1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
8 and 3.4.4 - once the server is enabled to respond to an OCSP request
9 it does even when not requested, resulting in a stapling non-aware
10 client dropping the TLS connection.
12 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
13 support variable-length bit vectors. No functional change.
15 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
16 The I= interface field on outgoing lines is now after the H= remote
17 host field, same as incoming lines. There is a separate
18 outgoing_interface log selector which allows you to disable the
21 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
22 If not running log_selector +smtp_connection the mainlog would be held
23 open indefinitely after a "too many connections" event, including to a
24 deleted file after a log rotate. Leave the per net connection logging
25 leaving it open for efficiency as that will be quickly detected by the
26 check on the next write.
28 HS/01 Bug 1671: Fix post transport crash.
29 Processing the wait-<transport> messages could crash the delivery
30 process if the message IDs didn't exist for some reason. When
31 using 'split_spool_directory=yes' the construction of the spool
32 file name failed already, exposing the same netto behaviour.
34 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
35 mime_regex ACL conditions.
37 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
38 to DSN fail messages (bounces): remote IP, remote greeting, remote response
39 to HELO, local diagnostic string.
41 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
42 log line to debug. Even when configured with a tls authenticator many
43 client connections are expected to not authenticate in this way, so
44 an authenticate fail is not an error.
46 HS/02 Add the Exim version string to the process info. This way exiwhat
47 gives some more detail about the running daemon.
49 JH/06 Bug 1395: time-limit cacheing of DNS lookups, to the TTL value. This may
50 matter for fast-change records such as DNSBLs.
52 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
53 retry record, even if constant. There may be multiple transports with
54 different interface settings and the retry behaviour needs to be kept
57 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
59 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
61 JH/10 Bug 840: fix log_defer_output option of pipe transport
66 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
69 JH/02 The smtp transport option "multi_domain" is now expanded.
71 JH/03 The smtp transport now requests PRDR by default, if the server offers
74 JH/04 Certificate name checking on server certificates, when exim is a client,
75 is now done by default. The transport option tls_verify_cert_hostnames
76 can be used to disable this per-host. The build option
77 EXPERIMENTAL_CERTNAMES is withdrawn.
79 JH/05 The value of the tls_verify_certificates smtp transport and main options
80 default to the word "system" to access the system default CA bundle.
81 For GnuTLS, only version 3.0.20 or later.
83 JH/06 Verification of the server certificate for a TLS connection is now tried
84 (but not required) by default. The verification status is now logged by
85 default, for both outbound TLS and client-certificate supplying inbound
88 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
91 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
92 Status Notification (bounce) messages are now MIME format per RFC 3464.
93 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
94 under the control of the dsn_advertise_hosts option, and routers may
95 have a dsn_lasthop option.
97 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
98 default, modifiable by a malware= option. The list separator for
99 the options can now be changed in the usual way. Bug 68.
101 JH/10 The smtp_receive_timeout main option is now expanded before use.
103 JH/11 The incoming_interface log option now also enables logging of the
104 local interface on delivery outgoing connections.
106 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
107 if the interface and destination host and port all match.
109 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
112 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
113 Patch from Andrew Lewis.
115 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
116 now supports optional time-restrictions, weighting, and priority
117 modifiers per server. Patch originally by <rommer@active.by>.
119 JH/16 The spamd_address main option now supports a mixed list of local
120 and remote servers. Remote servers can be IPv6 addresses, and
121 specify a port-range.
123 JH/17 Bug 68: The spamd_address main option now supports an optional
124 timeout value per server.
126 JH/18 Bug 1581: Router and transport options headers_add/remove can
127 now have the list separator specified.
129 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
132 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
135 JH/21 Support for the A6 type of dns record is withdrawn.
137 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
138 rather than the verbs used.
140 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
141 from 255 to 1024 chars.
143 JH/24 Verification callouts now attempt to use TLS by default.
145 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
146 are generic router options now. The defaults didn't change.
148 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
149 Original patch from Alexander Shikoff, worked over by JH.
151 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
152 files if ZCAT_COMMAND is not executable.
154 JH/26 Bug 1539: Add timout/retry options on dnsdb lookups.
156 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
158 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
159 Normally benign, it bites when the pair was led to by a CNAME;
160 modern usage is to not canoicalize the domain to a CNAME target
161 (and we were inconsistent anyway for A-only vs AAAA+A).
163 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
165 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
166 when evaluating $sender_host_dnssec.
168 JH/31 Check the HELO verification lookup for DNSSEC, adding new
169 $sender_helo_dnssec variable.
171 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
173 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
175 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
177 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
178 documented as working, but never had. Support all but $spam_report.
180 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
181 added for tls authenticator.
186 TL/01 When running the test suite, the README says that variables such as
187 no_msglog_check are global and can be placed anywhere in a specific
188 test's script, however it was observed that placement needed to be near
189 the beginning for it to behave that way. Changed the runtest perl
190 script to read through the entire script once to detect and set these
191 variables, reset to the beginning of the script, and then run through
192 the script parsing/test process like normal.
194 TL/02 The BSD's have an arc4random API. One of the functions to induce
195 adding randomness was arc4random_stir(), but it has been removed in
196 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
197 function when detected.
199 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
200 cause callback expansion.
202 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
203 syntax errors in an expansion can be treated as a string instead of
204 logging or causing an error, due to the internal use of bool_lax
205 instead of bool when processing it.
207 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
208 server certificates when making smtp deliveries.
210 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
212 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
214 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
216 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
217 Merged patch from Sebastian Wiedenroth.
219 JH/05 Fix results-pipe from transport process. Several recipients, combined
220 with certificate use, exposed issues where response data items split
221 over buffer boundaries were not parsed properly. This eventually
222 resulted in duplicates being sent. This issue only became common enough
223 to notice due to the introduction of conection certificate information,
224 the item size being so much larger. Found and fixed by Wolfgang Breyha.
226 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
227 size buffer was used, resulting in syntax errors when an expansion
230 JH/07 Add support for directories of certificates when compiled with a GnuTLS
231 version 3.3.6 or later.
233 JH/08 Rename the TPDA expermimental facility to Event Actions. The #ifdef
234 is EXPERIMENTAL_EVENT, the main-configuration and transport options
235 both become "event_action", the variables become $event_name, $event_data
236 and $event_defer_errno. There is a new variable $verify_mode, usable in
237 routers, transports and related events. The tls:cert event is now also
238 raised for inbound connections, if the main configuration event_action
241 TL/06 In test suite, disable OCSP for old versions of openssl which contained
242 early OCSP support, but no stapling (appears to be less than 1.0.0).
244 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
245 server certificate names available under the smtp transport option
246 "tls_verify_cert_hostname" now do not permit multi-component wildcard
249 JH/10 Time-related extraction expansions from certificates now use the main
250 option "timezone" setting for output formatting, and are consistent
251 between OpenSSL and GnuTLS compilations. Bug 1541.
253 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
254 encoded parameter in the incoming message. Bug 1558.
256 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
257 include certificate info, eximon was claiming there were spoolfile
260 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
262 JH/14 Log delivery-related information more consistently, using the sequence
263 "H=<name> [<ip>]" wherever possible.
265 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
266 are problematic for Debian distribution, omit them from the release
269 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
271 JH/16 Fix string representation of time values on 64bit time_t anchitectures.
274 JH/17 Fix a null-indirection in certextract expansions when a nondefault
275 output list separator was used.
280 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
281 checkers that were complaining about end of non-void function with no
284 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
285 This was a regression intruduced in 4.83 by another bugfix.
287 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
289 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
290 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
296 TF/01 Correctly close the server side of TLS when forking for delivery.
298 When a message was received over SMTP with TLS, Exim failed to clear up
299 the incoming connection properly after forking off the child process to
300 deliver the message. In some situations the subsequent outgoing
301 delivery connection happened to have the same fd number as the incoming
302 connection previously had. Exim would try to use TLS and fail, logging
303 a "Bad file descriptor" error.
305 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
306 utilities have not been installed.
308 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
309 temporary space as the ACL may create new global variables.
311 TL/01 LDAP support uses per connection or global context settings, depending
312 upon the detected version of the libraries at build time.
314 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
315 to extract and use the src ip:port in logging and expansions as if it
316 were a direct connection from the outside internet. PPv2 support was
317 updated based on HAProxy spec change in May 2014.
319 JH/02 Add ${listextract {number}{list}{success}{fail}}.
321 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
322 Properly escape header and check for NULL return.
324 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
327 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
329 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
330 characters in header names, implemented as a verify condition.
331 Contributed by Michael Fischer v. Mollard.
333 TL/05 Rename SPF condition results err_perm and err_temp to standardized
334 results permerror and temperror. Previous values are deprecated but
335 still accepted. In a future release, err_perm and err_temp will be
336 completely removed, which will be a backward incompatibility if the
337 ACL tests for either of these two old results. Patch contributed by
338 user bes-internal on the mailing list.
340 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
342 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
343 selectors, in both main and reject logs.
345 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
348 JH/07 Add malware type "sock" for talking to simple daemon.
350 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
352 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
353 routers/transports under cutthrough routing.
355 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
356 numbers. Touch up "bool" conditional to keep the same definition.
358 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
360 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
362 JH/12 Expand items in router/transport headers_add or headers_remove lists
363 individually rather than the list as a whole. Bug 1452.
365 Required for reasonable handling of multiple headers_ options when
366 they may be empty; requires that headers_remove items with embedded
367 colons must have them doubled (or the list-separator changed).
369 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
370 view the policy declared in the DMARC record. Currently, $dmarc_status
371 is a combined value of both the record presence and the result of the
374 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
376 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
377 dnslookup router and the smtp transport (applying to the forward
380 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
381 of ldap servers used for a specific lookup. Patch provided by Heiko
384 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
385 New variable $lookup_dnssec_authenticated for observability.
387 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
388 Patch submitted by Lars Timman.
390 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
392 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
393 Requires trusted mode and valid format message id, aborts otherwise.
394 Patch contributed by Heiko Schlichting.
396 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
397 certextract with support for various fields. Bug 1358.
399 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
400 is requested by default, modifiable by smtp transport option
403 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
404 operate on certificate variables to give certificate fingerprints
405 Also new ${sha256:cert_variable}.
407 JH/23 The PRDR feature is moved from being Experimental into the mainline.
409 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
410 Christian Aistleitner.
412 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
414 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
415 file. Patch from Wolfgang Breyha.
417 JH/25 Expand the coverage of the delivery $host and $host_address to
418 client authenticators run in verify callout. Bug 1476.
420 JH/26 Port service names are now accepted for tls_on_connect_ports, to
421 align with daemon_smtp_ports. Bug 72.
423 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
424 support and error reporting did not work properly.
426 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
427 and is readable. Patch from Andrew Colin Kissa.
429 TL/14 Enhance documentation of ${run expansion and how it parses the
430 commandline after expansion, particularly in the case when an
431 unquoted variable expansion results in an empty value.
433 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
435 PP/02 Fix internal collision of T_APL on systems which support RFC3123
436 by renaming away from it. Addresses GH issue 15, reported by
439 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
441 TL/15 SECURITY: prevent double expansion in math comparison functions
442 (can expand unsanitized data). Not remotely exploitable.
449 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
451 PP/02 Make -n do something, by making it not do something.
452 When combined with -bP, the name of an option is not output.
454 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
457 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
458 $sender_host_name and config options to manage this, and basic check
461 PP/05 DSCP support for outbound connections and control modifier for inbound.
463 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
464 (Only plugin which currently uses this is kerberos4, which nobody should
465 be using, but we should make it available and other future plugins might
466 conceivably use it, even though it would break NAT; stuff *should* be
467 using channel bindings instead).
469 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
470 name; added for Sendmail compatibility; requires admin caller.
471 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
472 just ignore it); requires trusted caller.
473 Also parse but ignore: -Ac -Am -X<logfile>
476 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
478 TL/02 Add +smtp_confirmation as a default logging option.
480 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
481 Patch by Magnus Holmgren from 2007-02-20.
483 TL/04 Bugzilla 1281 - Spec typo.
484 Bugzilla 1283 - Spec typo.
485 Bugzilla 1290 - Spec grammar fixes.
487 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
489 TL/06 Add Experimental DMARC support using libopendmarc libraries.
491 TL/07 Fix an out of order global option causing a segfault. Reported to dev
492 mailing list by by Dmitry Isaikin.
494 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
496 JH/02 Support "G" suffix to numbers in ${if comparisons.
498 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
500 NM/01 Bugzilla 1197 - Spec typo
501 Bugzilla 1196 - Spec examples corrections
503 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
505 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
506 gnutls_enable_pkcs11, but renamed to more accurately indicate its
509 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
510 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
512 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
513 "acl {{name}{arg}...}", and optional args on acl condition
516 JH/05 Permit multiple router/transport headers_add/remove lines.
518 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
520 JH/07 Avoid using a waiting database for a single-message-only transport.
521 Performance patch from Paul Fisher. Bugzilla 1262.
523 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
526 JH/09 Add $headers_added variable, with content from use of ACL modifier
527 add_header (but not yet added to the message). Bugzilla 199.
529 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
530 Pulled from Bugzilla 817 by Wolfgang Breyha.
532 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
534 (nb: this is the same fix as in Exim 4.80.1)
536 JH/11 Add A= logging on delivery lines, and a client_set_id option on
539 JH/12 Add optional authenticated_sender logging to A= and a log_selector
542 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
544 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
545 advertise SMTP AUTH mechanism to us, instead of a generic
546 protocol violation error. Also, make Exim more robust to bad
547 data from the Dovecot auth socket.
549 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
551 When a queue runner is handling a message, Exim first routes the
552 recipient addresses, during which it prunes them based on the retry
553 hints database. After that it attempts to deliver the message to
554 any remaining recipients. It then updates the hints database using
557 So if a recipient address works intermittently, it can get repeatedly
558 deferred at routing time. The retry hints record remains fresh so the
559 address never reaches the final cutoff time.
561 This is a fairly common occurrence when a user is bumping up against
562 their storage quota. Exim had some logic in its local delivery code
563 to deal with this. However it did not apply to per-recipient defers
564 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
566 This change adds a proper retry rule check during routing so that the
567 final cutoff time is checked against the message's age. We only do
568 this check if there is an address retry record and there is not a
569 domain retry record; this implies that previous attempts to handle
570 the address had the retry_use_local_parts option turned on. We use
571 this as an approximation for the destination being like a local
572 delivery, as in LMTP.
574 I suspect this new check makes the old local delivery cutoff check
575 redundant, but I have not verified this so I left the code in place.
577 TF/02 Correct gecos expansion when From: is a prefix of the username.
579 Test 0254 submits a message to Exim with the header
583 When I ran the test suite under the user fanf2, Exim expanded
584 the header to contain my full name, whereas it should have added
585 a Resent-Sender: header. It erroneously treats any prefix of the
586 username as equal to the username.
588 This change corrects that bug.
590 GF/01 DCC debug and logging tidyup
591 Error conditions log to paniclog rather than rejectlog.
592 Debug lines prefixed by "DCC: " to remove any ambiguity.
594 TF/03 Avoid unnecessary rebuilds of lookup-related code.
596 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
597 Bug spotted by Jeremy Harris; was flawed since initial commit.
598 Would have resulted in OCSP responses post-SNI triggering an Exim
599 NULL dereference and crash.
601 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
603 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
604 Bug detection, analysis and fix by Samuel Thibault.
605 Bugzilla 1331, Debian bug #698092.
607 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
609 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
610 Server implementation by Todd Lyons, client by JH.
611 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
612 config variable "prdr_enable" controls whether the server
613 advertises the facility. If the client requests PRDR a new
614 acl_data_smtp_prdr ACL is called once for each recipient, after
615 the body content is received and before the acl_smtp_data ACL.
616 The client is controlled by bolth of: a hosts_try_prdr option
617 on the smtp transport, and the server advertisement.
618 Default client logging of deliveries and rejections involving
619 PRDR are flagged with the string "PRDR".
621 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
622 fclose(). Diagnosis by Todd Lyons.
624 PP/17 Update configure.default to handle IPv6 localhost better.
625 Patch by Alain Williams (plus minor tweaks).
628 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
629 This is now consistent with GnuTLS, and is now documented: the
630 previous undocumented portable approach to treating the option as
631 unset was to force an expansion failure. That still works, and
632 an empty string is now equivalent.
634 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
635 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
636 not performing validation itself.
638 PP/20 Added force_command boolean option to pipe transport.
639 Patch from Nick Koston, of cPanel Inc.
641 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
644 TF/04 Added udpsend ACL modifer and hexquote expansion operator
646 PP/21 Fix eximon continuous updating with timestamped log-files.
647 Broken in a format-string cleanup in 4.80, missed when I repaired the
648 other false fix of the same issue.
649 Report and fix from Heiko Schlichting.
652 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
653 Report from Prashanth Katuri.
655 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
656 It's SecureTransport, so affects any MacOS clients which use the
657 system-integrated TLS libraries, including email clients.
659 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
660 using a MIME ACL for non-SMTP local injection.
661 Report and assistance in diagnosis by Warren Baker.
663 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
665 JH/16 Fix comparisons for 64b. Bugzilla 1385.
667 TL/09 Add expansion variable $authenticated_fail_id to keep track of
668 last id that failed so it may be referenced in subsequent ACL's.
670 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
673 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
674 ldap library initialization, allowing self-signed CA's to be
675 used. Also properly sets require_cert option later in code by
676 using NULL (global ldap config) instead of ldap handle (per
677 session). Bug diagnosis and testing by alxgomz.
679 TL/12 Enhanced documentation in the ratelimit.pl script provided in
680 the src/util/ subdirectory.
682 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
683 renamed to Transport Post Delivery Action by Jeremy Harris, as
686 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
687 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
688 redis_servers = needs to be configured which will be used by the redis
689 lookup. Patch from Warren Baker, of The Packet Hub.
691 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
693 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
694 hostname or reverse DNS when processing a host list. Used suggestions
695 from multiple comments on this bug.
697 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
699 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
700 Missed a few lines, added it to make the runtest require no keyboard
703 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
704 contains upper case chars. Make router use caseful_local_part.
706 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
707 support when GnuTLS has been built with p11-kit.
713 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
715 This, or similar/improved, will also be change PP/11 of 4.82.
721 PP/01 Handle short writes when writing local log-files.
722 In practice, only affects FreeBSD (8 onwards).
723 Bugzilla 1053, with thanks to Dmitry Isaikin.
725 NM/01 Bugzilla 949 - Documentation tweak
727 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
730 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
732 PP/02 Implemented gsasl authenticator.
734 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
736 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
737 `pkg-config foo` for cflags/libs.
739 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
740 with rest of GSASL and with heimdal_gssapi.
742 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
743 `pkg-config foo` for cflags/libs for the TLS implementation.
745 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
746 properties get this fed in as external SSF. A number of robustness
747 and debugging improvements to the cyrus_sasl authenticator.
749 PP/08 cyrus_sasl server now expands the server_realm option.
751 PP/09 Bugzilla 1214 - Log authentication information in reject log.
752 Patch by Jeremy Harris.
754 PP/10 Added dbmjz lookup type.
756 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
758 PP/12 MAIL args handles TAB as well as SP, for better interop with
759 non-compliant senders.
760 Analysis and variant patch by Todd Lyons.
762 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
763 Bug report from Lars Müller <lars@samba.org> (via SUSE),
764 Patch from Dirk Mueller <dmueller@suse.com>
766 PP/13 tls_peerdn now print-escaped for spool files.
767 Observed some $tls_peerdn in wild which contained \n, which resulted
768 in spool file corruption.
770 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
771 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
772 or write after TLS renegotiation, which otherwise led to messages
775 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
776 as a tracking header (ie: a signed header comes before the signature).
777 Patch from Wolfgang Breyha.
779 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
780 comma-sep list; embedded commas doubled.
782 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
784 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
786 Report and patch from Dmitry Banschikov.
788 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
789 Removed SSL_clear() after SSL_new() which led to protocol negotiation
790 failures. We appear to now support TLS1.1+ with Exim.
792 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
793 lets Exim select keys and certificates based upon TLS SNI from client.
794 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
795 before an outbound SMTP session. New log_selector, +tls_sni.
797 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
798 NULL dereference. Report and patch from Alun Jones.
800 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
801 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
802 Not seeing resolver debug output on NetBSD, but suspect this is a
803 resolver implementation change.
805 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
806 Left warnings. Added "eximon gdb" invocation mode.
808 PP/21 Defaulting "accept_8bitmime" to true, not false.
810 PP/22 Added -bw for inetd wait mode support.
812 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
813 locate the relevant includes and libraries. Made this the default.
815 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
816 Bugzilla 1246, report and most of solution from Tomasz Kusy.
818 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
819 This may cause build issues on older platforms.
821 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
822 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
823 gnutls_require_mac & gnutls_require_protocols (no longer supported).
824 Added SNI support via GnuTLS too.
825 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
827 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
829 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
830 Applied second patch from Janne, implementing suggestion to default
831 multiple-strings-in-record handling to match SPF spec.
833 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
835 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
836 read-only, out of scope).
837 Patch from Wolfgang Breyha, report from Stuart Northfield.
839 PP/29 Fix three issues highlighted by clang analyser static analysis.
840 Only crash-plausible issue would require the Cambridge-specific
841 iplookup router and a misconfiguration.
842 Report from Marcin Mirosław.
844 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
846 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
847 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
848 As part of this, removing so much warning spew let me fix some minor
849 real issues in debug logging.
851 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
852 assignment on my part. Fixed.
854 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
855 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
856 Janne Snabb (who went above and beyond: thank you).
858 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
859 string otherwise requires a connection and a bunch more work and it's
860 relatively easy to get wrong. Should also expose TLS library linkage
863 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
864 64-bit ${eval} (JH/03).
866 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
867 GNU libc to support some of the 64-bit stuff, should not lead to
868 conflicts. Defined before os.h is pulled in, so if a given platform
869 needs to override this, it can.
871 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
872 protection layer was required, which is not implemented.
873 Bugzilla 1254, patch from Wolfgang Breyha.
875 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
876 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
877 tls_dhparam take prime identifiers. Also unbreak combination of
878 OpenSSL+DH_params+TLSSNI.
880 PP/39 Disable SSLv2 by default in OpenSSL support.
886 PP/01 Solaris build fix for Oracle's LDAP libraries.
887 Bugzilla 1109, patch from Stephen Usher.
889 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
891 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
894 TF/02 Fix a couple more cases where we did not log the error message
895 when unlink() failed. See also change 4.74-TF/03.
897 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
898 lock up or crash if it happened to be inside a call to libc when it
899 got a SIGUSR1 from exiwhat.
901 The SIGUSR1 handler appends the current process status to the process
902 log which is later printed by exiwhat. It used to use the general
903 purpose logging code to do this, but several functions it calls are
904 not safe for signals.
906 The new output code in the SIGUSR1 handler is specific to the process
907 log, and simple enough that it's easy to inspect for signal safety.
908 Removing some special cases also simplifies the general logging code.
909 Removing the spurious timestamps from the process log simplifies
912 TF/04 Improved ratelimit ACL condition.
914 The /noupdate option has been deprecated in favour of /readonly which
915 has clearer semantics. The /leaky, /strict, and /readonly update modes
916 are mutually exclusive. The update mode is no longer included in the
917 database key; it just determines when the database is updated. (This
918 means that when you upgrde Exim will forget old rate measurements.)
920 Exim now checks that the per_* options are used with an update mode that
921 makes sense for the current ACL. For example, when Exim is processing a
922 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
923 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
924 must specify per_mail/readonly. If you omit the update mode it defaults to
925 /leaky where that makes sense (as before) or /readonly where required.
927 The /noupdate option is now undocumented but still supported for
928 backwards compatibility. It is equivalent to /readonly except that in
929 ACLs where /readonly is required you may specify /leaky/noupdate or
930 /strict/noupdate which are treated the same as /readonly.
932 A useful new feature is the /count= option. This is a generalization
933 of the per_byte option, so that you can measure the throughput of other
934 aggregate values. For example, the per_byte option is now equivalent
935 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
937 The per_rcpt option has been generalized using the /count= mechanism
938 (though it's more complicated than the per_byte equivalence). When it is
939 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
940 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
941 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
942 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
943 Note that using per_rcpt with a non-readonly update mode in more than
944 one ACL will cause the recipients to be double-counted. (The per_mail
945 and per_byte options don't have this problem.)
947 The handling of very low rates has changed slightly. If the computed rate
948 is less than the event's count (usually one) then this event is the first
949 after a long gap. In this case the rate is set to the same as this event's
950 count, so that the first message of a spam run is counted properly.
952 The major new feature is a mechanism for counting the rate of unique
953 events. The new per_addr option counts the number of different
954 recipients that someone has sent messages to in the last time period. It
955 behaves like per_rcpt if all the recipient addresses are different, but
956 duplicate recipient addresses do not increase the measured rate. Like
957 the /count= option this is a general mechanism, so the per_addr option
958 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
959 example, measure the rate that a client uses different sender addresses
960 with the options per_mail/unique=$sender_address. There are further
961 details in the main documentation.
963 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
965 TF/06 Removed a few PCRE remnants.
967 TF/07 Automatically extract Exim's version number from tags in the git
968 repository when doing development or release builds.
970 PP/02 Raise smtp_cmd_buffer_size to 16kB.
971 Bugzilla 879. Patch from Paul Fisher.
973 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
974 Heavily based on revision 40f9a89a from Simon Arlott's tree.
977 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
979 PP/05 Variable $av_failed, true if the AV scanner deferred.
980 Bugzilla 1078. Patch from John Horne.
982 PP/06 Stop make process more reliably on build failure.
983 Bugzilla 1087. Patch from Heiko Schlittermann.
985 PP/07 Make maildir_use_size_file an _expandable_ boolean.
986 Bugzilla 1089. Patch from Heiko Schlittermann.
988 PP/08 Handle ${run} returning more data than OS pipe buffer size.
989 Bugzilla 1131. Patch from Holger Weiß.
991 PP/09 Handle IPv6 addresses with SPF.
992 Bugzilla 860. Patch from Wolfgang Breyha.
994 PP/10 GnuTLS: support TLS 1.2 & 1.1.
996 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
999 PP/11 match_* no longer expand right-hand-side by default.
1000 New compile-time build option, EXPAND_LISTMATCH_RHS.
1001 New expansion conditions, "inlist", "inlisti".
1003 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1005 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1007 PP/14 fix log_write() format string regression from TF/03.
1008 Bugzilla 1152. Patch from Dmitry Isaikin.
1014 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1016 PP/02 Harmonised TLS library version reporting; only show if debugging.
1017 Layout now matches that introduced for other libraries in 4.74 PP/03.
1019 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1021 PP/04 New "dns_use_edns0" global option.
1023 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1026 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1027 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1029 TK/01 Updated PolarSSL code to 0.14.2.
1030 Bugzilla 1097. Patch from Andreas Metzler.
1032 PP/07 Catch divide-by-zero in ${eval:...}.
1033 Fixes bugzilla 1102.
1035 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1038 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1039 format-string attack -- SECURITY: remote arbitrary code execution.
1041 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1042 time unintentionally subject to list matching rules, letting the header
1043 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1044 arbitrary string expansion). This allowed for information disclosure.
1046 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1047 INT_MIN/-1 -- value coerced to INT_MAX.
1053 NM/01 Workround for PCRE version dependency in version reporting
1056 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1057 This fixes portability to compilers other than gcc, notably
1058 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1060 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1061 makefiles for portability to HP-UX and POSIX correctness.
1063 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1064 Also via indented variable definition in the Makefile.
1065 (Debugging by Oliver Heesakkers).
1067 PP/02 Restore caching of spamd results with expanded spamd_address.
1068 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1070 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1071 Improves build reliability. Fix from: Frank Elsner
1073 NM/02 Fix wide character breakage in the rfc2047 coding
1074 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1076 NM/03 Allow underscore in dnslist lookups
1077 Fixes bug 1026. Patch from Graeme Fowler
1079 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1080 Code patches from Adam Ciarcinski of NetBSD.
1082 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1085 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1086 is logged, to avoid truncation. Patch from John Horne.
1088 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1089 Patch from Jakob Hirsch.
1091 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1092 SQL string expansion failure details.
1093 Patch from Andrey Oktyabrski.
1095 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1096 Patch from Simon Arlott.
1098 PP/09 New lookups functionality failed to compile on old gcc which rejects
1099 extern declarations in function scope.
1100 Patch from Oliver Fleischmann
1102 PP/10 Use sig_atomic_t for flags set from signal handlers.
1103 Check getgroups() return and improve debugging.
1104 Fixed developed for diagnosis in bug 927 (which turned out to be
1107 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1108 Patch from Mark Zealey.
1110 PP/12 Bugzilla 1056: Improved spamd server selection.
1111 Patch from Mark Zealey.
1113 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1114 Based on patch from Heiko Schlittermann.
1116 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1117 Patch from Uwe Doering, sign-off by Michael Haardt.
1119 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1120 variable declaration deep within a block. Bug and patch from
1123 PP/15 lookups-Makefile IRIX compatibilty coercion.
1125 PP/16 Make DISABLE_DKIM build knob functional.
1127 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1128 Patch by Simon Arlott
1130 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1131 variable argument macros. Our copy now differs from upstream.
1137 TF/01 Failure to get a lock on a hints database can have serious
1138 consequences so log it to the panic log.
1140 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1141 controlled using the smtp_confirmation log selector.
1143 TF/03 Include the error message when we fail to unlink a spool file.
1145 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1146 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1147 for maintaining out-of-tree patches for some time.
1149 PP/01 Bugzilla 139: Documentation and portability issues.
1150 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1151 Handle per-OS dynamic-module compilation flags.
1153 PP/02 Let /dev/null have normal permissions.
1154 The 4.73 fixes were a little too stringent and complained about the
1155 permissions on /dev/null. Exempt it from some checks.
1156 Reported by Andreas M. Kirchwitz.
1158 PP/03 Report version information for many libraries, including
1159 Exim version information for dynamically loaded libraries. Created
1160 version.h, now support a version extension string for distributors
1161 who patch heavily. Dynamic module ABI change.
1163 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1164 privilege escalation vulnerability whereby the Exim run-time user
1165 can cause root to append content of the attacker's choosing to
1168 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1171 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1172 If dropping privileges for untrusted macros, we disabled normal logging
1173 on the basis that it would fail; for the Exim run-time user, this is not
1174 the case, and it resulted in successful deliveries going unlogged.
1175 Fixed. Reported by Andreas Metzler.
1181 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1182 only prepend for the Resent-* case. Fixes regression introduced in
1183 Exim 4.70 by NM/22 for Bugzilla 607.
1185 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1186 increasing numbers of administrators be bitten by this.
1188 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1190 PP/03 Bugzilla 994: added openssl_options main configuration option.
1192 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1194 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1196 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1198 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1199 without a peer certificate, leading to a segfault because of an
1200 assumption that peers always have certificates. Be a little more
1201 paranoid. Problem reported by Martin Tscholak.
1203 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1204 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1205 NB: ClamAV planning to remove STREAM in "middle of 2010".
1206 CL also introduces -bmalware, various -d+acl logging additions and
1207 more caution in buffer sizes.
1209 PP/09 Implemented reverse_ip expansion operator.
1211 PP/10 Bugzilla 937: provide a "debug" ACL control.
1213 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1215 PP/12 Bugzilla 973: Implement --version.
1217 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1219 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1221 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1223 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1224 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1225 ignore trailing whitespace.
1227 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1229 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1232 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1233 Notification from Dr Andrew Aitchison.
1235 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1236 ExtendedDetectionInfo response format.
1237 Notification from John Horne.
1239 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1242 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1243 XSL and documented dependency on system catalogs, with examples of how
1246 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1249 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1250 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1251 configuration file which is writeable by the Exim user or group.
1253 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1254 of configuration files to cover files specified with the -C option if
1255 they are going to be used with root privileges, not just the default
1258 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1259 option (effectively making it always true).
1261 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1262 files to be used while preserving root privileges.
1264 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1265 that rogue child processes cannot use them.
1267 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1268 run-time user, instead of root.
1270 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1271 Exim run-time user without dropping privileges.
1273 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1274 result string, instead of calling string_vformat() twice with the same
1277 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1278 for other users. Others should always drop root privileges if they use
1279 -C on the command line, even for a whitelisted configure file.
1281 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1283 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1289 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1290 $data_path, and $header_path variables; fixed documentation bugs and
1293 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1294 exipick to access non-standard spools, including the "frozen" queue
1297 NM/01 Bugzilla 965: Support mysql stored procedures.
1298 Patch from Alain Williams
1300 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1302 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1303 Patch from Andreas Metzler
1305 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1306 Patch from Kirill Miazine
1308 NM/05 Bugzilla 671: Added umask to procmail example.
1310 JJ/03 installed exipick 20100323.0, fixing doc bug
1312 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1313 directory. Notification and patch from Dan Rosenberg.
1315 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1317 TK/02 Improve log output when DKIM signing operation fails.
1319 MH/01 Treat the transport option dkim_domain as a colon separated
1320 list, not as a single string, and sign the message with each element,
1321 omitting multiple occurences of the same signer.
1323 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1324 Bugzilla 985, 986. Patch by Simon Arlott
1326 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1327 Patch by Simon Arlott
1329 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1330 MBX locking. Notification from Dan Rosenberg.
1336 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1338 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1340 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1342 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1344 NM/04 Bugzilla 915: Flag broken perl installation during build.
1350 TK/01 Added patch by Johannes Berg that expands the main option
1351 "spamd_address" if it starts with a dollar sign.
1353 TK/02 Write list of recipients to X-Envelope-Sender header when building
1354 the mbox-format spool file for content scanning (suggested by Jakob
1357 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1358 (http://www.dcc-servers.net/) support via dccifd. Activated by
1359 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1361 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1362 by Mark Daniel Reidel <mr@df.eu>.
1364 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1365 When building exim an external PCRE library is now needed -
1366 PCRE is a system library on the majority of modern systems.
1367 See entry on PCRE_LIBS in EDITME file.
1369 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1370 conversation. Added nologin parameter to request.
1371 Patch contributed by Kirill Miazine.
1373 TF/01 Do not log submission mode rewrites if they do not change the address.
1375 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1377 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1378 log files in place. Contributed by Roberto Lima.
1380 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1382 TF/03 Bugzilla 615: When checking the local_parts router precondition
1383 after a local_part_suffix or local_part_prefix option, Exim now
1384 does not use the address's named list lookup cache, since this
1385 contains cached lookups for the whole local part.
1387 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1388 Robert Millan. Documentation is in experimental-spec.txt.
1390 TF/04 Bugzilla 668: Fix parallel build (make -j).
1392 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1394 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1395 Patch provided by Jan Srzednicki.
1397 TF/05 Leading white space used to be stripped from $spam_report which
1398 wrecked the formatting. Now it is preserved.
1400 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1401 that they are available at delivery time.
1403 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1405 TF/08 TLS error reporting now respects the incoming_interface and
1406 incoming_port log selectors.
1408 TF/09 Produce a more useful error message if an SMTP transport's hosts
1409 setting expands to an empty string.
1411 NM/06 Bugzilla 744: EXPN did not work under TLS.
1412 Patch provided by Phil Pennock.
1414 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1415 Patch provided by Richard Godbee.
1417 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1418 acl_smtp_notquit, added index entry.
1420 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1421 Patch provided by Eugene Bujak.
1423 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1424 accept(). Patch provided by Maxim Dounin.
1426 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1427 Patch provided by Phil Pennock.
1429 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1431 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1432 Patch provided by Brad "anomie" Jorsch.
1434 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1435 Patch provided by Dean Brooks.
1437 TK/05 Add native DKIM support (does not depend on external libraries).
1439 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1440 Patch provided by Graeme Fowler.
1442 NM/16 Bugzilla 851: Documentation example syntax fix.
1444 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1446 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1449 NM/19 Bugzilla 745: TLS version reporting.
1450 Patch provided by Phil Pennock.
1452 NM/20 Bugzilla 167: bool: condition support.
1453 Patch provided by Phil Pennock.
1455 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1456 clients. Patch provided by Phil Pennock.
1458 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1459 Patch provided by Brad "anomie" Jorsch.
1461 NM/23 Bugzilla 687: Fix misparses in eximstats.
1462 Patch provided by Heiko Schlittermann.
1464 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1465 Patch provided by Heiko Schlittermann.
1467 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1468 plus update to original patch.
1470 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1472 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1473 Patch provided by David Brownlee.
1475 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1477 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1479 NM/30 Bugzilla 888: TLS documentation bugfixes.
1481 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1483 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1484 Unlike the original bugzilla I have changed all shell scripts in src tree.
1486 NM/33 Bugzilla 898: Transport filter timeout fix.
1487 Patch by Todd Rinaldo.
1489 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1490 Patch by Serge Demonchaux.
1492 NM/35 Bugzilla 39: Base64 decode bug fixes.
1493 Patch by Jakob Hirsch.
1495 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1497 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1499 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1501 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1507 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1508 ALT-N's libdkim that I have put here:
1509 http://duncanthrax.net/exim-experimental/
1511 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1512 were called 'true' and it seems that C99 defines that as a reserved
1513 keyword to be used with 'bool' variable types. That means you could
1514 not include C99-style headers which use bools without triggering
1515 build errors in sieve.c.
1517 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1518 as mailq or other aliases. Changed the --help handling significantly
1519 to do whats expected. exim_usage() emits usage/help information.
1521 SC/01 Added the -bylocaldomain option to eximstats.
1523 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1525 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1527 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1533 PH/01 Another patch from the Sieve maintainer.
1535 PH/02 When an IPv6 address is converted to a string for single-key lookup
1536 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1537 dots are used instead of colons so that keys in lsearch files need not
1538 contain colons. This was done some time before quoting was made available
1539 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1540 (notated using the quote facility) so as to distinguish them from IPv4
1541 keys. This meant that lookups for IP addresses in host lists did not work
1542 for iplsearch lookups.
1544 This has been fixed by arranging for IPv6 addresses to be expressed with
1545 colons if the lookup type is iplsearch. This is not incompatible, because
1546 previously such lookups could never work.
1548 The situation is now rather anomolous, since one *can* have colons in
1549 ordinary lsearch keys. However, making the change in all cases is
1550 incompatible and would probably break a number of configurations.
1552 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1555 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1556 conversion specification without a maximum field width, thereby enabling
1557 a rogue spamd server to cause a buffer overflow. While nobody in their
1558 right mind would setup Exim to query an untrusted spamd server, an
1559 attacker that gains access to a server running spamd could potentially
1560 exploit this vulnerability to run arbitrary code as the Exim user.
1562 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1563 $primary_hostname instead of what libspf2 thinks the hosts name is.
1565 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1566 a directory entry by the name of the lookup key. Previously, if a
1567 symlink pointed to a non-existing file or a file in a directory that
1568 Exim lacked permissions to read, a lookup for a key matching that
1569 symlink would fail. Now it is enough that a matching directory entry
1570 exists, symlink or not. (Bugzilla 503.)
1572 PH/03 The body_linecount and body_zerocount variables are now exported in the
1575 PH/04 Added the $dnslist_matched variable.
1577 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1578 This means they are set thereafter only if the connection becomes
1581 PH/06 Added the client_condition to authenticators so that some can be skipped
1582 by clients under certain conditions.
1584 PH/07 The error message for a badly-placed control=no_multiline_responses left
1585 "_responses" off the end of the name.
1587 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1589 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1590 (without spaces) instead of just copying the configuration text.
1592 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1594 PH/11 Added $max_received_linelength.
1596 PH/12 Added +ignore_defer and +include_defer to host lists.
1598 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1599 way in which PCRE > 7.0 is built.
1601 PH/14 Implemented queue_only_load_latch.
1603 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1604 MAIL command. The effect was to mangle the value on 64-bit systems.
1606 PH/16 Another patch from the Sieve maintainer.
1608 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1610 PH/18 If a system quota error occurred while trying to create the file for
1611 a maildir delivery, the message "Mailbox is full" was not appended to the
1612 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1613 only to a quota excession during the actual writing of the file.
1615 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1616 characters?) which causes problems in log lines. The DN values are now
1617 passed through string_printing() before being added to log lines.
1619 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1620 and InterBase are left for another time.)
1622 PH/21 Added message_body_newlines option.
1624 PH/22 Guard against possible overflow in moan_check_errorcopy().
1626 PH/23 POSIX allows open() to be a macro; guard against that.
1628 PH/24 If the recipient of an error message contained an @ in the local part
1629 (suitably quoted, of course), incorrect values were put in $domain and
1630 $local_part during the evaluation of errors_copy.
1636 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1637 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1640 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1641 issue a MAIL command.
1643 PH/02 In an ACL statement such as
1645 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1647 if a client was not listed at all, or was listed with a value other than
1648 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1649 the condition was not true (as it should be), so access was not denied.
1650 The bug was that the ! inversion was incorrectly passed on to the second
1651 item. This has been fixed.
1653 PH/03 Added additional dnslists conditions == and =& which are different from
1654 = and & when the dns lookup returns more than one IP address.
1656 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1657 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1659 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1660 FSYNC, which compiles an option called disable_fsync that allows for
1661 bypassing fsync(). The documentation is heavily laced with warnings.
1663 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1665 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1666 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1667 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1668 including adding "make clean"; (3) Added -fPIC when compiling the test
1669 dynamically loaded module, to get rid of a warning.
1671 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1672 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1673 = 0s. The bug is otherwise harmless.
1675 PH/07 There was a bug in the dovecot authenticator such that the value of
1676 $auth1 could be overwritten, and so not correctly preserved, after a
1677 successful authentication. This usually meant that the value preserved by
1678 the server_setid option was incorrect.
1680 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1682 PH/09 Installed PCRE release 7.0.
1684 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1685 run for batched SMTP input. It is now run at the start of every message
1686 in the batch. While fixing this I discovered that the process information
1687 (output by running exiwhat) was not always getting set for -bs and -bS
1688 input. This is fixed, and it now also says "batched" for BSMTP.
1690 PH/11 Added control=no_pipelining.
1692 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1693 patch, slightly modified), and move the expansion of helo_data till after
1694 the connection is made in the smtp transport (so it can use these
1697 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1699 PH/14 Added log_selector = +pid.
1701 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1703 PH/16 Add ${if forany and ${if forall.
1705 PH/17 Added dsn_from option to vary the From: line in DSNs.
1707 PH/18 Flush SMTP output before performing a callout, unless control =
1708 no_callout_flush is set.
1710 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1711 was true (the default) a successful delivery failed to delete the retry
1712 item, thus causing premature timeout of the address. The bug is now
1715 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1717 PH/21 Long custom messages for fakedefer and fakereject are now split up
1718 into multiline reponses in the same way that messages for "deny" and
1719 other ACL rejections are.
1721 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1722 with slight modification.
1724 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1725 draft, changing the syntax and factoring some duplicate code".
1727 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1728 for deliveries of the second and subsequent messages over the same SMTP
1731 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1732 ${reduce, with only minor "tidies".
1734 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1736 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1737 expansion side effects.
1739 PH/27 When a message times out after an over-quota error from an Exim-imposed
1740 quota, the bounce message says "mailbox is full". This message was not
1741 being given when it was a system quota that was exceeded. It now should
1744 MH/03 Made $recipients available in local_scan(). local_scan() already has
1745 better access to the recipient list through recipients_list[], but
1746 $recipients can be useful in postmaster-provided expansion strings.
1748 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1749 in the case of a MAIL command with additional options following the
1750 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1751 were accidentally chopped off.
1753 PH/29 SMTP synchronization checks are implemented when a command is read -
1754 there is a check that no more input is waiting when there shouldn't be
1755 any. However, for some commands, a delay in an ACL can mean that it is
1756 some time before the response is written. In this time, more input might
1757 arrive, invalidly. So now there are extra checks after an ACL has run for
1758 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1759 pipelining has not been advertised.
1761 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1763 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1764 returned in the SMTP response when smtp_return_error_details was false.
1765 This has been fixed.
1767 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1768 instead of the C I/O that was originally supplied, because problems were
1769 reported on Solaris.
1771 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1772 Exim which did not show up earlier: it was assuming that a call to
1773 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1774 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1775 was a macro that became an assignment, so it seemed to work. This has
1776 changed to a proper function call with a void return, hence the compile
1777 error. Exim's code has been fixed.
1779 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1782 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1784 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1786 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1787 because it turns out to be too big or there is a timeout) there is no
1788 "Completed" line in the log. When some messages of this type were
1789 selected by exigrep, they were listed as "not completed". Others were
1790 picked up by some special patterns. I have improved the selection
1791 criteria to be more general.
1793 PH/38 The host_find_failed option in the manualroute router can now be set
1794 to "ignore", to completely ignore a host whose IP address cannot be
1795 found. If all hosts are ignored, the behaviour is controlled by the new
1796 host_all_ignored option.
1798 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1799 homing or because of multiple MX records with /mx) generated more than
1800 one IP address, and the following item turned out to be the local host,
1801 all the secondary addresses of the first item were incorrectly removed
1802 from the list, along with the local host and any following hosts (which
1803 is what is supposed to happen).
1805 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1806 whoever called Exim into the -H file. In the case of the daemon it was
1807 behaving confusingly. When first started, it used values for whoever
1808 started the daemon, but after a SIGHUP it used the Exim user (because it
1809 calls itself on a restart). I have changed the code so that it now always
1812 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1813 message are rejected with the same error (e.g. no authentication or bad
1814 sender address), and a DATA command is nevertheless sent (as can happen
1815 with PIPELINING or a stupid MUA), the error message that was given to the
1816 RCPT commands is included in the rejection of the DATA command. This is
1817 intended to be helpful for MUAs that show only the final error to their
1820 PH/42 Another patch from the Sieve maintainer.
1822 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1823 Eximstats - Fixed some broken HTML links and added missing column headers
1825 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1826 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1828 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1830 PH/43 Yet another patch from the Sieve maintainer.
1832 PH/44 I found a way to check for a TCP/IP connection going away before sending
1833 the response to the final '.' that terminates a message, but only in the
1834 case where the client has not sent further data following the '.'
1835 (unfortunately, this is allowed). However, in many cases there won't be
1836 any further data because there won't be any more messages to send. A call
1837 to select() can be used: if it shows that the input is "ready", there is
1838 either input waiting, or the socket has been closed. An attempt to read
1839 the next input character can distinguish the two cases. Previously, Exim
1840 would have sent an OK response which the client would never have see.
1841 This could lead to message repetition. This fix should cure that, at
1842 least in a lot of common cases.
1844 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1845 advertised in response to EHLO.
1851 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1852 fixed by 4.65/MH/01 (is this a record?) are fixed:
1854 (i) An empty string was always treated as zero by the numeric comparison
1855 operators. This behaviour has been restored.
1857 (ii) It is documented that the numeric comparison operators always treat
1858 their arguments as decimal numbers. This was broken in that numbers
1859 starting with 0 were being interpreted as octal.
1861 While fixing these problems I realized that there was another issue that
1862 hadn't been noticed. Values of message_size_limit (both the global option
1863 and the transport option) were treated as octal if they started with 0.
1864 The documentation was vague. These values are now always treated as
1865 decimal, and I will make that clear in the documentation.
1871 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1872 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1875 MH/01 Don't check that the operands of numeric comparison operators are
1876 integers when their expansion is in "skipping" mode (fixes bug
1877 introduced by 4.64-PH/07).
1879 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1880 child addresses, Exim now panics and dies. Previously, because the count
1881 is held in a short int, deliveries were likely to be lost. As such a
1882 large number of recipients for a single message is ridiculous
1883 (performance will be very, very poor), I have chosen to impose a limit
1884 rather than extend the field.
1890 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1891 leftover -K file (the existence of which was triggered by #402).
1892 While we were at it, introduced process PID as part of the -K
1893 filename. This should rule out race conditions when creating
1896 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1897 processing considerably. Previous code took too long for large mails,
1898 triggering a timeout which in turn triggers #401.
1900 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1901 in the DK code in transports.c. sendfile() is not really portable,
1902 hence the _LINUX specificness.
1904 TF/01 In the add_headers option to the mail command in an Exim filter,
1905 there was a bug that Exim would claim a syntax error in any
1906 header after the first one which had an odd number of characters
1909 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1910 callout verification, Exim cached a "reject" for the entire domain. This
1911 is correct for most verifications, but it is not correct for a recipient
1912 verification with use_sender or use_postmaster set, because in that case
1913 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1914 case of MAIL FROM:<> rejection from other early rejections (e.g.
1915 rejection of HELO). When verifying a recipient using a non-null MAIL
1916 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1917 Whatever the result of the callout, the value of the domain cache is
1918 left unchanged (for any other kind of callout, getting as far as trying
1919 RCPT means that the domain itself is ok).
1921 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1924 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1925 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1926 session, the daemon ignores ECONNECT errors and logs others; it now
1927 ignores EPIPE as well.
1929 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1930 (quoted-printable decoding).
1932 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1933 later the small subsequent patch to fix an introduced bug.
1935 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1937 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1939 PH/08 An error is now given if message_size_limit is specified negative.
1941 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1942 to be given (somewhat) arbitrary names.
1944 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1947 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1948 miscellaneous code fixes
1950 PH/10 Added the log_reject_target ACL modifier to specify where to log
1953 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1954 hostname. This is wrong, because it relates to the incoming message (and
1955 probably the interface on which it is arriving) and not to the outgoing
1956 callout (which could be using a different interface). This has been
1957 changed to use the value of the helo_data option from the smtp transport
1958 instead - this is what is used when a message is actually being sent. If
1959 there is no remote transport (possible with a router that sets up host
1960 addresses), $smtp_active_hostname is used.
1962 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1963 tweaks were necessary in order to get it to work (see also 21 below):
1964 (a) The code assumed that strncpy() returns a negative number on buffer
1965 overflow, which isn't the case. Replaced with Exim's string_format()
1967 (b) There were several signed/unsigned issues. I just did the minimum
1968 hacking in of casts. There is scope for a larger refactoring.
1969 (c) The code used strcasecmp() which is not a standard C function.
1970 Replaced with Exim's strcmpic() function.
1971 (d) The code set only $1; it now sets $auth1 as well.
1972 (e) A simple test gave the error "authentication client didn't specify
1973 service in request". It would seem that Dovecot has changed its
1974 interface. Fortunately there's a specification; I followed it and
1975 changed what the client sends and it appears to be working now.
1977 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1980 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1981 address A is aliased to B and C, where B exists and C does not. Without
1982 -v the output is "A verified" because verification stops after a
1983 successful redirection if more than one address is generated. However,
1984 with -v the child addresses are also verified. Exim was outputting "A
1985 failed to verify" and then showing the successful verification for C,
1986 with its parentage. It now outputs "B failed to verify", showing B's
1987 parentage before showing the successful verification of C.
1989 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1990 look up a TXT record in a specific list after matching in a combined
1993 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1994 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1995 they consult the DNS. I had assumed they would set it the way they
1996 wanted; and indeed my experiments on Linux seem to show that in some
1997 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1998 To be on the safe side, however, I have now made the interface to
1999 host_find_byname() similar to host_find_bydns(), with an argument
2000 containing the DNS resolver options. The host_find_byname() function now
2001 sets these options at its start, just as host_find_bydns() does. The smtp
2002 transport options dns_qualify_single and dns_search_parents are passed to
2003 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2004 of host_find_byname() use the default settings of RES_DEFNAMES
2005 (qualify_single) but not RES_DNSRCH (search_parents).
2007 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2008 spool_read_header() do less string testing, by means of a preliminary
2009 switch on the second character of optional "-foo" lines. (This is
2010 overdue, caused by the large number of possibilities that now exist.
2011 Originally there were few.) While I was there, I also converted the
2012 str(n)cmp tests so they don't re-test the leading "-" and the first
2013 character, in the hope this might squeeze out yet more improvement.
2015 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2016 flag allowing group syntax was set by the header_syntax check but not
2017 turned off, possible causing trouble later; (2) The flag was not being
2018 set at all for the header_verify test, causing "group"-style headers to
2019 be rejected. I have now set it in this case, and also caused header_
2020 verify to ignore an empty address taken from a group. While doing this, I
2021 came across some other cases where the code for allowing group syntax
2022 while scanning a header line wasn't quite right (mostly, not resetting
2023 the flag correctly in the right place). These bugs could have caused
2024 trouble for malformed header lines. I hope it is now all correct.
2026 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2027 with the "reply" argument non-NULL. The code, however (which originally
2028 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2029 but it didn't always do it. This confused somebody who was copying the
2030 code for some other use. I have removed all the tests.
2032 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2033 feature that was used to support insecure browsers during the U.S. crypto
2034 embargo. It requires special client support, and Exim is probably the
2035 only MTA that supported it -- and would never use it because real RSA is
2036 always available. This code has been removed, because it had the bad
2037 effect of slowing Exim down by computing (never used) parameters for the
2038 RSA_EXPORT functionality.
2040 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2041 authenticator to fail if there's a tab character in the incoming data
2042 (there should never be unless someone is messing about, as it's supposed
2043 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2044 if the connection is using TLS or if the remote IP is the same as the
2045 local IP, and the "valid-client-cert option" if a client certificate has
2048 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2049 authenticators. This can be used for authorization after authentication
2050 succeeds. (In the case of plaintext, it servers for both authentication
2053 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2054 if any retry times were supplied.
2056 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2057 connection, where there is no client IP address to check. In this
2058 situation, the verify now always succeeds.
2060 PH/25 Applied John Jetmore's -Mset patch.
2062 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2064 PH/27 In a string expansion for a processed (not raw) header when multiple
2065 headers of the same name were present, leading whitespace was being
2066 removed from all of them, but trailing whitespace was being removed only
2067 from the last one. Now trailing whitespace is removed from each header
2068 before concatenation. Completely empty headers in a concatenation (as
2069 before) are ignored.
2071 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2072 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2074 PH/29 [Removed. This was a change that I later backed out, and forgot to
2075 correct the ChangeLog entry (that I had efficiently created) before
2076 committing the later change.]
2078 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2079 address errors (4xx response to RCPT) over the same connection as other
2080 messages routed to the same hosts. Such deliveries are always "forced",
2081 so retry times are not inspected. This resulted in far too many retries
2082 for the affected addresses. The effect occurred only when there were more
2083 hosts than the hosts_max_try setting in the smtp transport when it had
2084 the 4xx errors. Those hosts that it had tried were not added to the list
2085 of hosts for which the message was waiting, so if all were tried, there
2086 was no problem. Two fixes have been applied:
2088 (i) If there are any address or message errors in an SMTP delivery, none
2089 of the hosts (tried or untried) are now added to the list of hosts
2090 for which the message is waiting, so the message should not be a
2091 candidate for sending over the same connection that was used for a
2092 successful delivery of some other message. This seems entirely
2093 reasonable: after all the message is NOT "waiting for some host".
2094 This is so "obvious" that I'm not sure why it wasn't done
2095 previously. Hope I haven't missed anything, but it can't do any
2096 harm, as the worst effect is to miss an optimization.
2098 (ii) If, despite (i), such a delivery is accidentally attempted, the
2099 routing retry time is respected, so at least it doesn't keep
2100 hammering the server.
2102 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2103 in ${readsocket because some servers need this prod.
2105 PH/32 Added some extra debug output when updating a wait-xxx database.
2107 PH/33 The hint "could be header name not terminated by colon", which has been
2108 given for certain expansion errors for a long time, was not being given
2109 for the ${if def:h_colon_omitted{... case.
2111 PH/34 The spec says: "With one important exception, whenever a domain list is
2112 being scanned, $domain contains the subject domain." There was at least
2113 one case where this was not true.
2115 PH/35 The error "getsockname() failed: connection reset by peer" was being
2116 written to the panic log as well as the main log, but it isn't really
2117 panic-worthy as it just means the connection died rather early on. I have
2118 removed the panic log writing for the ECONNRESET error when getsockname()
2121 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2122 runs only) independently of the message's sender address. This meant
2123 that, if the 4xx error was in fact related to the sender, a different
2124 message to the same recipient with a different sender could confuse
2125 things. In particualar, this can happen when sending to a greylisting
2126 server, but other circumstances could also provoke similar problems.
2127 I have changed the default so that the retry time for these errors is now
2128 based a combination of the sender and recipient addresses. This change
2129 can be overridden by setting address_retry_include_sender=false in the
2132 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2133 remote server are returned as part of bounce messages. This was not
2134 happening for LMTP over a pipe (the lmtp transport), but now it is the
2135 same for both kinds of LMTP.
2137 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2138 in header lines that were in fact CNAMEs. This is no longer the case.
2140 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2141 and queue runs started by the daemon processed all messages. This has
2142 been fixed so that -R and -S can now usefully be given with -q<time>.
2144 PH/40 Import PCRE release 6.7 (fixes some bugs).
2146 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2148 PH/42 Give an error if -q is specified more than once.
2150 PH/43 Renamed the variables $interface_address and $interface_port as
2151 $received_ip_address and $received_port, to make it clear that these
2152 values apply to message reception, and not to the outgoing interface when
2153 a message is delivered. (The old names remain recognized, of course.)
2155 PH/44 There was no timeout on the connect() call when using a Unix domain
2156 socket in the ${readsocket expansion. There now is.
2158 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2159 be meaningful with "accept".
2161 SC/01 Eximstats V1.43
2162 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2164 SC/02 Eximstats V1.44
2165 Use a glob alias rather than an array ref in the generated
2166 parser. This improves both readability and performance.
2168 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2169 Collect SpamAssassin and rejection statistics.
2170 Don't display local sender or destination tables unless
2171 there is data to show.
2172 Added average volumes into the top table text output.
2174 SC/04 Eximstats V1.46
2175 Collect data on the number of addresses (recipients)
2176 as well as the number of messages.
2178 SC/05 Eximstats V1.47
2179 Added 'Message too big' to the list of mail rejection
2180 reasons (thanks to Marco Gaiarin).
2182 SC/06 Eximstats V1.48
2183 Mainlog lines which have GMT offsets and are too short to
2184 have a flag are now skipped.
2186 SC/07 Eximstats V1.49 (Alain Williams)
2187 Added the -emptyok flag.
2189 SC/08 Eximstats V1.50
2190 Fixes for obtaining the IP address from reject messages.
2192 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2193 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2194 whitesspace changes from 4.64-PH/27
2196 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2199 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2200 are found, allow negative numbers in numeric criteria)
2202 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2204 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2205 to match changes made in 4.64-PH/43
2207 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2209 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2210 "domain" is actually an IP address, and force a failure. This locks out
2211 those revolvers/nameservers that support "A-for-A" lookups, in
2212 contravention of the specifications.
2214 PH/48 When a host name was looked up from an IP address, and the subsequent
2215 forward lookup of the name timed out, the host name was left in
2216 $sender_host_name, contrary to the specification.
2218 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2219 restricted to single-key lookups, Exim was not diagnosing an error if
2220 * or *@ was used with a query-style lookup.
2222 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2224 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2225 long ago, but noone interested enough thought of it. Let's just say that
2226 the "1.1" means that there are some new functions that weren't there at
2227 some point in the past.
2229 PH/51 Error processing for expansion failure of helo_data from an smtp
2230 transport during callout processing was broken.
2232 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2233 tested/used via the -bh/-bhc/-bs options.
2235 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2236 bug, fixed in subsequent PCRE releases).
2238 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2239 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2241 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2247 SC/01 Use a glob alias rather than an array ref in eximstats generated
2248 parser. This improves both readability and performance.
2250 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2251 Don't display local sender or destination tables in eximstats unless
2252 there is data to show.
2253 Added average volumes into the eximstats top table text output.
2255 SC/03 Collect data on the number of addresses (recipients) as well
2256 as the number of messages in eximstats.
2258 TF/01 Correct an error in the documentation for the redirect router. Exim
2259 does (usually) call initgroups() when daemonizing.
2261 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2262 with consistent privilege compared to when running as a daemon.
2264 TF/03 Note in the spec that $authenticated_id is not set for local
2265 submissions from trusted users.
2267 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2268 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2270 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2271 by adding some example configuration directives to the default
2272 configuration file. A little bit of work is required to uncomment the
2273 directives and define how usernames and passwords are checked, but
2274 there is now a framework to start from.
2276 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2277 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2278 without this. I don't know how relevant this is to other LDAP libraries.
2280 PH/02 Add the verb name to the "unknown ACL verb" error.
2282 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2284 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2286 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2287 directory not expanded when it should be if an expanded home directory
2288 was set for the address (which is overridden by the transport).
2290 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2293 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2294 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2295 because it is too late at that time, and has no effect.
2297 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2298 security issue with \' (bugzilla #107). I could not use the
2299 PQescapeStringConn() function, because it needs a PGconn value as one of
2302 PH/08 When testing addresses using -bt, indicate those final addresses that
2303 are duplicates that would not cause an additional delivery. At least one
2304 person was confused, thinking that -bt output corresponded to deliveries.
2305 (Suppressing duplicates isn't a good idea as you lose the information
2306 about possibly different redirections that led to the duplicates.)
2308 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2309 systems where poll() doesn't work, in particular OS X.
2311 PH/10 Added more information to debugging output for retry time not reached.
2313 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2314 operations in malware.c.
2316 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2319 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2320 syslog_duplication set false, log lines that would normally be written
2321 both the the main log and to the reject log were not written to syslog at
2324 PH/14 In the default configuration, change the use of "message" in ACL warn
2325 statements to "add_header".
2327 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2328 not followed by a command (e.g. "seen endif").
2330 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2331 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2334 PH/17 Added extra conditions to the default value of delay_warning_condition
2338 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2339 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2340 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2343 The Auto-Submitted: and various List- headers are standardised, whereas I
2344 don't think Precedence: ever was.
2346 PH/18 Refactored debugging code in route_finduser() to show more information,
2347 in particular, the error code if getpwnam() issues one.
2349 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2350 This is apparently needed in addition to the PH/07 change above to avoid
2351 any possible encoding problems.
2353 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2354 but not after initializing Perl.
2356 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2357 output them only if debugging. By default they are written stderr,
2358 apparently, which is not desirable.
2360 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2363 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2366 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2368 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2369 authenticated or an ident call has been made. Suppress the default
2370 values for $authenticated_id and $authenticated_sender (but permit -oMai
2371 and -oMas) when testing with -bh.
2373 PH/24 Re-jigged the order of the tests in the default configuration so that the
2374 tests for valid domains and recipients precede the DNS black list and CSA
2375 tests, on the grounds that those ones are more expensive.
2377 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2378 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2379 This bug exists in every version of Exim that I still have, right back to
2382 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2383 However, an attempt to turn on case-sensitivity in a regex key by
2384 including (?-i) didn't work because the subject string was already
2385 lowercased, and the effects were non-intuitive. It turns out that a
2386 one-line patch can be used to allow (?-i) to work as expected.
2392 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2393 other effects) broke the use of negated acl sub-conditions.
2395 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2398 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2399 "Deny" causes Exim to reject the incoming connection with a 554 error.
2400 Unfortunately, if there is a major crisis, such as a disk failure,
2401 tcp-wrappers gives "deny", whereas what one would like would be some
2402 kind of temporary error. A kludge has been added to help with this.
2403 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2404 554 error is used if errno is still zero or contains ENOENT (which occurs
2405 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2408 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2410 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2411 errors. Otherwise a message that provokes a temporary error (when other
2412 messages do not) can cause a whole host to time out.
2414 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2415 addresses were routed directly to files or pipes from a redirect router.
2416 File deliveries just didn't batch; pipe deliveries might have suffered
2419 PH/06 A failure to get a lock for a hints database would erroneously always say
2420 "Failed to get write lock", even when it was really a read lock.
2422 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2423 of 0600. This has been changed to use the value of the lockfile_mode
2424 option (which defaults to 0600).
2426 PH/08 Applied small patch from the Sieve maintainer.
2428 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2429 folder from quota calculations, a direct delivery into this folder messed
2430 up the contents of the maildirsize file. This was because the regex was
2431 used only to exclude .Trash (or whatever) when the size of the mailbox
2432 was calculated. There was no check that a delivery was happening into an
2433 excluded directory. This bug has been fixed by ignoring all quota
2434 processing for deliveries into excluded directories.
2436 PH/10 Added the maildirfolder_create_regex option to appendfile.
2442 PH/01 The code for finding all the local interface addresses on a FreeBSD
2443 system running IPv6 was broken. This may well have applied to all BSD
2444 systems, as well as to others that have similar system calls. The broken
2445 code found IPv4 interfaces correctly, but gave incorrect values for the
2446 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2447 that it would not match correctly against @[] and not recognize the IPv6
2450 PH/02 The ipliteral router was not recognizing addresses of the form user@
2451 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2453 PH/03 Added disable_ipv6.
2455 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2456 decoded form, because it is most often used to construct To: headers
2457 lines in autoreplies, and the decoded form may well be syntactically
2458 invalid. However, $reply_address has leading white space removed, and all
2459 newlines turned into spaces so that the autoreply transport does not
2462 PH/05 If group was specified without a user on a router, and no group or user
2463 was specified on a transport, the group from the router was ignored.
2465 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2466 for visible compile-time settings that can be used to change these
2467 numbers, for those that want even more. Backwards compatibility with old
2468 spool files has been maintained. However, going back to a previous Exim
2469 release will lost any variables that are in spool files.
2471 PH/07 Two small changes when running in the test harness: increase delay when
2472 passing a TCP/IP connection to a new process, in case the original
2473 process has to generate a bounce, and remove special handling of
2474 127.0.0.2 (sic), which is no longer necessary.
2476 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2477 be the same on different OS.
2479 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2482 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2483 whether --show-vars was specified or not
2485 JJ/02 exipick: Added support for new ACL variable spool format introduced
2488 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2489 syntactically invalid From: or Reply-to: line, and a filter used this to
2490 generate an autoreply, and therefore failed to obtain an address for the
2491 autoreply, Exim could try to deliver to a non-existent relative file
2492 name, causing unrelated and misleading errors. What now happens is that
2493 it logs this as a hard delivery error, but does not attempt to create a
2496 PH/11 The exinext utility has a -C option for testing purposes, but although
2497 the given file was scanned by exinext itself; it wasn't being passed on
2498 when Exim was called.
2500 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2501 an end-of-file indication when reading a command response.
2503 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2504 compiled. In many other places in Exim, IPv6 addresses are always
2505 recognized, so I have changed this. It also means that IPv4 domain
2506 literals of the form [IPV4:n.n.n.n] are now always recognized.
2508 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2509 used if the router is not running as root, for example, when verifying at
2510 ACL time, or when using -bh. The debugging output from this situation was
2511 non-existent - all you got was a failure to exec. I have made two
2514 (a) Failures to set uid/gid, the current directory, or a process leader
2515 in a subprocess such as that created by queryprogram now generate
2516 suitable debugging ouput when -d is set.
2518 (b) The queryprogram router detects when it is not running as root,
2519 outputs suitable debugging information if -d is set, and then runs
2520 the subprocess without attempting to change uid/gid.
2522 PH/15 Minor change to Makefile for building test_host (undocumented testing
2525 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2526 additional section of a DNS packet that returns MX or SRV records.
2527 Instead, it always explicitly searches for A/AAAA records. This avoids
2528 major problems that occur when a DNS server includes only records of one
2529 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2530 fixed another bug: if SRV records were looked up and the corresponding
2531 address records were *not* found in the additional section, the port
2532 values from the SRV records were lost.
2534 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2535 using the correct key (the original address) when searching the retry
2536 rules in order to find which one to use for generating the retry hint.
2538 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2539 adding the default one. Similarly, if it contains a Reply-To: header, the
2540 errors_reply_to option, if set, is not used.
2542 PH/19 When calculating a retry time, Exim used to measure the "time since
2543 failure" by looking at the "first failed" field in the retry record. Now
2544 it does not use this if it is later than than the arrival time of the
2545 message. Instead it uses the arrival time. This makes for better
2546 behaviour in cases where some deliveries succeed, thus re-setting the
2547 "first failed" field. An example is a quota failure for a huge message
2548 when small messages continue to be delivered. Without this change, the
2549 "time since failure" will always be short, possible causing more frequent
2550 delivery attempts for the huge message than are intended.
2551 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2553 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2554 $1, $2, $3) because the numerical variables can be reset during some
2555 expansion items (e.g. "match"), thereby losing the authentication data.
2557 PH/21 Make -bV show the size of off_t variables so that the test suite can
2558 decide whether to run tests for quotas > 2G.
2560 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2561 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2562 filecount value is greater than 2G or if a quota value is greater than 2G
2563 on a system where the size of off_t is not greater than 4, a panic error
2566 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2567 never match. The debug and -bh output now contains an explicit error
2568 message indicating a malformed IPv4 address or mask.
2570 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2571 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2572 PH/23 above applies.
2574 PH/25 Do not write to syslog when running in the test harness. The only
2575 occasion when this arises is a failure to open the main or panic logs
2576 (for which there is an explicit test).
2578 PH/26 Added the /no_tell option to "control=freeze".
2580 PH/27 If a host name lookup failed very early in a connection, for example, if
2581 the IP address matched host_lookup and the reverse lookup yielded a name
2582 that did not have a forward lookup, an error message of the form "no IP
2583 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2584 could be logged. Now it outputs the IP address instead of "NULL".
2586 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2587 allows the sender and the authenticated sender to be set when
2588 submitting a message from within Exim. Since child_open_exim() is
2589 documented for local_scan(), the new function should be too.
2591 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2592 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2593 results in an empty string is now treated as unset.
2595 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2597 PH/31 Added sender_verify_fail logging option.
2599 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2600 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2601 tidied the source and removed it altogether.
2603 PH/33 When a queue run was abandoned because the load average was too high, a
2604 log line was always written; now it is written only if the queue_run log
2605 selector is set. In addition, the log line for abandonment now contains
2606 information about the queue run such as the pid. This is always present
2607 in "start" and "stop" lines but was omitted from the "abandon" line.
2609 PH/34 Omit spaces between a header name and the colon in the error message that
2610 is given when verify = headers_syntax fails (if there are lots of them,
2611 the message gets confusing).
2613 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2614 names, as there are now some PTR records that contain slashes. This check
2615 is only to protect against broken name servers that fall over on strange
2616 characters, so the fact that it applies to all lookups doesn't matter.
2618 PH/36 Now that the new test suite is complete, we can remove some of the
2619 special code in Exim that was needed for the old test suite. For example,
2620 sorting DNS records because real resolvers return them in an arbitrary
2621 order. The new test suite's fake resolver always returns records in the
2624 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2625 bounces) except when queue_only is set, to avoid logging races between
2626 the different processes.
2628 PH/38 Panic-die if .include specifies a non-absolute path.
2630 PH/39 A tweak to the "H" retry rule from its user.
2632 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2633 a label. They prevented compilation on older perls.
2635 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2636 a warning to be raised on newish perls.
2638 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2639 on queue. Changes to match documented behaviour of showing count of
2640 messages matching specified criteria.
2642 PH/40 Changed the default ident timeout from 30s to 5s.
2644 PH/41 Added support for the use of login_cap features, on those BSD systems
2645 that have them, for controlling the resources used by pipe deliveries.
2647 PH/42 The content-scanning code uses fopen() to create files in which to put
2648 message data. Previously it was not paying any attention to the mode of
2649 the files. Exim runs with umask(0) because the rest of the code creates
2650 files with open(), and sets the required mode explicitly. Thus, these
2651 files were ending up world-writeable. This was not a big issue, because,
2652 being within the spool directory, they were not world-accessible. I have
2653 created a function called modefopen, which takes an additional mode
2654 argument. It sets umask(777), creates the file, chmods it to the required
2655 mode, then resets the umask. All the relevant calls to fopen() in the
2656 content scanning code have been changed to use this function.
2658 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2659 to 24 hours. This avoids potential overflow problems when processing G
2660 and H retry rules. I suspect nobody ever tinkers with this value.
2662 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2664 PH/45 When the plaintext authenticator is running as a client, the server's
2665 challenges are checked to ensure they are valid base64 strings. By
2666 default, the authentication attempt is cancelled if an invalid string is
2667 received. Setting client_ignore_invalid_base64 true ignores these errors.
2668 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2669 they are received. Thus, the responses can be made to depend on the
2670 challenges. If an invalid string is ignored, an empty string is placed in
2673 PH/46 Messages that are created by the autoreply transport now contains a
2674 References: header, in accordance with RFCs 2822 and 3834.
2676 PH/47 Added authenticated_sender_force to the smtp transport.
2678 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2680 PH/49 Installed latest patch from the Sieve maintainer.
2682 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2683 was also set, the appendfile transport was unnecessarily scanning a
2684 directory of message files (e.g. for maildir delivery) to find the count
2685 of files (along with the size), even though it did not need this
2686 information. It now does the scan only if it needs to find either the
2687 size of the count of files.
2689 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2691 PH/52 Two bugs concerned with error handling when the smtp transport is
2694 (i) Exim was not creating retry information for temporary errors given
2695 for individual recipients after the DATA command when the smtp transport
2696 was used in LMTP mode. This meant that they could be retried too
2697 frequently, and not timed out correctly.
2699 (ii) Exim was setting the flag that allows error details to be returned
2700 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2701 recipients that were returned after the DATA command.
2703 PH/53 This is related to PH/52, but is more general: for any failing address,
2704 when detailed error information was permitted to be returned to the
2705 sender, but the error was temporary, then after the final timeout, only
2706 "retry timeout exceeded" was returned. Now it returns the full error as
2707 well as "retry timeout exceeded".
2709 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2710 do this, and (what is worse) MTAs that accept it.
2712 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2713 will now be deprecated.
2715 PH/56 New os.c-cygwin from the Cygwin maintainer.
2717 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2718 formats (previously only available in exim formats via -bpr, -bpru,
2719 and -bpra. Now also available in native and exiqgrep formats)
2721 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2722 with very large, slow to parse queues
2724 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2726 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2728 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2729 responses to authentication challenges, though it was showing the
2730 challenges; (ii) I've removed the CR characters from the debug output for
2733 PH/58 Allow for the insertion of a newline as well as a space when a string
2734 is turned into more than one encoded-word during RFC 2047 encoding. The
2735 Sieve code now uses this.
2737 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2738 data_4xx, lost_connection, tls_required.
2740 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2741 message was being sent as an SMTP response.
2743 PH/61 Add -l and -k options to exicyclog.
2745 PH/62 When verifying, if an address was redirected to one new address, so that
2746 verification continued, and the new address failed or deferred after
2747 having set something in $address_data, the value of $address_data was not
2748 passed back to the ACL. This was different to the case when no
2749 redirection occurred. The value is now passed back in both cases.
2751 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2752 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2753 use login_cap.h, so on its own it isn't the distinguishing feature. The
2754 new name refers directly to the setclassresources() function.
2756 PH/65 Added configuration files for NetBSD3.
2758 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2760 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2761 is preferred over IPv4.
2763 PH/68 The bounce_return_message and bounce_return_body options were not being
2764 honoured for bounces generated during the reception of non-SMTP messages.
2765 In particular, this applied to messages rejected by the ACL. This bug has
2766 been fixed. However, if bounce_return_message is true and bounce_return_
2767 body is false, the headers that are returned for a non-SMTP message
2768 include only those that have been read before the error was detected.
2769 (In the case of an ACL rejection, they have all been read.)
2771 PH/69 The HTML version of the specification is now built in a directory called
2772 spec_html instead of spec.html, because the latter looks like a path with
2773 a MIME-type, and this confuses some software.
2775 PH/70 Catch two compiler warnings in sieve.c.
2777 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2778 function verify_get_ident() calls ip_connect() to connect a socket, but
2779 if the "connect()" function timed out, ip_connect() used to close the
2780 socket. However, verify_get_ident() also closes the socket later, and in
2781 between Exim writes to the log, which may get opened at this point. When
2782 the socket was closed in ip_connect(), the log could get the same file
2783 descriptor number as the socket. This naturally causes chaos. The fix is
2784 not to close the socket in ip_connect(); the socket should be closed by
2785 the function that creates it. There was only one place in the code where
2786 this was missing, in the iplookup router, which I don't think anybody now
2787 uses, but I've fixed it anyway.
2789 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2790 well as to direct DNS lookups. Otherwise the handling of names in host
2791 lists is inconsistent and therefore confusing.
2797 PH/01 Two changes to the default runtime configuration:
2799 (1) Move the checks for relay_from_hosts and authenticated clients from
2800 after to before the (commented out) DNS black list checks.
2802 (2) Add control=submission to the relay_from_hosts and authenticated
2803 clients checks, on the grounds that messages accepted by these
2804 statements are most likely to be submissions.
2806 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2808 (1) Generate an error if the third argument for the ${prvs expansion is
2811 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2814 (3) Reset the variables that are obtained from the first argument of
2815 ${prvscheck and used in the second argument before leaving the code,
2816 because their memory is reclaimed, so using them afterwards may do
2819 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2820 one (it's much easier than Tom thought :-).
2822 (5) Because of (4), we can now allow for the use of $prvscheck_result
2823 inside the third argument.
2825 PH/03 For some reason, the default setting of PATH when running a command from
2826 a pipe transport was just "/usr/bin". I have changed it to
2829 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2830 anything to be listed in the output from -bV.
2832 PH/05 When a filter generated an autoreply, the entire To: header line was
2833 quoted in the delivery log line, like this:
2835 => >A.N.Other <ano@some.domain> <original@ddress> ...
2837 This has been changed so that it extracts the operative address. There
2838 may be more than one such address. If so, they are comma-separated, like
2841 => >ano@some.domain,ona@other.domain <original@ddress> ...
2843 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2844 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2845 not being looked up in the rDNS to get a host name, Exim was showing the
2846 IP address twice in Received: lines, even though the IP addresses were
2847 identical. For example:
2849 Received: from [1.2.3.4] (helo=[1.2.3.4])
2851 However, if the real host name was known, it was omitting the HELO data
2852 if it matched the actual IP address. This has been tidied up so that it
2853 doesn't show the same IP address twice.
2855 PH/07 When both +timestamp and +memory debugging was on, the value given by
2856 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2857 called by the debug printing, thereby overwriting the timestamp buffer.
2858 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2860 PH/08 When the original message was included in an autoreply transport, it
2861 always said "this is a copy of the message, including all the headers",
2862 even if body_only or headers_only was set. It now gives an appropriate
2865 PH/09 Applied a patch from the Sieve maintainer which:
2867 o fixes some comments
2868 o adds the (disabled) notify extension core
2869 o adds some debug output for the result of if/elsif tests
2870 o points to the current vacation draft in the documentation
2871 and documents the missing references header update
2875 o fixes a bug in processing the envelope test (when testing
2876 multiple envelope elements, the last element determinted the
2879 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2880 Electronic Mail") by including:
2882 Auto-submitted: auto-generated
2884 in the messages that it generates (bounce messages and others, such as
2885 warnings). In the case of bounce messages for non-SMTP mesages, there was
2886 also a typo: it was using "Auto_submitted" (underscore instead of
2887 hyphen). Since every message generated by Exim is necessarily in response
2888 to another message, thes have all been changed to:
2890 Auto-Submitted: auto-replied
2892 in accordance with these statements in the RFC:
2894 The auto-replied keyword:
2896 - SHOULD be used on messages sent in direct response to another
2897 message by an automatic process,
2899 - MUST NOT be used on manually-generated messages,
2901 - MAY be used on Delivery Status Notifications (DSNs) and Message
2902 Disposition Notifications (MDNs),
2904 - MUST NOT be used on messages generated by automatic or periodic
2905 processes, except for messages which are automatic responses to
2908 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2909 to the default Received: header definition.
2911 PH/12 Added log selector acl_warn_skipped (default on).
2913 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2914 variables because (a) they are in the wrong storage pool and (b) even if
2915 they were copied, it wouldn't work properly because of the caching.
2917 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2918 checking when decoding. Apparently there are clients that generate
2919 overlong encoded strings. Why am I not surprised?
2921 PH/15 If the first argument of "${if match_address" was not empty, but did not
2922 contain an "@" character, Exim crashed. Now it writes a panic log message
2923 and treats the condition as false.
2925 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2927 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2928 extension "envelope-auth". The code is finished and in agreement with
2929 other implementations, but there is no documentation so far and in fact,
2930 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2931 not changing the active code.
2933 Print executed "if" and "elsif" statements when debugging is used. This
2934 helps a great deal to understand what a filter does.
2936 Document more things not specified clearly in RFC3028. I had all this
2937 sorted out, when out of a sudden new issues came to my mind. Oops."
2939 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2942 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2943 canonical form (as documented). However, after a host name lookup from
2944 the IP address, check_host() was doing a simple string comparison with
2945 addresses acquired from the DNS when checking that the found name did
2946 have the original IP as one of its addresses. Since any found IPv6
2947 addresses are likely to be in abbreviated form, the comparison could
2948 fail. Luckily, there already exists a function for doing the comparison
2949 by converting both addresses to binary, so now that is used instead of
2950 the text comparison.
2952 PH/20 There was another similar case to PH/19, when a complete host name was
2953 given in a host list; looking up its IP address could give an abbreviated
2954 form, whereas the current host's name might or might not be abbreviated.
2955 The same fix has been applied.
2961 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2962 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2965 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2966 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2968 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2970 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2971 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2972 Also, for "Auto-Submitted", treat anything other than "no" as
2973 non-personal, in accordance with RFC 3834. (Previously it treated
2974 anything starting "auto-" as non-personal.)
2976 TF/01 The control=submission/name=... option had a problem with syntax
2977 errors if the name included a slash character. The /name= option
2978 now slurps the rest of the string, so it can include any characters
2979 but it must come last in the list of options (after /sender_retain
2982 PH/05 Some modifications to the interface to the fake nameserver for the new
2990 TK/01 Added the "success_on_redirect" address verification option. See
2991 NewStuff for rationale and an example.
2993 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2995 PH/02 Patch to exigrep to allow it to work on syslog lines.
2997 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2998 fread() to skip over the body file's header line, because in Cygwin the
2999 header line is locked and is inaccessible.
3001 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3002 co-exist for some time) to make it clear that it is the Exim ID that is
3003 referenced, not the Message-ID: header line.
3005 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3006 string_format() function, because snprintf() does not exist on all
3009 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3010 ${stat: expansion item.
3012 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3013 protocol synchronization error", to keep the pedants happy.
3015 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3016 well as for IRIX systems, when gcc is being used. See the host.c source
3019 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3021 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3024 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3025 transport and to the smtp transport in LMTP mode.
3027 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3029 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3030 run a filter in a subprocess. This could lead to confusion in subsequent
3031 lookups in the parent process. There should also be a search_tidyup() at
3032 the end of the subprocess.
3034 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3035 only if the host matched helo_try_verify_hosts, which caused the
3036 verification to occur when the EHLO/HELO command was issued. The ACL just
3037 tested the remembered result. Now, if a previous verification attempt has
3038 not happened, "verify = helo" does it there and then.
3040 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3042 TK/03 Fix log output including CR from clamd.
3044 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3045 exist provoked a memory error which could cause a segfault.
3047 PH/15 Installed PCRE 6.2
3049 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3051 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3052 of the problem. Specifically, suggested +O2 rather than +O1 for the
3055 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3057 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3058 record, the port was not correctly logged when the outgoing_port log
3059 selector was set (it logged the transort's default port).
3061 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3062 fallback_hosts, and "hosts" in the smtp transport.
3064 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3065 host errors such as "Connection refused".
3067 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3068 authentication with radiusclient 0.4.9:
3070 - Error returned from rc_read_config was caught wrongly
3071 - Username/password not passed on to radius server due to wrong length.
3073 The presumption is that some radiusclient API changes for 4.51/PH/17
3074 were not taken care of correctly. The code is still untested by me (my
3075 Linux distribution still has 0.3.2 of radiusclient), but it was
3076 contributed by a Radius user.
3078 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3079 expanding the "port" option of the smtp transport.
3081 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3082 while reading a MIME header. Thanks to Tom Hughes for a patch.
3084 PH/24 Include config.h inside local_scan.h so that configuration settings are
3087 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3088 that in an ACL for RCPT (for example), you can examine exactly what was
3091 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3092 commands, but it was not correctly comparing the address with the actual
3093 client host address. Thus, it would show the EHLO address in Received:
3094 header lines when this was not necessary.
3096 PH/27 Added the % operator to ${eval:}.
3098 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3099 it should be ignoring failures (because with -C, for example, it has lost
3100 privilege). It wasn't ignoring creation failures other than "already
3103 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3106 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3107 because an input line was too long, either on its own, or by virtue of
3108 too many continuations, the temporary file was not being removed, and the
3109 return code was incorrect.
3111 PH/31 Missing "BOOL" in function definition in filtertest.c.
3113 PH/32 Applied Sieve patches from the maintainer.
3115 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3117 PH/33 Added "verify = not_blind".
3119 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3120 Local/Makefile (with some defaults set). These are used in built scripts
3121 such as exicyclog, but they have never been used in the exim_install
3122 script (though there are many overriding facilities there). I have
3123 arranged that the exim_install script now takes note of these two
3126 PH/35 Installed configuration files for Dragonfly.
3128 PH/36 When a locally submitted message by a trusted user did not contain a
3129 From: header, and the sender address was obtained from -f or from an SMTP
3130 MAIL command, and the trusted user did not use -F to supply a sender
3131 name, $originator_name was incorrectly used when constructing a From:
3132 header. Furthermore, $originator_name was used for submission mode
3133 messages from external hosts without From: headers in a similar way,
3134 which is clearly wrong.
3136 PH/37 Added control=suppress_local_fixups.
3138 PH/38 When log_selector = +received_sender was set, and the addition of the
3139 sender made the log line's construction buffer exactly full, or one byte
3140 less than full, an overflow happened when the terminating "\n" was
3143 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3144 when the result of a list match is failure because a DNS lookup failed.
3146 PH/40 RM_COMMAND is now used in the building process.
3148 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3149 the "build-* directories that it finds.
3151 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3152 address in a domain literal was a prefix of an interface address.
3154 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3155 when verifying a sender address, unless rewrite_headers is false.
3157 PH/44 Wrote a long comment about why errors_to addresses are verified as
3158 recipients, not senders.
3160 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3161 the ratelimit ACL was added.
3163 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3165 PH/46 Added extra information about PostgreSQL errors to the error string.
3167 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3168 avoiding the need to install special zones in a real server. This is
3169 backwards compatible; if it can't find the fake resolver, it drops back.
3170 Thus, both old and new test suites can be run.
3172 TF/02 Added util/ratelimit.pl
3174 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3175 clock is set back in time.
3177 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3178 Candler <B.Candler@pobox.com>.
3180 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3181 OK for addresses that are the result of redirections.
3183 PH/48 A number of further additions for the benefit of the new test suite,
3184 including a fake gethostbyname() that interfaces to the fake DNS resolver
3187 TF/06 The fix for widen_domains has also been applied to qualify_single and
3188 search_parents which are the other dnslookup options that can cause
3191 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3194 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3196 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3197 service" if it encountered an MX record with an empty target hostname.
3198 The message is now "an MX or SRV record indicated no SMTP service".
3200 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3201 if the DNS of the sending site is misconfigured. This is quite a
3202 common situation. This change restores the behaviour of treating a
3203 helo verification defer as a failure.
3205 PH/51 If self=fail was set on a router, the bounce message did not include the
3206 actual error message.
3212 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3214 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3215 command itself ended in error, the underlying message about the transport
3216 filter timeout was being overwritten with the pipe command error. Now the
3217 underlying error message should be appended to the second error message.
3219 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3221 PH/02 Reduce the amount of output that "make" produces by default. Full output
3222 can still be requested.
3224 PH/03 The warning log line about a condition test deferring for a "warn" verb
3225 was being output only once per connection, rather than after each
3226 occurrence (because it was using the same function as for successful
3227 "warn" verbs). This seems wrong, so I have changed it.
3229 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3230 it should not have, which might have caused a crash in the right
3231 circumstances, but probably never did.
3233 PH/04 Installed a modified version of Tony Finch's patch to make submission
3234 mode fix the return path as well as the Sender: header line, and to
3235 add a /name= option so that you can make the user's friendly name appear
3238 TF/03 Added the control = fakedefer ACL modifier.
3240 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3241 Mark Lowes for thorough testing.
3243 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3245 TK/03 Merged latest SRS patch from Miles Wilton.
3247 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3248 with the definition in sysexits.h (which is #included earlier).
3249 Fortunately, Exim does not actually use EX_OK. The code used to try to
3250 preserve the sysexits.h value, by assumimg that macro definitions were
3251 scanned for macro replacements. I have been disabused of this notion,
3252 so now the code just undefines EX_OK before #including unistd.h.
3254 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3255 in the smtp transport. When a block could not be written in a single
3256 write() function, the timeout was being re-applied to each part-write.
3257 This seems wrong - if the receiver was accepting one byte at a time it
3258 would take for ever. The timeout is now adjusted when this happens. It
3259 doesn't have to be particularly precise.
3261 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3262 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3264 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3265 without a domain if the check to <postmaster@domain> fails.
3267 SC/01 Eximstats: added -xls and the ability to specify output files
3268 (patch written by Frank Heydlauf).
3270 SC/02 Eximstats: use FileHandles for outputing results.
3272 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3274 SC/04 Eximstats: fixed display of large numbers with -nvr option
3276 SC/05 Eximstats: fixed merging of reports with empty tables.
3278 SC/06 Eximstats: added the -include_original_destination flag
3280 SC/07 Eximstats: removed tabs and trailing whitespace.
3282 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3284 TK/06 MBOX spool code: Add real "From " MBOX separator line
3285 so the .eml file is really in mbox format (even though
3286 most programs do not really care). Patch from Alex Miller.
3288 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3289 The latter is generated from $received_to and is only set if the
3290 message has one envelope recipient. SA can use these headers,
3291 obviously out-of-the-box. Patch from Alex Miller.
3293 PH/08 The ${def test on a variable was returning false if the variable's
3294 value was "0", contrary to what the specification has always said!
3295 The result should be true unless the variable is empty.
3297 PH/09 The syntax error of a character other than { following "${if
3298 def:variable_name" (after optional whitespace) was not being diagnosed.
3299 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3300 accidental colon was present, for example, could give incorrect results.
3302 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3303 result is used (not including appendfile, where other changes are about
3306 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3307 This involved changing a lot of size variables from int to off_t. It
3308 should work with maildirs and everything.
3310 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3311 spamd dying while we are connected to it.
3313 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3316 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3317 function for BDB 4.3.
3319 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3321 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3322 the sources for PCRE have been re-arranged and more files are now
3325 PH/15 The code I had for printing potentially long long variables in PH/11
3326 above was not the best (it lost precision). The length of off_t variables
3327 is now inspected at build time, and an appropriate printing format (%ld
3328 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3329 to be "long long int" or "long int". This is needed for the internal
3330 formatting function string_vformat().
3332 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3333 the configuration file to be ":syslog", then the script "guesses" where
3334 the logs files are, rather than using the compiled in default. In our
3335 case the guess is not the same as the compiled default, so the script
3336 suddenly stopped working when I started to use syslog. The patch checks
3337 to see if log_file_path is "". If so, it attempts to read it from exim
3338 with no configuration file to get the compiled in version, before it
3339 falls back to the previous guessing code."
3341 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3342 implementing BATV in an Exim configuration. See NewStuff for the gory
3345 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3346 Makefile that are specific to HP-UX.
3348 PH/18 If the "use_postmaster" option was set for a recipient callout together
3349 with the "random" option, the postmaster address was used as the MAIL
3350 FROM address for the random test, but not for the subsequent recipient
3351 test. It is now used for both.
3353 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3354 patch removes a few documentation additions to RFC 3028, because the
3355 latest draft now contains them. It adds the new en;ascii-case comparator
3356 and a new error check for 8bit text in MIME parts. Comparator and
3357 require names are now matched exactly. I enabled the subaddress
3358 extension, but it is not well tested yet (read: it works for me)."
3360 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3361 rework some of the code of TK/09 above to avoid the hardwired use of
3362 "%lld" and "long long". Replaced the call to snprintf() with a call to
3365 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3366 records point to non-existent hosts", "retry timeout exceeded", and
3367 "retry time not reached for any host after a long failure period".
3369 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3370 experimental DomainKeys support:
3372 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3373 (2) On an error such as an illegally used "control", the wrong name for
3374 the control was given.
3376 These problems did NOT occur unless DomainKeys support was compiled.
3378 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3380 PH/24 Added ${if match_ip condition.
3382 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3383 databases so that it will be absolutely obvious if a crash occurs in the
3384 DB library. This is a regular occurrence (often caused by mis-matched
3387 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3388 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3389 picked up on a user's system that detects such things. There doesn't seem
3390 to be a gcc warning option for this - only an attribute that has to be
3391 put on the function's prototype. It seems that in Fedora Core 4 they have
3392 set this on a number of new functions. No doubt there will be more in due
3395 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3396 specify a transport. However, if an address that was verified by such a
3397 router was the subject of a callout, Exim crashed because it tried to
3398 read the rcpt_include_affixes from the non-existent transport. Now it
3399 just assumes that the setting of that option is false. This bug was
3400 introduced by 4.51/PH/31.
3402 PH/28 Changed -d+all to exclude +memory, because that information is very
3403 rarely of interest, but it makes the output a lot bigger. People tend to
3404 do -d+all out of habit.
3406 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3407 code in os-type was giving problems when libc.so lives in lib64, like on
3410 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3411 aren't the modern standard, and it seems that some systems' include files
3412 don't always have them. Exim was already checking for some of the newer
3413 ones like T_AAAA, and defining it itself. I've added checks for all the
3414 record types that Exim uses.
3416 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3417 not automatically generating a new one, as it is supposed to. This
3418 prevented TLS from working. If the file did exist, but contained invalid
3419 data, a new version was generated, as expected. It was only the case of a
3420 non-existent file that was broken.
3422 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3423 with a change in libdomainkeys > 0.64.
3425 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3426 from DNS. If the selector record carries the flag, it now has
3427 precedence over the domain-wide flag.
3429 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3431 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3432 the use of a port name that isn't defined in /etc/services) occurred, the
3433 message was deferred as in a normal delivery, and thus remained on the
3434 spool, instead of being failed because of the mua_wrapper setting. This
3435 is now fixed, and I tidied up some of the mua_wrapper messages at the
3438 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3439 the messages in a hash of arrays rather than using individual hashes.
3440 This is a bit cleaner and results in dramatic memory savings, albeit
3441 at a slight CPU cost.
3443 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3444 as requested by Marc Sherman.
3446 SC/10 Eximstats: added histograms for user specified patterns as requested
3449 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3451 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3452 fopen() in the content-scanning modules that did not already have it.
3458 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3459 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3461 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3463 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3465 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3466 to test Sieve filters that use "vacation".
3468 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3469 that changes the way the GnuTLS parameters are stored in the cache file.
3470 The new format can be generated externally. For backward compatibility,
3471 if the data in the cache doesn't make sense, Exim assumes it has read an
3472 old-format file, and it generates new data and writes a new file. This
3473 means that you can't go back to an older release without removing the
3476 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3477 work if there are any delivery delays because "one_time" forces the
3478 parent to be marked "delivered", so its unseen clone is never tried
3479 again. For this reason, Exim now forbids the simultaneous setting of
3482 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3483 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3484 entry if you want to know the details. The fix, however, neglected to
3485 consider the case when local delivery batching is involved. The test for
3486 "previously delivered" was not happening when checking to see if an
3487 address could be batched with a previous (undelivered) one; under
3488 certain circumstances this could lead to multiple deliveries to the same
3491 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3492 in its include files, and this causes problems building Exim.
3494 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3495 header_syntax) but Exim was just ignoring anything given after a slash.
3496 In particular, this caused confusion with an attempt to use "verify =
3497 reverse_host_lookup/defer_ok". An error is now given when options are
3498 supplied for verify items that do not have them. (Maybe reverse_host_
3499 lookup should have a defer_ok option, but that's a different point.)
3501 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3502 defined by RFC 821) to 2048, because there were problems with some AUTH
3503 commands, and RFC 1869 says the size should be increased for extended
3504 SMTP commands that take arguments.
3506 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3509 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3510 "unknown" error; now it says that the functionality isn't in the binary.
3512 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3513 an address' error message when a string expansion fails (syntax or
3514 whatever). Otherwise the password may appear in the log. Following change
3515 PH/42 below, there is no longer a chance of it appearing in a bounce
3518 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3520 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3521 of its addresses was ever tried. (Bugzilla bug #2.)
3523 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3524 the result incorrectly in the debug output. (It correctly added a newline
3525 to what was transported.)
3527 TF/01 Added $received_time.
3529 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3530 commented out examples of how to interface to a virus scanner and to
3531 SpamAssassin. Also added commented examples of av_scanner and
3532 spamd_address settings.
3534 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3535 and controls are allowed in which ACLs. There were a couple of minor
3536 errors. Some of the entries in the conditions table (which is a table of
3537 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3538 negation of where the condition IS allowed.
3540 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3542 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3543 header file does not have a version number, so I've had to invent a new
3544 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3545 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3546 radiusclient), but it was contributed by a Radius user.
3548 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3549 files or directories, for OpenSSL.
3551 PH/19 When an Exim process that is running as root has to create an Exim log
3552 file, it does so in a subprocess that runs as exim:exim so as to get the
3553 ownership right at creation (otherwise, other Exim processes might see
3554 the file with the wrong ownership). There was no test for failure of this
3555 fork() call, which would lead to the process getting stuck as it waited
3556 for a non-existent subprocess. Forks do occasionally fail when resources
3557 run out. I reviewed all the other calls to fork(); they all seem to check
3560 PH/20 When checking for unexpected SMTP input at connect time (before writing
3561 the banner), Exim was not dealing correctly with a non-positive return
3562 from the read() function. If the client had disconnected by this time,
3563 the result was a log entry for a synchronization error with an empty
3564 string after "input=" when read() returned zero. If read() returned -1
3565 (an event I could not check), uninitialized data bytes were printed.
3566 There were reports of junk text (parts of files, etc) appearing after
3569 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3571 PH/22 Added support for macro redefinition, and (re)definition in between
3572 driver and ACL definitions.
3574 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3575 forgetting to use the resulting value; it was using the unexpanded value.
3577 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3578 hadn't been configured. The fix is from Juergen Kreileder, who
3579 understands it better than I do:
3581 "Here's what I see happening with three configured cyrus_sasl
3582 authenticators configured (plain, login, cram-md5):
3584 On startup auth_cyrus_sasl_init() gets called for each of these.
3585 This means three calls to sasl_listmech() without a specified mech_list.
3586 => SASL tests which mechs of all available mechs actually work
3587 => three warnings about OTP not working
3588 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3590 With the patch, sasl_listmech() also gets called three times. But now
3591 SASL's mech_list option is set to the server_mech specified in the the
3592 authenticator. Or in other words, the answer from sasl_listmech()
3593 gets limited to just the mech you're testing for (which is different
3595 => the return list contains just 'plain' or 'login', 'cram-md5' or
3596 nothing depending on the value of ob->server_mech.
3598 I've just tested the patch: Authentication still works fine,
3599 unavailable mechs specified in the exim configuration are still
3600 caught, and the auth.log warnings about OTP are gone."
3602 PH/25 When debugging is enabled, the contents of the command line are added
3603 to the debugging output, even when log_selector=+arguments is not
3606 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3607 answer is "GNU", and only if the return is "GNU/something" is the answer
3610 PH/27 $acl_verify_message is now set immediately after the failure of a
3611 verification in an ACL, and so is available in subsequent modifiers. In
3612 particular, the message can be preserved by coding like this:
3614 warn !verify = sender
3615 set acl_m0 = $acl_verify_message
3617 Previously, $acl_verify_message was set only while expanding "message"
3618 and "log_message" when a very denied access.
3620 PH/28 Modified OS/os.c-Linux with
3622 -#ifndef OS_LOAD_AVERAGE
3623 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3625 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3626 nomenclature these days.)
3628 PH/29 Installed patch from the Sieve maintainer that adds the options
3629 sieve_useraddress and sieve_subaddress to the redirect router.
3631 PH/30 In these circumstances:
3632 . Two addresses routed to the same list of hosts;
3633 . First host does not offer TLS;
3634 . First host accepts first address;
3635 . First host gives temporary error to second address;
3636 . Second host offers TLS and a TLS session is established;
3637 . Second host accepts second address.
3638 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3639 and peerdn, if requested) that were in fact used only for the second
3642 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3643 attention to any local part prefix or suffix that was matched by the
3644 router that accepted the address. It now behaves in the same way as it
3645 does for delivery: the affixes are removed from the local part unless
3646 rcpt_include_affixes is set on the transport.
3648 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3649 timeout during the DATA phase of an incoming message.
3651 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3652 applied a patch sanctioned by the Sieve maintainer.
3654 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3655 the uid or gid is negative. A case of a negative gid caused this to be
3656 noticed. The fix allows for either to be negative.
3658 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3659 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3662 PH/36 The change PH/12 above was broken. Fixed it.
3664 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3665 the grounds that routing the same address twice would always produce the
3666 same answer. This might have been true once, but it is certainly no
3667 longer true now. Routing a child address may depend on the previous
3668 routing that produced that child. Some complicated redirection strategies
3669 went wrong when messages had multiple recipients, and made Exim's
3670 behaviour dependent on the order in which the addresses were given.
3672 I have moved the duplicate checking until after the routing is complete.
3673 Exim scans the addresses that are assigned to local and remote
3674 transports, and removes any duplicates. This means that more work will be
3675 done, as duplicates will always all be routed, but duplicates are
3676 presumably rare, so I don't expect this is of any significance.
3678 For deliveries to pipes, files, and autoreplies, the duplicate checking
3679 still happens during the routing process, since they are not going to be
3682 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3683 It corrects a timeout issue with spamd. This is Ian's comment: "The
3684 background is that sometimes spamd either never reads data from a
3685 connection it has accepted, or it never writes response data. The exiscan
3686 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3687 blindly assumes that writes won't block so it may never time out."
3689 PH/39 Allow G after quota size as well as K and M.
3691 PH/40 The value set for $authenticated_id in an authenticator may not contain
3692 binary zeroes or newlines because the value is written to log lines and
3693 to spool files. There was no check on this. Now the value is run through
3694 the string_printing() function so that such characters are converted to
3695 printable escape sequences.
3697 PH/41 $message_linecount is a new variable that contains the total number of
3698 lines in the message. Compare $body_linecount, which is the count for the
3701 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3702 bounce and delay warning messages, except in certain special cases, which
3705 (a) An SMTP error message from a remote host;
3706 (b) A message specified in a :fail: redirection;
3707 (c) A message specified in a "fail" command in a system filter;
3708 (d) A message specified in a FAIL return from the queryprogram router;
3709 (e) A message specified by the cannot_route_message router option.
3711 In these cases only, Exim does include the error details in bounce and
3712 warning messages. There are also a few cases where bland messages such
3713 as "unrouteable address" or "local delivery error" are given.
3715 PH/43 $value is now also set for the "else" part of a ${run expansion.
3717 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3718 being worked on, but at least Exim now implements the latest version to
3721 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3722 process to complete was treated as a delivery failure, a timeout while
3723 writing the message to the pipe was logged, but erroneously treated as a
3724 successful delivery. Such timeouts include transport filter timeouts. For
3725 consistency with the overall process timeout, these timeouts are now
3726 treated as errors, giving rise to delivery failures by default. However,
3727 there is now a new Boolean option for the pipe transport called
3728 timeout_defer, which, if set TRUE, converts the failures into defers for
3729 both kinds of timeout. A transport filter timeout is now identified in
3732 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3733 systems where "make" and "gmake" are different, calling "gmake" at top
3734 level broke things. I've arranged for the value of $(MAKE) to be passed
3735 from the Makefile to this script so that it can call the same version of
3739 A note about Exim versions 4.44 and 4.50
3740 ----------------------------------------
3742 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3743 changes of various kinds. As a consequence, a big documentation update was
3744 needed. This delayed the release for rather longer than seemed good, especially
3745 in the light of a couple of (minor) security issues. Therefore, the changes
3746 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3747 release. So 4.44 and 4.50 are in effect two different branches that both start
3750 I have left the 4.50 change log unchanged; it contains all the changes since
3751 4.43. The change log for 4.44 is below; many of its items are identical to
3752 those for 4.50. This seems to be the most sensible way to preserve the
3753 historical information.
3759 1. Minor wording change to the doc/README.SIEVE file.
3761 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3762 computation of the current number of files was incorrect.
3764 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3765 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3768 4. Give more explanation in the error message when the command for a transport
3769 filter fails to execute.
3771 5. There are several places where Exim runs a non-Exim command in a
3772 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3773 was being done only for the command run by the queryprogram router. It is
3774 now done for all such subprocesses. The other cases are: ${run, transport
3775 filters, and the commands run by the lmtp and pipe transports.
3777 6. Added CONFIGURE_GROUP build-time option.
3779 7. Some older OS have a limit of 256 on the maximum number of file
3780 descriptors. Exim was using setrlimit() to set 1000 as a large value
3781 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3782 systems. I've change it so that if it can't get 1000, it tries for 256.
3784 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3785 was an oversight, and furthermore, ever since the addition of extra
3786 controls (e.g. 4.43/32), the checks on when to allow different forms of
3787 "control" were broken. There should now be diagnostics for all cases when a
3788 control that does not make sense is encountered.
3790 9. Added the /retain_sender option to "control=submission".
3792 10. $recipients is now available in the predata ACL (oversight).
3794 11. Tidy the search cache before the fork to do a delivery from a message
3795 received from the command line. Otherwise the child will trigger a lookup
3796 failure and thereby defer the delivery if it tries to use (for example) a
3797 cached ldap connection that the parent has called unbind on.
3799 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3800 of $address_data from the recipient verification was clobbered by the
3801 sender verification.
3803 13. The value of address_data from a sender verification is now available in
3804 $sender_address_data in subsequent conditions in the ACL statement.
3806 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3808 15. Added a new option "connect=<time>" to callout options, to set a different
3811 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3812 was its contents. (It was OK if the option was not defined at all.)
3814 17. A "Completed" log line is now written for messages that are removed from
3815 the spool by the -Mrm option.
3817 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3818 information about exactly what failed.
3820 19. Added -dd to debug only the daemon process.
3822 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3823 handles timeouts, both on the server side and network timeouts. Renamed the
3824 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3826 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3827 It is now set to "smtps".
3829 22. $host_address is now set to the target address during the checking of
3830 ignore_target_hosts.
3832 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3833 being passed; this would have caused $sender_host_name to have been used if
3834 matching the list had actually called for a host name (not very likely,
3835 since this list is usually IP addresses). A host name is now passed as
3838 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3839 code that specifies a non-restarting handler (typically sigaction() in
3840 modern systems) in an attempt to fix a rare and obscure crash bug.
3842 25. Narrowed the window for a race in the daemon that could cause it to ignore
3843 SIGCHLD signals. This is not a major problem, because they are used only to
3844 wake it up if nothing else does.
3846 26. A malformed maildirsize file could cause Exim to calculate negative values
3847 for the mailbox size or file count. Odd effects could occur as a result.
3848 The maildirsize information is now recalculated if the size or filecount
3851 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3852 support for a long time. Removed HAVE_SYS_VFS_H.
3854 28. Installed the latest version of exipick from John Jetmore.
3856 29. In an address list, if the pattern was not a regular expression, an empty
3857 subject address (from a bounce message) matched only if the pattern was an
3858 empty string. Non-empty patterns were not even tested. This was the wrong
3859 because it is perfectly reasonable to use an empty address as part of a
3860 database query. An empty address is now tested by patterns that are
3861 lookups. However, all the other forms of pattern expect the subject to
3862 contain a local part and a domain, and therefore, for them, an empty
3863 address still always fails if the pattern is not itself empty.
3865 30. Exim went into a mad DNS loop when attempting to do a callout where the
3866 host was specified on an smtp transport, and looking it up yielded more
3867 than one IP address.
3869 31. Re-factored the code for checking spool and log partition space into a
3870 function that finds that data and another that does the check. The former
3871 is then used to implement four new variables: $spool_space, $log_space,
3872 $spool_inodes, and $log_inodes.
3874 32. The RFC2047 encoding function was originally intended for short strings
3875 such as real names; it was not keeping to the 75-character limit for
3876 encoded words that the RFC imposes. It now respects the limit, and
3877 generates multiple encoded words if necessary. To be on the safe side, I
3878 have increased the buffer size for the ${rfc2047: expansion operator from
3881 33. It is now permitted to omit both strings after an "if" condition; if the
3882 condition is true, the result is "true". As before, when the second string
3883 is omitted, a false condition yields an empty string. This makes it less
3884 cumbersome to write custom ACL and router conditions.
3886 34. Failure to deliver a bounce message always caused it to be frozen, even if
3887 there was an errors_to setting on the router. The errors_to setting is now
3890 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3891 canonical form (fully expanded) before being placed in
3892 $sender_host_address.
3894 36. The table in the code that translates DNS record types into text (T_A to
3895 "A" for instance) was missing entries for NS and CNAME. It is just possible
3896 that this could have caused confusion if both these types were looked up
3897 for the same domain, because the text type is used as part of Exim's
3898 per-process caching. But the chance of anyone hitting this buglet seems
3901 37. The dnsdb lookup has been extended in a number of ways.
3903 (1) There is a new type, "zns", which walks up the domain tree until it
3904 finds some nameserver records. It should be used with care.
3906 (2) There is a new type, "mxh", which is like "mx" except that it returns
3907 just the host names, not the priorities.
3909 (3) It is now possible to give a list of domains (or IP addresses) to be
3910 looked up. The behaviour when one of the lookups defers can be
3911 controlled by a keyword.
3913 (4) It is now possible to specify the separator character for use when
3914 multiple records are returned.
3916 38. The dnslists ACL condition has been extended: it is now possible to supply
3917 a list of IP addresses and/or domains to be looked up in a particular DNS
3920 39. Added log_selector=+queue_time_overall.
3922 40. When running the queue in the test harness, wait just a tad after forking a
3923 delivery process, to get repeatability of debugging output.
3925 41. Include certificate and key file names in error message when GnuTLS fails
3926 to set them up, because the GnuTLS error message doesn't include the name
3927 of the failing file when there is a problem reading it.
3929 42. Allow both -bf and -bF in the same test run.
3931 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3933 44. The "Exiscan patch" is now merged into the mainline Exim source.
3935 45. Sometimes the final signoff response after QUIT could fail to get
3936 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3937 < 0 before doing a fflush(). This bug looks as though it goes back to the
3938 introduction of TLS in release 3.20, but "sometimes" must have been rare
3939 because the tests only now provoked it.
3941 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3942 (this can affect the format of dates).
3944 47. exim_tidydb, when checking for the continued existence of a message for
3945 which it has found a message-specific retry record, was not finding
3946 messages that were in split spool directories. Consequently, it was
3947 deleting retry records that should have stayed in existence.
3949 48. Steve fixed some bugs in eximstats.
3951 49. The SPA authentication driver was not abandoning authentication and moving
3952 on to the next authenticator when an expansion was forced to fail,
3953 contradicting the general specification for all authenticators. Instead it
3954 was generating a temporary error. It now behaves as specified.
3956 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3957 (the order specifies the preference for clients). The order is now AES256,
3958 AES128, 3DES, ARCFOUR128.
3960 51. Small patch to Sieve code - explicitly set From: when generating an
3963 52. Exim crashed if a remote delivery caused a very long error message to be
3964 recorded - for instance if somebody sent an entire SpamAssassin report back
3965 as a large number of 550 error lines. This bug was coincidentally fixed by
3966 increasing the size of one of Exim's internal buffers (big_buffer) that
3967 happened as part of the Exiscan merge. However, to be on the safe side, I
3968 have made the code more robust (and fixed the comments that describe what
3971 53. Now that there can be additional text after "Completed" in log lines (if
3972 the queue_time_overall log selector is set), a one-byte patch to exigrep
3973 was needed to allow it to recognize "Completed" as not the last thing in
3976 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3977 patch that reportedly fixes this has been added. I am not expert enough to
3978 create a test for it. This is what the patch creator wrote:
3980 "I found a little strange behaviour of ldap code when working with
3981 Windows 2003 AD Domain, where users was placed in more than one
3982 Organization Units. When I tried to give exim partial DN, the exit code
3983 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3984 But simultaneously result of request was absolutely normal ldap result,
3985 so I produce this patch..."
3987 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3988 REFERENCE, so I have modified the code to exclude the patch when that macro
3991 55. Some experimental protocols are using DNS PTR records for new purposes. The
3992 keys for these records are domain names, not reversed IP addresses. The
3993 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3994 leaves it alone. Component reversal etc. now happens only for IP addresses.
3997 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3999 57. Double the size of the debug message buffer (to 2048) so that more of very
4000 long debug lines gets shown.
4002 58. The exicyclog utility now does better if the number of log files to keep
4003 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4005 59. Two changes related to the smtp_active_hostname option:
4007 (1) $smtp_active_hostname is now available as a variable.
4008 (2) The default for smtp_banner uses $smtp_active_hostname instead
4009 of $primary_hostname.
4011 60. The host_aton() function is supposed to be passed a string that is known
4012 to be a valid IP address. However, in the case of IPv6 addresses, it was
4013 not checking this. This is a hostage to fortune. Exim now panics and dies
4014 if the condition is not met. A case was found where this could be provoked
4015 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4016 components; fortuitously, this particular loophole had already been fixed
4017 by change 4.50/55 above.
4019 If there are any other similar loopholes, the new check in host_aton()
4020 itself should stop them being exploited. The report I received stated that
4021 data on the command line could provoke the exploit when Exim was running as
4022 exim, but did not say which command line option was involved. All I could
4023 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4024 running as the user.
4027 61. There was a buffer overflow vulnerability in the SPA authentication code
4028 (which came originally from the Samba project). I have added a test to the
4029 spa_base64_to_bits() function which I hope fixes it.
4032 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4033 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4035 63. The daemon start-up calls getloadavg() while still root for those OS that
4036 need the first call to be done as root, but it missed one case: when
4037 deliver_queue_load_max is set with deliver_drop_privilege. This is
4038 necessary for the benefit of the queue runner, because there is no re-exec
4039 when deliver_drop_privilege is set.
4041 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4042 This has been fixed.
4044 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4045 was in use, was not putting the data itself into the right store pool;
4046 consequently, it could be overwritten for a subsequent message in the same
4047 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4050 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4052 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4053 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4054 the function were treating the return as a boolean value, which happened to
4055 work because 0=false and not-0=true, but is not correct code.
4057 68. The host_aton() function was not handling scoped IPv6 addresses (those
4058 with, for example, "%eth0" on the end) correctly.
4060 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4061 negated items (that is, ~something) in unsigned ints. Some compilers
4062 apparently mutter when there is no cast.
4064 70. If an address verification called from an ACL failed, and did not produce a
4065 user-specific message (i.e. there was only a "system" message), nothing was
4066 put in $acl_verify_message. In this situation, it now puts the system
4069 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4070 session; change 4.31/43 added the unwanted input to the log line - except
4071 that it did not do this in the start of session case. It now does.
4073 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4074 This is wrong and can cause the other end to generate a synchronization
4075 error if it is another Exim or anything else that does the synchronization
4076 check. A QUIT command is no longer sent after a timeout.
4078 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4079 during host lookups.
4081 74. The defer_ok option of callout verification was not working if it was used
4082 when verifying addresses in header lines, that is, for this case:
4084 verify = header_sender/callout=defer_ok
4086 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4087 those file descriptors could be used for SMTP connections. If anything
4088 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4089 could be sent to the SMTP client, causing chaos. The daemon now opens
4090 stdin, stdout, and stderr to /dev/null when it puts itself into the
4093 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4094 log by default. The user can override this with suitable Perl magic.
4096 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4097 the log message when discard triggers, was not working for the DATA ACL or
4098 for the non-SMTP ACL.
4100 78. Error message wording change in sieve.c.
4102 79. If smtp_accept_max_per_host was set, the number of connections could be
4103 restricted to fewer than expected, because the daemon was trying to set up
4104 a new connection before checking whether the processes handling previous
4105 connections had finished. The check for completed processes is now done
4106 earlier. On busy systems, this bug wouldn't be noticed because something
4107 else would have woken the daemon, and it would have reaped the completed
4110 80. If a message was submitted locally by a user whose login name contained one
4111 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4112 It caused a spool format error. I have fixed the spool reading code. A
4113 related problem was that the "from" clause in the Received: line became
4114 illegal because of the space(s). It is now covered by ${quote_local_part.
4116 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4119 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4121 83. Patch from Sieve maintainer to fix unterminated string problem in
4122 "vacation" handling.
4124 84. Some minor changes to the Linux configuration files to help with other
4125 OS variants using glibc.
4127 85. One more patch for Sieve to update vacation handling to latest spec.
4130 ----------------------------------------------------
4131 See the note above about the 4.44 and 4.50 releases.
4132 ----------------------------------------------------
4138 1. Change 4.43/35 introduced a bug that caused file counts to be
4139 incorrectly computed when quota_filecount was set in an appendfile
4142 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4143 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4146 3. Give more explanation in the error message when the command for a transport
4147 filter fails to execute.
4149 4. There are several places where Exim runs a non-Exim command in a
4150 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4151 was being done only for the command run by the queryprogram router. It is
4152 now done for all such subprocesses. The other cases are: ${run, transport
4153 filters, and the commands run by the lmtp and pipe transports.
4155 5. Some older OS have a limit of 256 on the maximum number of file
4156 descriptors. Exim was using setrlimit() to set 1000 as a large value
4157 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4158 systems. I've change it so that if it can't get 1000, it tries for 256.
4160 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4161 was an oversight, and furthermore, ever since the addition of extra
4162 controls (e.g. 4.43/32), the checks on when to allow different forms of
4163 "control" were broken. There should now be diagnostics for all cases when a
4164 control that does not make sense is encountered.
4166 7. $recipients is now available in the predata ACL (oversight).
4168 8. Tidy the search cache before the fork to do a delivery from a message
4169 received from the command line. Otherwise the child will trigger a lookup
4170 failure and thereby defer the delivery if it tries to use (for example) a
4171 cached ldap connection that the parent has called unbind on.
4173 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4174 of $address_data from the recipient verification was clobbered by the
4175 sender verification.
4177 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4178 was its contents. (It was OK if the option was not defined at all.)
4180 11. A "Completed" log line is now written for messages that are removed from
4181 the spool by the -Mrm option.
4183 12. $host_address is now set to the target address during the checking of
4184 ignore_target_hosts.
4186 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4187 being passed; this would have caused $sender_host_name to have been used if
4188 matching the list had actually called for a host name (not very likely,
4189 since this list is usually IP addresses). A host name is now passed as
4192 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4193 code that specifies a non-restarting handler (typically sigaction() in
4194 modern systems) in an attempt to fix a rare and obscure crash bug.
4196 15. Narrowed the window for a race in the daemon that could cause it to ignore
4197 SIGCHLD signals. This is not a major problem, because they are used only to
4198 wake it up if nothing else does.
4200 16. A malformed maildirsize file could cause Exim to calculate negative values
4201 for the mailbox size or file count. Odd effects could occur as a result.
4202 The maildirsize information is now recalculated if the size or filecount
4205 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4206 support for a long time. Removed HAVE_SYS_VFS_H.
4208 18. Updated exipick to current release from John Jetmore.
4210 19. Allow an empty sender to be matched against a lookup in an address list.
4211 Previously the only cases considered were a regular expression, or an
4214 20. Exim went into a mad DNS lookup loop when doing a callout where the
4215 host was specified on the transport, if the DNS lookup yielded more than
4218 21. The RFC2047 encoding function was originally intended for short strings
4219 such as real names; it was not keeping to the 75-character limit for
4220 encoded words that the RFC imposes. It now respects the limit, and
4221 generates multiple encoded words if necessary. To be on the safe side, I
4222 have increased the buffer size for the ${rfc2047: expansion operator from
4225 22. Failure to deliver a bounce message always caused it to be frozen, even if
4226 there was an errors_to setting on the router. The errors_to setting is now
4229 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4230 canonical form (fully expanded) before being placed in
4231 $sender_host_address.
4233 24. Updated eximstats to version 1.33
4235 25. Include certificate and key file names in error message when GnuTLS fails
4236 to set them up, because the GnuTLS error message doesn't include the name
4237 of the failing file when there is a problem reading it.
4239 26. Expand error message when OpenSSL has problems setting up cert/key files.
4242 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4243 (this can affect the format of dates).
4245 28. exim_tidydb, when checking for the continued existence of a message for
4246 which it has found a message-specific retry record, was not finding
4247 messages that were in split spool directories. Consequently, it was
4248 deleting retry records that should have stayed in existence.
4250 29. eximstats updated to version 1.35
4251 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4252 1.35 - bugfix such that pie charts by volume are generated correctly
4254 30. The SPA authentication driver was not abandoning authentication and moving
4255 on to the next authenticator when an expansion was forced to fail,
4256 contradicting the general specification for all authenticators. Instead it
4257 was generating a temporary error. It now behaves as specified.
4259 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4260 (the order specifies the preference for clients). The order is now AES256,
4261 AES128, 3DES, ARCFOUR128.
4263 31. Small patch to Sieve code - explicitly set From: when generating an
4266 32. Exim crashed if a remote delivery caused a very long error message to be
4267 recorded - for instance if somebody sent an entire SpamAssassin report back
4268 as a large number of 550 error lines. This bug was coincidentally fixed by
4269 increasing the size of one of Exim's internal buffers (big_buffer) that
4270 happened as part of the Exiscan merge. However, to be on the safe side, I
4271 have made the code more robust (and fixed the comments that describe what
4274 33. Some experimental protocols are using DNS PTR records for new purposes. The
4275 keys for these records are domain names, not reversed IP addresses. The
4276 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4277 leaves it alone. Component reversal etc. now happens only for IP addresses.
4280 34. The host_aton() function is supposed to be passed a string that is known
4281 to be a valid IP address. However, in the case of IPv6 addresses, it was
4282 not checking this. This is a hostage to fortune. Exim now panics and dies
4283 if the condition is not met. A case was found where this could be provoked
4284 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4285 components; fortuitously, this particular loophole had already been fixed
4286 by change 4.50/55 or 4.44/33 above.
4288 If there are any other similar loopholes, the new check in host_aton()
4289 itself should stop them being exploited. The report I received stated that
4290 data on the command line could provoke the exploit when Exim was running as
4291 exim, but did not say which command line option was involved. All I could
4292 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4293 running as the user.
4296 35. There was a buffer overflow vulnerability in the SPA authentication code
4297 (which came originally from the Samba project). I have added a test to the
4298 spa_base64_to_bits() function which I hope fixes it.
4301 36. The daemon start-up calls getloadavg() while still root for those OS that
4302 need the first call to be done as root, but it missed one case: when
4303 deliver_queue_load_max is set with deliver_drop_privilege. This is
4304 necessary for the benefit of the queue runner, because there is no re-exec
4305 when deliver_drop_privilege is set.
4307 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4308 was in use, was not putting the data itself into the right store pool;
4309 consequently, it could be overwritten for a subsequent message in the same
4310 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4313 38. Sometimes the final signoff response after QUIT could fail to get
4314 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4315 < 0 before doing a fflush(). This bug looks as though it goes back to the
4316 introduction of TLS in release 3.20, but "sometimes" must have been rare
4317 because the tests only now provoked it.
4323 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4324 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4325 Now it returns an integer. A number of calls were still expecting a T/F
4326 return. Fortuitously, in all cases, the tests worked in OK situations,
4327 which is the norm. However, things would have gone wrong on any write
4328 failures on the smtp file descriptor. This function is used when sending
4329 messages over SMTP and also when doing verify callouts.
4331 2. When Exim is called to do synchronous delivery of a locally submitted
4332 message (the -odf or -odi options), it no longer closes stderr before doing
4335 3. Implemented the mua_wrapper option.
4337 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4339 5. Implemented the functions header_remove(), header_testname(),
4340 header_add_at_position(), and receive_remove_recipient(), and exported them
4343 6. If an ACL "warn" statement specified the addition of headers, Exim already
4344 inserted X-ACL-Warn: at the start if there was no header name. However, it
4345 was not making this test for the second and subsequent header lines if
4346 there were newlines in the string. This meant that an invalid header could
4347 be inserted if Exim was badly configured.
4349 7. Allow an ACL "warn" statement to add header lines at the start or after all
4350 the Received: headers, as well as at the end.
4352 8. Added the rcpt_4xx retry error code.
4354 9. Added postmaster_mailfrom=xxx to callout verification option.
4356 10. Added mailfrom=xxxx to the callout verification option, for verify=
4359 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4360 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4362 12. Inserted some casts to stop certain compilers warning when using pointer
4363 differences as field lengths or precisions in printf-type calls (mostly
4364 affecting debugging statements).
4366 13. Added optional readline() support for -be (dynamically loaded).
4368 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4369 same clock tick as a message's arrival, so that its received time was the
4370 same as the "first fail" time on the retry record, and that message
4371 remained on the queue past the ultimate address timeout, every queue runner
4372 would try a delivery (because it was past the ultimate address timeout) but
4373 after another failure, the ultimate address timeout, which should have then
4374 bounced the address, did not kick in. This was a "< instead of <=" error;
4375 in most cases the first failure would have been in the next clock tick
4376 after the received time, and all would be well.
4378 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4379 being recognized when the domain list was tested by the match_domain
4380 condition in an expansion string.
4382 16. Added the ${str2b64: operator.
4384 17. Exim was always calling setrlimit() to set a large limit for the number of
4385 processes, without checking whether the existing limit was already
4386 adequate. (It did check for the limit on file descriptors.) Furthermore,
4387 errors from getrlimit() and setrlimit() were being ignored. Now they are
4388 logged to the main and panic logs, but Exim does carry on, to try to do its
4389 job under whatever limits there are.
4391 18. Imported PCRE 5.0.
4393 19. Trivial typo in log message " temporarily refused connection" (the leading
4396 20. If the log selector return_path_on_delivery was set and an address was
4397 redirected to /dev/null, the delivery process crashed because it assumed
4398 that a return path would always be set for a "successful" delivery. In this
4399 case, the whole delivery is bypassed as an optimization, and therefore no
4402 21. Internal re-arrangement: the function for sending a challenge and reading
4403 a response while authentication was assuming a zero-terminated challenge
4404 string. It's now changed to take a pointer and a length, to allow for
4405 binary data in such strings.
4407 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4409 23. Exim was not respecting finduser_retries when seeking the login of the
4410 uid under which it was called; it was always trying 10 times. (The default
4411 setting of finduser_retries is zero.) Also, it was sleeping after the final
4412 failure, which is pointless.
4414 24. Implemented tls_on_connect_ports.
4416 25. Implemented acl_smtp_predata.
4418 26. If the domain in control=submission is set empty, Exim assumes that the
4419 authenticated id is a complete email address when it generates From: or
4420 Sender: header lines.
4422 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4423 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4424 chown and chgrp in /bin and hostname in /usr/bin.
4426 28. Exim was keeping the "process log" file open after each use, just as it
4427 does for the main log. This opens the possibility of it remaining open for
4428 long periods when the USR1 signal hits a daemon. Occasional processlog
4429 errors were reported, that could have been caused by this. Anyway, it seems
4430 much more sensible not to leave this file open at all, so that is what now
4433 29. The long-running daemon process does not normally write to the log once it
4434 has entered its main loop, and it closes the log before doing so. This is
4435 so that log files can straightforwardly be renamed and moved. However,
4436 there are a couple of unusual error situations where the daemon does write
4437 log entries, and I had neglected to close the log afterwards.
4439 30. The text of an SMTP error response that was received during a remote
4440 delivery was being truncated at 512 bytes. This is too short for some of
4441 the long messages that one sometimes sees. I've increased the limit to
4444 31. It is now possible to make retry rules that apply only when a message has a
4445 specific sender, in particular, an empty sender.
4447 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4448 it possible to be selective about when SMTP synchronization is enforced.
4450 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4452 32. Implemented hosts_connection_nolog.
4454 33. Added an ACL for QUIT.
4456 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4459 35. Added mailbox_size and mailbox_filecount to appendfile.
4461 36. Added control = no_multiline_responses to ACLs.
4463 37. There was a bug in the logic of the code that waits for the clock to tick
4464 in the case where the clock went backwards by a substantial amount such
4465 that the microsecond fraction of "now" was more than the microsecond
4466 fraction of "then" (but the whole seconds number was less).
4468 38. Added support for the libradius Radius client library this is found on
4469 FreeBSD (previously only the radiusclient library was supported).
4475 1. When certain lookups returned multiple values in the form name=value, the
4476 quoting of the values was not always being done properly. Specifically:
4477 (a) If the value started with a double quote, but contained no whitespace,
4479 (b) If the value contained whitespace other than a space character (i.e.
4480 tabs or newlines or carriage returns) it was not quoted.
4481 This fix has been applied to the mysql and pgsql lookups by writing a
4482 separate quoting function and calling it from the lookup code. The fix
4483 should probably also be applied to nisplus, ibase and oracle lookups, but
4484 since I cannot test any of those, I have not disturbed their existing code.
4486 2. A hit in the callout cache for a specific address caused a log line with no
4487 reason for rejecting RCPT. Now it says "Previous (cached) callout
4488 verification failure".
4490 3. There was an off-by-one bug in the queryprogram router. An over-long
4491 return line was truncated at 256 instead of 255 characters, thereby
4492 overflowing its buffer with the terminating zero. As well as fixing this, I
4493 have increased the buffer size to 1024 (and made a note to document this).
4495 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4496 when Exim is waiting for an SMTP response from a remote server, Exim
4497 restarts its select() call on the socket, thereby resetting its timeout.
4498 This is not a problem when such interrupts are rare. Somebody set up a cron
4499 job to run exiwhat every 2 minutes, which is less than the normal select()
4500 timeout (5 or 10 minutes). This meant that the select() timeout never
4501 kicked in because it was always reset. I have fixed this by comparing the
4502 time when an interrupt arrives with the time at the start of the first call
4503 to select(). If more time than the timeout has elapsed, the interrupt is
4504 treated as a timeout.
4506 5. Some internal re-factoring in preparation for the addition of Sieve
4507 extensions (by MH). In particular, the "personal" test is moved to a
4508 separate function, and given an option for scanning Cc: and Bcc: (which is
4509 not set for Exim filters).
4511 6. When Exim created an email address using the login of the caller as the
4512 local part (e.g. when creating a From: or Sender: header line), it was not
4513 quoting the local part when it contained special characters such as @.
4515 7. Installed new OpenBSD configuration files.
4517 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4518 try to make them clearer.
4520 9. Callout options, other than the timeout value, were being ignored when
4521 verifying sender addresses in header lines. For example, when using
4523 verify = header_sender/callout=no_cache
4525 the cache was (incorrectly) being used.
4527 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4528 only the Cygwin environment.
4530 11. When return_path_on_delivery was set as a log selector, if different remote
4531 addresses in the same message used different return paths and parallel
4532 remote delivery occurred, the wrong values would sometimes be logged.
4533 (Whenever a remote delivery process finished, the return path value from
4534 the most recently started remote delivery process was logged.)
4536 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4537 lines when AUTH and/or TLS are in use. This is the "received protocol"
4538 field. Exim used to use "asmtp" for authenticated SMTP, without any
4539 indication (in the protocol name) for TLS use. Now it follows the RFC and
4540 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4541 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4542 names appear in log lines as well as in Received: header lines.
4544 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4545 extensions, and comparison tests, and to fix some bugs.
4547 14. Changes to the "personal" filter test:
4549 (1) The test was buggy in that it was just doing the equivalent of
4550 "contains" tests on header lines. For example, if a user's address was
4551 anne@some.where, the "personal" test would incorrectly be true for
4553 To: susanne@some.where
4555 This test is now done by extracting each address from the header in turn,
4556 and checking the entire address. Other tests that are part of "personal"
4557 are now done using regular expressions (for example, to check local parts
4558 of addresses in From: header lines).
4560 (2) The list of non-personal local parts in From: addresses has been
4561 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4562 taken from the Sieve specification recommendations.
4564 (3) If the message contains any header line starting with "List-" it is
4565 treated as non-personal.
4567 (4) The test for "circular" in the Subject: header line has been removed
4568 because it now seems ill-conceived.
4570 15. Minor typos in src/EDITME comments corrected.
4572 16. Installed latest exipick from John Jetmore.
4574 17. If headers_add on a router specified a text string that was too long for
4575 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4576 of string_sprintf() is now avoided.
4578 18. $message_body_size was not set (it was always zero) when running the DATA
4579 ACL and the local_scan() function.
4581 19. For the "mail" command in an Exim filter, no default was being set for
4582 the once_repeat time, causing a random time value to be used if "once" was
4583 specified. (If the value happened to be <= 0, no repeat happened.) The
4584 default is now 0s, meaning "never repeat". The "vacation" command was OK
4585 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4586 (I found it when inspecting the code).
4588 20. There is now an overall timeout for performing a callout verification. It
4589 defaults to 4 times the callout timeout, which applies to individual SMTP
4590 commands during the callout. The overall timeout applies when there is more
4591 than one host that can be tried. The timeout is checked before trying the
4592 next host. This prevents very long delays if there are a large number of
4593 hosts and all are timing out (e.g. when the network connections are timing
4594 out). The value of the overall timeout can be changed by specifying an
4595 additional sub-option for "callout", called "maxwait". For example:
4597 verify = sender/callout=5s,maxwait=20s
4599 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4600 to the end before writing, but this should make it even safer).
4602 22. Exim was forgetting that it had advertised PIPELINING for the second and
4603 subsequent messages on an SMTP connection. It was also not resetting its
4604 memory on STARTTLS and an internal HELO.
4606 23. When Exim logs an SMTP synchronization error within a session, it now
4607 records whether PIPELINING has been advertised or not.
4609 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4610 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4611 rather than long int.
4613 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4615 26. Added the never_mail option to autoreply.
4621 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4622 crash if the getsockname() call failed; this can happen if a connection is
4623 closed very soon after it is established. The problem was simply in the
4624 order in which certain operations were done, causing Exim to try to write
4625 to the SMTP stream before it had set up the file descriptor. The bug has
4626 been fixed by making things happen in the correct order.
4632 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4633 before the connection was closed, thus losing the rejection response.
4635 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4636 some early Solaris releases, but causes trouble in current releases where
4637 socklen_t is defined.
4639 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4642 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4645 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4646 that releasing the top of it at the end releases what was used for sub-
4647 expansions (unless the block got too big). However, discard this block if
4648 the first thing is a variable or header, so that we can use its block when
4649 it is dynamic (useful for very large $message_headers, for example).
4651 6. Lookups now cache *every* query, not just the most recent. A new, separate
4652 store pool is used for this. It can be recovered when all lookup caches are
4653 flushed. Lookups now release memory at the end of their result strings.
4654 This has involved some general refactoring of the lookup sources.
4656 7. Some code has been added to the store_xxx() functions to reduce the amount
4657 of flapping under certain conditions.
4659 8. log_incoming_interface used to affect only the <= reception log lines. Now
4660 it causes the local interface and port to be added to several more SMTP log
4661 lines, for example "SMTP connection from", and rejection lines.
4663 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4665 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4667 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4668 could be overwritten at the end of the current message (or the start of a
4669 new message if it was set in a HELO ACL). The value is now preserved for
4670 the duration of the SMTP connection.
4672 12. If a transport had a headers_rewrite setting, and a matching header line
4673 contained an unqualified address, that address was qualified, even if it
4674 did not match any rewriting rules. The underlying bug was that the values
4675 of the flags that permit the existence of unqualified sender and recipient
4676 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4677 non-local messages, and by -bnq for local messages) were not being
4678 preserved with the message after it was received.
4680 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4681 "next input=" as part of the text comprising the host identity instead of
4682 the correct text. The code was using the same buffer for two different
4683 strings. However, depending on which order the printing function evaluated
4684 its arguments, the bug did not always show up. Under Linux, for example, my
4685 test suite worked just fine.
4687 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4688 doesn't work with some older versions of Perl. It has been changed to "my",
4689 which in any case is probably the better facility to use.
4691 15. A really picky compiler found some instances of statements for creating
4692 error messages that either had too many or two few arguments for the format
4695 16. The size of the buffer for calls to the DNS resolver has been increased
4696 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4697 for addresses that have a lot of PTR records. This alleviates a problem; it
4698 does not fully solve it.
4700 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4701 buffer now truncates the list and logs the incident, which is the same
4702 action as happens when Exim is looking up a host name and its aliases.
4703 Previously in this situation something unpredictable would happen;
4704 sometimes it was "internal error: store_reset failed".
4706 18. If a server dropped the connection unexpectedly when an Exim client was
4707 using GnuTLS and trying to read a response, the client delivery process
4708 crashed while trying to generate an error log message.
4710 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4711 string, for example:
4713 warn message = H1: something\nH2: something
4715 the text was added as a single header line from Exim's point of view
4716 though it ended up OK in the delivered message. However, searching for the
4717 second and subsequent header lines using $h_h2: did not work. This has been
4718 fixed. Similarly, if a system filter added multiple headers in this way,
4719 the routers could not see them.
4721 20. Expanded the error message when iplsearch is called with an invalid key to
4722 suggest using net-iplsearch in a host list.
4724 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4725 ACLs are no longer actually imposed (and a message to that effect is
4728 22. If a "gecos" field in a passwd entry contained escaped characters, in
4729 particular, if it contained a \" sequence, Exim got it wrong when building
4730 a From: or a Sender: header from that name. A second bug also caused
4731 incorrect handling when an unquoted " was present following a character
4732 that needed quoting.
4734 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4735 was not being matched caselessly.
4737 24. Arranged for all hyphens in the exim.8 source to be escaped with
4740 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4741 properly. Recipient callout cache records were still being keyed to include
4742 the sender, even when use_sender was set false. This led to far more
4743 callouts that were necessary. The sender is no longer included in the key
4744 when use_sender is false.
4746 26. Added "control = submission" modifier to ACLs.
4748 27. Added the ${base62d: operator to decode base 62 numbers.
4750 28. dnsdb lookups can now access SRV records.
4752 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4753 the configuration file.
4755 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4756 (-v) mode. This makes the output for a verbose queue run more intelligible.
4758 31. Added a use_postmaster feature to recipient callouts.
4760 32. Added the $body_zerocount variable, containing the number of binary zero
4761 bytes in the message body.
4763 33. The time of last modification of the "new" subdirectory is now used as the
4764 "mailbox time last read" when there is a quota error for a maildir
4767 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4769 35. Added +ignore_unknown as a special item in host lists.
4771 36. Code for decoding IPv6 addresses in host lists is now included, even if
4772 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4773 address was recognized as an IP address, but was then not correctly decoded
4774 into binary, causing unexpected and incorrect effects when compared with
4781 1. Very minor rewording of debugging text in manualroute to say "list of
4782 hosts" instead of "hostlist".
4784 2. If verify=header_syntax was set, and a header line with an unqualified
4785 address (no domain) and a large number of spaces between the end of the
4786 name and the colon was received, the reception process suffered a buffer
4787 overflow, and (when I tested it) crashed. This was caused by some obsolete
4788 code that should have been removed. The fix is to remove it!
4790 3. When running in the test harness, delay a bit after writing a bounce
4791 message to get a bit more predictability in the log output.
4793 4. Added a call to search_tidyup() just before forking a reception process. In
4794 theory, someone could use a lookup in the expansion of smtp_accept_max_
4795 per_host which, without the tidyup, could leave open a database connection.
4797 5. Added the variables $recipient_data and $sender_data which get set from a
4798 lookup success in an ACL "recipients" or "senders" condition, or a router
4799 "senders" option, similar to $domain_data and $local_part_data.
4801 6. Moved the writing of debug_print from before to after the "senders" test
4804 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4805 problems for message scanning, either using a data ACL, or using
4806 local_scan() because the Received: header was not generated till after they
4807 were called (in order to set the time as the time of reception completion).
4808 I have revised the way this works. The header is now generated after the
4809 body is received, but before the ACL or local_scan() are called. After they
4810 are run, the timestamp in the header is updated.
4816 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4817 before starting a queue runner without re-exec. This happened only when
4818 deliver_drop_privilege was set or when the Exim user was set to root. The
4819 effect of the bug was that timeouts during subsequent deliveries caused
4820 crashes instead of being properly handled. The handler is now left at its
4821 default (and expected) setting.
4823 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4824 message, again when deliver_drop_privilege is set or Exim is run as root.
4825 The bug described in (1) was not present in this case, but the tidying up
4826 of the other signals was missing. I have made the two cases consistent.
4828 3. The ignore_target_hosts setting on a manualroute router was being ignored
4829 for hosts that were looked up using the /MX notation.
4831 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4834 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4835 operated on the sender address. After changing the $sender_address to <>
4836 for the sender address verify, Exim was re-instated it as the original
4837 (before rewriting) address, but remembering that it had rewritten it, so it
4838 wasn't rewriting it again. This bug also had the effect of breaking the
4839 sender address verification caching when the sender address was rewritten.
4841 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4842 This has been changed so that if the ip literal address matches
4843 ignore_target_hosts, the router declines.
4845 7. Added expansion conditions match_domain, match_address, and match_local_
4846 part (NOT match_host).
4848 8. The placeholder for the Received: header didn't have a length field set.
4850 9. Added code to Exim itself and to exim_lock to test for a specific race
4851 condition that could lead to file corruption when using MBX delivery. The
4852 issue is with the lockfile that is created in /tmp. If this file is removed
4853 after a process has opened it but before that process has acquired a lock,
4854 there is the potential for a second process to recreate the file and also
4855 acquire a lock. This could lead to two Exim processes writing to the file
4856 at the same time. The added code performs the same test as UW imapd; it
4857 checks after acquiring the lock that its file descriptor still refers to
4858 the same named file.
4860 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4861 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4864 11. Added the smtp_active_hostname option. If used, this will typically be made
4865 to depend on the incoming interface address. Because $interface_address is
4866 not set up until the daemon has forked a reception process, error responses
4867 that can happen earlier (such as "too many connections") no longer contain
4870 12. If an expansion in a condition on a "warn" statement fails because a lookup
4871 defers, the "warn" statement is abandoned, and the next ACL statement is
4872 processed. Previously this caused the whole ACL to be aborted.
4874 13. Added the iplsearch lookup type.
4876 14. Added ident_timeout as a log selector.
4878 15. Added tls_certificate_verified as a log selector.
4880 16. Added a global option tls_require_ciphers (compare the smtp transport
4881 option of the same name). This controls incoming TLS connections.
4883 17. I finally figured out how to make tls_require_ciphers do a similar thing
4884 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4885 before starting the TLS session.
4887 18. Tabs are now shown as \t in -bP output.
4889 19. If the log selector return_path_on_delivery was set, Exim crashed when
4890 bouncing a message because it had too many Received: header lines.
4892 20. If two routers both had headers_remove settings, and the first one included
4893 a superfluous trailing colon, the final name in the first list and the
4894 first name in the second list were incorrectly joined into one item (with a
4895 colon in the middle).
4901 1. Added -C and -D options to the exinext utility, mainly to make it easier
4902 to include in the automated testing, but these could be helpful when
4903 multiple configurations are in use.
4905 2. The exinext utility was not formatting the output nicely when there was
4906 an alternate port involved in the retry record key, nor when there was a
4907 message id as well (for retries that were specific to a specific message
4908 and a specific host). It was also confused by IPv6 addresses, because of
4909 the additional colons they contain. I have fixed the IPv4 problem, and
4910 patched it up to do a reasonable job for IPv6.
4912 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4913 delivery, the log line now contains "pipelined" if PIPELINING was used.
4915 4. An SMTP transport process used to panic and die if the bind() call to set
4916 an explicit outgoing interface failed. This has been changed; it is now
4917 treated in the same way as a connect() failure.
4919 5. A reference to $sender_host_name in the part of a conditional expansion
4920 that was being skipped was still causing a DNS lookup. This no longer
4923 6. The def: expansion condition was not recognizing references to header lines
4924 that used bh_ and bheader_.
4926 7. Added the _cache feature to named lists.
4928 8. The code for checking quota_filecount in the appendfile transport was
4929 allowing one more file than it should have been.
4931 9. For compatibility with Sendmail, the command line option
4939 and sets the incoming protocol and host name (for trusted callers). The
4940 host name and its colon can be omitted when only the protocol is to be set.
4941 Note the Exim already has two private options, -pd and -ps, that refer to
4942 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4943 "s", but I don't think that's a major issue.
4945 10. A number of refactoring changes to the code, none of which should affect
4948 (a) The number of logging options was getting close to filling up the
4949 32-bit word that was used as a bit map. I have split them into two classes:
4950 those that are passed in the argument to log_write(), and those that are
4951 only ever tested independently outside of that function. These are now in
4952 separate 32-bit words, so there is plenty of room for expansion again.
4953 There is no change in the user interface or the logging behaviour.
4955 (b) When building, for example, log lines, the code previously used a
4956 macro that called string_cat() twice, in order to add two strings. This is
4957 not really sufficiently general. Furthermore, there was one instance where
4958 it was actually wrong because one of the argument was used twice, and in
4959 one call a function was used. (As it happened, calling the function twice
4960 did not affect the overall behaviour.) The macro has been replaced by a
4961 function that can join an arbitrary number of extra strings onto a growing
4964 (c) The code for expansion conditions now uses a table and a binary chop
4965 instead of a serial search (which was left over from when there were very
4966 few conditions). Also, it now recognizes conditions like "pam" even when
4967 the relevant support is not compiled in: a suitably worded error message is
4968 given if an attempt is made to use such a condition.
4970 11. Added ${time_interval:xxxxx}.
4972 12. A bug was causing one of the ddress fields not to be passed back correctly
4973 from remote delivery subprocesses. The field in question was not being
4974 subsequently used, so this caused to problems in practice.
4976 13. Added new log selectors queue_time and deliver_time.
4978 14. Might have fixed a bug in maildirsizefile handling that threw up
4979 "unexpected character" debug warnings, and recalculated the data
4980 unnecessarily. In any case, I expanded the warning message to give more
4983 15. Added the message "Restricted characters in address" to the statements in
4984 the default ACL that block characters like @ and % in local parts.
4986 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4987 Three changes have been made:
4989 (a) There was a serious bug; a negative response to MAIL caused the whole
4990 recipient domain to be cached as invalid, thereby blocking all messages
4991 to all local parts at the same domain, from all senders. This bug has
4992 been fixed. The domain is no longer cached after a negative response to
4993 MAIL if the sender used is not empty.
4995 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4998 (c) A new callout option, "use_sender" has been added for people who want
4999 the modified behaviour.
5005 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5008 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5009 indeed breaks things for older releases.
5011 3. Added additional logging to the case where there is a problem reading data
5012 from a filter that is running in a subprocess using a pipe, in order to
5013 try to track down a specific problem.
5015 4. Testing facility fudge: when running in the test harness and attempting
5016 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5017 getting "No route to host". Convert this to a timeout.
5019 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5022 6. Some OS don't have socklen_t but use size_t instead. This affects the
5023 fifth argument of getsockopt() amongst other things. This is now
5024 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5025 can be set for individual OS. I have set it for SunOS5, OSF1, and
5026 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5027 some earlier ones do not.
5029 7. Change 4.30/15 was not doing the test caselessly.
5031 8. The standard form for an IPv6 address literal was being rejected by address
5032 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5033 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5034 this, as well as the form without the "IPv6" on the front (but only when
5035 address literals are enabled, of course).
5037 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5039 10. Exim crashed if a message with an empty sender address specified by -f
5040 encountered a router with an errors_to setting. This could be provoked only
5041 by a command such as
5045 where an empty string was supplied; "<>" did not hit this bug.
5047 11. Installed PCRE release 4.5.
5049 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5050 remained set. It is now erased.
5052 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5053 times from message ids (which are base 36 rather than the normal 62).
5055 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5056 were being counted as actual protocol errors, and logged if the log
5057 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5058 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5059 and DATA following a set of rejected RCPTs do not count as protocol errors.
5060 In other words, Exim assumes they were pipelined, though this may not
5061 actually be the case. Of course, in all cases the client gets an
5062 appropriate error code.
5064 15. If a lookup fails in an ACL condition, a message about the failure may
5065 be available; it is used if testing the ACL cannot continue, because most
5066 such messages specify what the cause of the deferral is. However, some
5067 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5068 that caused an old message to be retained and used if a later statement
5069 caused a defer, replacing the real cause of the deferral.
5071 16. If an IP address had so many PTR records that the DNS lookup buffer
5072 was not large enough to hold them, Exim could crash while trying to process
5073 the truncated data. It now detects and logs this case.
5075 17. Further to 4.21/58, another change has been made: if (and only if) the
5076 first line of a message (the first header line) ends with CRLF, a bare LF
5077 in a subsequent header line has a space inserted after it, so as not to
5078 terminate the header.
5080 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5081 unnecessarily, used atoi() instead of strtol(), and didn't check the
5082 termination when getting file sizes from file names by regex.
5084 19. Completely re-implemented the support for maildirsize files, in the light
5085 of a number of problems with the previous contributed implementation
5086 (4.30/29). In particular:
5088 . If the quota is zero, the maildirsize file is maintained, but no quota is
5091 . If the maildir directory does not exist, it is created before any attempt
5092 to write a maildirsize file.
5094 . The quota value in the file is just a cache; if the quota is changed in
5095 the transport, the new value overrides.
5097 . A regular expression is available for excluding directories from the
5100 20. The autoreply transport checks the characters in options that define the
5101 message's headers; it allows continued headers, but it was checking with
5102 isspace() after an embedded newline instead of explicitly looking for a
5105 21. If all the "regular" hosts to which an address was routed had passed their
5106 expiry times, and had not reached their retry times, the address was
5107 bounced, even if fallback hosts were defined. Now Exim should go on to try
5110 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5111 equivalent code in the SMTP transport. Some hosts send humungous responses
5112 to HELO/EHLO, more than 1024 it seems.
5114 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5115 gives compiler warnings in some environments. I've now done it "properly",
5118 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5119 (because of problems with the built-in one) was declared to return uschar *
5120 instead of char *, causing compiler failure.
5122 25. Fixed a file descriptor leak when processing alias/forward files.
5124 26. Fixed a minor format string issue in dbfn.c.
5126 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5128 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5129 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5132 29. When a sender address is verified, it is cached, to save repeating the test
5133 when there is more than one recipient in a message. However, when the
5134 verification involves a callout, it is possible for different callout
5135 options to be set for different recipients. It is too complicated to keep
5136 track of this in the cache, so now Exim always runs a verification when a
5137 callout is required, relying on the callout cache for the optimization.
5138 The overhead is duplication of the address routing, but this should not be
5141 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5142 to be verified with callout=postmaster, and the main callout worked but the
5143 postmaster check failed, the verification correctly failed. However, if a
5144 subsequent RCPT command asked for sender verification *without* the
5145 postmaster check, incorrect caching caused this verification also to fail,
5148 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5149 it was not caching the DNS options (qualify_single, search_parents) that
5150 were used when the lookup failed. A subsequent lookup with different
5151 options therefore always gave the same answer, though there were cases
5152 where it should not have. (Example: a "domains = !$mx_any" option on a
5153 dnslookup router: the "domains" option is always processed without any
5154 widening, but the router might have qualify_single set.) Now Exim uses the
5155 cached value only when the same options are set.
5157 32. Added John Jetmore's "exipick" utility to the distribution.
5159 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5160 than a timeout (e.g. a certificate is required, and is not provided), an
5161 Exim server now closes the connection immediately. Previously it waited for
5162 the client to close - but if the client is SSL, it seems that they each
5163 wait for each other, leading to a delay before one of them times out.
5165 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5166 maintained 0.8.x compatibility because I don't think many are using it, and
5167 it is clearly obsolete.
5169 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5172 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5173 client certificate was expired. A simple patch fixes this, though I don't
5174 understand the full logic of why the verify callback is called multiple
5177 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5178 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5179 which causes problems with some clients (such as the Certicom SSL Plus
5180 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5181 disables the coutermeasure allowing Eudora to connect."
5183 38. Exim was not checking that a write() to a log file succeeded. This could
5184 lead to Bad Things if a log got too big, in particular if it hit a file
5185 size limit. Exim now panics and dies if it cannot write to a log file, just
5186 as it does if it cannot open a log file.
5188 39. Modified OS/Makefile-Linux so that it now contains
5190 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5192 The two -D definitions ensure that Exim is compiled with large file
5193 support, which makes it possible to handle log files that are bigger than
5196 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5197 instance) a domain was checked against a named list that involved a lookup,
5198 causing $domain_data to be set, then another domain was checked against the
5199 same list, then the first domain was re-checked, the value of $domain_data
5200 after the final check could be wrong. In particular, if the second check
5201 failed, it could be set empty. This bug probably also applied to
5204 41. The strip_trailing_dot option was not being applied to the address given
5205 with the -f command-line option.
5207 42. The code for reading a message's header from the spool was incrementing
5208 $received_count, but never initializing it. This meant that the value was
5209 incorrect (doubled) while delivering a message in the same process in which
5210 it was received. In the most common configuration of Exim, this never
5211 happens - a fresh exec is done - but it can happen when
5212 deliver_drop_privilege is set.
5214 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5215 it now includes up to 150 characters of the unexpected data in the log
5218 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5219 and building data strings. The size of both of these buffers was 10 000
5220 bytes - far larger than anybody would *ever* want, thought I. Needless to
5221 say, somebody hit the limit. I have increased the maximum line length to
5222 20 000 and the maximum data length of concatenated lines to 100 000. I have
5223 also fixed two bugs, because there was no checking on these buffers. Tsk,
5224 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5225 buffer is too small.
5227 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5228 lsearch lookups. Now it does.
5230 46. When parsing a route_list item in a manualroute router, a fixed-length
5231 buffer was used for the list of hosts. I made this 1024 bytes long,
5232 thinking that nobody would ever have a list of hosts that long. Wrong.
5233 Somebody had a whole pile of complicated expansion conditions, and the
5234 string was silently truncated, leading to an expansion error. It turns out
5235 that it is easier to change to an unlimited length (owing to other changes
5236 that have happened since this code was originally written) than to build
5237 structure for giving a limitation error. The length of the item that
5238 expands into the list of hosts is now unlimited.
5240 47. The lsearch lookup could not handle data where the length of text line was
5241 more than 4095 characters. Such lines were truncated, leading to shortened
5242 data being returned. It should now handle lines of any length.
5244 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5245 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5248 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5249 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5250 no longer happens in comment lines. A list of replacements is now placed
5251 at the head of all of the source files, except those whose only change is
5252 to replace PERL_COMMAND in the very first #! line.
5254 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5255 messages on the queue, with a bottom-up merge sort, using code contributed
5256 by Michael Haardt. This should make operations like -bp somewhat faster on
5257 large queues. It won't affect queue runners, except when queue_run_in_order
5260 51. Installed eximstats 1.31 in the distribution.
5262 52. Added support for SRV lookups to the dnslookup router.
5264 53. If an ACL referred to $message_body or $message_body_end, the value was not
5265 reset for any messages that followed in the same SMTP session.
5267 54. The store-handling optimization for building very long strings was not
5268 differentiating between the different store pools. I don't think this
5269 actually made any difference in practice, but I've tidied it.
5271 55. While running the routers to verify a sender address, $sender_address
5272 was still set to the sender address. This is wrong, because when routing to
5273 send a bounce to the sender, it would be empty. Therefore, I have changed
5274 it so that, while verifying a sender address, $sender_address is set to <>.
5275 (There is no change to what happens when verifying a recipient address.)
5277 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5278 target A or AAAA records (if not already returned) without resetting the
5279 qualify_single or search_parents options of the DNS resolver. These are
5280 inappropriate in this case because the targets of MX and SRV records must
5281 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5282 target that, when qualified, matched something in the local domain. These
5283 two options are now turned off when doing these lookups.
5285 57. It seems that at least some releases of Reiserfs (which does not have the
5286 concept of a fixed number of inodes) returns zero and not -1 for the
5287 number of available inodes. This interacted badly with check_spool_inodes,
5288 which assumed that -1 was the "no such thing" setting. What I have done is
5289 to check that the total number of inodes is greater than zero before doing
5290 the test of how many are available.
5292 58. When a "warn" ACL statement has a log_message modifier, the message is
5293 remembered, and not repeated. This is to avoid a lot of repetition when a
5294 message has many recipients that cause the same warning to be written.
5295 Howewer, Exim was preserving the list of already written lines for an
5296 entire SMTP session, which doesn't seem right. The memory is now reset if a
5297 new message is started.
5299 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5300 debugging output unless log_rewrite was also set.
5302 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5303 of (int)(handle) when we know that handle contains (void *)(-1).
5305 61. The Exim daemon panic-logs an error return when it closes the incoming
5306 connection. However "connection reset by peer" seems to be common, and
5307 isn't really an error worthy of noting specially, so that particular error
5310 62. When Exim is trying to find all the local interfaces, it used to panic and
5311 die if the ioctl to get the interface flags failed. However, it seems that
5312 on at least one OS (Solaris 9) it is possible to have an interface that is
5313 included in the list of interfaces, but for which you get a failure error
5314 for this call. This happens when the interface is not "plumbed" into a
5315 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5316 failure of the "get flags" call assumes that the interface is down.
5318 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5319 makes life easier for people who are doing arithmetic on fields extracted
5320 from dates, where you often get leading zeros that should not be
5321 interpreted as octal.
5323 64. Added qualify_domain to the redirect router, to override the global
5326 65. If a pathologically long header line contained very many addresses (the
5327 report of this problem mentioned 10 000) and each of them was rewritten,
5328 Exim could use up a very large amount of memory. (It kept on making new
5329 copies of the header line as it rewrote, and never released the old ones.)
5330 At the expense of a bit more processing, the header rewriting function has
5331 been changed so that it no longer eats memory in this way.
5333 66. The generation of the Received: header has been moved from the time that a
5334 message starts to be received, to the time that it finishes. The timestamp
5335 in the Received: header should now be very close to that of the <= log
5336 line. There are two side-effects of this change:
5338 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5339 logged header lines no longer include the local Received: line, because
5340 it has not yet been created. The same applies to a copy of the message
5341 that is returned to a non-SMTP sender when a message is rejected.
5343 (b) When a filter file is tested using -bf, no additional Received: header
5344 is added to the test message. After some thought, I decided that this
5347 This change does not affect the value of $received_for. It is still set
5348 after address rewriting, but before local_scan() is called.
5350 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5352 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5353 gave an unhelpful panic error message, and a defer error. I have managed to
5354 change this behaviour so that it now rejects any supplied certificate,
5355 which seems right, as the list of acceptable certificates is empty.
5357 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5358 gave an unhelpful defer error. I have not managed to make this reject any
5359 supplied certificates, but the error message it gives is "no certificate
5360 supplied", which is not helpful.
5362 70. exigrep's output now also includes lines that are not associated with any
5363 message, but which match the given pattern. Implemented by a patch from
5364 Martin Sluka, which also tidied up the Perl a bit.
5366 71. Recipient callout verification, like sender verification, was using <> in
5367 the MAIL FROM command. This isn't really the right thing, since the actual
5368 sender may affect whether the remote host accepts the recipient or not. I
5369 have changed it to use the actual sender in the callout; this means that
5370 the cache record is now keyed on a recipient/sender pair, not just the
5371 recipient address. There doesn't seem to be a real danger of callout loops,
5372 since a callout by the remote host to check the sender would use <>.
5373 [SEE ABOVE: changed after hitting problems.]
5375 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5376 temporary errors. However, in the case of such a code being given after
5377 the end of a data transmission (i.e. after ".") Exim was failing to write
5378 a retry record for the message. (Yes, there was some broken host that was
5379 actually sending 8xx at this point.)
5381 73. An unknown lookup type in a host list could cause Exim to panic-die when
5382 the list was checked. (An example that provoked this was putting <; in the
5383 middle of a list instead of at the start.) If this happened during a DATA
5384 ACL check, a -D file could be left lying around. This kind of configuration
5385 error no longer causes Exim to die; instead it causes a defer errror. The
5386 incident is still logged to the main and panic logs.
5388 74. Buglet left over from Exim 3 conversion. The message "too many messages
5389 in one connection" was written to the rejectlog but not the mainlog, except
5390 when address rewriting (yes!) was being logged.
5392 75. Added write_rejectlog option.
5394 76. When a system filter was run not as root (that is, when system_filter_user
5395 was set), the values of the $n variables were not being returned to the
5396 main process; thus, they were not subsequently available in the $sn
5399 77. Added +return_path_on_delivery log selector.
5401 78. A connection timeout was being treated differently from recipients deferred
5402 when testing hosts_max_try with a message that was older than the host's
5403 retry timeout. (The host should not be counted, thus allowing all hosts to
5404 be tried at least once before bouncing.) This may have been the cause of an
5405 occasionally reported bug whereby a message would remain on the queue
5406 longer than the retry timeout, but would be bounced if a delivery was
5407 forced. I say "may" because I never totally pinned down the problem;
5408 setting up timeout/retry tests is difficult. See also the next item.
5410 79. The ultimate address timeout was not being applied to errors that involved
5411 a combination of host plus message (for example, a timeout on a MAIL
5412 command). When an address resolved to a number of possible hosts, and they
5413 were not all tried for each delivery (e.g. because of hosts_max_try), a
5414 message could remain on the queue longer than the retry timeout.
5416 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5419 81. Fixed an obscure SMTP outgoing bug which required at least the following
5420 conditions: (a) there was another message waiting for the same server;
5421 (b) the server returned 5xx to all RCPT commands in the first message so
5422 that the message was not completed; (c) the server dropped the connection
5423 or gave a negative response to the RSET that Exim sends to abort the
5424 transaction. The observed case was a dropped connection after DATA that had
5425 been sent in pipelining mode. That is, the server had advertised PIPELINING
5426 but was not implementing it correctly. The effect of the bug was incorrect
5427 behaviour, such as trying another host, and this could lead to a crash.
5433 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5434 and daemon.c were passed as pointers to ints; they should have been
5435 pointers to socklen_t variables (which are typically unsigned ints).
5437 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5440 3. Fixed a really odd bug that affected only the testing scheme; patching a
5441 certain fixed string in the binary changed the value of another string that
5442 happened to be identical to the end of the original first string.
5444 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5445 name", it returns that address as the IP address. On some operating
5446 systems (e.g. Solaris), it also passes back the IP address string as the
5447 "host name". However, on others (e.g. Linux), it passes back an empty
5448 string. Exim wasn't checking for this, and was changing the host name to an
5449 empty string, assuming it had been canonicized.
5451 5. Although rare, it is permitted to have more than one PTR record for a given
5452 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5453 all the names associated with an address, because they do in Solaris.
5454 However, it seems that they do not in Linux for data that comes from the
5455 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5456 I found this out when I moved to a new Linux workstation and tried to run
5457 the Exim test suite.
5459 To get round this problem I have changed the code so that it now does its
5460 own call to the DNS to look up PTR records when searching for a host name.
5461 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5462 addresses that are only in /etc/hosts are still found.
5464 This behaviour is, however, controlled by an option called host_lookup_
5465 order, which defaults to "bydns:byaddr". If people want to use the other
5466 order, or indeed, just use one or the other means of lookup, they can
5467 specify it in this variable.
5469 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5470 some operating systems, this comes back from gethostbyaddr() as an empty
5471 string, and this is what Exim used to test for. However, it seems that in
5472 other systems, "." is yielded. Exim now tests for this case too.
5474 7. The values of check_spool_space and check_log_space are now held internally
5475 as a number of kilobytes instead of an absolute number of bytes. If a
5476 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5477 kilobyte. This means that much larger values can be stored.
5479 8. Exim monitor: an attempt to get the action menu when not actually pointing
5480 at a message produces an empty menu entitled "No message selected". This
5481 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5482 no entries in it ("Shell widget menu has zero width and/or height"). So I
5483 have added a single, blank menu entry in this case.
5485 9. Added ${quote_local_part.
5487 10. MIME decoding is now applied to the contents of Subject: header lines when
5490 11. Now that a reference to $sender_host_address automatically causes a reverse
5491 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5492 host lookup before query-style lookups in lists that might use this
5493 variable. This has therefore been abolished, and the "net-" prefix is no
5494 longer necessary for query-style lookups.
5496 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5497 have been a typo for LFLAGS, so it has been changed.
5499 13. The install script calls Exim with "-C /dev/null" in order to find the
5500 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5501 to be output. Howeve, since Exim outputs its version number before the
5502 error, it didn't break the script. It just looked ugly. I fixed this by
5503 always allowing "-C /dev/null" if the caller is root.
5505 14. Ignore overlarge ACL variable number when reading spool file - insurance
5506 against a later release with more variables having written the file.
5508 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5509 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5510 this, as well as the form without the "IPv6" on the front.
5512 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5513 OS/Makefile-Darwin file.
5515 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5516 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5518 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5519 variable. However, some OS use ints for inodes. I've added cast to long int
5520 to get rid of the compiler warning.
5522 19. I had forgotten to lock out "/../" in configuration file names when
5523 ALT_CONFIG_PREFIX was set.
5525 20. Routers used for verification do not need to specify transports. However,
5526 if such a router generated a host list, and callout was configured, Exim
5527 crashed, because it could not find a port number from the (non-existent)
5528 transport. It now assumes port 25 in this circumstance.
5530 21. Added the -t option to exigrep.
5532 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5533 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5534 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5535 them is set, LOOKUP_LSEARCH is forced.
5537 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5539 24. Added sender and host information to the "rejected by local_scan()" log
5540 line; previously there was no indication of these.
5542 25. Added .include_if_exists.
5544 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5545 Linux. It turns out that not all file systems support this. Apparently some
5546 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5547 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5548 not supported on the file descriptor, is now ignored when Exim is trying to
5549 sync a directory. This applies only to Linux.
5551 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5553 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5554 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5557 29. Added support for maildirsize files from supplied patch (modified a bit).
5559 30. The use of :fail: followed by an empty string could lead Exim to respond to
5560 sender verification failures with (e.g.):
5562 550 Verification failed for <xxx>
5563 550 Sender verify failed
5565 where the first response line was missing the '-' that indicates it is not
5566 the final line of the response.
5568 31. The loop for finding the name of the user that called Exim had a hardwired
5569 limit of 10; it now uses the value of finduser_retries, which is used for
5570 all other user lookups.
5572 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5575 33. Exim was neglecting to zero errno before one call of strtol() when
5576 expanding a string and expecting an integer value. On some systems this
5577 resulted in spurious "integer overflow" errors. Also, it was casting the
5578 result into an int without checking.
5580 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5581 did not work. The code looks as if it has *never* worked, though it appears
5582 to have been documented since at least releast 1.62. I have made it work.
5584 35. The "timeout_DNS" error in retry rules, also documented since at least
5585 1.62, also never worked. As it isn't clear exactly what this means, and
5586 clearly it isn't a major issue, I have abolished the feature by treating it
5587 as "timeout", and writing a warning to the main and panic logs.
5589 36. The display of retry rules for -brt wasn't always showing the error code
5592 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5593 timeout_connect_A, timeout_connect_MX.
5595 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5596 to the empty sender.
5598 39. The daemon was not analysing the content of -oX till after it had closed
5599 stderr and disconnected from the controlling terminal. This meant that any
5600 syntax errors were only noted on the panic log, and the return code from
5601 the command was 0. By re-arranging the code a little, I've made the
5602 decoding happen first, so such errors now appear on stderr, and the return
5603 code is 1. However, the actual setting up of the sockets still happens in
5604 the disconnected process, so errors there are still only recorded on the
5607 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5608 connections (as happens on some IP stacks) was logged at start up time as
5609 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5610 it from "IPv6 and IPv4", which means that two separate sockets are being
5613 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5614 says whether AF_INET or AF_INET6 was passed as an argument.
5616 42. Exiwhat output was messed up when time zones were included in log
5619 43. Exiwhat now gives more information about the daemon's listening ports,
5620 and whether -tls-on-connect was used.
5622 44. The "port" option of the smtp transport is now expanded.
5624 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5625 silently ignored. Now an error message is written to the main and panic
5628 46. There's a new ACL modifier called "logwrite" which writes to a log file
5629 as soon as it is encountered.
5631 47. Added $local_user_uid and $local_user_gid at routing time.
5633 48. Exim crashed when trying to verify a sender address that was being
5636 49. Exim was recognizing only a space character after ".include". It now also
5637 recognizes a tab character.
5639 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5640 extracting the relevant information from the specification. The man page no
5641 longer contains scrambled data for the -d option, and I've added a section
5642 at the front about calling Exim under different names.
5644 51. Added "extra_headers" argument to the "mail" command in filter files.
5646 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5649 53. Installed eximstats 1.29.
5651 54. Added transport_filter_timeout as a generic transport option.
5653 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5654 Cc: header lines. This was required by RFC 822, but it not required by RFC
5657 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5658 incoming messages that did not have them. Now it does so only if the
5659 message originates locally, that is, if there is no associated remote host
5660 address. When Resent- header lines are present, this applies to the Resent-
5661 lines rather than the non-Resent- lines.
5663 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5664 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5666 58. Messages for configuration errors now include the name of the main
5667 configuration file - useful now that there may be more than one file in a
5668 list (.included file names were always shown).
5670 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5671 for those rare installations that do not start the daemon as root or run it
5672 setuid root. I've cut out the call to initgroups() if the daemon is not
5675 60. The Exim user and group can now be bound into the binary as text strings
5676 that are looked up at the start of Exim's processing.
5678 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5680 62. Added $mailstore_basename variable.
5682 63. Installed patch to sieve.c from Michael Haardt.
5684 64. When Exim failed to open the panic log after failing to open the main log,
5685 the original message it was trying to log was written to stderr and debug
5686 output, but if they were not available (the usual case in production), it
5687 was lost. Now it is written to syslog before the two lines that record the
5688 failures to open the logs.
5690 65. Users' Exim filters run in subprocesses under the user's uid. It is
5691 possible for a "deliver" command or an alias in a "personal" command to
5692 provoke an address rewrite. If logging of address rewriting is configured,
5693 this fails because the process is not running as root or exim. There may be
5694 a better way of dealing with this, but for the moment (because 4.30 needs
5695 to be released), I have disabled address rewrite logging when running a
5696 filter in a non-root, non-exim process.
5702 1. The buildconfig auxiliary program wasn't quoting the value set for
5703 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5704 not defined. This bug was masked in 4.22 by the effect that was fixed in
5707 2. Some messages that were rejected after a message id was allocated were
5708 shown as "incomplete" by exigrep. It no longer does this for messages that
5709 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5711 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5712 have allow_domain_literals set, the ID did not get logged in the <= line.
5713 Domain literals are now always recognized in Message-ID: header lines.
5715 4. The first argument for a ${extract expansion item is the key name or field
5716 number. Leading and trailing spaces in this item were not being ignored,
5717 causing some misleading effects.
5719 5. When deliver_drop_privilege was set, single queue runner processes started
5720 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5721 same command in the process it spins off) were not dropping privilege.
5723 6. When the daemon running as "exim" started a queue runner, it always
5724 re-executed Exim in the spun-off process. This is a waste of effort when
5725 deliver_drop_privilege is set. The new process now just calls the
5726 queue-runner function directly.
5732 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5735 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5736 ignored. Though the use of -oP was forcing the writing of a pid file, it
5737 was always written to the default place.
5739 3. If the message "no IP address found for host xxxx" is generated during
5740 incoming verification, it is now followed by identification of the incoming
5741 connection (so you can more easily find what provoked it).
5743 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5745 5. Added some features to "harden" Exim a bit more against certain attacks:
5747 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5748 be put in Local/Makefile. This is like the never_users runtime option,
5749 but it cannot be overridden. The default setting is "root".
5751 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5752 prefix string with which any file named in a -C command line option
5755 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5756 is retained for -C and -D only if the caller of Exim is root. Without
5757 it, the exim user may also use -C and -D and retain privilege.
5759 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5760 command line option is disabled.
5762 6. Macro names set by the -D option must start with an upper case letter, just
5763 like macro names defined in the configuration file.
5765 7. Added "dereference=" facility to LDAP.
5767 8. Two instances of the typo "uknown" in the source files are fixed.
5769 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5770 the Configure-Makefile script screwed up while processing it.
5772 10. Incorporated PCRE 4.4.
5774 11. The SMTP synchronization check was not operating right at the start of an
5775 SMTP session. For example, it could not catch a HELO sent before the client
5776 waited for the greeting. There is now a check for outstanding input at the
5777 point when the greeting is written. Because of the duplex, asynchronous
5778 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5779 way, but not yet received, when the check is performed.
5781 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5782 on TCP/IP sockets, because this apparently causes some broken clients to
5785 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5786 unchanged) from the Cygwin maintainer.
5788 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5789 was supported instead of testing for mbx. Effectively a typo.
5791 15. The spa authenticator server code was not checking that the input it
5792 received was valid base64.
5794 16. The debug output line for the "set" modifier in ACLs was not showing the
5795 name of the variable that was being set.
5797 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5799 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5800 lookup on its own. Something else was needed to trigger the lookup. For
5801 example, a match in host_lookup or the need for a host name in a host list.
5802 Now, if $sender_host_name is referenced and the host name has not yet been
5803 looked up, a lookup is performed. If the lookup fails, the variable remains
5804 empty, and $host_lookup_failed is set to "1".
5806 19. Added "eqi" as a case-independent comparison operator.
5808 20. The saslauthd authentication condition could segfault if neither service
5809 nor realm was specified.
5811 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5812 error message that was logged was misleading, and incoming SMTP
5813 connections were dropped. The message is now more accurate, and temporary
5814 errors are given to SMTP connections.
5816 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5817 (except RSET) with a 421 error, until QUIT is received. However, it was
5818 failing to send a response to QUIT.
5820 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5821 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5822 ACL is now run after the helo_try_verify_hosts code.
5824 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5825 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5826 case-independent, so other case variants are also recognized). Apparently
5827 some systems use these upper case variants.
5829 25. If more than two messages were waiting for the same host, and a transport
5830 filter was specified for the transport, Exim sent two messages over the
5831 same TCP/IP connection, and then failed with "socket operation on non-
5832 socket" when it tried to send the third.
5834 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5836 27. The extern definition of crypt16() in expand.c was not being excluded when
5837 the OS had its own crypt16() function.
5839 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5840 as a preferred synonym for return_size_limit, both as an option and as an
5843 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5845 30. Changed the default configuration ACL to relax the local part checking rule
5846 for addresses that are not in any local domains. For these addresses,
5847 slashes and pipe symbols are allowed within local parts, but the sequence
5848 /../ is explicitly forbidden.
5850 31. SPA server authentication was not clearing the challenge buffer before
5853 32. log_message in a "warn" ACL statement was writing to the reject log as
5854 well as to the main log, which contradicts the documentation and doesn't
5855 seem right (because no rejection is happening). So I have stopped it.
5857 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5858 However, I am unable to do any testing of this.
5860 34. Fixed an infelicity in the appendfile transport. When checking directories
5861 for a mailbox, to see if any needed to be created, it was accidentally
5862 using path names with one or more superfluous leading slashes; tracing
5863 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5865 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5866 contents are added to the log line that is written for every discarded
5867 recipient. (Previously a log_message setting was ignored.)
5869 36. The ${quote: operator now quotes the string if it is empty.
5871 37. The install script runs exim in order to find its version number. If for
5872 some reason other than non-existence or emptiness, which it checks, it
5873 could not run './exim', it was installing it with an empty version number,
5874 i.e. as "exim-". This error state is now caught, and the installation is
5877 38. An argument was missing from the function that creates an error message
5878 when Exim fails to connect to the socket for saslauthd authentication.
5879 This could cause Exim to crash, or give a corrupted message.
5881 39. Added isip, isip4, and isip6 to ${if conditions.
5883 40. The ACL variables $acl_xx are now saved with the message, and can be
5884 accessed later in routers, transports, and filters.
5886 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5887 strings in the file are not string-expanded.
5889 42. If a MAIL command specified a SIZE value that was too large to fit into an
5890 int variable, the check against message_size_limit failed. Such values are
5891 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5892 day this will have to be increased, but I don't think I want to be around
5893 when emails are that large.
5900 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5901 iconv() is not standard in FreeBSD.
5903 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5904 IPv6 enabled. The observed symptom was a segmentation fault on return from
5905 the function os_common_find_running_interfaces() in src/os.c.
5907 3. In the check_special_case() function in daemon.c I had used "errno" as an
5908 argument name, which causes warnings on some systems. This was basically a
5909 typo, since it was named "eno" in the comments!
5911 4. The code that waits for the clock to tick (at a resolution of some fraction
5912 of a second) so as to ensure message-id uniqueness was always waiting for
5913 at least one whole tick, when it could have waited for less. [This is
5914 almost certainly not relevant at current processor speeds, where it is
5915 unlikely to ever wait at all. But we try to future-proof.]
5917 5. The function that sleeps for a time interval that includes fractions of a
5918 second contained a race. It did not block SIGALRM between setting the
5919 timer, and suspending (a couple of lines later). If the interval was short
5920 and the sigsuspend() was delayed until after it had expired, the suspension
5921 never ended. On busy systems this could lead to processes getting stuck for
5924 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5925 process, before it forks any delivery processes. The open lookup caching
5926 mechanism meant that the open file or database connection was passed into
5927 the delivery process. The problem was that delivery processes always tidy
5928 up cached lookup data. This could cause a problem for the next delivery
5929 process started by the queue runner, because the external queue runner
5930 process does not know about the closure. So the next delivery process
5931 still has data in the lookup cache. In the case of a file lookup, there was
5932 no problem because closing a file descriptor in a subprocess doesn't affect
5933 the parent. However, if the lookup was caching a connection to a database,
5934 the connection was closed, and the second delivery process was likely to
5935 see errors such as "PGSQL: query failed: server closed the connection
5936 unexpectedly". The problem has been fixed by closing all cached lookups
5937 in a queue runner before running a delivery process.
5939 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5940 seem to have the "const" qualifier which it has on other OS. I've
5943 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5944 *of the same type* (client or server) with the same public name that an
5945 error should be diagnosed.
5947 9. When Exim looked up a host name for an IP address, but failed to find the
5948 original IP address when looking up the host name (a safety check), it
5949 output the message "<ip address> does not match any IP for NULL", which was
5950 confusing, to say the least. The bug was that the host name should have
5951 appeared instead of "NULL".
5953 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5954 user that is built into the binary, and the -C or -D options is used, root
5955 privilege is dropped before the configuration file is read. In addition,
5956 logging is switched to stderr instead of the normal log files. If the
5957 configuration then re-defines the Exim user, the unprivileged environment
5958 is probably not what is expected, so Exim logs a panic warning message (but
5961 However, if deliver_drop_privilege is set, the unprivileged state may well
5962 be exactly what is intended, so the warning has been cut out in that case,
5963 and Exim is allowed to try to write to its normal log files.
5969 1. smtp_return_error_details was not giving details for temporary sender
5970 or receiver verification errors.
5972 2. Diagnose a configuration error if two authenticators have the same public
5975 3. Exim used not to create the message log file for a message until the first
5976 delivery attempt. This could be confusing when incoming messages were held
5977 for policy or load reasons. The message log file is now created at the time
5978 the message is received, and an initial "Received" line is written to it.
5980 4. The automatically generated man page for command line options had a minor
5981 bug that caused no ill effects; however, a more serious problem was that
5982 the procedure for building the man page automatically didn't always
5983 operate. Consequently, release 4.20 contains an out-of-date version. This
5984 shouldn't happen again.
5986 5. When building Exim with embedded Perl support, the script that builds the
5987 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5988 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5990 6. The freeze_tell option was not being used for messages that were frozen on
5991 arrival, either by an ACL or by local_scan().
5993 7. Added the smtp_incomplete_transaction log selector.
5995 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5996 was accepting AUTH without a new EHLO.
5998 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5999 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6002 10. Logging of TCP/IP connections (when configured) now happens in the main
6003 daemon process instead of the child process, so that the TCP/IP connection
6004 count is more accurate (but it can never be perfect).
6006 11. The use of "drop" in a nested ACL was not being handled correctly in the
6007 outer ACL. Now, if condition failure induced by the nested "drop" causes
6008 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6009 or "require"), the connection is dropped.
6011 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6012 that yield "discard" can now be used with an "accept" or a "discard" verb,
6013 but an error is generated for any others (because I can't see a useful way
6014 to define what should happen).
6016 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6017 are now processed in the same way as lines in the Exim configuration file.
6018 In particular, continuation lines are supported.
6020 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6022 15. Added -ti meaning -t -i.
6024 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6025 domains, and warn by logging if others are found.
6027 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
6028 returned from GIFCONF when Exim is trying to find the list of interfaces on
6029 a host. The code in os.c has been modified to copy each ifreq to an aligned
6030 structure in all cases.
6032 Also, in some cases, the returned ifreq's were being copied to a 'struct
6033 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6034 means the last couple of bytes of an IPv6 address could be chopped if the
6035 ifreq contained only a normal sockaddr (14 bytes storage).
6037 18. Named domain lists were not supported in the hosts_treat_as_local option.
6038 An entry such as +xxxx was not recognized, and was treated as a literal
6041 19. Ensure that header lines added by a DATA ACL are included in the reject log
6042 if the ACL subsequently rejects the message.
6044 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6045 MD5 (which is deprecated).
6047 21. When testing a filter file using -bf, Exim was writing a message when it
6048 took the sender from a "From " line in the message, but it was not doing so
6049 when it took $return_path from a Return-Path: header line. It now does.
6051 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6052 with a valid header line field name (a series of printing characters
6053 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6055 23. Changed "disc" in the source to "disk" to conform to the documentation and
6056 the book and for uniformity.
6058 24. Ignore Sendmail's -Ooption=value command line item.
6060 25. When execve() failed while trying to run a command in a pipe transport,
6061 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
6062 could be confused with a return value of 69 from the command itself. This
6063 has been changed to 127, the value the shell returns if it is asked to run
6064 a non-existent command. The wording for the related log line suggests a
6065 non-existent command as the problem.
6067 26. If received_header_text expands to an empty string, do not add a Received:
6068 header line to the message. (Well, it adds a token one on the spool, but
6069 marks it "old" so that it doesn't get used or transmitted.)
6071 27. Installed eximstats 1.28 (addition of -nt option).
6073 28. There was no check for failure on the call to getsockname() in the daemon
6074 code. This can fail if there is a shortage of resources on the system, with
6075 ENOMEM, for example. A temporary error is now given on failure.
6077 29. Contrary to the C standard, it seems that in some environments, the
6078 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6079 program. Exim now does this explicitly; it affects the formatting of
6080 timestamps using strftime().
6082 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6083 complaints. I've now initialized all the variables, to avoid this.
6085 32. Header lines added by a system filter were not being "seen" during
6086 transport-time rewrites.
6088 33. The info_callback() function passed to OpenSSL is set up with type void
6089 (*)(SSL *, int, int), as described somewhere. However, when calling the
6090 function (actually a macro) that sets it up, the type void(*)() is
6091 expected. I've put in a cast to prevent warnings from picky compilers.
6093 34. If a DNS black list lookup found a CNAME record, but there were no A
6094 records associated with the domain it pointed at, Exim crashed.
6096 35. If a DNS black list lookup returned more than one A record, Exim ignored
6097 all but the first. It now scans all returned addresses if a particular IP
6098 value is being sought. In this situation, the contents of the
6099 $dnslist_value variable are a list of all the addresses, separated by a
6102 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6103 to accept a host name and all its aliases if the forward lookup for any of
6104 them yielded the IP address of the incoming connection. Now it accepts only
6105 those names whose forward lookup yields the correct IP address. Any other
6106 names are discarded. This closes a loophole whereby a rogue DNS
6107 administrator could create reverse DNS records to break through a
6108 wildcarded host restriction in an ACL.
6110 37. If a user filter or a system filter that ran in a subprocess used any of
6111 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6112 the wrong values were passed to the pipe command ($thisaddress had the
6113 value of $0, $0 had the value of $1, etc). This bug was introduced by
6114 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6116 38. Improved the line breaking for long SMTP error messages from ACLs.
6117 Previously, if there was no break point between 40 and 75 characters, Exim
6118 left the rest of the message alone. Two changes have been made: (a) I've
6119 reduced the minimum length to 35 characters; (b) if it can't find a break
6120 point between 35 and 75 characters, it looks ahead and uses the first one
6121 that it finds. This may give the occasional overlong line, but at least the
6122 remaining text gets split now.
6124 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6125 file descriptors might be low, and that setting 1000 would always raise it.
6126 It turns out that in some environments, the limit is already over 1000 and
6127 that lowering it causes trouble. So now Exim takes care not to decrease it.
6129 40. When delivering a message, the value of $return_path is set to $sender_
6130 address at the start of routing (routers may change the value). By an
6131 oversight, this default was not being set up when an address was tested by
6132 -bt or -bv, which affected the outcome if any router or filter referred to
6135 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6136 log only when "message" or "log_message" are set. However, if one of the
6137 conditions was an address verification, or a call to a nested ACL, the
6138 messages generated by the underlying test were being passed through. This
6139 no longer happens. The underlying message is available in $acl_verify_
6140 message for both "message" and "log_message" expansions, so it can be
6141 passed through if needed.
6143 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6144 new expansion $bh_ to give the encoded byte string without charset
6145 translation. Translation happens only if iconv() is available; HAVE_ICONV
6146 indicates this at build time. HEADERS_CHARSET gives the charset to
6147 translate to; headers_charset can change it in the configuration, and
6148 "headers charset" can change it in an individual filter file.
6150 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6151 that creates RFC 2047 encoded "words" labels them as that charset instead
6152 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6153 expansion operator; (ii) when Exim creates a From: line for a local
6154 message; (iii) when a header line is rewritten to include a "phrase" part.
6156 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6157 buggy, causing it to skip the first lines of messages whose message ID
6158 ended in 'D'. This would not have bitten before Exim release 4.14, because
6159 message IDs were unlikely to end in 'D' before then. The effect was to have
6160 incorrect size information for certain domains.
6162 45. #include "config.h" was missing at the start of the crypt16.c module. This
6163 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6166 46. If there was a timeout during a "random" callout check, Exim treated it as
6167 a failure of the random address, and carried on sending RSET and the real
6168 address. If the delay was just some slowness somewhere, the response to the
6169 original RCPT would be taken as a response to RSET and so on, causing
6170 mayhem of various kinds.
6172 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6173 when I implemented it. It didn't allow for the fact that some option values
6174 may legitimatetly be negative (e.g. size_addition), and it didn't even do
6175 the right test for positive values.
6177 48. Domain names in DNS records are case-independent. Exim always looks them up
6178 in lower case. Some resolvers return domain names in exactly the case they
6179 appear in the zone file, that is, they may contain uppercase letters. Not
6180 all resolvers do this - some return always lower case. Exim was treating a
6181 change of case by a resolver as a change of domain, similar to a widening
6182 of a domain abbreviation. This triggered its re-routing code and so it was
6183 trying to route what was effectively the same domain again. This normally
6184 caused routing to fail (because the router wouldn't handle the domain
6185 twice). Now Exim checks for this case specially, and just changes the
6186 casing of the domain that it ultimately uses when it transmits the message
6189 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6192 50. If a filter generated a file delivery with a non-absolute name (possible if
6193 no home directory exists for the router), the forbid_file option was not
6196 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6197 the existing equality matching.
6199 52. Exim was using ints instead of ino_t variables in some places where it was
6200 dealing with inode numbers.
6202 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6203 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6204 called TMPDIR, and if it finds it is different, it changes its value.
6206 54. The smtp_printf() function is now made available to local_scan() so
6207 additional output lines can be written before returning. There is also an
6208 smtp_fflush() function to enable the detection of a dropped connection.
6209 The variables smtp_input and smtp_batched_input are exported to
6212 55. Changed the default runtime configuration: the message "Unknown user"
6213 has been removed from the ACL, and instead placed on the localuser router,
6214 using the cannot_route_message feature. This means that any verification
6215 failures that generate their own messages won't get overridden. Similarly,
6216 the "Unrouteable address" message that was in the ACL for unverifiable
6217 relay addresses has also been removed.
6219 56. Added hosts_avoid_esmtp to the smtp transport.
6221 57. The exicyclog script was not checking for the esoteric option
6222 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6223 will work only if exicyclog is run under the appropriate euid.
6225 58. Following a discussion on the list, the rules by which Exim recognises line
6226 endings on incoming messages have been changed. The -dropcr and drop_cr
6227 options are now no-ops, retained only for backwards compatibility. The
6228 following line terminators are recognized: LF CRLF CR. However, special
6229 processing applies to CR:
6231 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6232 nor a local message in the state where . is a terminator.
6234 (ii) If a bare CR is encountered in a header line, an extra space is added
6235 after the line terminator so as not to end the header. The reasoning
6236 behind this is that bare CRs in header lines are most likely either
6237 to be mistakes, or people trying to play silly games.
6239 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6240 was being incorrectly given as 18 bytes larger than it should have been.
6241 This is a VOB (very old bug).
6243 60. This may never have affected anything current, but just in case it has:
6244 When the local host is found other than at the start of a list of hosts,
6245 the local host, those with the same MX, and any that follow, are discarded.
6246 When the list in question was part of a longer list of hosts, the following
6247 hosts (not currently being processed) were also being discarded. This no
6248 longer happens. I'm not sure if this situation could ever has previously
6251 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6254 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6255 header. This is something that is recommended in a new Internet Draft, and
6256 is something that is documented as being done by Sendmail. There are two
6257 possible values. For messages generated by the autoreply transport, Exim
6260 Auto-Submitted: auto-replied
6262 whereas for all other generated messages (e.g. bounces) it adds
6264 Auto-Submitted: auto-generated
6266 63. The "personal" condition in filters now includes a test for the
6267 Auto-Submitted: header. If it contains the string "auto-" the message it
6268 not considered personal.
6270 64. Added rcpt_include_affixes as a generic transport option.
6272 65. Added queue_only_override (default true).
6274 66. Added the syslog_duplication option.
6276 67. If what should have been the first header line of a message consisted of
6277 a space followed by a colon, Exim was mis-interpreting it as a header line.
6278 It isn't of course - it is syntactically invalid and should therefore be
6279 treated as the start of the message body. The misbehaviour could have
6280 caused a number of strange effects, including loss of data in subsequent
6281 header lines, and spool format errors.
6283 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6284 client host had authenticated. This control can now be exercised by an ACL
6285 for more flexibility.
6287 69. By default, callouts do not happen when testing with -bh. There is now a
6288 variant, -bhc, which does actually run the callout code, including
6289 consulting and updating the callout cache.
6291 70. Added support for saslauthd authentication, courtesy of Alexander
6294 71. If statvfs() failed on the spool or log directories while checking their
6295 size for availability, Exim confusingly gave the error "space shortage".
6296 Furthermore, in debugging mode it crashed with a floating point exception.
6297 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6298 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6299 serious problem, Exim now writes to the main and panic logs when this
6300 happens, with details of the failure. It then refuses to accept the
6301 incoming message, giving the message "spool directory problem" or "log
6302 directory problem" with a 421 code for SMTP messages.
6304 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6305 0, 1, and 2 exist, because some OS complain for execs without them (see
6306 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6307 descriptors. However, the code omitted to check that the open succeeded,
6308 causing mysterious errors if for some reason the permissions on /dev/null
6309 got screwed. Now Exim writes a message to the main and panic logs, and
6310 bombs out if it can't open /dev/null.
6312 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6313 interact so that it is all more flexible. It is supposed to remain
6314 backwards compatible. Also added extra_local_interfaces.
6316 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6317 to bomb out with an assertion failure - to the client this appears as a
6318 connection drop. This problem occurs in the part of the code that was taken
6319 from the Samba project. Fortunately, the assertion is in a very simple
6320 function, so I have fixed this by reproducing the function inline in the
6321 one place where it is called, and arranging for authentication to fail
6322 instead of killing the process with assert().
6324 75. The SPA client code was not working when the server requested OEM rather
6325 than Unicode encoding.
6327 76. Added code to make require_files with a specific uid setting more usable in
6328 the case where statting the file as root fails - usually a non-root-mounted
6329 NFS file system. When this happens and the failure is EACCES, Exim now
6330 forks a subprocess and does the per-uid checking as the relevant uid.
6332 77. Added process_log_path.
6334 78. If log_file_path was not explicitly set, a setting of check_log_space or
6335 check_log_inodes was ignored.
6337 79. If a space check for the spool or log partitions fails, the incident is now
6338 logged. Of course, in the latter case the data may get lost...
6340 80. Added the %p formatting code to string_format() so that it can be used to
6341 print addresses in debug_print(). Adjusted all the address printing in the
6342 debugging in store.c to use %p rather than %d.
6344 81. There was a concern that a line of code in smtp_in.c could overflow a
6345 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6346 initially expressed, the concern was not well-founded, because trailing
6347 spaces are removed early. However, if the trailing spaces were followed by
6348 a NULL, they did not get removed, so the overflow was possible. Two fixes
6351 (a) I re-wrote the offending code in a cleaner fashion.
6352 (b) If an incoming SMTP command contains a NULL character, it is rejected
6355 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6356 runs initgroups(), so that if the Exim user is in any additional groups,
6357 they will be used during message reception.
6363 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob