had the effect of breaking the
4804 sender address verification caching when the sender address was rewritten.
4806 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4807 This has been changed so that if the ip literal address matches
4808 ignore_target_hosts, the router declines.
4810 7. Added expansion conditions match_domain, match_address, and match_local_
4811 part (NOT match_host).
4813 8. The placeholder for the Received: header didn't have a length field set.
4815 9. Added code to Exim itself and to exim_lock to test for a specific race
4816 condition that could lead to file corruption when using MBX delivery. The
4817 issue is with the lockfile that is created in /tmp. If this file is removed
4818 after a process has opened it but before that process has acquired a lock,
4819 there is the potential for a second process to recreate the file and also
4820 acquire a lock. This could lead to two Exim processes writing to the file
4821 at the same time. The added code performs the same test as UW imapd; it
4822 checks after acquiring the lock that its file descriptor still refers to
4823 the same named file.
4825 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4826 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4829 11. Added the smtp_active_hostname option. If used, this will typically be made
4830 to depend on the incoming interface address. Because $interface_address is
4831 not set up until the daemon has forked a reception process, error responses
4832 that can happen earlier (such as "too many connections") no longer contain
4835 12. If an expansion in a condition on a "warn" statement fails because a lookup
4836 defers, the "warn" statement is abandoned, and the next ACL statement is
4837 processed. Previously this caused the whole ACL to be aborted.
4839 13. Added the iplsearch lookup type.
4841 14. Added ident_timeout as a log selector.
4843 15. Added tls_certificate_verified as a log selector.
4845 16. Added a global option tls_require_ciphers (compare the smtp transport
4846 option of the same name). This controls incoming TLS connections.
4848 17. I finally figured out how to make tls_require_ciphers do a similar thing
4849 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4850 before starting the TLS session.
4852 18. Tabs are now shown as \t in -bP output.
4854 19. If the log selector return_path_on_delivery was set, Exim crashed when
4855 bouncing a message because it had too many Received: header lines.
4857 20. If two routers both had headers_remove settings, and the first one included
4858 a superfluous trailing colon, the final name in the first list and the
4859 first name in the second list were incorrectly joined into one item (with a
4860 colon in the middle).
4866 1. Added -C and -D options to the exinext utility, mainly to make it easier
4867 to include in the automated testing, but these could be helpful when
4868 multiple configurations are in use.
4870 2. The exinext utility was not formatting the output nicely when there was
4871 an alternate port involved in the retry record key, nor when there was a
4872 message id as well (for retries that were specific to a specific message
4873 and a specific host). It was also confused by IPv6 addresses, because of
4874 the additional colons they contain. I have fixed the IPv4 problem, and
4875 patched it up to do a reasonable job for IPv6.
4877 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4878 delivery, the log line now contains "pipelined" if PIPELINING was used.
4880 4. An SMTP transport process used to panic and die if the bind() call to set
4881 an explicit outgoing interface failed. This has been changed; it is now
4882 treated in the same way as a connect() failure.
4884 5. A reference to $sender_host_name in the part of a conditional expansion
4885 that was being skipped was still causing a DNS lookup. This no longer
4888 6. The def: expansion condition was not recognizing references to header lines
4889 that used bh_ and bheader_.
4891 7. Added the _cache feature to named lists.
4893 8. The code for checking quota_filecount in the appendfile transport was
4894 allowing one more file than it should have been.
4896 9. For compatibility with Sendmail, the command line option
4904 and sets the incoming protocol and host name (for trusted callers). The
4905 host name and its colon can be omitted when only the protocol is to be set.
4906 Note the Exim already has two private options, -pd and -ps, that refer to
4907 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4908 "s", but I don't think that's a major issue.
4910 10. A number of refactoring changes to the code, none of which should affect
4913 (a) The number of logging options was getting close to filling up the
4914 32-bit word that was used as a bit map. I have split them into two classes:
4915 those that are passed in the argument to log_write(), and those that are
4916 only ever tested independently outside of that function. These are now in
4917 separate 32-bit words, so there is plenty of room for expansion again.
4918 There is no change in the user interface or the logging behaviour.
4920 (b) When building, for example, log lines, the code previously used a
4921 macro that called string_cat() twice, in order to add two strings. This is
4922 not really sufficiently general. Furthermore, there was one instance where
4923 it was actually wrong because one of the argument was used twice, and in
4924 one call a function was used. (As it happened, calling the function twice
4925 did not affect the overall behaviour.) The macro has been replaced by a
4926 function that can join an arbitrary number of extra strings onto a growing
4929 (c) The code for expansion conditions now uses a table and a binary chop
4930 instead of a serial search (which was left over from when there were very
4931 few conditions). Also, it now recognizes conditions like "pam" even when
4932 the relevant support is not compiled in: a suitably worded error message is
4933 given if an attempt is made to use such a condition.
4935 11. Added ${time_interval:xxxxx}.
4937 12. A bug was causing one of the ddress fields not to be passed back correctly
4938 from remote delivery subprocesses. The field in question was not being
4939 subsequently used, so this caused to problems in practice.
4941 13. Added new log selectors queue_time and deliver_time.
4943 14. Might have fixed a bug in maildirsizefile handling that threw up
4944 "unexpected character" debug warnings, and recalculated the data
4945 unnecessarily. In any case, I expanded the warning message to give more
4948 15. Added the message "Restricted characters in address" to the statements in
4949 the default ACL that block characters like @ and % in local parts.
4951 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4952 Three changes have been made:
4954 (a) There was a serious bug; a negative response to MAIL caused the whole
4955 recipient domain to be cached as invalid, thereby blocking all messages
4956 to all local parts at the same domain, from all senders. This bug has
4957 been fixed. The domain is no longer cached after a negative response to
4958 MAIL if the sender used is not empty.
4960 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4963 (c) A new callout option, "use_sender" has been added for people who want
4964 the modified behaviour.
4970 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4973 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4974 indeed breaks things for older releases.
4976 3. Added additional logging to the case where there is a problem reading data
4977 from a filter that is running in a subprocess using a pipe, in order to
4978 try to track down a specific problem.
4980 4. Testing facility fudge: when running in the test harness and attempting
4981 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4982 getting "No route to host". Convert this to a timeout.
4984 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4987 6. Some OS don't have socklen_t but use size_t instead. This affects the
4988 fifth argument of getsockopt() amongst other things. This is now
4989 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4990 can be set for individual OS. I have set it for SunOS5, OSF1, and
4991 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4992 some earlier ones do not.
4994 7. Change 4.30/15 was not doing the test caselessly.
4996 8. The standard form for an IPv6 address literal was being rejected by address
4997 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4998 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4999 this, as well as the form without the "IPv6" on the front (but only when
5000 address literals are enabled, of course).
5002 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5004 10. Exim crashed if a message with an empty sender address specified by -f
5005 encountered a router with an errors_to setting. This could be provoked only
5006 by a command such as
5010 where an empty string was supplied; "<>" did not hit this bug.
5012 11. Installed PCRE release 4.5.
5014 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5015 remained set. It is now erased.
5017 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5018 times from message ids (which are base 36 rather than the normal 62).
5020 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5021 were being counted as actual protocol errors, and logged if the log
5022 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5023 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5024 and DATA following a set of rejected RCPTs do not count as protocol errors.
5025 In other words, Exim assumes they were pipelined, though this may not
5026 actually be the case. Of course, in all cases the client gets an
5027 appropriate error code.
5029 15. If a lookup fails in an ACL condition, a message about the failure may
5030 be available; it is used if testing the ACL cannot continue, because most
5031 such messages specify what the cause of the deferral is. However, some
5032 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5033 that caused an old message to be retained and used if a later statement
5034 caused a defer, replacing the real cause of the deferral.
5036 16. If an IP address had so many PTR records that the DNS lookup buffer
5037 was not large enough to hold them, Exim could crash while trying to process
5038 the truncated data. It now detects and logs this case.
5040 17. Further to 4.21/58, another change has been made: if (and only if) the
5041 first line of a message (the first header line) ends with CRLF, a bare LF
5042 in a subsequent header line has a space inserted after it, so as not to
5043 terminate the header.
5045 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5046 unnecessarily, used atoi() instead of strtol(), and didn't check the
5047 termination when getting file sizes from file names by regex.
5049 19. Completely re-implemented the support for maildirsize files, in the light
5050 of a number of problems with the previous contributed implementation
5051 (4.30/29). In particular:
5053 . If the quota is zero, the maildirsize file is maintained, but no quota is
5056 . If the maildir directory does not exist, it is created before any attempt
5057 to write a maildirsize file.
5059 . The quota value in the file is just a cache; if the quota is changed in
5060 the transport, the new value overrides.
5062 . A regular expression is available for excluding directories from the
5065 20. The autoreply transport checks the characters in options that define the
5066 message's headers; it allows continued headers, but it was checking with
5067 isspace() after an embedded newline instead of explicitly looking for a
5070 21. If all the "regular" hosts to which an address was routed had passed their
5071 expiry times, and had not reached their retry times, the address was
5072 bounced, even if fallback hosts were defined. Now Exim should go on to try
5075 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5076 equivalent code in the SMTP transport. Some hosts send humungous responses
5077 to HELO/EHLO, more than 1024 it seems.
5079 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5080 gives compiler warnings in some environments. I've now done it "properly",
5083 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5084 (because of problems with the built-in one) was declared to return uschar *
5085 instead of char *, causing compiler failure.
5087 25. Fixed a file descriptor leak when processing alias/forward files.
5089 26. Fixed a minor format string issue in dbfn.c.
5091 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5093 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5094 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5097 29. When a sender address is verified, it is cached, to save repeating the test
5098 when there is more than one recipient in a message. However, when the
5099 verification involves a callout, it is possible for different callout
5100 options to be set for different recipients. It is too complicated to keep
5101 track of this in the cache, so now Exim always runs a verification when a
5102 callout is required, relying on the callout cache for the optimization.
5103 The overhead is duplication of the address routing, but this should not be
5106 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5107 to be verified with callout=postmaster, and the main callout worked but the
5108 postmaster check failed, the verification correctly failed. However, if a
5109 subsequent RCPT command asked for sender verification *without* the
5110 postmaster check, incorrect caching caused this verification also to fail,
5113 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5114 it was not caching the DNS options (qualify_single, search_parents) that
5115 were used when the lookup failed. A subsequent lookup with different
5116 options therefore always gave the same answer, though there were cases
5117 where it should not have. (Example: a "domains = !$mx_any" option on a
5118 dnslookup router: the "domains" option is always processed without any
5119 widening, but the router might have qualify_single set.) Now Exim uses the
5120 cached value only when the same options are set.
5122 32. Added John Jetmore's "exipick" utility to the distribution.
5124 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5125 than a timeout (e.g. a certificate is required, and is not provided), an
5126 Exim server now closes the connection immediately. Previously it waited for
5127 the client to close - but if the client is SSL, it seems that they each
5128 wait for each other, leading to a delay before one of them times out.
5130 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5131 maintained 0.8.x compatibility because I don't think many are using it, and
5132 it is clearly obsolete.
5134 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5137 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5138 client certificate was expired. A simple patch fixes this, though I don't
5139 understand the full logic of why the verify callback is called multiple
5142 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5143 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5144 which causes problems with some clients (such as the Certicom SSL Plus
5145 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5146 disables the coutermeasure allowing Eudora to connect."
5148 38. Exim was not checking that a write() to a log file succeeded. This could
5149 lead to Bad Things if a log got too big, in particular if it hit a file
5150 size limit. Exim now panics and dies if it cannot write to a log file, just
5151 as it does if it cannot open a log file.
5153 39. Modified OS/Makefile-Linux so that it now contains
5155 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5157 The two -D definitions ensure that Exim is compiled with large file
5158 support, which makes it possible to handle log files that are bigger than
5161 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5162 instance) a domain was checked against a named list that involved a lookup,
5163 causing $domain_data to be set, then another domain was checked against the
5164 same list, then the first domain was re-checked, the value of $domain_data
5165 after the final check could be wrong. In particular, if the second check
5166 failed, it could be set empty. This bug probably also applied to
5169 41. The strip_trailing_dot option was not being applied to the address given
5170 with the -f command-line option.
5172 42. The code for reading a message's header from the spool was incrementing
5173 $received_count, but never initializing it. This meant that the value was
5174 incorrect (doubled) while delivering a message in the same process in which
5175 it was received. In the most common configuration of Exim, this never
5176 happens - a fresh exec is done - but it can happen when
5177 deliver_drop_privilege is set.
5179 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5180 it now includes up to 150 characters of the unexpected data in the log
5183 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5184 and building data strings. The size of both of these buffers was 10 000
5185 bytes - far larger than anybody would *ever* want, thought I. Needless to
5186 say, somebody hit the limit. I have increased the maximum line length to
5187 20 000 and the maximum data length of concatenated lines to 100 000. I have
5188 also fixed two bugs, because there was no checking on these buffers. Tsk,
5189 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5190 buffer is too small.
5192 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5193 lsearch lookups. Now it does.
5195 46. When parsing a route_list item in a manualroute router, a fixed-length
5196 buffer was used for the list of hosts. I made this 1024 bytes long,
5197 thinking that nobody would ever have a list of hosts that long. Wrong.
5198 Somebody had a whole pile of complicated expansion conditions, and the
5199 string was silently truncated, leading to an expansion error. It turns out
5200 that it is easier to change to an unlimited length (owing to other changes
5201 that have happened since this code was originally written) than to build
5202 structure for giving a limitation error. The length of the item that
5203 expands into the list of hosts is now unlimited.
5205 47. The lsearch lookup could not handle data where the length of text line was
5206 more than 4095 characters. Such lines were truncated, leading to shortened
5207 data being returned. It should now handle lines of any length.
5209 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5210 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5213 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5214 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5215 no longer happens in comment lines. A list of replacements is now placed
5216 at the head of all of the source files, except those whose only change is
5217 to replace PERL_COMMAND in the very first #! line.
5219 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5220 messages on the queue, with a bottom-up merge sort, using code contributed
5221 by Michael Haardt. This should make operations like -bp somewhat faster on
5222 large queues. It won't affect queue runners, except when queue_run_in_order
5225 51. Installed eximstats 1.31 in the distribution.
5227 52. Added support for SRV lookups to the dnslookup router.
5229 53. If an ACL referred to $message_body or $message_body_end, the value was not
5230 reset for any messages that followed in the same SMTP session.
5232 54. The store-handling optimization for building very long strings was not
5233 differentiating between the different store pools. I don't think this
5234 actually made any difference in practice, but I've tidied it.
5236 55. While running the routers to verify a sender address, $sender_address
5237 was still set to the sender address. This is wrong, because when routing to
5238 send a bounce to the sender, it would be empty. Therefore, I have changed
5239 it so that, while verifying a sender address, $sender_address is set to <>.
5240 (There is no change to what happens when verifying a recipient address.)
5242 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5243 target A or AAAA records (if not already returned) without resetting the
5244 qualify_single or search_parents options of the DNS resolver. These are
5245 inappropriate in this case because the targets of MX and SRV records must
5246 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5247 target that, when qualified, matched something in the local domain. These
5248 two options are now turned off when doing these lookups.
5250 57. It seems that at least some releases of Reiserfs (which does not have the
5251 concept of a fixed number of inodes) returns zero and not -1 for the
5252 number of available inodes. This interacted badly with check_spool_inodes,
5253 which assumed that -1 was the "no such thing" setting. What I have done is
5254 to check that the total number of inodes is greater than zero before doing
5255 the test of how many are available.
5257 58. When a "warn" ACL statement has a log_message modifier, the message is
5258 remembered, and not repeated. This is to avoid a lot of repetition when a
5259 message has many recipients that cause the same warning to be written.
5260 Howewer, Exim was preserving the list of already written lines for an
5261 entire SMTP session, which doesn't seem right. The memory is now reset if a
5262 new message is started.
5264 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5265 debugging output unless log_rewrite was also set.
5267 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5268 of (int)(handle) when we know that handle contains (void *)(-1).
5270 61. The Exim daemon panic-logs an error return when it closes the incoming
5271 connection. However "connection reset by peer" seems to be common, and
5272 isn't really an error worthy of noting specially, so that particular error
5275 62. When Exim is trying to find all the local interfaces, it used to panic and
5276 die if the ioctl to get the interface flags failed. However, it seems that
5277 on at least one OS (Solaris 9) it is possible to have an interface that is
5278 included in the list of interfaces, but for which you get a failure error
5279 for this call. This happens when the interface is not "plumbed" into a
5280 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5281 failure of the "get flags" call assumes that the interface is down.
5283 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5284 makes life easier for people who are doing arithmetic on fields extracted
5285 from dates, where you often get leading zeros that should not be
5286 interpreted as octal.
5288 64. Added qualify_domain to the redirect router, to override the global
5291 65. If a pathologically long header line contained very many addresses (the
5292 report of this problem mentioned 10 000) and each of them was rewritten,
5293 Exim could use up a very large amount of memory. (It kept on making new
5294 copies of the header line as it rewrote, and never released the old ones.)
5295 At the expense of a bit more processing, the header rewriting function has
5296 been changed so that it no longer eats memory in this way.
5298 66. The generation of the Received: header has been moved from the time that a
5299 message starts to be received, to the time that it finishes. The timestamp
5300 in the Received: header should now be very close to that of the <= log
5301 line. There are two side-effects of this change:
5303 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5304 logged header lines no longer include the local Received: line, because
5305 it has not yet been created. The same applies to a copy of the message
5306 that is returned to a non-SMTP sender when a message is rejected.
5308 (b) When a filter file is tested using -bf, no additional Received: header
5309 is added to the test message. After some thought, I decided that this
5312 This change does not affect the value of $received_for. It is still set
5313 after address rewriting, but before local_scan() is called.
5315 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5317 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5318 gave an unhelpful panic error message, and a defer error. I have managed to
5319 change this behaviour so that it now rejects any supplied certificate,
5320 which seems right, as the list of acceptable certificates is empty.
5322 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5323 gave an unhelpful defer error. I have not managed to make this reject any
5324 supplied certificates, but the error message it gives is "no certificate
5325 supplied", which is not helpful.
5327 70. exigrep's output now also includes lines that are not associated with any
5328 message, but which match the given pattern. Implemented by a patch from
5329 Martin Sluka, which also tidied up the Perl a bit.
5331 71. Recipient callout verification, like sender verification, was using <> in
5332 the MAIL FROM command. This isn't really the right thing, since the actual
5333 sender may affect whether the remote host accepts the recipient or not. I
5334 have changed it to use the actual sender in the callout; this means that
5335 the cache record is now keyed on a recipient/sender pair, not just the
5336 recipient address. There doesn't seem to be a real danger of callout loops,
5337 since a callout by the remote host to check the sender would use <>.
5338 [SEE ABOVE: changed after hitting problems.]
5340 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5341 temporary errors. However, in the case of such a code being given after
5342 the end of a data transmission (i.e. after ".") Exim was failing to write
5343 a retry record for the message. (Yes, there was some broken host that was
5344 actually sending 8xx at this point.)
5346 73. An unknown lookup type in a host list could cause Exim to panic-die when
5347 the list was checked. (An example that provoked this was putting <; in the
5348 middle of a list instead of at the start.) If this happened during a DATA
5349 ACL check, a -D file could be left lying around. This kind of configuration
5350 error no longer causes Exim to die; instead it causes a defer errror. The
5351 incident is still logged to the main and panic logs.
5353 74. Buglet left over from Exim 3 conversion. The message "too many messages
5354 in one connection" was written to the rejectlog but not the mainlog, except
5355 when address rewriting (yes!) was being logged.
5357 75. Added write_rejectlog option.
5359 76. When a system filter was run not as root (that is, when system_filter_user
5360 was set), the values of the $n variables were not being returned to the
5361 main process; thus, they were not subsequently available in the $sn
5364 77. Added +return_path_on_delivery log selector.
5366 78. A connection timeout was being treated differently from recipients deferred
5367 when testing hosts_max_try with a message that was older than the host's
5368 retry timeout. (The host should not be counted, thus allowing all hosts to
5369 be tried at least once before bouncing.) This may have been the cause of an
5370 occasionally reported bug whereby a message would remain on the queue
5371 longer than the retry timeout, but would be bounced if a delivery was
5372 forced. I say "may" because I never totally pinned down the problem;
5373 setting up timeout/retry tests is difficult. See also the next item.
5375 79. The ultimate address timeout was not being applied to errors that involved
5376 a combination of host plus message (for example, a timeout on a MAIL
5377 command). When an address resolved to a number of possible hosts, and they
5378 were not all tried for each delivery (e.g. because of hosts_max_try), a
5379 message could remain on the queue longer than the retry timeout.
5381 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5384 81. Fixed an obscure SMTP outgoing bug which required at least the following
5385 conditions: (a) there was another message waiting for the same server;
5386 (b) the server returned 5xx to all RCPT commands in the first message so
5387 that the message was not completed; (c) the server dropped the connection
5388 or gave a negative response to the RSET that Exim sends to abort the
5389 transaction. The observed case was a dropped connection after DATA that had
5390 been sent in pipelining mode. That is, the server had advertised PIPELINING
5391 but was not implementing it correctly. The effect of the bug was incorrect
5392 behaviour, such as trying another host, and this could lead to a crash.
5398 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5399 and daemon.c were passed as pointers to ints; they should have been
5400 pointers to socklen_t variables (which are typically unsigned ints).
5402 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5405 3. Fixed a really odd bug that affected only the testing scheme; patching a
5406 certain fixed string in the binary changed the value of another string that
5407 happened to be identical to the end of the original first string.
5409 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5410 name", it returns that address as the IP address. On some operating
5411 systems (e.g. Solaris), it also passes back the IP address string as the
5412 "host name". However, on others (e.g. Linux), it passes back an empty
5413 string. Exim wasn't checking for this, and was changing the host name to an
5414 empty string, assuming it had been canonicized.
5416 5. Although rare, it is permitted to have more than one PTR record for a given
5417 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5418 all the names associated with an address, because they do in Solaris.
5419 However, it seems that they do not in Linux for data that comes from the
5420 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5421 I found this out when I moved to a new Linux workstation and tried to run
5422 the Exim test suite.
5424 To get round this problem I have changed the code so that it now does its
5425 own call to the DNS to look up PTR records when searching for a host name.
5426 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5427 addresses that are only in /etc/hosts are still found.
5429 This behaviour is, however, controlled by an option called host_lookup_
5430 order, which defaults to "bydns:byaddr". If people want to use the other
5431 order, or indeed, just use one or the other means of lookup, they can
5432 specify it in this variable.
5434 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5435 some operating systems, this comes back from gethostbyaddr() as an empty
5436 string, and this is what Exim used to test for. However, it seems that in
5437 other systems, "." is yielded. Exim now tests for this case too.
5439 7. The values of check_spool_space and check_log_space are now held internally
5440 as a number of kilobytes instead of an absolute number of bytes. If a
5441 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5442 kilobyte. This means that much larger values can be stored.
5444 8. Exim monitor: an attempt to get the action menu when not actually pointing
5445 at a message produces an empty menu entitled "No message selected". This
5446 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5447 no entries in it ("Shell widget menu has zero width and/or height"). So I
5448 have added a single, blank menu entry in this case.
5450 9. Added ${quote_local_part.
5452 10. MIME decoding is now applied to the contents of Subject: header lines when
5455 11. Now that a reference to $sender_host_address automatically causes a reverse
5456 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5457 host lookup before query-style lookups in lists that might use this
5458 variable. This has therefore been abolished, and the "net-" prefix is no
5459 longer necessary for query-style lookups.
5461 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5462 have been a typo for LFLAGS, so it has been changed.
5464 13. The install script calls Exim with "-C /dev/null" in order to find the
5465 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5466 to be output. Howeve, since Exim outputs its version number before the
5467 error, it didn't break the script. It just looked ugly. I fixed this by
5468 always allowing "-C /dev/null" if the caller is root.
5470 14. Ignore overlarge ACL variable number when reading spool file - insurance
5471 against a later release with more variables having written the file.
5473 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5474 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5475 this, as well as the form without the "IPv6" on the front.
5477 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5478 OS/Makefile-Darwin file.
5480 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5481 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5483 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5484 variable. However, some OS use ints for inodes. I've added cast to long int
5485 to get rid of the compiler warning.
5487 19. I had forgotten to lock out "/../" in configuration file names when
5488 ALT_CONFIG_PREFIX was set.
5490 20. Routers used for verification do not need to specify transports. However,
5491 if such a router generated a host list, and callout was configured, Exim
5492 crashed, because it could not find a port number from the (non-existent)
5493 transport. It now assumes port 25 in this circumstance.
5495 21. Added the -t option to exigrep.
5497 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5498 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5499 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5500 them is set, LOOKUP_LSEARCH is forced.
5502 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5504 24. Added sender and host information to the "rejected by local_scan()" log
5505 line; previously there was no indication of these.
5507 25. Added .include_if_exists.
5509 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5510 Linux. It turns out that not all file systems support this. Apparently some
5511 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5512 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5513 not supported on the file descriptor, is now ignored when Exim is trying to
5514 sync a directory. This applies only to Linux.
5516 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5518 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5519 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5522 29. Added support for maildirsize files from supplied patch (modified a bit).
5524 30. The use of :fail: followed by an empty string could lead Exim to respond to
5525 sender verification failures with (e.g.):
5527 550 Verification failed for <xxx>
5528 550 Sender verify failed
5530 where the first response line was missing the '-' that indicates it is not
5531 the final line of the response.
5533 31. The loop for finding the name of the user that called Exim had a hardwired
5534 limit of 10; it now uses the value of finduser_retries, which is used for
5535 all other user lookups.
5537 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5540 33. Exim was neglecting to zero errno before one call of strtol() when
5541 expanding a string and expecting an integer value. On some systems this
5542 resulted in spurious "integer overflow" errors. Also, it was casting the
5543 result into an int without checking.
5545 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5546 did not work. The code looks as if it has *never* worked, though it appears
5547 to have been documented since at least releast 1.62. I have made it work.
5549 35. The "timeout_DNS" error in retry rules, also documented since at least
5550 1.62, also never worked. As it isn't clear exactly what this means, and
5551 clearly it isn't a major issue, I have abolished the feature by treating it
5552 as "timeout", and writing a warning to the main and panic logs.
5554 36. The display of retry rules for -brt wasn't always showing the error code
5557 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5558 timeout_connect_A, timeout_connect_MX.
5560 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5561 to the empty sender.
5563 39. The daemon was not analysing the content of -oX till after it had closed
5564 stderr and disconnected from the controlling terminal. This meant that any
5565 syntax errors were only noted on the panic log, and the return code from
5566 the command was 0. By re-arranging the code a little, I've made the
5567 decoding happen first, so such errors now appear on stderr, and the return
5568 code is 1. However, the actual setting up of the sockets still happens in
5569 the disconnected process, so errors there are still only recorded on the
5572 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5573 connections (as happens on some IP stacks) was logged at start up time as
5574 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5575 it from "IPv6 and IPv4", which means that two separate sockets are being
5578 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5579 says whether AF_INET or AF_INET6 was passed as an argument.
5581 42. Exiwhat output was messed up when time zones were included in log
5584 43. Exiwhat now gives more information about the daemon's listening ports,
5585 and whether -tls-on-connect was used.
5587 44. The "port" option of the smtp transport is now expanded.
5589 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5590 silently ignored. Now an error message is written to the main and panic
5593 46. There's a new ACL modifier called "logwrite" which writes to a log file
5594 as soon as it is encountered.
5596 47. Added $local_user_uid and $local_user_gid at routing time.
5598 48. Exim crashed when trying to verify a sender address that was being
5601 49. Exim was recognizing only a space character after ".include". It now also
5602 recognizes a tab character.
5604 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5605 extracting the relevant information from the specification. The man page no
5606 longer contains scrambled data for the -d option, and I've added a section
5607 at the front about calling Exim under different names.
5609 51. Added "extra_headers" argument to the "mail" command in filter files.
5611 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5614 53. Installed eximstats 1.29.
5616 54. Added transport_filter_timeout as a generic transport option.
5618 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5619 Cc: header lines. This was required by RFC 822, but it not required by RFC
5622 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5623 incoming messages that did not have them. Now it does so only if the
5624 message originates locally, that is, if there is no associated remote host
5625 address. When Resent- header lines are present, this applies to the Resent-
5626 lines rather than the non-Resent- lines.
5628 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5629 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5631 58. Messages for configuration errors now include the name of the main
5632 configuration file - useful now that there may be more than one file in a
5633 list (.included file names were always shown).
5635 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5636 for those rare installations that do not start the daemon as root or run it
5637 setuid root. I've cut out the call to initgroups() if the daemon is not
5640 60. The Exim user and group can now be bound into the binary as text strings
5641 that are looked up at the start of Exim's processing.
5643 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5645 62. Added $mailstore_basename variable.
5647 63. Installed patch to sieve.c from Michael Haardt.
5649 64. When Exim failed to open the panic log after failing to open the main log,
5650 the original message it was trying to log was written to stderr and debug
5651 output, but if they were not available (the usual case in production), it
5652 was lost. Now it is written to syslog before the two lines that record the
5653 failures to open the logs.
5655 65. Users' Exim filters run in subprocesses under the user's uid. It is
5656 possible for a "deliver" command or an alias in a "personal" command to
5657 provoke an address rewrite. If logging of address rewriting is configured,
5658 this fails because the process is not running as root or exim. There may be
5659 a better way of dealing with this, but for the moment (because 4.30 needs
5660 to be released), I have disabled address rewrite logging when running a
5661 filter in a non-root, non-exim process.
5667 1. The buildconfig auxiliary program wasn't quoting the value set for
5668 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5669 not defined. This bug was masked in 4.22 by the effect that was fixed in
5672 2. Some messages that were rejected after a message id was allocated were
5673 shown as "incomplete" by exigrep. It no longer does this for messages that
5674 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5676 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5677 have allow_domain_literals set, the ID did not get logged in the <= line.
5678 Domain literals are now always recognized in Message-ID: header lines.
5680 4. The first argument for a ${extract expansion item is the key name or field
5681 number. Leading and trailing spaces in this item were not being ignored,
5682 causing some misleading effects.
5684 5. When deliver_drop_privilege was set, single queue runner processes started
5685 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5686 same command in the process it spins off) were not dropping privilege.
5688 6. When the daemon running as "exim" started a queue runner, it always
5689 re-executed Exim in the spun-off process. This is a waste of effort when
5690 deliver_drop_privilege is set. The new process now just calls the
5691 queue-runner function directly.
5697 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5700 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5701 ignored. Though the use of -oP was forcing the writing of a pid file, it
5702 was always written to the default place.
5704 3. If the message "no IP address found for host xxxx" is generated during
5705 incoming verification, it is now followed by identification of the incoming
5706 connection (so you can more easily find what provoked it).
5708 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5710 5. Added some features to "harden" Exim a bit more against certain attacks:
5712 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5713 be put in Local/Makefile. This is like the never_users runtime option,
5714 but it cannot be overridden. The default setting is "root".
5716 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5717 prefix string with which any file named in a -C command line option
5720 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5721 is retained for -C and -D only if the caller of Exim is root. Without
5722 it, the exim user may also use -C and -D and retain privilege.
5724 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5725 command line option is disabled.
5727 6. Macro names set by the -D option must start with an upper case letter, just
5728 like macro names defined in the configuration file.
5730 7. Added "dereference=" facility to LDAP.
5732 8. Two instances of the typo "uknown" in the source files are fixed.
5734 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5735 the Configure-Makefile script screwed up while processing it.
5737 10. Incorporated PCRE 4.4.
5739 11. The SMTP synchronization check was not operating right at the start of an
5740 SMTP session. For example, it could not catch a HELO sent before the client
5741 waited for the greeting. There is now a check for outstanding input at the
5742 point when the greeting is written. Because of the duplex, asynchronous
5743 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5744 way, but not yet received, when the check is performed.
5746 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5747 on TCP/IP sockets, because this apparently causes some broken clients to
5750 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5751 unchanged) from the Cygwin maintainer.
5753 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5754 was supported instead of testing for mbx. Effectively a typo.
5756 15. The spa authenticator server code was not checking that the input it
5757 received was valid base64.
5759 16. The debug output line for the "set" modifier in ACLs was not showing the
5760 name of the variable that was being set.
5762 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5764 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5765 lookup on its own. Something else was needed to trigger the lookup. For
5766 example, a match in host_lookup or the need for a host name in a host list.
5767 Now, if $sender_host_name is referenced and the host name has not yet been
5768 looked up, a lookup is performed. If the lookup fails, the variable remains
5769 empty, and $host_lookup_failed is set to "1".
5771 19. Added "eqi" as a case-independent comparison operator.
5773 20. The saslauthd authentication condition could segfault if neither service
5774 nor realm was specified.
5776 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5777 error message that was logged was misleading, and incoming SMTP
5778 connections were dropped. The message is now more accurate, and temporary
5779 errors are given to SMTP connections.
5781 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5782 (except RSET) with a 421 error, until QUIT is received. However, it was
5783 failing to send a response to QUIT.
5785 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5786 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5787 ACL is now run after the helo_try_verify_hosts code.
5789 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5790 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5791 case-independent, so other case variants are also recognized). Apparently
5792 some systems use these upper case variants.
5794 25. If more than two messages were waiting for the same host, and a transport
5795 filter was specified for the transport, Exim sent two messages over the
5796 same TCP/IP connection, and then failed with "socket operation on non-
5797 socket" when it tried to send the third.
5799 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5801 27. The extern definition of crypt16() in expand.c was not being excluded when
5802 the OS had its own crypt16() function.
5804 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5805 as a preferred synonym for return_size_limit, both as an option and as an
5808 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5810 30. Changed the default configuration ACL to relax the local part checking rule
5811 for addresses that are not in any local domains. For these addresses,
5812 slashes and pipe symbols are allowed within local parts, but the sequence
5813 /../ is explicitly forbidden.
5815 31. SPA server authentication was not clearing the challenge buffer before
5818 32. log_message in a "warn" ACL statement was writing to the reject log as
5819 well as to the main log, which contradicts the documentation and doesn't
5820 seem right (because no rejection is happening). So I have stopped it.
5822 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5823 However, I am unable to do any testing of this.
5825 34. Fixed an infelicity in the appendfile transport. When checking directories
5826 for a mailbox, to see if any needed to be created, it was accidentally
5827 using path names with one or more superfluous leading slashes; tracing
5828 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5830 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5831 contents are added to the log line that is written for every discarded
5832 recipient. (Previously a log_message setting was ignored.)
5834 36. The ${quote: operator now quotes the string if it is empty.
5836 37. The install script runs exim in order to find its version number. If for
5837 some reason other than non-existence or emptiness, which it checks, it
5838 could not run './exim', it was installing it with an empty version number,
5839 i.e. as "exim-". This error state is now caught, and the installation is
5842 38. An argument was missing from the function that creates an error message
5843 when Exim fails to connect to the socket for saslauthd authentication.
5844 This could cause Exim to crash, or give a corrupted message.
5846 39. Added isip, isip4, and isip6 to ${if conditions.
5848 40. The ACL variables $acl_xx are now saved with the message, and can be
5849 accessed later in routers, transports, and filters.
5851 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5852 strings in the file are not string-expanded.
5854 42. If a MAIL command specified a SIZE value that was too large to fit into an
5855 int variable, the check against message_size_limit failed. Such values are
5856 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5857 day this will have to be increased, but I don't think I want to be around
5858 when emails are that large.
5865 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5866 iconv() is not standard in FreeBSD.
5868 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5869 IPv6 enabled. The observed symptom was a segmentation fault on return from
5870 the function os_common_find_running_interfaces() in src/os.c.
5872 3. In the check_special_case() function in daemon.c I had used "errno" as an
5873 argument name, which causes warnings on some systems. This was basically a
5874 typo, since it was named "eno" in the comments!
5876 4. The code that waits for the clock to tick (at a resolution of some fraction
5877 of a second) so as to ensure message-id uniqueness was always waiting for
5878 at least one whole tick, when it could have waited for less. [This is
5879 almost certainly not relevant at current processor speeds, where it is
5880 unlikely to ever wait at all. But we try to future-proof.]
5882 5. The function that sleeps for a time interval that includes fractions of a
5883 second contained a race. It did not block SIGALRM between setting the
5884 timer, and suspending (a couple of lines later). If the interval was short
5885 and the sigsuspend() was delayed until after it had expired, the suspension
5886 never ended. On busy systems this could lead to processes getting stuck for
5889 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5890 process, before it forks any delivery processes. The open lookup caching
5891 mechanism meant that the open file or database connection was passed into
5892 the delivery process. The problem was that delivery processes always tidy
5893 up cached lookup data. This could cause a problem for the next delivery
5894 process started by the queue runner, because the external queue runner
5895 process does not know about the closure. So the next delivery process
5896 still has data in the lookup cache. In the case of a file lookup, there was
5897 no problem because closing a file descriptor in a subprocess doesn't affect
5898 the parent. However, if the lookup was caching a connection to a database,
5899 the connection was closed, and the second delivery process was likely to
5900 see errors such as "PGSQL: query failed: server closed the connection
5901 unexpectedly". The problem has been fixed by closing all cached lookups
5902 in a queue runner before running a delivery process.
5904 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5905 seem to have the "const" qualifier which it has on other OS. I've
5908 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5909 *of the same type* (client or server) with the same public name that an
5910 error should be diagnosed.
5912 9. When Exim looked up a host name for an IP address, but failed to find the
5913 original IP address when looking up the host name (a safety check), it
5914 output the message "<ip address> does not match any IP for NULL", which was
5915 confusing, to say the least. The bug was that the host name should have
5916 appeared instead of "NULL".
5918 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5919 user that is built into the binary, and the -C or -D options is used, root
5920 privilege is dropped before the configuration file is read. In addition,
5921 logging is switched to stderr instead of the normal log files. If the
5922 configuration then re-defines the Exim user, the unprivileged environment
5923 is probably not what is expected, so Exim logs a panic warning message (but
5926 However, if deliver_drop_privilege is set, the unprivileged state may well
5927 be exactly what is intended, so the warning has been cut out in that case,
5928 and Exim is allowed to try to write to its normal log files.
5934 1. smtp_return_error_details was not giving details for temporary sender
5935 or receiver verification errors.
5937 2. Diagnose a configuration error if two authenticators have the same public
5940 3. Exim used not to create the message log file for a message until the first
5941 delivery attempt. This could be confusing when incoming messages were held
5942 for policy or load reasons. The message log file is now created at the time
5943 the message is received, and an initial "Received" line is written to it.
5945 4. The automatically generated man page for command line options had a minor
5946 bug that caused no ill effects; however, a more serious problem was that
5947 the procedure for building the man page automatically didn't always
5948 operate. Consequently, release 4.20 contains an out-of-date version. This
5949 shouldn't happen again.
5951 5. When building Exim with embedded Perl support, the script that builds the
5952 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5953 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5955 6. The freeze_tell option was not being used for messages that were frozen on
5956 arrival, either by an ACL or by local_scan().
5958 7. Added the smtp_incomplete_transaction log selector.
5960 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5961 was accepting AUTH without a new EHLO.
5963 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5964 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5967 10. Logging of TCP/IP connections (when configured) now happens in the main
5968 daemon process instead of the child process, so that the TCP/IP connection
5969 count is more accurate (but it can never be perfect).
5971 11. The use of "drop" in a nested ACL was not being handled correctly in the
5972 outer ACL. Now, if condition failure induced by the nested "drop" causes
5973 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5974 or "require"), the connection is dropped.
5976 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5977 that yield "discard" can now be used with an "accept" or a "discard" verb,
5978 but an error is generated for any others (because I can't see a useful way
5979 to define what should happen).
5981 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5982 are now processed in the same way as lines in the Exim configuration file.
5983 In particular, continuation lines are supported.
5985 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5987 15. Added -ti meaning -t -i.
5989 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5990 domains, and warn by logging if others are found.
5992 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5993 returned from GIFCONF when Exim is trying to find the list of interfaces on
5994 a host. The code in os.c has been modified to copy each ifreq to an aligned
5995 structure in all cases.
5997 Also, in some cases, the returned ifreq's were being copied to a 'struct
5998 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5999 means the last couple of bytes of an IPv6 address could be chopped if the
6000 ifreq contained only a normal sockaddr (14 bytes storage).
6002 18. Named domain lists were not supported in the hosts_treat_as_local option.
6003 An entry such as +xxxx was not recognized, and was treated as a literal
6006 19. Ensure that header lines added by a DATA ACL are included in the reject log
6007 if the ACL subsequently rejects the message.
6009 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6010 MD5 (which is deprecated).
6012 21. When testing a filter file using -bf, Exim was writing a message when it
6013 took the sender from a "From " line in the message, but it was not doing so
6014 when it took $return_path from a Return-Path: header line. It now does.
6016 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6017 with a valid header line field name (a series of printing characters
6018 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6020 23. Changed "disc" in the source to "disk" to conform to the documentation and
6021 the book and for uniformity.
6023 24. Ignore Sendmail's -Ooption=value command line item.
6025 25. When execve() failed while trying to run a command in a pipe transport,
6026 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6027 could be confused with a return value of 69 from the command itself. This
6028 has been changed to 127, the value the shell returns if it is asked to run
6029 a non-existent command. The wording for the related log line suggests a
6030 non-existent command as the problem.
6032 26. If received_header_text expands to an empty string, do not add a Received:
6033 header line to the message. (Well, it adds a token one on the spool, but
6034 marks it "old" so that it doesn't get used or transmitted.)
6036 27. Installed eximstats 1.28 (addition of -nt option).
6038 28. There was no check for failure on the call to getsockname() in the daemon
6039 code. This can fail if there is a shortage of resources on the system, with
6040 ENOMEM, for example. A temporary error is now given on failure.
6042 29. Contrary to the C standard, it seems that in some environments, the
6043 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6044 program. Exim now does this explicitly; it affects the formatting of
6045 timestamps using strftime().
6047 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6048 complaints. I've now initialized all the variables, to avoid this.
6050 32. Header lines added by a system filter were not being "seen" during
6051 transport-time rewrites.
6053 33. The info_callback() function passed to OpenSSL is set up with type void
6054 (*)(SSL *, int, int), as described somewhere. However, when calling the
6055 function (actually a macro) that sets it up, the type void(*)() is
6056 expected. I've put in a cast to prevent warnings from picky compilers.
6058 34. If a DNS black list lookup found a CNAME record, but there were no A
6059 records associated with the domain it pointed at, Exim crashed.
6061 35. If a DNS black list lookup returned more than one A record, Exim ignored
6062 all but the first. It now scans all returned addresses if a particular IP
6063 value is being sought. In this situation, the contents of the
6064 $dnslist_value variable are a list of all the addresses, separated by a
6067 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6068 to accept a host name and all its aliases if the forward lookup for any of
6069 them yielded the IP address of the incoming connection. Now it accepts only
6070 those names whose forward lookup yields the correct IP address. Any other
6071 names are discarded. This closes a loophole whereby a rogue DNS
6072 administrator could create reverse DNS records to break through a
6073 wildcarded host restriction in an ACL.
6075 37. If a user filter or a system filter that ran in a subprocess used any of
6076 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6077 the wrong values were passed to the pipe command ($thisaddress had the
6078 value of $0, $0 had the value of $1, etc). This bug was introduced by
6079 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6081 38. Improved the line breaking for long SMTP error messages from ACLs.
6082 Previously, if there was no break point between 40 and 75 characters, Exim
6083 left the rest of the message alone. Two changes have been made: (a) I've
6084 reduced the minimum length to 35 characters; (b) if it can't find a break
6085 point between 35 and 75 characters, it looks ahead and uses the first one
6086 that it finds. This may give the occasional overlong line, but at least the
6087 remaining text gets split now.
6089 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6090 file descriptors might be low, and that setting 1000 would always raise it.
6091 It turns out that in some environments, the limit is already over 1000 and
6092 that lowering it causes trouble. So now Exim takes care not to decrease it.
6094 40. When delivering a message, the value of $return_path is set to $sender_
6095 address at the start of routing (routers may change the value). By an
6096 oversight, this default was not being set up when an address was tested by
6097 -bt or -bv, which affected the outcome if any router or filter referred to
6100 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6101 log only when "message" or "log_message" are set. However, if one of the
6102 conditions was an address verification, or a call to a nested ACL, the
6103 messages generated by the underlying test were being passed through. This
6104 no longer happens. The underlying message is available in $acl_verify_
6105 message for both "message" and "log_message" expansions, so it can be
6106 passed through if needed.
6108 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6109 new expansion $bh_ to give the encoded byte string without charset
6110 translation. Translation happens only if iconv() is available; HAVE_ICONV
6111 indicates this at build time. HEADERS_CHARSET gives the charset to
6112 translate to; headers_charset can change it in the configuration, and
6113 "headers charset" can change it in an individual filter file.
6115 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6116 that creates RFC 2047 encoded "words" labels them as that charset instead
6117 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6118 expansion operator; (ii) when Exim creates a From: line for a local
6119 message; (iii) when a header line is rewritten to include a "phrase" part.
6121 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6122 buggy, causing it to skip the first lines of messages whose message ID
6123 ended in 'D'. This would not have bitten before Exim release 4.14, because
6124 message IDs were unlikely to end in 'D' before then. The effect was to have
6125 incorrect size information for certain domains.
6127 45. #include "config.h" was missing at the start of the crypt16.c module. This
6128 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6131 46. If there was a timeout during a "random" callout check, Exim treated it as
6132 a failure of the random address, and carried on sending RSET and the real
6133 address. If the delay was just some slowness somewhere, the response to the
6134 original RCPT would be taken as a response to RSET and so on, causing
6135 mayhem of various kinds.
6137 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6138 when I implemented it. It didn't allow for the fact that some option values
6139 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6140 the right test for positive values.
6142 48. Domain names in DNS records are case-independent. Exim always looks them up
6143 in lower case. Some resolvers return domain names in exactly the case they
6144 appear in the zone file, that is, they may contain uppercase letters. Not
6145 all resolvers do this - some return always lower case. Exim was treating a
6146 change of case by a resolver as a change of domain, similar to a widening
6147 of a domain abbreviation. This triggered its re-routing code and so it was
6148 trying to route what was effectively the same domain again. This normally
6149 caused routing to fail (because the router wouldn't handle the domain
6150 twice). Now Exim checks for this case specially, and just changes the
6151 casing of the domain that it ultimately uses when it transmits the message
6154 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6157 50. If a filter generated a file delivery with a non-absolute name (possible if
6158 no home directory exists for the router), the forbid_file option was not
6161 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6162 the existing equality matching.
6164 52. Exim was using ints instead of ino_t variables in some places where it was
6165 dealing with inode numbers.
6167 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6168 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6169 called TMPDIR, and if it finds it is different, it changes its value.
6171 54. The smtp_printf() function is now made available to local_scan() so
6172 additional output lines can be written before returning. There is also an
6173 smtp_fflush() function to enable the detection of a dropped connection.
6174 The variables smtp_input and smtp_batched_input are exported to
6177 55. Changed the default runtime configuration: the message "Unknown user"
6178 has been removed from the ACL, and instead placed on the localuser router,
6179 using the cannot_route_message feature. This means that any verification
6180 failures that generate their own messages won't get overridden. Similarly,
6181 the "Unrouteable address" message that was in the ACL for unverifiable
6182 relay addresses has also been removed.
6184 56. Added hosts_avoid_esmtp to the smtp transport.
6186 57. The exicyclog script was not checking for the esoteric option
6187 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6188 will work only if exicyclog is run under the appropriate euid.
6190 58. Following a discussion on the list, the rules by which Exim recognises line
6191 endings on incoming messages have been changed. The -dropcr and drop_cr
6192 options are now no-ops, retained only for backwards compatibility. The
6193 following line terminators are recognized: LF CRLF CR. However, special
6194 processing applies to CR:
6196 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6197 nor a local message in the state where . is a terminator.
6199 (ii) If a bare CR is encountered in a header line, an extra space is added
6200 after the line terminator so as not to end the header. The reasoning
6201 behind this is that bare CRs in header lines are most likely either
6202 to be mistakes, or people trying to play silly games.
6204 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6205 was being incorrectly given as 18 bytes larger than it should have been.
6206 This is a VOB (very old bug).
6208 60. This may never have affected anything current, but just in case it has:
6209 When the local host is found other than at the start of a list of hosts,
6210 the local host, those with the same MX, and any that follow, are discarded.
6211 When the list in question was part of a longer list of hosts, the following
6212 hosts (not currently being processed) were also being discarded. This no
6213 longer happens. I'm not sure if this situation could ever has previously
6216 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6219 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6220 header. This is something that is recommended in a new Internet Draft, and
6221 is something that is documented as being done by Sendmail. There are two
6222 possible values. For messages generated by the autoreply transport, Exim
6225 Auto-Submitted: auto-replied
6227 whereas for all other generated messages (e.g. bounces) it adds
6229 Auto-Submitted: auto-generated
6231 63. The "personal" condition in filters now includes a test for the
6232 Auto-Submitted: header. If it contains the string "auto-" the message it
6233 not considered personal.
6235 64. Added rcpt_include_affixes as a generic transport option.
6237 65. Added queue_only_override (default true).
6239 66. Added the syslog_duplication option.
6241 67. If what should have been the first header line of a message consisted of
6242 a space followed by a colon, Exim was mis-interpreting it as a header line.
6243 It isn't of course - it is syntactically invalid and should therefore be
6244 treated as the start of the message body. The misbehaviour could have
6245 caused a number of strange effects, including loss of data in subsequent
6246 header lines, and spool format errors.
6248 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6249 client host had authenticated. This control can now be exercised by an ACL
6250 for more flexibility.
6252 69. By default, callouts do not happen when testing with -bh. There is now a
6253 variant, -bhc, which does actually run the callout code, including
6254 consulting and updating the callout cache.
6256 70. Added support for saslauthd authentication, courtesy of Alexander
6259 71. If statvfs() failed on the spool or log directories while checking their
6260 size for availability, Exim confusingly gave the error "space shortage".
6261 Furthermore, in debugging mode it crashed with a floating point exception.
6262 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6263 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6264 serious problem, Exim now writes to the main and panic logs when this
6265 happens, with details of the failure. It then refuses to accept the
6266 incoming message, giving the message "spool directory problem" or "log
6267 directory problem" with a 421 code for SMTP messages.
6269 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6270 0, 1, and 2 exist, because some OS complain for execs without them (see
6271 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6272 descriptors. However, the code omitted to check that the open succeeded,
6273 causing mysterious errors if for some reason the permissions on /dev/null
6274 got screwed. Now Exim writes a message to the main and panic logs, and
6275 bombs out if it can't open /dev/null.
6277 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6278 interact so that it is all more flexible. It is supposed to remain
6279 backwards compatible. Also added extra_local_interfaces.
6281 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6282 to bomb out with an assertion failure - to the client this appears as a
6283 connection drop. This problem occurs in the part of the code that was taken
6284 from the Samba project. Fortunately, the assertion is in a very simple
6285 function, so I have fixed this by reproducing the function inline in the
6286 one place where it is called, and arranging for authentication to fail
6287 instead of killing the process with assert().
6289 75. The SPA client code was not working when the server requested OEM rather
6290 than Unicode encoding.
6292 76. Added code to make require_files with a specific uid setting more usable in
6293 the case where statting the file as root fails - usually a non-root-mounted
6294 NFS file system. When this happens and the failure is EACCES, Exim now
6295 forks a subprocess and does the per-uid checking as the relevant uid.
6297 77. Added process_log_path.
6299 78. If log_file_path was not explicitly set, a setting of check_log_space or
6300 check_log_inodes was ignored.
6302 79. If a space check for the spool or log partitions fails, the incident is now
6303 logged. Of course, in the latter case the data may get lost...
6305 80. Added the %p formatting code to string_format() so that it can be used to
6306 print addresses in debug_print(). Adjusted all the address printing in the
6307 debugging in store.c to use %p rather than %d.
6309 81. There was a concern that a line of code in smtp_in.c could overflow a
6310 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6311 initially expressed, the concern was not well-founded, because trailing
6312 spaces are removed early. However, if the trailing spaces were followed by
6313 a NULL, they did not get removed, so the overflow was possible. Two fixes
6316 (a) I re-wrote the offending code in a cleaner fashion.
6317 (b) If an incoming SMTP command contains a NULL character, it is rejected
6320 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6321 runs initgroups(), so that if the Exim user is in any additional groups,
6322 they will be used during message reception.
6328 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob