1 Change log file for Exim from version 4.21
2 -------------------------------------------
8 TF/01 Correctly close the server side of TLS when forking for delivery.
10 When a message was received over SMTP with TLS, Exim failed to clear up
11 the incoming connection properly after forking off the child process to
12 deliver the message. In some situations the subsequent outgoing
13 delivery connection happened to have the same fd number as the incoming
14 connection previously had. Exim would try to use TLS and fail, logging
15 a "Bad file descriptor" error.
17 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
18 utilities have not been installed.
20 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
21 temporary space as the ACL may create new global variables.
23 TL/01 LDAP support uses per connection or global context settings, depending
24 upon the detected version of the libraries at build time.
26 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
27 to extract and use the src ip:port in logging and expansions as if it
28 were a direct connection from the outside internet.
30 JH/02 Add ${listextract {number}{list}{success}{fail}}.
32 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
33 Properly escape header and check for NULL return.
35 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
38 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
40 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
41 characters in header names, implemented as a verify condition.
42 Contributed by Michael Fischer v. Mollard.
44 TL/05 Rename SPF condition results err_perm and err_temp to standardized
45 results permerror and temperror. Previous values are deprecated but
46 still accepted. In a future release, err_perm and err_temp will be
47 completely removed, which will be a backward incompatibility if the
48 ACL tests for either of these two old results. Patch contributed by
49 user bes-internal on the mailing list.
51 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
53 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
54 selectors, in both main and reject logs.
56 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
59 JH/07 Add malware type "sock" for talking to simple daemon.
61 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
63 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
64 routers/transports under cutthrough routing.
66 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
67 numbers. Touch up "bool" conditional to keep the same definition.
69 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
71 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
73 JH/12 Expand items in router/transport headers_add or headers_remove lists
74 individually rather than the list as a whole. Bug 1452.
76 Required for reasonable handling of multiple headers_ options when
77 they may be empty; requires that headers_remove items with embedded
78 colons must have them doubled (or the list-separator changed).
80 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
81 view the policy declared in the DMARC record. Currently, $dmarc_status
82 is a combined value of both the record presence and the result of the
85 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
87 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
88 dnslookup router (applying to the forward lookup).
94 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
96 PP/02 Make -n do something, by making it not do something.
97 When combined with -bP, the name of an option is not output.
99 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
102 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
103 $sender_host_name and config options to manage this, and basic check
106 PP/05 DSCP support for outbound connections and control modifier for inbound.
108 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
109 (Only plugin which currently uses this is kerberos4, which nobody should
110 be using, but we should make it available and other future plugins might
111 conceivably use it, even though it would break NAT; stuff *should* be
112 using channel bindings instead).
114 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
115 name; added for Sendmail compatibility; requires admin caller.
116 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
117 just ignore it); requires trusted caller.
118 Also parse but ignore: -Ac -Am -X<logfile>
121 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
123 TL/02 Add +smtp_confirmation as a default logging option.
125 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
126 Patch by Magnus Holmgren from 2007-02-20.
128 TL/04 Bugzilla 1281 - Spec typo.
129 Bugzilla 1283 - Spec typo.
130 Bugzilla 1290 - Spec grammar fixes.
132 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
134 TL/06 Add Experimental DMARC support using libopendmarc libraries.
136 TL/07 Fix an out of order global option causing a segfault. Reported to dev
137 mailing list by by Dmitry Isaikin.
139 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
141 JH/02 Support "G" suffix to numbers in ${if comparisons.
143 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
145 NM/01 Bugzilla 1197 - Spec typo
146 Bugzilla 1196 - Spec examples corrections
148 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
150 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
151 gnutls_enable_pkcs11, but renamed to more accurately indicate its
154 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
155 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
157 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
158 "acl {{name}{arg}...}", and optional args on acl condition
161 JH/05 Permit multiple router/transport headers_add/remove lines.
163 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
165 JH/07 Avoid using a waiting database for a single-message-only transport.
166 Performance patch from Paul Fisher. Bugzilla 1262.
168 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
171 JH/09 Add $headers_added variable, with content from use of ACL modifier
172 add_header (but not yet added to the message). Bugzilla 199.
174 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
175 Pulled from Bugzilla 817 by Wolfgang Breyha.
177 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
179 (nb: this is the same fix as in Exim 4.80.1)
181 JH/11 Add A= logging on delivery lines, and a client_set_id option on
184 JH/12 Add optional authenticated_sender logging to A= and a log_selector
187 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
189 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
190 advertise SMTP AUTH mechanism to us, instead of a generic
191 protocol violation error. Also, make Exim more robust to bad
192 data from the Dovecot auth socket.
194 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
196 When a queue runner is handling a message, Exim first routes the
197 recipient addresses, during which it prunes them based on the retry
198 hints database. After that it attempts to deliver the message to
199 any remaining recipients. It then updates the hints database using
202 So if a recipient address works intermittently, it can get repeatedly
203 deferred at routing time. The retry hints record remains fresh so the
204 address never reaches the final cutoff time.
206 This is a fairly common occurrence when a user is bumping up against
207 their storage quota. Exim had some logic in its local delivery code
208 to deal with this. However it did not apply to per-recipient defers
209 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
211 This change adds a proper retry rule check during routing so that the
212 final cutoff time is checked against the message's age. We only do
213 this check if there is an address retry record and there is not a
214 domain retry record; this implies that previous attempts to handle
215 the address had the retry_use_local_parts option turned on. We use
216 this as an approximation for the destination being like a local
217 delivery, as in LMTP.
219 I suspect this new check makes the old local delivery cutoff check
220 redundant, but I have not verified this so I left the code in place.
222 TF/02 Correct gecos expansion when From: is a prefix of the username.
224 Test 0254 submits a message to Exim with the header
228 When I ran the test suite under the user fanf2, Exim expanded
229 the header to contain my full name, whereas it should have added
230 a Resent-Sender: header. It erroneously treats any prefix of the
231 username as equal to the username.
233 This change corrects that bug.
235 GF/01 DCC debug and logging tidyup
236 Error conditions log to paniclog rather than rejectlog.
237 Debug lines prefixed by "DCC: " to remove any ambiguity.
239 TF/03 Avoid unnecessary rebuilds of lookup-related code.
241 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
242 Bug spotted by Jeremy Harris; was flawed since initial commit.
243 Would have resulted in OCSP responses post-SNI triggering an Exim
244 NULL dereference and crash.
246 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
248 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
249 Bug detection, analysis and fix by Samuel Thibault.
250 Bugzilla 1331, Debian bug #698092.
252 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
254 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
255 Server implementation by Todd Lyons, client by JH.
256 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
257 config variable "prdr_enable" controls whether the server
258 advertises the facility. If the client requests PRDR a new
259 acl_data_smtp_prdr ACL is called once for each recipient, after
260 the body content is received and before the acl_smtp_data ACL.
261 The client is controlled by bolth of: a hosts_try_prdr option
262 on the smtp transport, and the server advertisement.
263 Default client logging of deliveries and rejections involving
264 PRDR are flagged with the string "PRDR".
266 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
267 fclose(). Diagnosis by Todd Lyons.
269 PP/17 Update configure.default to handle IPv6 localhost better.
270 Patch by Alain Williams (plus minor tweaks).
273 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
274 This is now consistent with GnuTLS, and is now documented: the
275 previous undocumented portable approach to treating the option as
276 unset was to force an expansion failure. That still works, and
277 an empty string is now equivalent.
279 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
280 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
281 not performing validation itself.
283 PP/20 Added force_command boolean option to pipe transport.
284 Patch from Nick Koston, of cPanel Inc.
286 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
289 TF/04 Added udpsend ACL modifer and hexquote expansion operator
291 PP/21 Fix eximon continuous updating with timestamped log-files.
292 Broken in a format-string cleanup in 4.80, missed when I repaired the
293 other false fix of the same issue.
294 Report and fix from Heiko Schlichting.
297 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
298 Report from Prashanth Katuri.
300 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
301 It's SecureTransport, so affects any MacOS clients which use the
302 system-integrated TLS libraries, including email clients.
304 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
305 using a MIME ACL for non-SMTP local injection.
306 Report and assistance in diagnosis by Warren Baker.
308 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
310 JH/16 Fix comparisons for 64b. Bugzilla 1385.
312 TL/09 Add expansion variable $authenticated_fail_id to keep track of
313 last id that failed so it may be referenced in subsequent ACL's.
315 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
318 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
319 ldap library initialization, allowing self-signed CA's to be
320 used. Also properly sets require_cert option later in code by
321 using NULL (global ldap config) instead of ldap handle (per
322 session). Bug diagnosis and testing by alxgomz.
324 TL/12 Enhanced documentation in the ratelimit.pl script provided in
325 the src/util/ subdirectory.
327 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
328 renamed to Transport Post Delivery Action by Jeremy Harris, as
331 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
332 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
333 redis_servers = needs to be configured which will be used by the redis
334 lookup. Patch from Warren Baker, of The Packet Hub.
336 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
338 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
339 hostname or reverse DNS when processing a host list. Used suggestions
340 from multiple comments on this bug.
342 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
344 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
345 Missed a few lines, added it to make the runtest require no keyboard
348 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
349 contains upper case chars. Make router use caseful_local_part.
351 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
352 support when GnuTLS has been built with p11-kit.
358 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
360 This, or similar/improved, will also be change PP/11 of 4.82.
366 PP/01 Handle short writes when writing local log-files.
367 In practice, only affects FreeBSD (8 onwards).
368 Bugzilla 1053, with thanks to Dmitry Isaikin.
370 NM/01 Bugzilla 949 - Documentation tweak
372 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
375 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
377 PP/02 Implemented gsasl authenticator.
379 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
381 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
382 `pkg-config foo` for cflags/libs.
384 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
385 with rest of GSASL and with heimdal_gssapi.
387 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
388 `pkg-config foo` for cflags/libs for the TLS implementation.
390 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
391 properties get this fed in as external SSF. A number of robustness
392 and debugging improvements to the cyrus_sasl authenticator.
394 PP/08 cyrus_sasl server now expands the server_realm option.
396 PP/09 Bugzilla 1214 - Log authentication information in reject log.
397 Patch by Jeremy Harris.
399 PP/10 Added dbmjz lookup type.
401 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
403 PP/12 MAIL args handles TAB as well as SP, for better interop with
404 non-compliant senders.
405 Analysis and variant patch by Todd Lyons.
407 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
408 Bug report from Lars Müller <lars@samba.org> (via SUSE),
409 Patch from Dirk Mueller <dmueller@suse.com>
411 PP/13 tls_peerdn now print-escaped for spool files.
412 Observed some $tls_peerdn in wild which contained \n, which resulted
413 in spool file corruption.
415 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
416 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
417 or write after TLS renegotiation, which otherwise led to messages
420 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
421 as a tracking header (ie: a signed header comes before the signature).
422 Patch from Wolfgang Breyha.
424 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
425 comma-sep list; embedded commas doubled.
427 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
429 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
431 Report and patch from Dmitry Banschikov.
433 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
434 Removed SSL_clear() after SSL_new() which led to protocol negotiation
435 failures. We appear to now support TLS1.1+ with Exim.
437 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
438 lets Exim select keys and certificates based upon TLS SNI from client.
439 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
440 before an outbound SMTP session. New log_selector, +tls_sni.
442 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
443 NULL dereference. Report and patch from Alun Jones.
445 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
446 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
447 Not seeing resolver debug output on NetBSD, but suspect this is a
448 resolver implementation change.
450 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
451 Left warnings. Added "eximon gdb" invocation mode.
453 PP/21 Defaulting "accept_8bitmime" to true, not false.
455 PP/22 Added -bw for inetd wait mode support.
457 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
458 locate the relevant includes and libraries. Made this the default.
460 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
461 Bugzilla 1246, report and most of solution from Tomasz Kusy.
463 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
464 This may cause build issues on older platforms.
466 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
467 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
468 gnutls_require_mac & gnutls_require_protocols (no longer supported).
469 Added SNI support via GnuTLS too.
470 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
472 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
474 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
475 Applied second patch from Janne, implementing suggestion to default
476 multiple-strings-in-record handling to match SPF spec.
478 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
480 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
481 read-only, out of scope).
482 Patch from Wolfgang Breyha, report from Stuart Northfield.
484 PP/29 Fix three issues highlighted by clang analyser static analysis.
485 Only crash-plausible issue would require the Cambridge-specific
486 iplookup router and a misconfiguration.
487 Report from Marcin Mirosław.
489 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
491 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
492 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
493 As part of this, removing so much warning spew let me fix some minor
494 real issues in debug logging.
496 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
497 assignment on my part. Fixed.
499 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
500 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
501 Janne Snabb (who went above and beyond: thank you).
503 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
504 string otherwise requires a connection and a bunch more work and it's
505 relatively easy to get wrong. Should also expose TLS library linkage
508 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
509 64-bit ${eval} (JH/03).
511 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
512 GNU libc to support some of the 64-bit stuff, should not lead to
513 conflicts. Defined before os.h is pulled in, so if a given platform
514 needs to override this, it can.
516 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
517 protection layer was required, which is not implemented.
518 Bugzilla 1254, patch from Wolfgang Breyha.
520 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
521 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
522 tls_dhparam take prime identifiers. Also unbreak combination of
523 OpenSSL+DH_params+TLSSNI.
525 PP/39 Disable SSLv2 by default in OpenSSL support.
531 PP/01 Solaris build fix for Oracle's LDAP libraries.
532 Bugzilla 1109, patch from Stephen Usher.
534 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
536 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
539 TF/02 Fix a couple more cases where we did not log the error message
540 when unlink() failed. See also change 4.74-TF/03.
542 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
543 lock up or crash if it happened to be inside a call to libc when it
544 got a SIGUSR1 from exiwhat.
546 The SIGUSR1 handler appends the current process status to the process
547 log which is later printed by exiwhat. It used to use the general
548 purpose logging code to do this, but several functions it calls are
549 not safe for signals.
551 The new output code in the SIGUSR1 handler is specific to the process
552 log, and simple enough that it's easy to inspect for signal safety.
553 Removing some special cases also simplifies the general logging code.
554 Removing the spurious timestamps from the process log simplifies
557 TF/04 Improved ratelimit ACL condition.
559 The /noupdate option has been deprecated in favour of /readonly which
560 has clearer semantics. The /leaky, /strict, and /readonly update modes
561 are mutually exclusive. The update mode is no longer included in the
562 database key; it just determines when the database is updated. (This
563 means that when you upgrde Exim will forget old rate measurements.)
565 Exim now checks that the per_* options are used with an update mode that
566 makes sense for the current ACL. For example, when Exim is processing a
567 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
568 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
569 must specify per_mail/readonly. If you omit the update mode it defaults to
570 /leaky where that makes sense (as before) or /readonly where required.
572 The /noupdate option is now undocumented but still supported for
573 backwards compatibility. It is equivalent to /readonly except that in
574 ACLs where /readonly is required you may specify /leaky/noupdate or
575 /strict/noupdate which are treated the same as /readonly.
577 A useful new feature is the /count= option. This is a generalization
578 of the per_byte option, so that you can measure the throughput of other
579 aggregate values. For example, the per_byte option is now equivalent
580 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
582 The per_rcpt option has been generalized using the /count= mechanism
583 (though it's more complicated than the per_byte equivalence). When it is
584 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
585 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
586 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
587 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
588 Note that using per_rcpt with a non-readonly update mode in more than
589 one ACL will cause the recipients to be double-counted. (The per_mail
590 and per_byte options don't have this problem.)
592 The handling of very low rates has changed slightly. If the computed rate
593 is less than the event's count (usually one) then this event is the first
594 after a long gap. In this case the rate is set to the same as this event's
595 count, so that the first message of a spam run is counted properly.
597 The major new feature is a mechanism for counting the rate of unique
598 events. The new per_addr option counts the number of different
599 recipients that someone has sent messages to in the last time period. It
600 behaves like per_rcpt if all the recipient addresses are different, but
601 duplicate recipient addresses do not increase the measured rate. Like
602 the /count= option this is a general mechanism, so the per_addr option
603 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
604 example, measure the rate that a client uses different sender addresses
605 with the options per_mail/unique=$sender_address. There are further
606 details in the main documentation.
608 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
610 TF/06 Removed a few PCRE remnants.
612 TF/07 Automatically extract Exim's version number from tags in the git
613 repository when doing development or release builds.
615 PP/02 Raise smtp_cmd_buffer_size to 16kB.
616 Bugzilla 879. Patch from Paul Fisher.
618 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
619 Heavily based on revision 40f9a89a from Simon Arlott's tree.
622 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
624 PP/05 Variable $av_failed, true if the AV scanner deferred.
625 Bugzilla 1078. Patch from John Horne.
627 PP/06 Stop make process more reliably on build failure.
628 Bugzilla 1087. Patch from Heiko Schlittermann.
630 PP/07 Make maildir_use_size_file an _expandable_ boolean.
631 Bugzilla 1089. Patch from Heiko Schlittermann.
633 PP/08 Handle ${run} returning more data than OS pipe buffer size.
634 Bugzilla 1131. Patch from Holger Weiß.
636 PP/09 Handle IPv6 addresses with SPF.
637 Bugzilla 860. Patch from Wolfgang Breyha.
639 PP/10 GnuTLS: support TLS 1.2 & 1.1.
641 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
644 PP/11 match_* no longer expand right-hand-side by default.
645 New compile-time build option, EXPAND_LISTMATCH_RHS.
646 New expansion conditions, "inlist", "inlisti".
648 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
650 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
652 PP/14 fix log_write() format string regression from TF/03.
653 Bugzilla 1152. Patch from Dmitry Isaikin.
659 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
661 PP/02 Harmonised TLS library version reporting; only show if debugging.
662 Layout now matches that introduced for other libraries in 4.74 PP/03.
664 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
666 PP/04 New "dns_use_edns0" global option.
668 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
671 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
672 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
674 TK/01 Updated PolarSSL code to 0.14.2.
675 Bugzilla 1097. Patch from Andreas Metzler.
677 PP/07 Catch divide-by-zero in ${eval:...}.
680 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
683 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
684 format-string attack -- SECURITY: remote arbitrary code execution.
686 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
687 time unintentionally subject to list matching rules, letting the header
688 cause arbitrary Exim lookups (of items which can occur in lists, *not*
689 arbitrary string expansion). This allowed for information disclosure.
691 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
692 INT_MIN/-1 -- value coerced to INT_MAX.
698 NM/01 Workround for PCRE version dependency in version reporting
701 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
702 This fixes portability to compilers other than gcc, notably
703 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
705 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
706 makefiles for portability to HP-UX and POSIX correctness.
708 PP/01 Permit LOOKUP_foo enabling on the make command-line.
709 Also via indented variable definition in the Makefile.
710 (Debugging by Oliver Heesakkers).
712 PP/02 Restore caching of spamd results with expanded spamd_address.
713 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
715 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
716 Improves build reliability. Fix from: Frank Elsner
718 NM/02 Fix wide character breakage in the rfc2047 coding
719 Fixes bug 1064. Patch from Andrey N. Oktyabrski
721 NM/03 Allow underscore in dnslist lookups
722 Fixes bug 1026. Patch from Graeme Fowler
724 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
725 Code patches from Adam Ciarcinski of NetBSD.
727 NM/04 Fixed exiqgrep to cope with mailq missing size issue
730 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
731 is logged, to avoid truncation. Patch from John Horne.
733 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
734 Patch from Jakob Hirsch.
736 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
737 SQL string expansion failure details.
738 Patch from Andrey Oktyabrski.
740 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
741 Patch from Simon Arlott.
743 PP/09 New lookups functionality failed to compile on old gcc which rejects
744 extern declarations in function scope.
745 Patch from Oliver Fleischmann
747 PP/10 Use sig_atomic_t for flags set from signal handlers.
748 Check getgroups() return and improve debugging.
749 Fixed developed for diagnosis in bug 927 (which turned out to be
752 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
753 Patch from Mark Zealey.
755 PP/12 Bugzilla 1056: Improved spamd server selection.
756 Patch from Mark Zealey.
758 PP/13 Bugzilla 1086: Deal with maildir quota file races.
759 Based on patch from Heiko Schlittermann.
761 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
762 Patch from Uwe Doering, sign-off by Michael Haardt.
764 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
765 variable declaration deep within a block. Bug and patch from
768 PP/15 lookups-Makefile IRIX compatibilty coercion.
770 PP/16 Make DISABLE_DKIM build knob functional.
772 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
773 Patch by Simon Arlott
775 TF/03 Fix valgrind.h portability to C89 compilers that do not support
776 variable argument macros. Our copy now differs from upstream.
782 TF/01 Failure to get a lock on a hints database can have serious
783 consequences so log it to the panic log.
785 TF/02 Log LMTP confirmation messages in the same way as SMTP,
786 controlled using the smtp_confirmation log selector.
788 TF/03 Include the error message when we fail to unlink a spool file.
790 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
791 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
792 for maintaining out-of-tree patches for some time.
794 PP/01 Bugzilla 139: Documentation and portability issues.
795 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
796 Handle per-OS dynamic-module compilation flags.
798 PP/02 Let /dev/null have normal permissions.
799 The 4.73 fixes were a little too stringent and complained about the
800 permissions on /dev/null. Exempt it from some checks.
801 Reported by Andreas M. Kirchwitz.
803 PP/03 Report version information for many libraries, including
804 Exim version information for dynamically loaded libraries. Created
805 version.h, now support a version extension string for distributors
806 who patch heavily. Dynamic module ABI change.
808 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
809 privilege escalation vulnerability whereby the Exim run-time user
810 can cause root to append content of the attacker's choosing to
813 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
816 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
817 If dropping privileges for untrusted macros, we disabled normal logging
818 on the basis that it would fail; for the Exim run-time user, this is not
819 the case, and it resulted in successful deliveries going unlogged.
820 Fixed. Reported by Andreas Metzler.
826 PP/01 Date: & Message-Id: revert to normally being appended to a message,
827 only prepend for the Resent-* case. Fixes regression introduced in
828 Exim 4.70 by NM/22 for Bugzilla 607.
830 PP/02 Include check_rfc2047_length in configure.default because we're seeing
831 increasing numbers of administrators be bitten by this.
833 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
835 PP/03 Bugzilla 994: added openssl_options main configuration option.
837 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
839 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
841 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
843 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
844 without a peer certificate, leading to a segfault because of an
845 assumption that peers always have certificates. Be a little more
846 paranoid. Problem reported by Martin Tscholak.
848 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
849 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
850 NB: ClamAV planning to remove STREAM in "middle of 2010".
851 CL also introduces -bmalware, various -d+acl logging additions and
852 more caution in buffer sizes.
854 PP/09 Implemented reverse_ip expansion operator.
856 PP/10 Bugzilla 937: provide a "debug" ACL control.
858 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
860 PP/12 Bugzilla 973: Implement --version.
862 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
864 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
866 PP/15 Bugzilla 816: support multiple condition rules on Routers.
868 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
869 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
870 ignore trailing whitespace.
872 JJ/02 prevent non-panic DKIM error from being sent to paniclog
874 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
877 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
878 Notification from Dr Andrew Aitchison.
880 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
881 ExtendedDetectionInfo response format.
882 Notification from John Horne.
884 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
887 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
888 XSL and documented dependency on system catalogs, with examples of how
891 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
894 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
895 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
896 configuration file which is writeable by the Exim user or group.
898 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
899 of configuration files to cover files specified with the -C option if
900 they are going to be used with root privileges, not just the default
903 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
904 option (effectively making it always true).
906 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
907 files to be used while preserving root privileges.
909 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
910 that rogue child processes cannot use them.
912 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
913 run-time user, instead of root.
915 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
916 Exim run-time user without dropping privileges.
918 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
919 result string, instead of calling string_vformat() twice with the same
922 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
923 for other users. Others should always drop root privileges if they use
924 -C on the command line, even for a whitelisted configure file.
926 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
928 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
934 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
935 $data_path, and $header_path variables; fixed documentation bugs and
938 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
939 exipick to access non-standard spools, including the "frozen" queue
942 NM/01 Bugzilla 965: Support mysql stored procedures.
943 Patch from Alain Williams
945 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
947 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
948 Patch from Andreas Metzler
950 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
951 Patch from Kirill Miazine
953 NM/05 Bugzilla 671: Added umask to procmail example.
955 JJ/03 installed exipick 20100323.0, fixing doc bug
957 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
958 directory. Notification and patch from Dan Rosenberg.
960 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
962 TK/02 Improve log output when DKIM signing operation fails.
964 MH/01 Treat the transport option dkim_domain as a colon separated
965 list, not as a single string, and sign the message with each element,
966 omitting multiple occurences of the same signer.
968 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
969 Bugzilla 985, 986. Patch by Simon Arlott
971 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
972 Patch by Simon Arlott
974 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
975 MBX locking. Notification from Dan Rosenberg.
981 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
983 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
985 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
987 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
989 NM/04 Bugzilla 915: Flag broken perl installation during build.
995 TK/01 Added patch by Johannes Berg that expands the main option
996 "spamd_address" if it starts with a dollar sign.
998 TK/02 Write list of recipients to X-Envelope-Sender header when building
999 the mbox-format spool file for content scanning (suggested by Jakob
1002 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1003 (http://www.dcc-servers.net/) support via dccifd. Activated by
1004 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1006 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1007 by Mark Daniel Reidel <mr@df.eu>.
1009 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1010 When building exim an external PCRE library is now needed -
1011 PCRE is a system library on the majority of modern systems.
1012 See entry on PCRE_LIBS in EDITME file.
1014 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1015 conversation. Added nologin parameter to request.
1016 Patch contributed by Kirill Miazine.
1018 TF/01 Do not log submission mode rewrites if they do not change the address.
1020 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1022 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1023 log files in place. Contributed by Roberto Lima.
1025 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1027 TF/03 Bugzilla 615: When checking the local_parts router precondition
1028 after a local_part_suffix or local_part_prefix option, Exim now
1029 does not use the address's named list lookup cache, since this
1030 contains cached lookups for the whole local part.
1032 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1033 Robert Millan. Documentation is in experimental-spec.txt.
1035 TF/04 Bugzilla 668: Fix parallel build (make -j).
1037 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1039 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1040 Patch provided by Jan Srzednicki.
1042 TF/05 Leading white space used to be stripped from $spam_report which
1043 wrecked the formatting. Now it is preserved.
1045 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1046 that they are available at delivery time.
1048 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1050 TF/08 TLS error reporting now respects the incoming_interface and
1051 incoming_port log selectors.
1053 TF/09 Produce a more useful error message if an SMTP transport's hosts
1054 setting expands to an empty string.
1056 NM/06 Bugzilla 744: EXPN did not work under TLS.
1057 Patch provided by Phil Pennock.
1059 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1060 Patch provided by Richard Godbee.
1062 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1063 acl_smtp_notquit, added index entry.
1065 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1066 Patch provided by Eugene Bujak.
1068 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1069 accept(). Patch provided by Maxim Dounin.
1071 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1072 Patch provided by Phil Pennock.
1074 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1076 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1077 Patch provided by Brad "anomie" Jorsch.
1079 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1080 Patch provided by Dean Brooks.
1082 TK/05 Add native DKIM support (does not depend on external libraries).
1084 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1085 Patch provided by Graeme Fowler.
1087 NM/16 Bugzilla 851: Documentation example syntax fix.
1089 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1091 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1094 NM/19 Bugzilla 745: TLS version reporting.
1095 Patch provided by Phil Pennock.
1097 NM/20 Bugzilla 167: bool: condition support.
1098 Patch provided by Phil Pennock.
1100 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1101 clients. Patch provided by Phil Pennock.
1103 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1104 Patch provided by Brad "anomie" Jorsch.
1106 NM/23 Bugzilla 687: Fix misparses in eximstats.
1107 Patch provided by Heiko Schlittermann.
1109 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1110 Patch provided by Heiko Schlittermann.
1112 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1113 plus update to original patch.
1115 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1117 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1118 Patch provided by David Brownlee.
1120 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1122 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1124 NM/30 Bugzilla 888: TLS documentation bugfixes.
1126 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1128 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1129 Unlike the original bugzilla I have changed all shell scripts in src tree.
1131 NM/33 Bugzilla 898: Transport filter timeout fix.
1132 Patch by Todd Rinaldo.
1134 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1135 Patch by Serge Demonchaux.
1137 NM/35 Bugzilla 39: Base64 decode bug fixes.
1138 Patch by Jakob Hirsch.
1140 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1142 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1144 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1146 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1152 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1153 ALT-N's libdkim that I have put here:
1154 http://duncanthrax.net/exim-experimental/
1156 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1157 were called 'true' and it seems that C99 defines that as a reserved
1158 keyword to be used with 'bool' variable types. That means you could
1159 not include C99-style headers which use bools without triggering
1160 build errors in sieve.c.
1162 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1163 as mailq or other aliases. Changed the --help handling significantly
1164 to do whats expected. exim_usage() emits usage/help information.
1166 SC/01 Added the -bylocaldomain option to eximstats.
1168 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1170 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1172 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1178 PH/01 Another patch from the Sieve maintainer.
1180 PH/02 When an IPv6 address is converted to a string for single-key lookup
1181 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1182 dots are used instead of colons so that keys in lsearch files need not
1183 contain colons. This was done some time before quoting was made available
1184 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1185 (notated using the quote facility) so as to distinguish them from IPv4
1186 keys. This meant that lookups for IP addresses in host lists did not work
1187 for iplsearch lookups.
1189 This has been fixed by arranging for IPv6 addresses to be expressed with
1190 colons if the lookup type is iplsearch. This is not incompatible, because
1191 previously such lookups could never work.
1193 The situation is now rather anomolous, since one *can* have colons in
1194 ordinary lsearch keys. However, making the change in all cases is
1195 incompatible and would probably break a number of configurations.
1197 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1200 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1201 conversion specification without a maximum field width, thereby enabling
1202 a rogue spamd server to cause a buffer overflow. While nobody in their
1203 right mind would setup Exim to query an untrusted spamd server, an
1204 attacker that gains access to a server running spamd could potentially
1205 exploit this vulnerability to run arbitrary code as the Exim user.
1207 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1208 $primary_hostname instead of what libspf2 thinks the hosts name is.
1210 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1211 a directory entry by the name of the lookup key. Previously, if a
1212 symlink pointed to a non-existing file or a file in a directory that
1213 Exim lacked permissions to read, a lookup for a key matching that
1214 symlink would fail. Now it is enough that a matching directory entry
1215 exists, symlink or not. (Bugzilla 503.)
1217 PH/03 The body_linecount and body_zerocount variables are now exported in the
1220 PH/04 Added the $dnslist_matched variable.
1222 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1223 This means they are set thereafter only if the connection becomes
1226 PH/06 Added the client_condition to authenticators so that some can be skipped
1227 by clients under certain conditions.
1229 PH/07 The error message for a badly-placed control=no_multiline_responses left
1230 "_responses" off the end of the name.
1232 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1234 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1235 (without spaces) instead of just copying the configuration text.
1237 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1239 PH/11 Added $max_received_linelength.
1241 PH/12 Added +ignore_defer and +include_defer to host lists.
1243 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1244 way in which PCRE > 7.0 is built.
1246 PH/14 Implemented queue_only_load_latch.
1248 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1249 MAIL command. The effect was to mangle the value on 64-bit systems.
1251 PH/16 Another patch from the Sieve maintainer.
1253 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1255 PH/18 If a system quota error occurred while trying to create the file for
1256 a maildir delivery, the message "Mailbox is full" was not appended to the
1257 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1258 only to a quota excession during the actual writing of the file.
1260 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1261 characters?) which causes problems in log lines. The DN values are now
1262 passed through string_printing() before being added to log lines.
1264 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1265 and InterBase are left for another time.)
1267 PH/21 Added message_body_newlines option.
1269 PH/22 Guard against possible overflow in moan_check_errorcopy().
1271 PH/23 POSIX allows open() to be a macro; guard against that.
1273 PH/24 If the recipient of an error message contained an @ in the local part
1274 (suitably quoted, of course), incorrect values were put in $domain and
1275 $local_part during the evaluation of errors_copy.
1281 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1282 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1285 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1286 issue a MAIL command.
1288 PH/02 In an ACL statement such as
1290 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1292 if a client was not listed at all, or was listed with a value other than
1293 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1294 the condition was not true (as it should be), so access was not denied.
1295 The bug was that the ! inversion was incorrectly passed on to the second
1296 item. This has been fixed.
1298 PH/03 Added additional dnslists conditions == and =& which are different from
1299 = and & when the dns lookup returns more than one IP address.
1301 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1302 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1304 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1305 FSYNC, which compiles an option called disable_fsync that allows for
1306 bypassing fsync(). The documentation is heavily laced with warnings.
1308 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1310 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1311 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1312 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1313 including adding "make clean"; (3) Added -fPIC when compiling the test
1314 dynamically loaded module, to get rid of a warning.
1316 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1317 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1318 = 0s. The bug is otherwise harmless.
1320 PH/07 There was a bug in the dovecot authenticator such that the value of
1321 $auth1 could be overwritten, and so not correctly preserved, after a
1322 successful authentication. This usually meant that the value preserved by
1323 the server_setid option was incorrect.
1325 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1327 PH/09 Installed PCRE release 7.0.
1329 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1330 run for batched SMTP input. It is now run at the start of every message
1331 in the batch. While fixing this I discovered that the process information
1332 (output by running exiwhat) was not always getting set for -bs and -bS
1333 input. This is fixed, and it now also says "batched" for BSMTP.
1335 PH/11 Added control=no_pipelining.
1337 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1338 patch, slightly modified), and move the expansion of helo_data till after
1339 the connection is made in the smtp transport (so it can use these
1342 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1344 PH/14 Added log_selector = +pid.
1346 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1348 PH/16 Add ${if forany and ${if forall.
1350 PH/17 Added dsn_from option to vary the From: line in DSNs.
1352 PH/18 Flush SMTP output before performing a callout, unless control =
1353 no_callout_flush is set.
1355 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1356 was true (the default) a successful delivery failed to delete the retry
1357 item, thus causing premature timeout of the address. The bug is now
1360 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1362 PH/21 Long custom messages for fakedefer and fakereject are now split up
1363 into multiline reponses in the same way that messages for "deny" and
1364 other ACL rejections are.
1366 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1367 with slight modification.
1369 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1370 draft, changing the syntax and factoring some duplicate code".
1372 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1373 for deliveries of the second and subsequent messages over the same SMTP
1376 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1377 ${reduce, with only minor "tidies".
1379 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1381 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1382 expansion side effects.
1384 PH/27 When a message times out after an over-quota error from an Exim-imposed
1385 quota, the bounce message says "mailbox is full". This message was not
1386 being given when it was a system quota that was exceeded. It now should
1389 MH/03 Made $recipients available in local_scan(). local_scan() already has
1390 better access to the recipient list through recipients_list[], but
1391 $recipients can be useful in postmaster-provided expansion strings.
1393 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1394 in the case of a MAIL command with additional options following the
1395 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1396 were accidentally chopped off.
1398 PH/29 SMTP synchronization checks are implemented when a command is read -
1399 there is a check that no more input is waiting when there shouldn't be
1400 any. However, for some commands, a delay in an ACL can mean that it is
1401 some time before the response is written. In this time, more input might
1402 arrive, invalidly. So now there are extra checks after an ACL has run for
1403 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1404 pipelining has not been advertised.
1406 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1408 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1409 returned in the SMTP response when smtp_return_error_details was false.
1410 This has been fixed.
1412 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1413 instead of the C I/O that was originally supplied, because problems were
1414 reported on Solaris.
1416 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1417 Exim which did not show up earlier: it was assuming that a call to
1418 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1419 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1420 was a macro that became an assignment, so it seemed to work. This has
1421 changed to a proper function call with a void return, hence the compile
1422 error. Exim's code has been fixed.
1424 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1427 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1429 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1431 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1432 because it turns out to be too big or there is a timeout) there is no
1433 "Completed" line in the log. When some messages of this type were
1434 selected by exigrep, they were listed as "not completed". Others were
1435 picked up by some special patterns. I have improved the selection
1436 criteria to be more general.
1438 PH/38 The host_find_failed option in the manualroute router can now be set
1439 to "ignore", to completely ignore a host whose IP address cannot be
1440 found. If all hosts are ignored, the behaviour is controlled by the new
1441 host_all_ignored option.
1443 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1444 homing or because of multiple MX records with /mx) generated more than
1445 one IP address, and the following item turned out to be the local host,
1446 all the secondary addresses of the first item were incorrectly removed
1447 from the list, along with the local host and any following hosts (which
1448 is what is supposed to happen).
1450 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1451 whoever called Exim into the -H file. In the case of the daemon it was
1452 behaving confusingly. When first started, it used values for whoever
1453 started the daemon, but after a SIGHUP it used the Exim user (because it
1454 calls itself on a restart). I have changed the code so that it now always
1457 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1458 message are rejected with the same error (e.g. no authentication or bad
1459 sender address), and a DATA command is nevertheless sent (as can happen
1460 with PIPELINING or a stupid MUA), the error message that was given to the
1461 RCPT commands is included in the rejection of the DATA command. This is
1462 intended to be helpful for MUAs that show only the final error to their
1465 PH/42 Another patch from the Sieve maintainer.
1467 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1468 Eximstats - Fixed some broken HTML links and added missing column headers
1470 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1471 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1473 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1475 PH/43 Yet another patch from the Sieve maintainer.
1477 PH/44 I found a way to check for a TCP/IP connection going away before sending
1478 the response to the final '.' that terminates a message, but only in the
1479 case where the client has not sent further data following the '.'
1480 (unfortunately, this is allowed). However, in many cases there won't be
1481 any further data because there won't be any more messages to send. A call
1482 to select() can be used: if it shows that the input is "ready", there is
1483 either input waiting, or the socket has been closed. An attempt to read
1484 the next input character can distinguish the two cases. Previously, Exim
1485 would have sent an OK response which the client would never have see.
1486 This could lead to message repetition. This fix should cure that, at
1487 least in a lot of common cases.
1489 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1490 advertised in response to EHLO.
1496 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1497 fixed by 4.65/MH/01 (is this a record?) are fixed:
1499 (i) An empty string was always treated as zero by the numeric comparison
1500 operators. This behaviour has been restored.
1502 (ii) It is documented that the numeric comparison operators always treat
1503 their arguments as decimal numbers. This was broken in that numbers
1504 starting with 0 were being interpreted as octal.
1506 While fixing these problems I realized that there was another issue that
1507 hadn't been noticed. Values of message_size_limit (both the global option
1508 and the transport option) were treated as octal if they started with 0.
1509 The documentation was vague. These values are now always treated as
1510 decimal, and I will make that clear in the documentation.
1516 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1517 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1520 MH/01 Don't check that the operands of numeric comparison operators are
1521 integers when their expansion is in "skipping" mode (fixes bug
1522 introduced by 4.64-PH/07).
1524 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1525 child addresses, Exim now panics and dies. Previously, because the count
1526 is held in a short int, deliveries were likely to be lost. As such a
1527 large number of recipients for a single message is ridiculous
1528 (performance will be very, very poor), I have chosen to impose a limit
1529 rather than extend the field.
1535 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1536 leftover -K file (the existence of which was triggered by #402).
1537 While we were at it, introduced process PID as part of the -K
1538 filename. This should rule out race conditions when creating
1541 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1542 processing considerably. Previous code took too long for large mails,
1543 triggering a timeout which in turn triggers #401.
1545 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1546 in the DK code in transports.c. sendfile() is not really portable,
1547 hence the _LINUX specificness.
1549 TF/01 In the add_headers option to the mail command in an Exim filter,
1550 there was a bug that Exim would claim a syntax error in any
1551 header after the first one which had an odd number of characters
1554 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1555 callout verification, Exim cached a "reject" for the entire domain. This
1556 is correct for most verifications, but it is not correct for a recipient
1557 verification with use_sender or use_postmaster set, because in that case
1558 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1559 case of MAIL FROM:<> rejection from other early rejections (e.g.
1560 rejection of HELO). When verifying a recipient using a non-null MAIL
1561 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1562 Whatever the result of the callout, the value of the domain cache is
1563 left unchanged (for any other kind of callout, getting as far as trying
1564 RCPT means that the domain itself is ok).
1566 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1569 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1570 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1571 session, the daemon ignores ECONNECT errors and logs others; it now
1572 ignores EPIPE as well.
1574 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1575 (quoted-printable decoding).
1577 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1578 later the small subsequent patch to fix an introduced bug.
1580 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1582 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1584 PH/08 An error is now given if message_size_limit is specified negative.
1586 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1587 to be given (somewhat) arbitrary names.
1589 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1592 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1593 miscellaneous code fixes
1595 PH/10 Added the log_reject_target ACL modifier to specify where to log
1598 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1599 hostname. This is wrong, because it relates to the incoming message (and
1600 probably the interface on which it is arriving) and not to the outgoing
1601 callout (which could be using a different interface). This has been
1602 changed to use the value of the helo_data option from the smtp transport
1603 instead - this is what is used when a message is actually being sent. If
1604 there is no remote transport (possible with a router that sets up host
1605 addresses), $smtp_active_hostname is used.
1607 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1608 tweaks were necessary in order to get it to work (see also 21 below):
1609 (a) The code assumed that strncpy() returns a negative number on buffer
1610 overflow, which isn't the case. Replaced with Exim's string_format()
1612 (b) There were several signed/unsigned issues. I just did the minimum
1613 hacking in of casts. There is scope for a larger refactoring.
1614 (c) The code used strcasecmp() which is not a standard C function.
1615 Replaced with Exim's strcmpic() function.
1616 (d) The code set only $1; it now sets $auth1 as well.
1617 (e) A simple test gave the error "authentication client didn't specify
1618 service in request". It would seem that Dovecot has changed its
1619 interface. Fortunately there's a specification; I followed it and
1620 changed what the client sends and it appears to be working now.
1622 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1625 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1626 address A is aliased to B and C, where B exists and C does not. Without
1627 -v the output is "A verified" because verification stops after a
1628 successful redirection if more than one address is generated. However,
1629 with -v the child addresses are also verified. Exim was outputting "A
1630 failed to verify" and then showing the successful verification for C,
1631 with its parentage. It now outputs "B failed to verify", showing B's
1632 parentage before showing the successful verification of C.
1634 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1635 look up a TXT record in a specific list after matching in a combined
1638 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1639 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1640 they consult the DNS. I had assumed they would set it the way they
1641 wanted; and indeed my experiments on Linux seem to show that in some
1642 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1643 To be on the safe side, however, I have now made the interface to
1644 host_find_byname() similar to host_find_bydns(), with an argument
1645 containing the DNS resolver options. The host_find_byname() function now
1646 sets these options at its start, just as host_find_bydns() does. The smtp
1647 transport options dns_qualify_single and dns_search_parents are passed to
1648 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1649 of host_find_byname() use the default settings of RES_DEFNAMES
1650 (qualify_single) but not RES_DNSRCH (search_parents).
1652 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1653 spool_read_header() do less string testing, by means of a preliminary
1654 switch on the second character of optional "-foo" lines. (This is
1655 overdue, caused by the large number of possibilities that now exist.
1656 Originally there were few.) While I was there, I also converted the
1657 str(n)cmp tests so they don't re-test the leading "-" and the first
1658 character, in the hope this might squeeze out yet more improvement.
1660 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1661 flag allowing group syntax was set by the header_syntax check but not
1662 turned off, possible causing trouble later; (2) The flag was not being
1663 set at all for the header_verify test, causing "group"-style headers to
1664 be rejected. I have now set it in this case, and also caused header_
1665 verify to ignore an empty address taken from a group. While doing this, I
1666 came across some other cases where the code for allowing group syntax
1667 while scanning a header line wasn't quite right (mostly, not resetting
1668 the flag correctly in the right place). These bugs could have caused
1669 trouble for malformed header lines. I hope it is now all correct.
1671 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1672 with the "reply" argument non-NULL. The code, however (which originally
1673 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1674 but it didn't always do it. This confused somebody who was copying the
1675 code for some other use. I have removed all the tests.
1677 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1678 feature that was used to support insecure browsers during the U.S. crypto
1679 embargo. It requires special client support, and Exim is probably the
1680 only MTA that supported it -- and would never use it because real RSA is
1681 always available. This code has been removed, because it had the bad
1682 effect of slowing Exim down by computing (never used) parameters for the
1683 RSA_EXPORT functionality.
1685 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1686 authenticator to fail if there's a tab character in the incoming data
1687 (there should never be unless someone is messing about, as it's supposed
1688 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1689 if the connection is using TLS or if the remote IP is the same as the
1690 local IP, and the "valid-client-cert option" if a client certificate has
1693 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1694 authenticators. This can be used for authorization after authentication
1695 succeeds. (In the case of plaintext, it servers for both authentication
1698 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1699 if any retry times were supplied.
1701 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1702 connection, where there is no client IP address to check. In this
1703 situation, the verify now always succeeds.
1705 PH/25 Applied John Jetmore's -Mset patch.
1707 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1709 PH/27 In a string expansion for a processed (not raw) header when multiple
1710 headers of the same name were present, leading whitespace was being
1711 removed from all of them, but trailing whitespace was being removed only
1712 from the last one. Now trailing whitespace is removed from each header
1713 before concatenation. Completely empty headers in a concatenation (as
1714 before) are ignored.
1716 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1717 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1719 PH/29 [Removed. This was a change that I later backed out, and forgot to
1720 correct the ChangeLog entry (that I had efficiently created) before
1721 committing the later change.]
1723 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1724 address errors (4xx response to RCPT) over the same connection as other
1725 messages routed to the same hosts. Such deliveries are always "forced",
1726 so retry times are not inspected. This resulted in far too many retries
1727 for the affected addresses. The effect occurred only when there were more
1728 hosts than the hosts_max_try setting in the smtp transport when it had
1729 the 4xx errors. Those hosts that it had tried were not added to the list
1730 of hosts for which the message was waiting, so if all were tried, there
1731 was no problem. Two fixes have been applied:
1733 (i) If there are any address or message errors in an SMTP delivery, none
1734 of the hosts (tried or untried) are now added to the list of hosts
1735 for which the message is waiting, so the message should not be a
1736 candidate for sending over the same connection that was used for a
1737 successful delivery of some other message. This seems entirely
1738 reasonable: after all the message is NOT "waiting for some host".
1739 This is so "obvious" that I'm not sure why it wasn't done
1740 previously. Hope I haven't missed anything, but it can't do any
1741 harm, as the worst effect is to miss an optimization.
1743 (ii) If, despite (i), such a delivery is accidentally attempted, the
1744 routing retry time is respected, so at least it doesn't keep
1745 hammering the server.
1747 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1748 in ${readsocket because some servers need this prod.
1750 PH/32 Added some extra debug output when updating a wait-xxx database.
1752 PH/33 The hint "could be header name not terminated by colon", which has been
1753 given for certain expansion errors for a long time, was not being given
1754 for the ${if def:h_colon_omitted{... case.
1756 PH/34 The spec says: "With one important exception, whenever a domain list is
1757 being scanned, $domain contains the subject domain." There was at least
1758 one case where this was not true.
1760 PH/35 The error "getsockname() failed: connection reset by peer" was being
1761 written to the panic log as well as the main log, but it isn't really
1762 panic-worthy as it just means the connection died rather early on. I have
1763 removed the panic log writing for the ECONNRESET error when getsockname()
1766 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1767 runs only) independently of the message's sender address. This meant
1768 that, if the 4xx error was in fact related to the sender, a different
1769 message to the same recipient with a different sender could confuse
1770 things. In particualar, this can happen when sending to a greylisting
1771 server, but other circumstances could also provoke similar problems.
1772 I have changed the default so that the retry time for these errors is now
1773 based a combination of the sender and recipient addresses. This change
1774 can be overridden by setting address_retry_include_sender=false in the
1777 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1778 remote server are returned as part of bounce messages. This was not
1779 happening for LMTP over a pipe (the lmtp transport), but now it is the
1780 same for both kinds of LMTP.
1782 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1783 in header lines that were in fact CNAMEs. This is no longer the case.
1785 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1786 and queue runs started by the daemon processed all messages. This has
1787 been fixed so that -R and -S can now usefully be given with -q<time>.
1789 PH/40 Import PCRE release 6.7 (fixes some bugs).
1791 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1793 PH/42 Give an error if -q is specified more than once.
1795 PH/43 Renamed the variables $interface_address and $interface_port as
1796 $received_ip_address and $received_port, to make it clear that these
1797 values apply to message reception, and not to the outgoing interface when
1798 a message is delivered. (The old names remain recognized, of course.)
1800 PH/44 There was no timeout on the connect() call when using a Unix domain
1801 socket in the ${readsocket expansion. There now is.
1803 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1804 be meaningful with "accept".
1806 SC/01 Eximstats V1.43
1807 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1809 SC/02 Eximstats V1.44
1810 Use a glob alias rather than an array ref in the generated
1811 parser. This improves both readability and performance.
1813 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1814 Collect SpamAssassin and rejection statistics.
1815 Don't display local sender or destination tables unless
1816 there is data to show.
1817 Added average volumes into the top table text output.
1819 SC/04 Eximstats V1.46
1820 Collect data on the number of addresses (recipients)
1821 as well as the number of messages.
1823 SC/05 Eximstats V1.47
1824 Added 'Message too big' to the list of mail rejection
1825 reasons (thanks to Marco Gaiarin).
1827 SC/06 Eximstats V1.48
1828 Mainlog lines which have GMT offsets and are too short to
1829 have a flag are now skipped.
1831 SC/07 Eximstats V1.49 (Alain Williams)
1832 Added the -emptyok flag.
1834 SC/08 Eximstats V1.50
1835 Fixes for obtaining the IP address from reject messages.
1837 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1838 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1839 whitesspace changes from 4.64-PH/27
1841 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1844 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1845 are found, allow negative numbers in numeric criteria)
1847 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1849 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1850 to match changes made in 4.64-PH/43
1852 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1854 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1855 "domain" is actually an IP address, and force a failure. This locks out
1856 those revolvers/nameservers that support "A-for-A" lookups, in
1857 contravention of the specifications.
1859 PH/48 When a host name was looked up from an IP address, and the subsequent
1860 forward lookup of the name timed out, the host name was left in
1861 $sender_host_name, contrary to the specification.
1863 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1864 restricted to single-key lookups, Exim was not diagnosing an error if
1865 * or *@ was used with a query-style lookup.
1867 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1869 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1870 long ago, but noone interested enough thought of it. Let's just say that
1871 the "1.1" means that there are some new functions that weren't there at
1872 some point in the past.
1874 PH/51 Error processing for expansion failure of helo_data from an smtp
1875 transport during callout processing was broken.
1877 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1878 tested/used via the -bh/-bhc/-bs options.
1880 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1881 bug, fixed in subsequent PCRE releases).
1883 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1884 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1886 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1892 SC/01 Use a glob alias rather than an array ref in eximstats generated
1893 parser. This improves both readability and performance.
1895 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1896 Don't display local sender or destination tables in eximstats unless
1897 there is data to show.
1898 Added average volumes into the eximstats top table text output.
1900 SC/03 Collect data on the number of addresses (recipients) as well
1901 as the number of messages in eximstats.
1903 TF/01 Correct an error in the documentation for the redirect router. Exim
1904 does (usually) call initgroups() when daemonizing.
1906 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1907 with consistent privilege compared to when running as a daemon.
1909 TF/03 Note in the spec that $authenticated_id is not set for local
1910 submissions from trusted users.
1912 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1913 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1915 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1916 by adding some example configuration directives to the default
1917 configuration file. A little bit of work is required to uncomment the
1918 directives and define how usernames and passwords are checked, but
1919 there is now a framework to start from.
1921 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1922 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1923 without this. I don't know how relevant this is to other LDAP libraries.
1925 PH/02 Add the verb name to the "unknown ACL verb" error.
1927 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1929 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1931 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1932 directory not expanded when it should be if an expanded home directory
1933 was set for the address (which is overridden by the transport).
1935 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1938 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1939 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1940 because it is too late at that time, and has no effect.
1942 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1943 security issue with \' (bugzilla #107). I could not use the
1944 PQescapeStringConn() function, because it needs a PGconn value as one of
1947 PH/08 When testing addresses using -bt, indicate those final addresses that
1948 are duplicates that would not cause an additional delivery. At least one
1949 person was confused, thinking that -bt output corresponded to deliveries.
1950 (Suppressing duplicates isn't a good idea as you lose the information
1951 about possibly different redirections that led to the duplicates.)
1953 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1954 systems where poll() doesn't work, in particular OS X.
1956 PH/10 Added more information to debugging output for retry time not reached.
1958 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1959 operations in malware.c.
1961 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1964 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1965 syslog_duplication set false, log lines that would normally be written
1966 both the the main log and to the reject log were not written to syslog at
1969 PH/14 In the default configuration, change the use of "message" in ACL warn
1970 statements to "add_header".
1972 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1973 not followed by a command (e.g. "seen endif").
1975 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1976 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1979 PH/17 Added extra conditions to the default value of delay_warning_condition
1983 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1984 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1985 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1988 The Auto-Submitted: and various List- headers are standardised, whereas I
1989 don't think Precedence: ever was.
1991 PH/18 Refactored debugging code in route_finduser() to show more information,
1992 in particular, the error code if getpwnam() issues one.
1994 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1995 This is apparently needed in addition to the PH/07 change above to avoid
1996 any possible encoding problems.
1998 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1999 but not after initializing Perl.
2001 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2002 output them only if debugging. By default they are written stderr,
2003 apparently, which is not desirable.
2005 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2008 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2011 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2013 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2014 authenticated or an ident call has been made. Suppress the default
2015 values for $authenticated_id and $authenticated_sender (but permit -oMai
2016 and -oMas) when testing with -bh.
2018 PH/24 Re-jigged the order of the tests in the default configuration so that the
2019 tests for valid domains and recipients precede the DNS black list and CSA
2020 tests, on the grounds that those ones are more expensive.
2022 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2023 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2024 This bug exists in every version of Exim that I still have, right back to
2027 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2028 However, an attempt to turn on case-sensitivity in a regex key by
2029 including (?-i) didn't work because the subject string was already
2030 lowercased, and the effects were non-intuitive. It turns out that a
2031 one-line patch can be used to allow (?-i) to work as expected.
2037 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2038 other effects) broke the use of negated acl sub-conditions.
2040 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2043 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2044 "Deny" causes Exim to reject the incoming connection with a 554 error.
2045 Unfortunately, if there is a major crisis, such as a disk failure,
2046 tcp-wrappers gives "deny", whereas what one would like would be some
2047 kind of temporary error. A kludge has been added to help with this.
2048 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2049 554 error is used if errno is still zero or contains ENOENT (which occurs
2050 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2053 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2055 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2056 errors. Otherwise a message that provokes a temporary error (when other
2057 messages do not) can cause a whole host to time out.
2059 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2060 addresses were routed directly to files or pipes from a redirect router.
2061 File deliveries just didn't batch; pipe deliveries might have suffered
2064 PH/06 A failure to get a lock for a hints database would erroneously always say
2065 "Failed to get write lock", even when it was really a read lock.
2067 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2068 of 0600. This has been changed to use the value of the lockfile_mode
2069 option (which defaults to 0600).
2071 PH/08 Applied small patch from the Sieve maintainer.
2073 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2074 folder from quota calculations, a direct delivery into this folder messed
2075 up the contents of the maildirsize file. This was because the regex was
2076 used only to exclude .Trash (or whatever) when the size of the mailbox
2077 was calculated. There was no check that a delivery was happening into an
2078 excluded directory. This bug has been fixed by ignoring all quota
2079 processing for deliveries into excluded directories.
2081 PH/10 Added the maildirfolder_create_regex option to appendfile.
2087 PH/01 The code for finding all the local interface addresses on a FreeBSD
2088 system running IPv6 was broken. This may well have applied to all BSD
2089 systems, as well as to others that have similar system calls. The broken
2090 code found IPv4 interfaces correctly, but gave incorrect values for the
2091 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2092 that it would not match correctly against @[] and not recognize the IPv6
2095 PH/02 The ipliteral router was not recognizing addresses of the form user@
2096 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2098 PH/03 Added disable_ipv6.
2100 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2101 decoded form, because it is most often used to construct To: headers
2102 lines in autoreplies, and the decoded form may well be syntactically
2103 invalid. However, $reply_address has leading white space removed, and all
2104 newlines turned into spaces so that the autoreply transport does not
2107 PH/05 If group was specified without a user on a router, and no group or user
2108 was specified on a transport, the group from the router was ignored.
2110 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2111 for visible compile-time settings that can be used to change these
2112 numbers, for those that want even more. Backwards compatibility with old
2113 spool files has been maintained. However, going back to a previous Exim
2114 release will lost any variables that are in spool files.
2116 PH/07 Two small changes when running in the test harness: increase delay when
2117 passing a TCP/IP connection to a new process, in case the original
2118 process has to generate a bounce, and remove special handling of
2119 127.0.0.2 (sic), which is no longer necessary.
2121 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2122 be the same on different OS.
2124 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2127 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2128 whether --show-vars was specified or not
2130 JJ/02 exipick: Added support for new ACL variable spool format introduced
2133 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2134 syntactically invalid From: or Reply-to: line, and a filter used this to
2135 generate an autoreply, and therefore failed to obtain an address for the
2136 autoreply, Exim could try to deliver to a non-existent relative file
2137 name, causing unrelated and misleading errors. What now happens is that
2138 it logs this as a hard delivery error, but does not attempt to create a
2141 PH/11 The exinext utility has a -C option for testing purposes, but although
2142 the given file was scanned by exinext itself; it wasn't being passed on
2143 when Exim was called.
2145 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2146 an end-of-file indication when reading a command response.
2148 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2149 compiled. In many other places in Exim, IPv6 addresses are always
2150 recognized, so I have changed this. It also means that IPv4 domain
2151 literals of the form [IPV4:n.n.n.n] are now always recognized.
2153 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2154 used if the router is not running as root, for example, when verifying at
2155 ACL time, or when using -bh. The debugging output from this situation was
2156 non-existent - all you got was a failure to exec. I have made two
2159 (a) Failures to set uid/gid, the current directory, or a process leader
2160 in a subprocess such as that created by queryprogram now generate
2161 suitable debugging ouput when -d is set.
2163 (b) The queryprogram router detects when it is not running as root,
2164 outputs suitable debugging information if -d is set, and then runs
2165 the subprocess without attempting to change uid/gid.
2167 PH/15 Minor change to Makefile for building test_host (undocumented testing
2170 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2171 additional section of a DNS packet that returns MX or SRV records.
2172 Instead, it always explicitly searches for A/AAAA records. This avoids
2173 major problems that occur when a DNS server includes only records of one
2174 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2175 fixed another bug: if SRV records were looked up and the corresponding
2176 address records were *not* found in the additional section, the port
2177 values from the SRV records were lost.
2179 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2180 using the correct key (the original address) when searching the retry
2181 rules in order to find which one to use for generating the retry hint.
2183 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2184 adding the default one. Similarly, if it contains a Reply-To: header, the
2185 errors_reply_to option, if set, is not used.
2187 PH/19 When calculating a retry time, Exim used to measure the "time since
2188 failure" by looking at the "first failed" field in the retry record. Now
2189 it does not use this if it is later than than the arrival time of the
2190 message. Instead it uses the arrival time. This makes for better
2191 behaviour in cases where some deliveries succeed, thus re-setting the
2192 "first failed" field. An example is a quota failure for a huge message
2193 when small messages continue to be delivered. Without this change, the
2194 "time since failure" will always be short, possible causing more frequent
2195 delivery attempts for the huge message than are intended.
2196 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2198 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2199 $1, $2, $3) because the numerical variables can be reset during some
2200 expansion items (e.g. "match"), thereby losing the authentication data.
2202 PH/21 Make -bV show the size of off_t variables so that the test suite can
2203 decide whether to run tests for quotas > 2G.
2205 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2206 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2207 filecount value is greater than 2G or if a quota value is greater than 2G
2208 on a system where the size of off_t is not greater than 4, a panic error
2211 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2212 never match. The debug and -bh output now contains an explicit error
2213 message indicating a malformed IPv4 address or mask.
2215 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2216 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2217 PH/23 above applies.
2219 PH/25 Do not write to syslog when running in the test harness. The only
2220 occasion when this arises is a failure to open the main or panic logs
2221 (for which there is an explicit test).
2223 PH/26 Added the /no_tell option to "control=freeze".
2225 PH/27 If a host name lookup failed very early in a connection, for example, if
2226 the IP address matched host_lookup and the reverse lookup yielded a name
2227 that did not have a forward lookup, an error message of the form "no IP
2228 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2229 could be logged. Now it outputs the IP address instead of "NULL".
2231 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2232 allows the sender and the authenticated sender to be set when
2233 submitting a message from within Exim. Since child_open_exim() is
2234 documented for local_scan(), the new function should be too.
2236 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2237 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2238 results in an empty string is now treated as unset.
2240 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2242 PH/31 Added sender_verify_fail logging option.
2244 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2245 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2246 tidied the source and removed it altogether.
2248 PH/33 When a queue run was abandoned because the load average was too high, a
2249 log line was always written; now it is written only if the queue_run log
2250 selector is set. In addition, the log line for abandonment now contains
2251 information about the queue run such as the pid. This is always present
2252 in "start" and "stop" lines but was omitted from the "abandon" line.
2254 PH/34 Omit spaces between a header name and the colon in the error message that
2255 is given when verify = headers_syntax fails (if there are lots of them,
2256 the message gets confusing).
2258 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2259 names, as there are now some PTR records that contain slashes. This check
2260 is only to protect against broken name servers that fall over on strange
2261 characters, so the fact that it applies to all lookups doesn't matter.
2263 PH/36 Now that the new test suite is complete, we can remove some of the
2264 special code in Exim that was needed for the old test suite. For example,
2265 sorting DNS records because real resolvers return them in an arbitrary
2266 order. The new test suite's fake resolver always returns records in the
2269 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2270 bounces) except when queue_only is set, to avoid logging races between
2271 the different processes.
2273 PH/38 Panic-die if .include specifies a non-absolute path.
2275 PH/39 A tweak to the "H" retry rule from its user.
2277 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2278 a label. They prevented compilation on older perls.
2280 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2281 a warning to be raised on newish perls.
2283 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2284 on queue. Changes to match documented behaviour of showing count of
2285 messages matching specified criteria.
2287 PH/40 Changed the default ident timeout from 30s to 5s.
2289 PH/41 Added support for the use of login_cap features, on those BSD systems
2290 that have them, for controlling the resources used by pipe deliveries.
2292 PH/42 The content-scanning code uses fopen() to create files in which to put
2293 message data. Previously it was not paying any attention to the mode of
2294 the files. Exim runs with umask(0) because the rest of the code creates
2295 files with open(), and sets the required mode explicitly. Thus, these
2296 files were ending up world-writeable. This was not a big issue, because,
2297 being within the spool directory, they were not world-accessible. I have
2298 created a function called modefopen, which takes an additional mode
2299 argument. It sets umask(777), creates the file, chmods it to the required
2300 mode, then resets the umask. All the relevant calls to fopen() in the
2301 content scanning code have been changed to use this function.
2303 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2304 to 24 hours. This avoids potential overflow problems when processing G
2305 and H retry rules. I suspect nobody ever tinkers with this value.
2307 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2309 PH/45 When the plaintext authenticator is running as a client, the server's
2310 challenges are checked to ensure they are valid base64 strings. By
2311 default, the authentication attempt is cancelled if an invalid string is
2312 received. Setting client_ignore_invalid_base64 true ignores these errors.
2313 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2314 they are received. Thus, the responses can be made to depend on the
2315 challenges. If an invalid string is ignored, an empty string is placed in
2318 PH/46 Messages that are created by the autoreply transport now contains a
2319 References: header, in accordance with RFCs 2822 and 3834.
2321 PH/47 Added authenticated_sender_force to the smtp transport.
2323 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2325 PH/49 Installed latest patch from the Sieve maintainer.
2327 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2328 was also set, the appendfile transport was unnecessarily scanning a
2329 directory of message files (e.g. for maildir delivery) to find the count
2330 of files (along with the size), even though it did not need this
2331 information. It now does the scan only if it needs to find either the
2332 size of the count of files.
2334 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2336 PH/52 Two bugs concerned with error handling when the smtp transport is
2339 (i) Exim was not creating retry information for temporary errors given
2340 for individual recipients after the DATA command when the smtp transport
2341 was used in LMTP mode. This meant that they could be retried too
2342 frequently, and not timed out correctly.
2344 (ii) Exim was setting the flag that allows error details to be returned
2345 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2346 recipients that were returned after the DATA command.
2348 PH/53 This is related to PH/52, but is more general: for any failing address,
2349 when detailed error information was permitted to be returned to the
2350 sender, but the error was temporary, then after the final timeout, only
2351 "retry timeout exceeded" was returned. Now it returns the full error as
2352 well as "retry timeout exceeded".
2354 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2355 do this, and (what is worse) MTAs that accept it.
2357 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2358 will now be deprecated.
2360 PH/56 New os.c-cygwin from the Cygwin maintainer.
2362 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2363 formats (previously only available in exim formats via -bpr, -bpru,
2364 and -bpra. Now also available in native and exiqgrep formats)
2366 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2367 with very large, slow to parse queues
2369 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2371 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2373 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2374 responses to authentication challenges, though it was showing the
2375 challenges; (ii) I've removed the CR characters from the debug output for
2378 PH/58 Allow for the insertion of a newline as well as a space when a string
2379 is turned into more than one encoded-word during RFC 2047 encoding. The
2380 Sieve code now uses this.
2382 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2383 data_4xx, lost_connection, tls_required.
2385 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2386 message was being sent as an SMTP response.
2388 PH/61 Add -l and -k options to exicyclog.
2390 PH/62 When verifying, if an address was redirected to one new address, so that
2391 verification continued, and the new address failed or deferred after
2392 having set something in $address_data, the value of $address_data was not
2393 passed back to the ACL. This was different to the case when no
2394 redirection occurred. The value is now passed back in both cases.
2396 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2397 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2398 use login_cap.h, so on its own it isn't the distinguishing feature. The
2399 new name refers directly to the setclassresources() function.
2401 PH/65 Added configuration files for NetBSD3.
2403 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2405 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2406 is preferred over IPv4.
2408 PH/68 The bounce_return_message and bounce_return_body options were not being
2409 honoured for bounces generated during the reception of non-SMTP messages.
2410 In particular, this applied to messages rejected by the ACL. This bug has
2411 been fixed. However, if bounce_return_message is true and bounce_return_
2412 body is false, the headers that are returned for a non-SMTP message
2413 include only those that have been read before the error was detected.
2414 (In the case of an ACL rejection, they have all been read.)
2416 PH/69 The HTML version of the specification is now built in a directory called
2417 spec_html instead of spec.html, because the latter looks like a path with
2418 a MIME-type, and this confuses some software.
2420 PH/70 Catch two compiler warnings in sieve.c.
2422 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2423 function verify_get_ident() calls ip_connect() to connect a socket, but
2424 if the "connect()" function timed out, ip_connect() used to close the
2425 socket. However, verify_get_ident() also closes the socket later, and in
2426 between Exim writes to the log, which may get opened at this point. When
2427 the socket was closed in ip_connect(), the log could get the same file
2428 descriptor number as the socket. This naturally causes chaos. The fix is
2429 not to close the socket in ip_connect(); the socket should be closed by
2430 the function that creates it. There was only one place in the code where
2431 this was missing, in the iplookup router, which I don't think anybody now
2432 uses, but I've fixed it anyway.
2434 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2435 well as to direct DNS lookups. Otherwise the handling of names in host
2436 lists is inconsistent and therefore confusing.
2442 PH/01 Two changes to the default runtime configuration:
2444 (1) Move the checks for relay_from_hosts and authenticated clients from
2445 after to before the (commented out) DNS black list checks.
2447 (2) Add control=submission to the relay_from_hosts and authenticated
2448 clients checks, on the grounds that messages accepted by these
2449 statements are most likely to be submissions.
2451 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2453 (1) Generate an error if the third argument for the ${prvs expansion is
2456 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2459 (3) Reset the variables that are obtained from the first argument of
2460 ${prvscheck and used in the second argument before leaving the code,
2461 because their memory is reclaimed, so using them afterwards may do
2464 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2465 one (it's much easier than Tom thought :-).
2467 (5) Because of (4), we can now allow for the use of $prvscheck_result
2468 inside the third argument.
2470 PH/03 For some reason, the default setting of PATH when running a command from
2471 a pipe transport was just "/usr/bin". I have changed it to
2474 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2475 anything to be listed in the output from -bV.
2477 PH/05 When a filter generated an autoreply, the entire To: header line was
2478 quoted in the delivery log line, like this:
2480 => >A.N.Other <ano@some.domain> <original@ddress> ...
2482 This has been changed so that it extracts the operative address. There
2483 may be more than one such address. If so, they are comma-separated, like
2486 => >ano@some.domain,ona@other.domain <original@ddress> ...
2488 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2489 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2490 not being looked up in the rDNS to get a host name, Exim was showing the
2491 IP address twice in Received: lines, even though the IP addresses were
2492 identical. For example:
2494 Received: from [1.2.3.4] (helo=[1.2.3.4])
2496 However, if the real host name was known, it was omitting the HELO data
2497 if it matched the actual IP address. This has been tidied up so that it
2498 doesn't show the same IP address twice.
2500 PH/07 When both +timestamp and +memory debugging was on, the value given by
2501 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2502 called by the debug printing, thereby overwriting the timestamp buffer.
2503 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2505 PH/08 When the original message was included in an autoreply transport, it
2506 always said "this is a copy of the message, including all the headers",
2507 even if body_only or headers_only was set. It now gives an appropriate
2510 PH/09 Applied a patch from the Sieve maintainer which:
2512 o fixes some comments
2513 o adds the (disabled) notify extension core
2514 o adds some debug output for the result of if/elsif tests
2515 o points to the current vacation draft in the documentation
2516 and documents the missing references header update
2520 o fixes a bug in processing the envelope test (when testing
2521 multiple envelope elements, the last element determinted the
2524 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2525 Electronic Mail") by including:
2527 Auto-submitted: auto-generated
2529 in the messages that it generates (bounce messages and others, such as
2530 warnings). In the case of bounce messages for non-SMTP mesages, there was
2531 also a typo: it was using "Auto_submitted" (underscore instead of
2532 hyphen). Since every message generated by Exim is necessarily in response
2533 to another message, thes have all been changed to:
2535 Auto-Submitted: auto-replied
2537 in accordance with these statements in the RFC:
2539 The auto-replied keyword:
2541 - SHOULD be used on messages sent in direct response to another
2542 message by an automatic process,
2544 - MUST NOT be used on manually-generated messages,
2546 - MAY be used on Delivery Status Notifications (DSNs) and Message
2547 Disposition Notifications (MDNs),
2549 - MUST NOT be used on messages generated by automatic or periodic
2550 processes, except for messages which are automatic responses to
2553 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2554 to the default Received: header definition.
2556 PH/12 Added log selector acl_warn_skipped (default on).
2558 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2559 variables because (a) they are in the wrong storage pool and (b) even if
2560 they were copied, it wouldn't work properly because of the caching.
2562 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2563 checking when decoding. Apparently there are clients that generate
2564 overlong encoded strings. Why am I not surprised?
2566 PH/15 If the first argument of "${if match_address" was not empty, but did not
2567 contain an "@" character, Exim crashed. Now it writes a panic log message
2568 and treats the condition as false.
2570 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2572 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2573 extension "envelope-auth". The code is finished and in agreement with
2574 other implementations, but there is no documentation so far and in fact,
2575 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2576 not changing the active code.
2578 Print executed "if" and "elsif" statements when debugging is used. This
2579 helps a great deal to understand what a filter does.
2581 Document more things not specified clearly in RFC3028. I had all this
2582 sorted out, when out of a sudden new issues came to my mind. Oops."
2584 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2587 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2588 canonical form (as documented). However, after a host name lookup from
2589 the IP address, check_host() was doing a simple string comparison with
2590 addresses acquired from the DNS when checking that the found name did
2591 have the original IP as one of its addresses. Since any found IPv6
2592 addresses are likely to be in abbreviated form, the comparison could
2593 fail. Luckily, there already exists a function for doing the comparison
2594 by converting both addresses to binary, so now that is used instead of
2595 the text comparison.
2597 PH/20 There was another similar case to PH/19, when a complete host name was
2598 given in a host list; looking up its IP address could give an abbreviated
2599 form, whereas the current host's name might or might not be abbreviated.
2600 The same fix has been applied.
2606 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2607 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2610 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2611 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2613 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2615 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2616 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2617 Also, for "Auto-Submitted", treat anything other than "no" as
2618 non-personal, in accordance with RFC 3834. (Previously it treated
2619 anything starting "auto-" as non-personal.)
2621 TF/01 The control=submission/name=... option had a problem with syntax
2622 errors if the name included a slash character. The /name= option
2623 now slurps the rest of the string, so it can include any characters
2624 but it must come last in the list of options (after /sender_retain
2627 PH/05 Some modifications to the interface to the fake nameserver for the new
2635 TK/01 Added the "success_on_redirect" address verification option. See
2636 NewStuff for rationale and an example.
2638 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2640 PH/02 Patch to exigrep to allow it to work on syslog lines.
2642 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2643 fread() to skip over the body file's header line, because in Cygwin the
2644 header line is locked and is inaccessible.
2646 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2647 co-exist for some time) to make it clear that it is the Exim ID that is
2648 referenced, not the Message-ID: header line.
2650 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2651 string_format() function, because snprintf() does not exist on all
2654 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2655 ${stat: expansion item.
2657 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2658 protocol synchronization error", to keep the pedants happy.
2660 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2661 well as for IRIX systems, when gcc is being used. See the host.c source
2664 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2666 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2669 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2670 transport and to the smtp transport in LMTP mode.
2672 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2674 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2675 run a filter in a subprocess. This could lead to confusion in subsequent
2676 lookups in the parent process. There should also be a search_tidyup() at
2677 the end of the subprocess.
2679 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2680 only if the host matched helo_try_verify_hosts, which caused the
2681 verification to occur when the EHLO/HELO command was issued. The ACL just
2682 tested the remembered result. Now, if a previous verification attempt has
2683 not happened, "verify = helo" does it there and then.
2685 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2687 TK/03 Fix log output including CR from clamd.
2689 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2690 exist provoked a memory error which could cause a segfault.
2692 PH/15 Installed PCRE 6.2
2694 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2696 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2697 of the problem. Specifically, suggested +O2 rather than +O1 for the
2700 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2702 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2703 record, the port was not correctly logged when the outgoing_port log
2704 selector was set (it logged the transort's default port).
2706 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2707 fallback_hosts, and "hosts" in the smtp transport.
2709 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2710 host errors such as "Connection refused".
2712 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2713 authentication with radiusclient 0.4.9:
2715 - Error returned from rc_read_config was caught wrongly
2716 - Username/password not passed on to radius server due to wrong length.
2718 The presumption is that some radiusclient API changes for 4.51/PH/17
2719 were not taken care of correctly. The code is still untested by me (my
2720 Linux distribution still has 0.3.2 of radiusclient), but it was
2721 contributed by a Radius user.
2723 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2724 expanding the "port" option of the smtp transport.
2726 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2727 while reading a MIME header. Thanks to Tom Hughes for a patch.
2729 PH/24 Include config.h inside local_scan.h so that configuration settings are
2732 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2733 that in an ACL for RCPT (for example), you can examine exactly what was
2736 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2737 commands, but it was not correctly comparing the address with the actual
2738 client host address. Thus, it would show the EHLO address in Received:
2739 header lines when this was not necessary.
2741 PH/27 Added the % operator to ${eval:}.
2743 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2744 it should be ignoring failures (because with -C, for example, it has lost
2745 privilege). It wasn't ignoring creation failures other than "already
2748 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2751 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2752 because an input line was too long, either on its own, or by virtue of
2753 too many continuations, the temporary file was not being removed, and the
2754 return code was incorrect.
2756 PH/31 Missing "BOOL" in function definition in filtertest.c.
2758 PH/32 Applied Sieve patches from the maintainer.
2760 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2762 PH/33 Added "verify = not_blind".
2764 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2765 Local/Makefile (with some defaults set). These are used in built scripts
2766 such as exicyclog, but they have never been used in the exim_install
2767 script (though there are many overriding facilities there). I have
2768 arranged that the exim_install script now takes note of these two
2771 PH/35 Installed configuration files for Dragonfly.
2773 PH/36 When a locally submitted message by a trusted user did not contain a
2774 From: header, and the sender address was obtained from -f or from an SMTP
2775 MAIL command, and the trusted user did not use -F to supply a sender
2776 name, $originator_name was incorrectly used when constructing a From:
2777 header. Furthermore, $originator_name was used for submission mode
2778 messages from external hosts without From: headers in a similar way,
2779 which is clearly wrong.
2781 PH/37 Added control=suppress_local_fixups.
2783 PH/38 When log_selector = +received_sender was set, and the addition of the
2784 sender made the log line's construction buffer exactly full, or one byte
2785 less than full, an overflow happened when the terminating "\n" was
2788 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2789 when the result of a list match is failure because a DNS lookup failed.
2791 PH/40 RM_COMMAND is now used in the building process.
2793 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2794 the "build-* directories that it finds.
2796 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2797 address in a domain literal was a prefix of an interface address.
2799 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2800 when verifying a sender address, unless rewrite_headers is false.
2802 PH/44 Wrote a long comment about why errors_to addresses are verified as
2803 recipients, not senders.
2805 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2806 the ratelimit ACL was added.
2808 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2810 PH/46 Added extra information about PostgreSQL errors to the error string.
2812 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2813 avoiding the need to install special zones in a real server. This is
2814 backwards compatible; if it can't find the fake resolver, it drops back.
2815 Thus, both old and new test suites can be run.
2817 TF/02 Added util/ratelimit.pl
2819 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2820 clock is set back in time.
2822 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2823 Candler <B.Candler@pobox.com>.
2825 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2826 OK for addresses that are the result of redirections.
2828 PH/48 A number of further additions for the benefit of the new test suite,
2829 including a fake gethostbyname() that interfaces to the fake DNS resolver
2832 TF/06 The fix for widen_domains has also been applied to qualify_single and
2833 search_parents which are the other dnslookup options that can cause
2836 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2839 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2841 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2842 service" if it encountered an MX record with an empty target hostname.
2843 The message is now "an MX or SRV record indicated no SMTP service".
2845 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2846 if the DNS of the sending site is misconfigured. This is quite a
2847 common situation. This change restores the behaviour of treating a
2848 helo verification defer as a failure.
2850 PH/51 If self=fail was set on a router, the bounce message did not include the
2851 actual error message.
2857 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2859 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2860 command itself ended in error, the underlying message about the transport
2861 filter timeout was being overwritten with the pipe command error. Now the
2862 underlying error message should be appended to the second error message.
2864 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2866 PH/02 Reduce the amount of output that "make" produces by default. Full output
2867 can still be requested.
2869 PH/03 The warning log line about a condition test deferring for a "warn" verb
2870 was being output only once per connection, rather than after each
2871 occurrence (because it was using the same function as for successful
2872 "warn" verbs). This seems wrong, so I have changed it.
2874 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2875 it should not have, which might have caused a crash in the right
2876 circumstances, but probably never did.
2878 PH/04 Installed a modified version of Tony Finch's patch to make submission
2879 mode fix the return path as well as the Sender: header line, and to
2880 add a /name= option so that you can make the user's friendly name appear
2883 TF/03 Added the control = fakedefer ACL modifier.
2885 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2886 Mark Lowes for thorough testing.
2888 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2890 TK/03 Merged latest SRS patch from Miles Wilton.
2892 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2893 with the definition in sysexits.h (which is #included earlier).
2894 Fortunately, Exim does not actually use EX_OK. The code used to try to
2895 preserve the sysexits.h value, by assumimg that macro definitions were
2896 scanned for macro replacements. I have been disabused of this notion,
2897 so now the code just undefines EX_OK before #including unistd.h.
2899 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2900 in the smtp transport. When a block could not be written in a single
2901 write() function, the timeout was being re-applied to each part-write.
2902 This seems wrong - if the receiver was accepting one byte at a time it
2903 would take for ever. The timeout is now adjusted when this happens. It
2904 doesn't have to be particularly precise.
2906 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2907 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2909 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2910 without a domain if the check to <postmaster@domain> fails.
2912 SC/01 Eximstats: added -xls and the ability to specify output files
2913 (patch written by Frank Heydlauf).
2915 SC/02 Eximstats: use FileHandles for outputing results.
2917 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2919 SC/04 Eximstats: fixed display of large numbers with -nvr option
2921 SC/05 Eximstats: fixed merging of reports with empty tables.
2923 SC/06 Eximstats: added the -include_original_destination flag
2925 SC/07 Eximstats: removed tabs and trailing whitespace.
2927 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2929 TK/06 MBOX spool code: Add real "From " MBOX separator line
2930 so the .eml file is really in mbox format (even though
2931 most programs do not really care). Patch from Alex Miller.
2933 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2934 The latter is generated from $received_to and is only set if the
2935 message has one envelope recipient. SA can use these headers,
2936 obviously out-of-the-box. Patch from Alex Miller.
2938 PH/08 The ${def test on a variable was returning false if the variable's
2939 value was "0", contrary to what the specification has always said!
2940 The result should be true unless the variable is empty.
2942 PH/09 The syntax error of a character other than { following "${if
2943 def:variable_name" (after optional whitespace) was not being diagnosed.
2944 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2945 accidental colon was present, for example, could give incorrect results.
2947 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2948 result is used (not including appendfile, where other changes are about
2951 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2952 This involved changing a lot of size variables from int to off_t. It
2953 should work with maildirs and everything.
2955 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2956 spamd dying while we are connected to it.
2958 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2961 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2962 function for BDB 4.3.
2964 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2966 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2967 the sources for PCRE have been re-arranged and more files are now
2970 PH/15 The code I had for printing potentially long long variables in PH/11
2971 above was not the best (it lost precision). The length of off_t variables
2972 is now inspected at build time, and an appropriate printing format (%ld
2973 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2974 to be "long long int" or "long int". This is needed for the internal
2975 formatting function string_vformat().
2977 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2978 the configuration file to be ":syslog", then the script "guesses" where
2979 the logs files are, rather than using the compiled in default. In our
2980 case the guess is not the same as the compiled default, so the script
2981 suddenly stopped working when I started to use syslog. The patch checks
2982 to see if log_file_path is "". If so, it attempts to read it from exim
2983 with no configuration file to get the compiled in version, before it
2984 falls back to the previous guessing code."
2986 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2987 implementing BATV in an Exim configuration. See NewStuff for the gory
2990 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2991 Makefile that are specific to HP-UX.
2993 PH/18 If the "use_postmaster" option was set for a recipient callout together
2994 with the "random" option, the postmaster address was used as the MAIL
2995 FROM address for the random test, but not for the subsequent recipient
2996 test. It is now used for both.
2998 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2999 patch removes a few documentation additions to RFC 3028, because the
3000 latest draft now contains them. It adds the new en;ascii-case comparator
3001 and a new error check for 8bit text in MIME parts. Comparator and
3002 require names are now matched exactly. I enabled the subaddress
3003 extension, but it is not well tested yet (read: it works for me)."
3005 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3006 rework some of the code of TK/09 above to avoid the hardwired use of
3007 "%lld" and "long long". Replaced the call to snprintf() with a call to
3010 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3011 records point to non-existent hosts", "retry timeout exceeded", and
3012 "retry time not reached for any host after a long failure period".
3014 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3015 experimental DomainKeys support:
3017 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3018 (2) On an error such as an illegally used "control", the wrong name for
3019 the control was given.
3021 These problems did NOT occur unless DomainKeys support was compiled.
3023 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3025 PH/24 Added ${if match_ip condition.
3027 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3028 databases so that it will be absolutely obvious if a crash occurs in the
3029 DB library. This is a regular occurrence (often caused by mis-matched
3032 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3033 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3034 picked up on a user's system that detects such things. There doesn't seem
3035 to be a gcc warning option for this - only an attribute that has to be
3036 put on the function's prototype. It seems that in Fedora Core 4 they have
3037 set this on a number of new functions. No doubt there will be more in due
3040 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3041 specify a transport. However, if an address that was verified by such a
3042 router was the subject of a callout, Exim crashed because it tried to
3043 read the rcpt_include_affixes from the non-existent transport. Now it
3044 just assumes that the setting of that option is false. This bug was
3045 introduced by 4.51/PH/31.
3047 PH/28 Changed -d+all to exclude +memory, because that information is very
3048 rarely of interest, but it makes the output a lot bigger. People tend to
3049 do -d+all out of habit.
3051 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3052 code in os-type was giving problems when libc.so lives in lib64, like on
3055 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3056 aren't the modern standard, and it seems that some systems' include files
3057 don't always have them. Exim was already checking for some of the newer
3058 ones like T_AAAA, and defining it itself. I've added checks for all the
3059 record types that Exim uses.
3061 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3062 not automatically generating a new one, as it is supposed to. This
3063 prevented TLS from working. If the file did exist, but contained invalid
3064 data, a new version was generated, as expected. It was only the case of a
3065 non-existent file that was broken.
3067 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3068 with a change in libdomainkeys > 0.64.
3070 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3071 from DNS. If the selector record carries the flag, it now has
3072 precedence over the domain-wide flag.
3074 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3076 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3077 the use of a port name that isn't defined in /etc/services) occurred, the
3078 message was deferred as in a normal delivery, and thus remained on the
3079 spool, instead of being failed because of the mua_wrapper setting. This
3080 is now fixed, and I tidied up some of the mua_wrapper messages at the
3083 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3084 the messages in a hash of arrays rather than using individual hashes.
3085 This is a bit cleaner and results in dramatic memory savings, albeit
3086 at a slight CPU cost.
3088 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3089 as requested by Marc Sherman.
3091 SC/10 Eximstats: added histograms for user specified patterns as requested
3094 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3096 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3097 fopen() in the content-scanning modules that did not already have it.
3103 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3104 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3106 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3108 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3110 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3111 to test Sieve filters that use "vacation".
3113 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3114 that changes the way the GnuTLS parameters are stored in the cache file.
3115 The new format can be generated externally. For backward compatibility,
3116 if the data in the cache doesn't make sense, Exim assumes it has read an
3117 old-format file, and it generates new data and writes a new file. This
3118 means that you can't go back to an older release without removing the
3121 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3122 work if there are any delivery delays because "one_time" forces the
3123 parent to be marked "delivered", so its unseen clone is never tried
3124 again. For this reason, Exim now forbids the simultaneous setting of
3127 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3128 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3129 entry if you want to know the details. The fix, however, neglected to
3130 consider the case when local delivery batching is involved. The test for
3131 "previously delivered" was not happening when checking to see if an
3132 address could be batched with a previous (undelivered) one; under
3133 certain circumstances this could lead to multiple deliveries to the same
3136 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3137 in its include files, and this causes problems building Exim.
3139 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3140 header_syntax) but Exim was just ignoring anything given after a slash.
3141 In particular, this caused confusion with an attempt to use "verify =
3142 reverse_host_lookup/defer_ok". An error is now given when options are
3143 supplied for verify items that do not have them. (Maybe reverse_host_
3144 lookup should have a defer_ok option, but that's a different point.)
3146 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3147 defined by RFC 821) to 2048, because there were problems with some AUTH
3148 commands, and RFC 1869 says the size should be increased for extended
3149 SMTP commands that take arguments.
3151 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3154 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3155 "unknown" error; now it says that the functionality isn't in the binary.
3157 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3158 an address' error message when a string expansion fails (syntax or
3159 whatever). Otherwise the password may appear in the log. Following change
3160 PH/42 below, there is no longer a chance of it appearing in a bounce
3163 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3165 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3166 of its addresses was ever tried. (Bugzilla bug #2.)
3168 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3169 the result incorrectly in the debug output. (It correctly added a newline
3170 to what was transported.)
3172 TF/01 Added $received_time.
3174 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3175 commented out examples of how to interface to a virus scanner and to
3176 SpamAssassin. Also added commented examples of av_scanner and
3177 spamd_address settings.
3179 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3180 and controls are allowed in which ACLs. There were a couple of minor
3181 errors. Some of the entries in the conditions table (which is a table of
3182 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3183 negation of where the condition IS allowed.
3185 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3187 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3188 header file does not have a version number, so I've had to invent a new
3189 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3190 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3191 radiusclient), but it was contributed by a Radius user.
3193 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3194 files or directories, for OpenSSL.
3196 PH/19 When an Exim process that is running as root has to create an Exim log
3197 file, it does so in a subprocess that runs as exim:exim so as to get the
3198 ownership right at creation (otherwise, other Exim processes might see
3199 the file with the wrong ownership). There was no test for failure of this
3200 fork() call, which would lead to the process getting stuck as it waited
3201 for a non-existent subprocess. Forks do occasionally fail when resources
3202 run out. I reviewed all the other calls to fork(); they all seem to check
3205 PH/20 When checking for unexpected SMTP input at connect time (before writing
3206 the banner), Exim was not dealing correctly with a non-positive return
3207 from the read() function. If the client had disconnected by this time,
3208 the result was a log entry for a synchronization error with an empty
3209 string after "input=" when read() returned zero. If read() returned -1
3210 (an event I could not check), uninitialized data bytes were printed.
3211 There were reports of junk text (parts of files, etc) appearing after
3214 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3216 PH/22 Added support for macro redefinition, and (re)definition in between
3217 driver and ACL definitions.
3219 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3220 forgetting to use the resulting value; it was using the unexpanded value.
3222 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3223 hadn't been configured. The fix is from Juergen Kreileder, who
3224 understands it better than I do:
3226 "Here's what I see happening with three configured cyrus_sasl
3227 authenticators configured (plain, login, cram-md5):
3229 On startup auth_cyrus_sasl_init() gets called for each of these.
3230 This means three calls to sasl_listmech() without a specified mech_list.
3231 => SASL tests which mechs of all available mechs actually work
3232 => three warnings about OTP not working
3233 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3235 With the patch, sasl_listmech() also gets called three times. But now
3236 SASL's mech_list option is set to the server_mech specified in the the
3237 authenticator. Or in other words, the answer from sasl_listmech()
3238 gets limited to just the mech you're testing for (which is different
3240 => the return list contains just 'plain' or 'login', 'cram-md5' or
3241 nothing depending on the value of ob->server_mech.
3243 I've just tested the patch: Authentication still works fine,
3244 unavailable mechs specified in the exim configuration are still
3245 caught, and the auth.log warnings about OTP are gone."
3247 PH/25 When debugging is enabled, the contents of the command line are added
3248 to the debugging output, even when log_selector=+arguments is not
3251 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3252 answer is "GNU", and only if the return is "GNU/something" is the answer
3255 PH/27 $acl_verify_message is now set immediately after the failure of a
3256 verification in an ACL, and so is available in subsequent modifiers. In
3257 particular, the message can be preserved by coding like this:
3259 warn !verify = sender
3260 set acl_m0 = $acl_verify_message
3262 Previously, $acl_verify_message was set only while expanding "message"
3263 and "log_message" when a very denied access.
3265 PH/28 Modified OS/os.c-Linux with
3267 -#ifndef OS_LOAD_AVERAGE
3268 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3270 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3271 nomenclature these days.)
3273 PH/29 Installed patch from the Sieve maintainer that adds the options
3274 sieve_useraddress and sieve_subaddress to the redirect router.
3276 PH/30 In these circumstances:
3277 . Two addresses routed to the same list of hosts;
3278 . First host does not offer TLS;
3279 . First host accepts first address;
3280 . First host gives temporary error to second address;
3281 . Second host offers TLS and a TLS session is established;
3282 . Second host accepts second address.
3283 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3284 and peerdn, if requested) that were in fact used only for the second
3287 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3288 attention to any local part prefix or suffix that was matched by the
3289 router that accepted the address. It now behaves in the same way as it
3290 does for delivery: the affixes are removed from the local part unless
3291 rcpt_include_affixes is set on the transport.
3293 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3294 timeout during the DATA phase of an incoming message.
3296 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3297 applied a patch sanctioned by the Sieve maintainer.
3299 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3300 the uid or gid is negative. A case of a negative gid caused this to be
3301 noticed. The fix allows for either to be negative.
3303 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3304 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3307 PH/36 The change PH/12 above was broken. Fixed it.
3309 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3310 the grounds that routing the same address twice would always produce the
3311 same answer. This might have been true once, but it is certainly no
3312 longer true now. Routing a child address may depend on the previous
3313 routing that produced that child. Some complicated redirection strategies
3314 went wrong when messages had multiple recipients, and made Exim's
3315 behaviour dependent on the order in which the addresses were given.
3317 I have moved the duplicate checking until after the routing is complete.
3318 Exim scans the addresses that are assigned to local and remote
3319 transports, and removes any duplicates. This means that more work will be
3320 done, as duplicates will always all be routed, but duplicates are
3321 presumably rare, so I don't expect this is of any significance.
3323 For deliveries to pipes, files, and autoreplies, the duplicate checking
3324 still happens during the routing process, since they are not going to be
3327 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3328 It corrects a timeout issue with spamd. This is Ian's comment: "The
3329 background is that sometimes spamd either never reads data from a
3330 connection it has accepted, or it never writes response data. The exiscan
3331 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3332 blindly assumes that writes won't block so it may never time out."
3334 PH/39 Allow G after quota size as well as K and M.
3336 PH/40 The value set for $authenticated_id in an authenticator may not contain
3337 binary zeroes or newlines because the value is written to log lines and
3338 to spool files. There was no check on this. Now the value is run through
3339 the string_printing() function so that such characters are converted to
3340 printable escape sequences.
3342 PH/41 $message_linecount is a new variable that contains the total number of
3343 lines in the message. Compare $body_linecount, which is the count for the
3346 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3347 bounce and delay warning messages, except in certain special cases, which
3350 (a) An SMTP error message from a remote host;
3351 (b) A message specified in a :fail: redirection;
3352 (c) A message specified in a "fail" command in a system filter;
3353 (d) A message specified in a FAIL return from the queryprogram router;
3354 (e) A message specified by the cannot_route_message router option.
3356 In these cases only, Exim does include the error details in bounce and
3357 warning messages. There are also a few cases where bland messages such
3358 as "unrouteable address" or "local delivery error" are given.
3360 PH/43 $value is now also set for the "else" part of a ${run expansion.
3362 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3363 being worked on, but at least Exim now implements the latest version to
3366 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3367 process to complete was treated as a delivery failure, a timeout while
3368 writing the message to the pipe was logged, but erroneously treated as a
3369 successful delivery. Such timeouts include transport filter timeouts. For
3370 consistency with the overall process timeout, these timeouts are now
3371 treated as errors, giving rise to delivery failures by default. However,
3372 there is now a new Boolean option for the pipe transport called
3373 timeout_defer, which, if set TRUE, converts the failures into defers for
3374 both kinds of timeout. A transport filter timeout is now identified in
3377 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3378 systems where "make" and "gmake" are different, calling "gmake" at top
3379 level broke things. I've arranged for the value of $(MAKE) to be passed
3380 from the Makefile to this script so that it can call the same version of
3384 A note about Exim versions 4.44 and 4.50
3385 ----------------------------------------
3387 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3388 changes of various kinds. As a consequence, a big documentation update was
3389 needed. This delayed the release for rather longer than seemed good, especially
3390 in the light of a couple of (minor) security issues. Therefore, the changes
3391 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3392 release. So 4.44 and 4.50 are in effect two different branches that both start
3395 I have left the 4.50 change log unchanged; it contains all the changes since
3396 4.43. The change log for 4.44 is below; many of its items are identical to
3397 those for 4.50. This seems to be the most sensible way to preserve the
3398 historical information.
3404 1. Minor wording change to the doc/README.SIEVE file.
3406 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3407 computation of the current number of files was incorrect.
3409 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3410 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3413 4. Give more explanation in the error message when the command for a transport
3414 filter fails to execute.
3416 5. There are several places where Exim runs a non-Exim command in a
3417 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3418 was being done only for the command run by the queryprogram router. It is
3419 now done for all such subprocesses. The other cases are: ${run, transport
3420 filters, and the commands run by the lmtp and pipe transports.
3422 6. Added CONFIGURE_GROUP build-time option.
3424 7. Some older OS have a limit of 256 on the maximum number of file
3425 descriptors. Exim was using setrlimit() to set 1000 as a large value
3426 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3427 systems. I've change it so that if it can't get 1000, it tries for 256.
3429 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3430 was an oversight, and furthermore, ever since the addition of extra
3431 controls (e.g. 4.43/32), the checks on when to allow different forms of
3432 "control" were broken. There should now be diagnostics for all cases when a
3433 control that does not make sense is encountered.
3435 9. Added the /retain_sender option to "control=submission".
3437 10. $recipients is now available in the predata ACL (oversight).
3439 11. Tidy the search cache before the fork to do a delivery from a message
3440 received from the command line. Otherwise the child will trigger a lookup
3441 failure and thereby defer the delivery if it tries to use (for example) a
3442 cached ldap connection that the parent has called unbind on.
3444 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3445 of $address_data from the recipient verification was clobbered by the
3446 sender verification.
3448 13. The value of address_data from a sender verification is now available in
3449 $sender_address_data in subsequent conditions in the ACL statement.
3451 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3453 15. Added a new option "connect=<time>" to callout options, to set a different
3456 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3457 was its contents. (It was OK if the option was not defined at all.)
3459 17. A "Completed" log line is now written for messages that are removed from
3460 the spool by the -Mrm option.
3462 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3463 information about exactly what failed.
3465 19. Added -dd to debug only the daemon process.
3467 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3468 handles timeouts, both on the server side and network timeouts. Renamed the
3469 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3471 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3472 It is now set to "smtps".
3474 22. $host_address is now set to the target address during the checking of
3475 ignore_target_hosts.
3477 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3478 being passed; this would have caused $sender_host_name to have been used if
3479 matching the list had actually called for a host name (not very likely,
3480 since this list is usually IP addresses). A host name is now passed as
3483 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3484 code that specifies a non-restarting handler (typically sigaction() in
3485 modern systems) in an attempt to fix a rare and obscure crash bug.
3487 25. Narrowed the window for a race in the daemon that could cause it to ignore
3488 SIGCHLD signals. This is not a major problem, because they are used only to
3489 wake it up if nothing else does.
3491 26. A malformed maildirsize file could cause Exim to calculate negative values
3492 for the mailbox size or file count. Odd effects could occur as a result.
3493 The maildirsize information is now recalculated if the size or filecount
3496 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3497 support for a long time. Removed HAVE_SYS_VFS_H.
3499 28. Installed the latest version of exipick from John Jetmore.
3501 29. In an address list, if the pattern was not a regular expression, an empty
3502 subject address (from a bounce message) matched only if the pattern was an
3503 empty string. Non-empty patterns were not even tested. This was the wrong
3504 because it is perfectly reasonable to use an empty address as part of a
3505 database query. An empty address is now tested by patterns that are
3506 lookups. However, all the other forms of pattern expect the subject to
3507 contain a local part and a domain, and therefore, for them, an empty
3508 address still always fails if the pattern is not itself empty.
3510 30. Exim went into a mad DNS loop when attempting to do a callout where the
3511 host was specified on an smtp transport, and looking it up yielded more
3512 than one IP address.
3514 31. Re-factored the code for checking spool and log partition space into a
3515 function that finds that data and another that does the check. The former
3516 is then used to implement four new variables: $spool_space, $log_space,
3517 $spool_inodes, and $log_inodes.
3519 32. The RFC2047 encoding function was originally intended for short strings
3520 such as real names; it was not keeping to the 75-character limit for
3521 encoded words that the RFC imposes. It now respects the limit, and
3522 generates multiple encoded words if necessary. To be on the safe side, I
3523 have increased the buffer size for the ${rfc2047: expansion operator from
3526 33. It is now permitted to omit both strings after an "if" condition; if the
3527 condition is true, the result is "true". As before, when the second string
3528 is omitted, a false condition yields an empty string. This makes it less
3529 cumbersome to write custom ACL and router conditions.
3531 34. Failure to deliver a bounce message always caused it to be frozen, even if
3532 there was an errors_to setting on the router. The errors_to setting is now
3535 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3536 canonical form (fully expanded) before being placed in
3537 $sender_host_address.
3539 36. The table in the code that translates DNS record types into text (T_A to
3540 "A" for instance) was missing entries for NS and CNAME. It is just possible
3541 that this could have caused confusion if both these types were looked up
3542 for the same domain, because the text type is used as part of Exim's
3543 per-process caching. But the chance of anyone hitting this buglet seems
3546 37. The dnsdb lookup has been extended in a number of ways.
3548 (1) There is a new type, "zns", which walks up the domain tree until it
3549 finds some nameserver records. It should be used with care.
3551 (2) There is a new type, "mxh", which is like "mx" except that it returns
3552 just the host names, not the priorities.
3554 (3) It is now possible to give a list of domains (or IP addresses) to be
3555 looked up. The behaviour when one of the lookups defers can be
3556 controlled by a keyword.
3558 (4) It is now possible to specify the separator character for use when
3559 multiple records are returned.
3561 38. The dnslists ACL condition has been extended: it is now possible to supply
3562 a list of IP addresses and/or domains to be looked up in a particular DNS
3565 39. Added log_selector=+queue_time_overall.
3567 40. When running the queue in the test harness, wait just a tad after forking a
3568 delivery process, to get repeatability of debugging output.
3570 41. Include certificate and key file names in error message when GnuTLS fails
3571 to set them up, because the GnuTLS error message doesn't include the name
3572 of the failing file when there is a problem reading it.
3574 42. Allow both -bf and -bF in the same test run.
3576 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3578 44. The "Exiscan patch" is now merged into the mainline Exim source.
3580 45. Sometimes the final signoff response after QUIT could fail to get
3581 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3582 < 0 before doing a fflush(). This bug looks as though it goes back to the
3583 introduction of TLS in release 3.20, but "sometimes" must have been rare
3584 because the tests only now provoked it.
3586 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3587 (this can affect the format of dates).
3589 47. exim_tidydb, when checking for the continued existence of a message for
3590 which it has found a message-specific retry record, was not finding
3591 messages that were in split spool directories. Consequently, it was
3592 deleting retry records that should have stayed in existence.
3594 48. Steve fixed some bugs in eximstats.
3596 49. The SPA authentication driver was not abandoning authentication and moving
3597 on to the next authenticator when an expansion was forced to fail,
3598 contradicting the general specification for all authenticators. Instead it
3599 was generating a temporary error. It now behaves as specified.
3601 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3602 (the order specifies the preference for clients). The order is now AES256,
3603 AES128, 3DES, ARCFOUR128.
3605 51. Small patch to Sieve code - explicitly set From: when generating an
3608 52. Exim crashed if a remote delivery caused a very long error message to be
3609 recorded - for instance if somebody sent an entire SpamAssassin report back
3610 as a large number of 550 error lines. This bug was coincidentally fixed by
3611 increasing the size of one of Exim's internal buffers (big_buffer) that
3612 happened as part of the Exiscan merge. However, to be on the safe side, I
3613 have made the code more robust (and fixed the comments that describe what
3616 53. Now that there can be additional text after "Completed" in log lines (if
3617 the queue_time_overall log selector is set), a one-byte patch to exigrep
3618 was needed to allow it to recognize "Completed" as not the last thing in
3621 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3622 patch that reportedly fixes this has been added. I am not expert enough to
3623 create a test for it. This is what the patch creator wrote:
3625 "I found a little strange behaviour of ldap code when working with
3626 Windows 2003 AD Domain, where users was placed in more than one
3627 Organization Units. When I tried to give exim partial DN, the exit code
3628 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3629 But simultaneously result of request was absolutely normal ldap result,
3630 so I produce this patch..."
3632 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3633 REFERENCE, so I have modified the code to exclude the patch when that macro
3636 55. Some experimental protocols are using DNS PTR records for new purposes. The
3637 keys for these records are domain names, not reversed IP addresses. The
3638 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3639 leaves it alone. Component reversal etc. now happens only for IP addresses.
3642 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3644 57. Double the size of the debug message buffer (to 2048) so that more of very
3645 long debug lines gets shown.
3647 58. The exicyclog utility now does better if the number of log files to keep
3648 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3650 59. Two changes related to the smtp_active_hostname option:
3652 (1) $smtp_active_hostname is now available as a variable.
3653 (2) The default for smtp_banner uses $smtp_active_hostname instead
3654 of $primary_hostname.
3656 60. The host_aton() function is supposed to be passed a string that is known
3657 to be a valid IP address. However, in the case of IPv6 addresses, it was
3658 not checking this. This is a hostage to fortune. Exim now panics and dies
3659 if the condition is not met. A case was found where this could be provoked
3660 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3661 components; fortuitously, this particular loophole had already been fixed
3662 by change 4.50/55 above.
3664 If there are any other similar loopholes, the new check in host_aton()
3665 itself should stop them being exploited. The report I received stated that
3666 data on the command line could provoke the exploit when Exim was running as
3667 exim, but did not say which command line option was involved. All I could
3668 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3669 running as the user.
3672 61. There was a buffer overflow vulnerability in the SPA authentication code
3673 (which came originally from the Samba project). I have added a test to the
3674 spa_base64_to_bits() function which I hope fixes it.
3677 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3678 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3680 63. The daemon start-up calls getloadavg() while still root for those OS that
3681 need the first call to be done as root, but it missed one case: when
3682 deliver_queue_load_max is set with deliver_drop_privilege. This is
3683 necessary for the benefit of the queue runner, because there is no re-exec
3684 when deliver_drop_privilege is set.
3686 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3687 This has been fixed.
3689 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3690 was in use, was not putting the data itself into the right store pool;
3691 consequently, it could be overwritten for a subsequent message in the same
3692 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3695 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3697 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3698 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3699 the function were treating the return as a boolean value, which happened to
3700 work because 0=false and not-0=true, but is not correct code.
3702 68. The host_aton() function was not handling scoped IPv6 addresses (those
3703 with, for example, "%eth0" on the end) correctly.
3705 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3706 negated items (that is, ~something) in unsigned ints. Some compilers
3707 apparently mutter when there is no cast.
3709 70. If an address verification called from an ACL failed, and did not produce a
3710 user-specific message (i.e. there was only a "system" message), nothing was
3711 put in $acl_verify_message. In this situation, it now puts the system
3714 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3715 session; change 4.31/43 added the unwanted input to the log line - except
3716 that it did not do this in the start of session case. It now does.
3718 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3719 This is wrong and can cause the other end to generate a synchronization
3720 error if it is another Exim or anything else that does the synchronization
3721 check. A QUIT command is no longer sent after a timeout.
3723 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3724 during host lookups.
3726 74. The defer_ok option of callout verification was not working if it was used
3727 when verifying addresses in header lines, that is, for this case:
3729 verify = header_sender/callout=defer_ok
3731 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3732 those file descriptors could be used for SMTP connections. If anything
3733 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3734 could be sent to the SMTP client, causing chaos. The daemon now opens
3735 stdin, stdout, and stderr to /dev/null when it puts itself into the
3738 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3739 log by default. The user can override this with suitable Perl magic.
3741 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3742 the log message when discard triggers, was not working for the DATA ACL or
3743 for the non-SMTP ACL.
3745 78. Error message wording change in sieve.c.
3747 79. If smtp_accept_max_per_host was set, the number of connections could be
3748 restricted to fewer than expected, because the daemon was trying to set up
3749 a new connection before checking whether the processes handling previous
3750 connections had finished. The check for completed processes is now done
3751 earlier. On busy systems, this bug wouldn't be noticed because something
3752 else would have woken the daemon, and it would have reaped the completed
3755 80. If a message was submitted locally by a user whose login name contained one
3756 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3757 It caused a spool format error. I have fixed the spool reading code. A
3758 related problem was that the "from" clause in the Received: line became
3759 illegal because of the space(s). It is now covered by ${quote_local_part.
3761 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3764 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3766 83. Patch from Sieve maintainer to fix unterminated string problem in
3767 "vacation" handling.
3769 84. Some minor changes to the Linux configuration files to help with other
3770 OS variants using glibc.
3772 85. One more patch for Sieve to update vacation handling to latest spec.
3775 ----------------------------------------------------
3776 See the note above about the 4.44 and 4.50 releases.
3777 ----------------------------------------------------
3783 1. Change 4.43/35 introduced a bug that caused file counts to be
3784 incorrectly computed when quota_filecount was set in an appendfile
3787 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3788 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3791 3. Give more explanation in the error message when the command for a transport
3792 filter fails to execute.
3794 4. There are several places where Exim runs a non-Exim command in a
3795 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3796 was being done only for the command run by the queryprogram router. It is
3797 now done for all such subprocesses. The other cases are: ${run, transport
3798 filters, and the commands run by the lmtp and pipe transports.
3800 5. Some older OS have a limit of 256 on the maximum number of file
3801 descriptors. Exim was using setrlimit() to set 1000 as a large value
3802 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3803 systems. I've change it so that if it can't get 1000, it tries for 256.
3805 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3806 was an oversight, and furthermore, ever since the addition of extra
3807 controls (e.g. 4.43/32), the checks on when to allow different forms of
3808 "control" were broken. There should now be diagnostics for all cases when a
3809 control that does not make sense is encountered.
3811 7. $recipients is now available in the predata ACL (oversight).
3813 8. Tidy the search cache before the fork to do a delivery from a message
3814 received from the command line. Otherwise the child will trigger a lookup
3815 failure and thereby defer the delivery if it tries to use (for example) a
3816 cached ldap connection that the parent has called unbind on.
3818 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3819 of $address_data from the recipient verification was clobbered by the
3820 sender verification.
3822 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3823 was its contents. (It was OK if the option was not defined at all.)
3825 11. A "Completed" log line is now written for messages that are removed from
3826 the spool by the -Mrm option.
3828 12. $host_address is now set to the target address during the checking of
3829 ignore_target_hosts.
3831 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3832 being passed; this would have caused $sender_host_name to have been used if
3833 matching the list had actually called for a host name (not very likely,
3834 since this list is usually IP addresses). A host name is now passed as
3837 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3838 code that specifies a non-restarting handler (typically sigaction() in
3839 modern systems) in an attempt to fix a rare and obscure crash bug.
3841 15. Narrowed the window for a race in the daemon that could cause it to ignore
3842 SIGCHLD signals. This is not a major problem, because they are used only to
3843 wake it up if nothing else does.
3845 16. A malformed maildirsize file could cause Exim to calculate negative values
3846 for the mailbox size or file count. Odd effects could occur as a result.
3847 The maildirsize information is now recalculated if the size or filecount
3850 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3851 support for a long time. Removed HAVE_SYS_VFS_H.
3853 18. Updated exipick to current release from John Jetmore.
3855 19. Allow an empty sender to be matched against a lookup in an address list.
3856 Previously the only cases considered were a regular expression, or an
3859 20. Exim went into a mad DNS lookup loop when doing a callout where the
3860 host was specified on the transport, if the DNS lookup yielded more than
3863 21. The RFC2047 encoding function was originally intended for short strings
3864 such as real names; it was not keeping to the 75-character limit for
3865 encoded words that the RFC imposes. It now respects the limit, and
3866 generates multiple encoded words if necessary. To be on the safe side, I
3867 have increased the buffer size for the ${rfc2047: expansion operator from
3870 22. Failure to deliver a bounce message always caused it to be frozen, even if
3871 there was an errors_to setting on the router. The errors_to setting is now
3874 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3875 canonical form (fully expanded) before being placed in
3876 $sender_host_address.
3878 24. Updated eximstats to version 1.33
3880 25. Include certificate and key file names in error message when GnuTLS fails
3881 to set them up, because the GnuTLS error message doesn't include the name
3882 of the failing file when there is a problem reading it.
3884 26. Expand error message when OpenSSL has problems setting up cert/key files.
3887 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3888 (this can affect the format of dates).
3890 28. exim_tidydb, when checking for the continued existence of a message for
3891 which it has found a message-specific retry record, was not finding
3892 messages that were in split spool directories. Consequently, it was
3893 deleting retry records that should have stayed in existence.
3895 29. eximstats updated to version 1.35
3896 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3897 1.35 - bugfix such that pie charts by volume are generated correctly
3899 30. The SPA authentication driver was not abandoning authentication and moving
3900 on to the next authenticator when an expansion was forced to fail,
3901 contradicting the general specification for all authenticators. Instead it
3902 was generating a temporary error. It now behaves as specified.
3904 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3905 (the order specifies the preference for clients). The order is now AES256,
3906 AES128, 3DES, ARCFOUR128.
3908 31. Small patch to Sieve code - explicitly set From: when generating an
3911 32. Exim crashed if a remote delivery caused a very long error message to be
3912 recorded - for instance if somebody sent an entire SpamAssassin report back
3913 as a large number of 550 error lines. This bug was coincidentally fixed by
3914 increasing the size of one of Exim's internal buffers (big_buffer) that
3915 happened as part of the Exiscan merge. However, to be on the safe side, I
3916 have made the code more robust (and fixed the comments that describe what
3919 33. Some experimental protocols are using DNS PTR records for new purposes. The
3920 keys for these records are domain names, not reversed IP addresses. The
3921 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3922 leaves it alone. Component reversal etc. now happens only for IP addresses.
3925 34. The host_aton() function is supposed to be passed a string that is known
3926 to be a valid IP address. However, in the case of IPv6 addresses, it was
3927 not checking this. This is a hostage to fortune. Exim now panics and dies
3928 if the condition is not met. A case was found where this could be provoked
3929 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3930 components; fortuitously, this particular loophole had already been fixed
3931 by change 4.50/55 or 4.44/33 above.
3933 If there are any other similar loopholes, the new check in host_aton()
3934 itself should stop them being exploited. The report I received stated that
3935 data on the command line could provoke the exploit when Exim was running as
3936 exim, but did not say which command line option was involved. All I could
3937 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3938 running as the user.
3941 35. There was a buffer overflow vulnerability in the SPA authentication code
3942 (which came originally from the Samba project). I have added a test to the
3943 spa_base64_to_bits() function which I hope fixes it.
3946 36. The daemon start-up calls getloadavg() while still root for those OS that
3947 need the first call to be done as root, but it missed one case: when
3948 deliver_queue_load_max is set with deliver_drop_privilege. This is
3949 necessary for the benefit of the queue runner, because there is no re-exec
3950 when deliver_drop_privilege is set.
3952 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3953 was in use, was not putting the data itself into the right store pool;
3954 consequently, it could be overwritten for a subsequent message in the same
3955 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3958 38. Sometimes the final signoff response after QUIT could fail to get
3959 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3960 < 0 before doing a fflush(). This bug looks as though it goes back to the
3961 introduction of TLS in release 3.20, but "sometimes" must have been rare
3962 because the tests only now provoked it.
3968 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3969 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3970 Now it returns an integer. A number of calls were still expecting a T/F
3971 return. Fortuitously, in all cases, the tests worked in OK situations,
3972 which is the norm. However, things would have gone wrong on any write
3973 failures on the smtp file descriptor. This function is used when sending
3974 messages over SMTP and also when doing verify callouts.
3976 2. When Exim is called to do synchronous delivery of a locally submitted
3977 message (the -odf or -odi options), it no longer closes stderr before doing
3980 3. Implemented the mua_wrapper option.
3982 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3984 5. Implemented the functions header_remove(), header_testname(),
3985 header_add_at_position(), and receive_remove_recipient(), and exported them
3988 6. If an ACL "warn" statement specified the addition of headers, Exim already
3989 inserted X-ACL-Warn: at the start if there was no header name. However, it
3990 was not making this test for the second and subsequent header lines if
3991 there were newlines in the string. This meant that an invalid header could
3992 be inserted if Exim was badly configured.
3994 7. Allow an ACL "warn" statement to add header lines at the start or after all
3995 the Received: headers, as well as at the end.
3997 8. Added the rcpt_4xx retry error code.
3999 9. Added postmaster_mailfrom=xxx to callout verification option.
4001 10. Added mailfrom=xxxx to the callout verification option, for verify=
4004 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4005 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4007 12. Inserted some casts to stop certain compilers warning when using pointer
4008 differences as field lengths or precisions in printf-type calls (mostly
4009 affecting debugging statements).
4011 13. Added optional readline() support for -be (dynamically loaded).
4013 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4014 same clock tick as a message's arrival, so that its received time was the
4015 same as the "first fail" time on the retry record, and that message
4016 remained on the queue past the ultimate address timeout, every queue runner
4017 would try a delivery (because it was past the ultimate address timeout) but
4018 after another failure, the ultimate address timeout, which should have then
4019 bounced the address, did not kick in. This was a "< instead of <=" error;
4020 in most cases the first failure would have been in the next clock tick
4021 after the received time, and all would be well.
4023 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4024 being recognized when the domain list was tested by the match_domain
4025 condition in an expansion string.
4027 16. Added the ${str2b64: operator.
4029 17. Exim was always calling setrlimit() to set a large limit for the number of
4030 processes, without checking whether the existing limit was already
4031 adequate. (It did check for the limit on file descriptors.) Furthermore,
4032 errors from getrlimit() and setrlimit() were being ignored. Now they are
4033 logged to the main and panic logs, but Exim does carry on, to try to do its
4034 job under whatever limits there are.
4036 18. Imported PCRE 5.0.
4038 19. Trivial typo in log message " temporarily refused connection" (the leading
4041 20. If the log selector return_path_on_delivery was set and an address was
4042 redirected to /dev/null, the delivery process crashed because it assumed
4043 that a return path would always be set for a "successful" delivery. In this
4044 case, the whole delivery is bypassed as an optimization, and therefore no
4047 21. Internal re-arrangement: the function for sending a challenge and reading
4048 a response while authentication was assuming a zero-terminated challenge
4049 string. It's now changed to take a pointer and a length, to allow for
4050 binary data in such strings.
4052 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4054 23. Exim was not respecting finduser_retries when seeking the login of the
4055 uid under which it was called; it was always trying 10 times. (The default
4056 setting of finduser_retries is zero.) Also, it was sleeping after the final
4057 failure, which is pointless.
4059 24. Implemented tls_on_connect_ports.
4061 25. Implemented acl_smtp_predata.
4063 26. If the domain in control=submission is set empty, Exim assumes that the
4064 authenticated id is a complete email address when it generates From: or
4065 Sender: header lines.
4067 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4068 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4069 chown and chgrp in /bin and hostname in /usr/bin.
4071 28. Exim was keeping the "process log" file open after each use, just as it
4072 does for the main log. This opens the possibility of it remaining open for
4073 long periods when the USR1 signal hits a daemon. Occasional processlog
4074 errors were reported, that could have been caused by this. Anyway, it seems
4075 much more sensible not to leave this file open at all, so that is what now
4078 29. The long-running daemon process does not normally write to the log once it
4079 has entered its main loop, and it closes the log before doing so. This is
4080 so that log files can straightforwardly be renamed and moved. However,
4081 there are a couple of unusual error situations where the daemon does write
4082 log entries, and I had neglected to close the log afterwards.
4084 30. The text of an SMTP error response that was received during a remote
4085 delivery was being truncated at 512 bytes. This is too short for some of
4086 the long messages that one sometimes sees. I've increased the limit to
4089 31. It is now possible to make retry rules that apply only when a message has a
4090 specific sender, in particular, an empty sender.
4092 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4093 it possible to be selective about when SMTP synchronization is enforced.
4095 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4097 32. Implemented hosts_connection_nolog.
4099 33. Added an ACL for QUIT.
4101 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4104 35. Added mailbox_size and mailbox_filecount to appendfile.
4106 36. Added control = no_multiline_responses to ACLs.
4108 37. There was a bug in the logic of the code that waits for the clock to tick
4109 in the case where the clock went backwards by a substantial amount such
4110 that the microsecond fraction of "now" was more than the microsecond
4111 fraction of "then" (but the whole seconds number was less).
4113 38. Added support for the libradius Radius client library this is found on
4114 FreeBSD (previously only the radiusclient library was supported).
4120 1. When certain lookups returned multiple values in the form name=value, the
4121 quoting of the values was not always being done properly. Specifically:
4122 (a) If the value started with a double quote, but contained no whitespace,
4124 (b) If the value contained whitespace other than a space character (i.e.
4125 tabs or newlines or carriage returns) it was not quoted.
4126 This fix has been applied to the mysql and pgsql lookups by writing a
4127 separate quoting function and calling it from the lookup code. The fix
4128 should probably also be applied to nisplus, ibase and oracle lookups, but
4129 since I cannot test any of those, I have not disturbed their existing code.
4131 2. A hit in the callout cache for a specific address caused a log line with no
4132 reason for rejecting RCPT. Now it says "Previous (cached) callout
4133 verification failure".
4135 3. There was an off-by-one bug in the queryprogram router. An over-long
4136 return line was truncated at 256 instead of 255 characters, thereby
4137 overflowing its buffer with the terminating zero. As well as fixing this, I
4138 have increased the buffer size to 1024 (and made a note to document this).
4140 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4141 when Exim is waiting for an SMTP response from a remote server, Exim
4142 restarts its select() call on the socket, thereby resetting its timeout.
4143 This is not a problem when such interrupts are rare. Somebody set up a cron
4144 job to run exiwhat every 2 minutes, which is less than the normal select()
4145 timeout (5 or 10 minutes). This meant that the select() timeout never
4146 kicked in because it was always reset. I have fixed this by comparing the
4147 time when an interrupt arrives with the time at the start of the first call
4148 to select(). If more time than the timeout has elapsed, the interrupt is
4149 treated as a timeout.
4151 5. Some internal re-factoring in preparation for the addition of Sieve
4152 extensions (by MH). In particular, the "personal" test is moved to a
4153 separate function, and given an option for scanning Cc: and Bcc: (which is
4154 not set for Exim filters).
4156 6. When Exim created an email address using the login of the caller as the
4157 local part (e.g. when creating a From: or Sender: header line), it was not
4158 quoting the local part when it contained special characters such as @.
4160 7. Installed new OpenBSD configuration files.
4162 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4163 try to make them clearer.
4165 9. Callout options, other than the timeout value, were being ignored when
4166 verifying sender addresses in header lines. For example, when using
4168 verify = header_sender/callout=no_cache
4170 the cache was (incorrectly) being used.
4172 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4173 only the Cygwin environment.
4175 11. When return_path_on_delivery was set as a log selector, if different remote
4176 addresses in the same message used different return paths and parallel
4177 remote delivery occurred, the wrong values would sometimes be logged.
4178 (Whenever a remote delivery process finished, the return path value from
4179 the most recently started remote delivery process was logged.)
4181 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4182 lines when AUTH and/or TLS are in use. This is the "received protocol"
4183 field. Exim used to use "asmtp" for authenticated SMTP, without any
4184 indication (in the protocol name) for TLS use. Now it follows the RFC and
4185 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4186 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4187 names appear in log lines as well as in Received: header lines.
4189 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4190 extensions, and comparison tests, and to fix some bugs.
4192 14. Changes to the "personal" filter test:
4194 (1) The test was buggy in that it was just doing the equivalent of
4195 "contains" tests on header lines. For example, if a user's address was
4196 anne@some.where, the "personal" test would incorrectly be true for
4198 To: susanne@some.where
4200 This test is now done by extracting each address from the header in turn,
4201 and checking the entire address. Other tests that are part of "personal"
4202 are now done using regular expressions (for example, to check local parts
4203 of addresses in From: header lines).
4205 (2) The list of non-personal local parts in From: addresses has been
4206 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4207 taken from the Sieve specification recommendations.
4209 (3) If the message contains any header line starting with "List-" it is
4210 treated as non-personal.
4212 (4) The test for "circular" in the Subject: header line has been removed
4213 because it now seems ill-conceived.
4215 15. Minor typos in src/EDITME comments corrected.
4217 16. Installed latest exipick from John Jetmore.
4219 17. If headers_add on a router specified a text string that was too long for
4220 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4221 of string_sprintf() is now avoided.
4223 18. $message_body_size was not set (it was always zero) when running the DATA
4224 ACL and the local_scan() function.
4226 19. For the "mail" command in an Exim filter, no default was being set for
4227 the once_repeat time, causing a random time value to be used if "once" was
4228 specified. (If the value happened to be <= 0, no repeat happened.) The
4229 default is now 0s, meaning "never repeat". The "vacation" command was OK
4230 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4231 (I found it when inspecting the code).
4233 20. There is now an overall timeout for performing a callout verification. It
4234 defaults to 4 times the callout timeout, which applies to individual SMTP
4235 commands during the callout. The overall timeout applies when there is more
4236 than one host that can be tried. The timeout is checked before trying the
4237 next host. This prevents very long delays if there are a large number of
4238 hosts and all are timing out (e.g. when the network connections are timing
4239 out). The value of the overall timeout can be changed by specifying an
4240 additional sub-option for "callout", called "maxwait". For example:
4242 verify = sender/callout=5s,maxwait=20s
4244 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4245 to the end before writing, but this should make it even safer).
4247 22. Exim was forgetting that it had advertised PIPELINING for the second and
4248 subsequent messages on an SMTP connection. It was also not resetting its
4249 memory on STARTTLS and an internal HELO.
4251 23. When Exim logs an SMTP synchronization error within a session, it now
4252 records whether PIPELINING has been advertised or not.
4254 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4255 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4256 rather than long int.
4258 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4260 26. Added the never_mail option to autoreply.
4266 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4267 crash if the getsockname() call failed; this can happen if a connection is
4268 closed very soon after it is established. The problem was simply in the
4269 order in which certain operations were done, causing Exim to try to write
4270 to the SMTP stream before it had set up the file descriptor. The bug has
4271 been fixed by making things happen in the correct order.
4277 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4278 before the connection was closed, thus losing the rejection response.
4280 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4281 some early Solaris releases, but causes trouble in current releases where
4282 socklen_t is defined.
4284 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4287 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4290 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4291 that releasing the top of it at the end releases what was used for sub-
4292 expansions (unless the block got too big). However, discard this block if
4293 the first thing is a variable or header, so that we can use its block when
4294 it is dynamic (useful for very large $message_headers, for example).
4296 6. Lookups now cache *every* query, not just the most recent. A new, separate
4297 store pool is used for this. It can be recovered when all lookup caches are
4298 flushed. Lookups now release memory at the end of their result strings.
4299 This has involved some general refactoring of the lookup sources.
4301 7. Some code has been added to the store_xxx() functions to reduce the amount
4302 of flapping under certain conditions.
4304 8. log_incoming_interface used to affect only the <= reception log lines. Now
4305 it causes the local interface and port to be added to several more SMTP log
4306 lines, for example "SMTP connection from", and rejection lines.
4308 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4310 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4312 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4313 could be overwritten at the end of the current message (or the start of a
4314 new message if it was set in a HELO ACL). The value is now preserved for
4315 the duration of the SMTP connection.
4317 12. If a transport had a headers_rewrite setting, and a matching header line
4318 contained an unqualified address, that address was qualified, even if it
4319 did not match any rewriting rules. The underlying bug was that the values
4320 of the flags that permit the existence of unqualified sender and recipient
4321 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4322 non-local messages, and by -bnq for local messages) were not being
4323 preserved with the message after it was received.
4325 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4326 "next input=" as part of the text comprising the host identity instead of
4327 the correct text. The code was using the same buffer for two different
4328 strings. However, depending on which order the printing function evaluated
4329 its arguments, the bug did not always show up. Under Linux, for example, my
4330 test suite worked just fine.
4332 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4333 doesn't work with some older versions of Perl. It has been changed to "my",
4334 which in any case is probably the better facility to use.
4336 15. A really picky compiler found some instances of statements for creating
4337 error messages that either had too many or two few arguments for the format
4340 16. The size of the buffer for calls to the DNS resolver has been increased
4341 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4342 for addresses that have a lot of PTR records. This alleviates a problem; it
4343 does not fully solve it.
4345 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4346 buffer now truncates the list and logs the incident, which is the same
4347 action as happens when Exim is looking up a host name and its aliases.
4348 Previously in this situation something unpredictable would happen;
4349 sometimes it was "internal error: store_reset failed".
4351 18. If a server dropped the connection unexpectedly when an Exim client was
4352 using GnuTLS and trying to read a response, the client delivery process
4353 crashed while trying to generate an error log message.
4355 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4356 string, for example:
4358 warn message = H1: something\nH2: something
4360 the text was added as a single header line from Exim's point of view
4361 though it ended up OK in the delivered message. However, searching for the
4362 second and subsequent header lines using $h_h2: did not work. This has been
4363 fixed. Similarly, if a system filter added multiple headers in this way,
4364 the routers could not see them.
4366 20. Expanded the error message when iplsearch is called with an invalid key to
4367 suggest using net-iplsearch in a host list.
4369 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4370 ACLs are no longer actually imposed (and a message to that effect is
4373 22. If a "gecos" field in a passwd entry contained escaped characters, in
4374 particular, if it contained a \" sequence, Exim got it wrong when building
4375 a From: or a Sender: header from that name. A second bug also caused
4376 incorrect handling when an unquoted " was present following a character
4377 that needed quoting.
4379 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4380 was not being matched caselessly.
4382 24. Arranged for all hyphens in the exim.8 source to be escaped with
4385 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4386 properly. Recipient callout cache records were still being keyed to include
4387 the sender, even when use_sender was set false. This led to far more
4388 callouts that were necessary. The sender is no longer included in the key
4389 when use_sender is false.
4391 26. Added "control = submission" modifier to ACLs.
4393 27. Added the ${base62d: operator to decode base 62 numbers.
4395 28. dnsdb lookups can now access SRV records.
4397 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4398 the configuration file.
4400 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4401 (-v) mode. This makes the output for a verbose queue run more intelligible.
4403 31. Added a use_postmaster feature to recipient callouts.
4405 32. Added the $body_zerocount variable, containing the number of binary zero
4406 bytes in the message body.
4408 33. The time of last modification of the "new" subdirectory is now used as the
4409 "mailbox time last read" when there is a quota error for a maildir
4412 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4414 35. Added +ignore_unknown as a special item in host lists.
4416 36. Code for decoding IPv6 addresses in host lists is now included, even if
4417 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4418 address was recognized as an IP address, but was then not correctly decoded
4419 into binary, causing unexpected and incorrect effects when compared with
4426 1. Very minor rewording of debugging text in manualroute to say "list of
4427 hosts" instead of "hostlist".
4429 2. If verify=header_syntax was set, and a header line with an unqualified
4430 address (no domain) and a large number of spaces between the end of the
4431 name and the colon was received, the reception process suffered a buffer
4432 overflow, and (when I tested it) crashed. This was caused by some obsolete
4433 code that should have been removed. The fix is to remove it!
4435 3. When running in the test harness, delay a bit after writing a bounce
4436 message to get a bit more predictability in the log output.
4438 4. Added a call to search_tidyup() just before forking a reception process. In
4439 theory, someone could use a lookup in the expansion of smtp_accept_max_
4440 per_host which, without the tidyup, could leave open a database connection.
4442 5. Added the variables $recipient_data and $sender_data which get set from a
4443 lookup success in an ACL "recipients" or "senders" condition, or a router
4444 "senders" option, similar to $domain_data and $local_part_data.
4446 6. Moved the writing of debug_print from before to after the "senders" test
4449 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4450 problems for message scanning, either using a data ACL, or using
4451 local_scan() because the Received: header was not generated till after they
4452 were called (in order to set the time as the time of reception completion).
4453 I have revised the way this works. The header is now generated after the
4454 body is received, but before the ACL or local_scan() are called. After they
4455 are run, the timestamp in the header is updated.
4461 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4462 before starting a queue runner without re-exec. This happened only when
4463 deliver_drop_privilege was set or when the Exim user was set to root. The
4464 effect of the bug was that timeouts during subsequent deliveries caused
4465 crashes instead of being properly handled. The handler is now left at its
4466 default (and expected) setting.
4468 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4469 message, again when deliver_drop_privilege is set or Exim is run as root.
4470 The bug described in (1) was not present in this case, but the tidying up
4471 of the other signals was missing. I have made the two cases consistent.
4473 3. The ignore_target_hosts setting on a manualroute router was being ignored
4474 for hosts that were looked up using the /MX notation.
4476 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4479 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4480 operated on the sender address. After changing the $sender_address to <>
4481 for the sender address verify, Exim was re-instated it as the original
4482 (before rewriting) address, but remembering that it had rewritten it, so it
4483 wasn't rewriting it again. This bug also had the effect of breaking the
4484 sender address verification caching when the sender address was rewritten.
4486 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4487 This has been changed so that if the ip literal address matches
4488 ignore_target_hosts, the router declines.
4490 7. Added expansion conditions match_domain, match_address, and match_local_
4491 part (NOT match_host).
4493 8. The placeholder for the Received: header didn't have a length field set.
4495 9. Added code to Exim itself and to exim_lock to test for a specific race
4496 condition that could lead to file corruption when using MBX delivery. The
4497 issue is with the lockfile that is created in /tmp. If this file is removed
4498 after a process has opened it but before that process has acquired a lock,
4499 there is the potential for a second process to recreate the file and also
4500 acquire a lock. This could lead to two Exim processes writing to the file
4501 at the same time. The added code performs the same test as UW imapd; it
4502 checks after acquiring the lock that its file descriptor still refers to
4503 the same named file.
4505 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4506 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4509 11. Added the smtp_active_hostname option. If used, this will typically be made
4510 to depend on the incoming interface address. Because $interface_address is
4511 not set up until the daemon has forked a reception process, error responses
4512 that can happen earlier (such as "too many connections") no longer contain
4515 12. If an expansion in a condition on a "warn" statement fails because a lookup
4516 defers, the "warn" statement is abandoned, and the next ACL statement is
4517 processed. Previously this caused the whole ACL to be aborted.
4519 13. Added the iplsearch lookup type.
4521 14. Added ident_timeout as a log selector.
4523 15. Added tls_certificate_verified as a log selector.
4525 16. Added a global option tls_require_ciphers (compare the smtp transport
4526 option of the same name). This controls incoming TLS connections.
4528 17. I finally figured out how to make tls_require_ciphers do a similar thing
4529 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4530 before starting the TLS session.
4532 18. Tabs are now shown as \t in -bP output.
4534 19. If the log selector return_path_on_delivery was set, Exim crashed when
4535 bouncing a message because it had too many Received: header lines.
4537 20. If two routers both had headers_remove settings, and the first one included
4538 a superfluous trailing colon, the final name in the first list and the
4539 first name in the second list were incorrectly joined into one item (with a
4540 colon in the middle).
4546 1. Added -C and -D options to the exinext utility, mainly to make it easier
4547 to include in the automated testing, but these could be helpful when
4548 multiple configurations are in use.
4550 2. The exinext utility was not formatting the output nicely when there was
4551 an alternate port involved in the retry record key, nor when there was a
4552 message id as well (for retries that were specific to a specific message
4553 and a specific host). It was also confused by IPv6 addresses, because of
4554 the additional colons they contain. I have fixed the IPv4 problem, and
4555 patched it up to do a reasonable job for IPv6.
4557 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4558 delivery, the log line now contains "pipelined" if PIPELINING was used.
4560 4. An SMTP transport process used to panic and die if the bind() call to set
4561 an explicit outgoing interface failed. This has been changed; it is now
4562 treated in the same way as a connect() failure.
4564 5. A reference to $sender_host_name in the part of a conditional expansion
4565 that was being skipped was still causing a DNS lookup. This no longer
4568 6. The def: expansion condition was not recognizing references to header lines
4569 that used bh_ and bheader_.
4571 7. Added the _cache feature to named lists.
4573 8. The code for checking quota_filecount in the appendfile transport was
4574 allowing one more file than it should have been.
4576 9. For compatibility with Sendmail, the command line option
4584 and sets the incoming protocol and host name (for trusted callers). The
4585 host name and its colon can be omitted when only the protocol is to be set.
4586 Note the Exim already has two private options, -pd and -ps, that refer to
4587 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4588 "s", but I don't think that's a major issue.
4590 10. A number of refactoring changes to the code, none of which should affect
4593 (a) The number of logging options was getting close to filling up the
4594 32-bit word that was used as a bit map. I have split them into two classes:
4595 those that are passed in the argument to log_write(), and those that are
4596 only ever tested independently outside of that function. These are now in
4597 separate 32-bit words, so there is plenty of room for expansion again.
4598 There is no change in the user interface or the logging behaviour.
4600 (b) When building, for example, log lines, the code previously used a
4601 macro that called string_cat() twice, in order to add two strings. This is
4602 not really sufficiently general. Furthermore, there was one instance where
4603 it was actually wrong because one of the argument was used twice, and in
4604 one call a function was used. (As it happened, calling the function twice
4605 did not affect the overall behaviour.) The macro has been replaced by a
4606 function that can join an arbitrary number of extra strings onto a growing
4609 (c) The code for expansion conditions now uses a table and a binary chop
4610 instead of a serial search (which was left over from when there were very
4611 few conditions). Also, it now recognizes conditions like "pam" even when
4612 the relevant support is not compiled in: a suitably worded error message is
4613 given if an attempt is made to use such a condition.
4615 11. Added ${time_interval:xxxxx}.
4617 12. A bug was causing one of the ddress fields not to be passed back correctly
4618 from remote delivery subprocesses. The field in question was not being
4619 subsequently used, so this caused to problems in practice.
4621 13. Added new log selectors queue_time and deliver_time.
4623 14. Might have fixed a bug in maildirsizefile handling that threw up
4624 "unexpected character" debug warnings, and recalculated the data
4625 unnecessarily. In any case, I expanded the warning message to give more
4628 15. Added the message "Restricted characters in address" to the statements in
4629 the default ACL that block characters like @ and % in local parts.
4631 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4632 Three changes have been made:
4634 (a) There was a serious bug; a negative response to MAIL caused the whole
4635 recipient domain to be cached as invalid, thereby blocking all messages
4636 to all local parts at the same domain, from all senders. This bug has
4637 been fixed. The domain is no longer cached after a negative response to
4638 MAIL if the sender used is not empty.
4640 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4643 (c) A new callout option, "use_sender" has been added for people who want
4644 the modified behaviour.
4650 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4653 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4654 indeed breaks things for older releases.
4656 3. Added additional logging to the case where there is a problem reading data
4657 from a filter that is running in a subprocess using a pipe, in order to
4658 try to track down a specific problem.
4660 4. Testing facility fudge: when running in the test harness and attempting
4661 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4662 getting "No route to host". Convert this to a timeout.
4664 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4667 6. Some OS don't have socklen_t but use size_t instead. This affects the
4668 fifth argument of getsockopt() amongst other things. This is now
4669 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4670 can be set for individual OS. I have set it for SunOS5, OSF1, and
4671 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4672 some earlier ones do not.
4674 7. Change 4.30/15 was not doing the test caselessly.
4676 8. The standard form for an IPv6 address literal was being rejected by address
4677 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4678 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4679 this, as well as the form without the "IPv6" on the front (but only when
4680 address literals are enabled, of course).
4682 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4684 10. Exim crashed if a message with an empty sender address specified by -f
4685 encountered a router with an errors_to setting. This could be provoked only
4686 by a command such as
4690 where an empty string was supplied; "<>" did not hit this bug.
4692 11. Installed PCRE release 4.5.
4694 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4695 remained set. It is now erased.
4697 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4698 times from message ids (which are base 36 rather than the normal 62).
4700 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4701 were being counted as actual protocol errors, and logged if the log
4702 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4703 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4704 and DATA following a set of rejected RCPTs do not count as protocol errors.
4705 In other words, Exim assumes they were pipelined, though this may not
4706 actually be the case. Of course, in all cases the client gets an
4707 appropriate error code.
4709 15. If a lookup fails in an ACL condition, a message about the failure may
4710 be available; it is used if testing the ACL cannot continue, because most
4711 such messages specify what the cause of the deferral is. However, some
4712 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4713 that caused an old message to be retained and used if a later statement
4714 caused a defer, replacing the real cause of the deferral.
4716 16. If an IP address had so many PTR records that the DNS lookup buffer
4717 was not large enough to hold them, Exim could crash while trying to process
4718 the truncated data. It now detects and logs this case.
4720 17. Further to 4.21/58, another change has been made: if (and only if) the
4721 first line of a message (the first header line) ends with CRLF, a bare LF
4722 in a subsequent header line has a space inserted after it, so as not to
4723 terminate the header.
4725 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4726 unnecessarily, used atoi() instead of strtol(), and didn't check the
4727 termination when getting file sizes from file names by regex.
4729 19. Completely re-implemented the support for maildirsize files, in the light
4730 of a number of problems with the previous contributed implementation
4731 (4.30/29). In particular:
4733 . If the quota is zero, the maildirsize file is maintained, but no quota is
4736 . If the maildir directory does not exist, it is created before any attempt
4737 to write a maildirsize file.
4739 . The quota value in the file is just a cache; if the quota is changed in
4740 the transport, the new value overrides.
4742 . A regular expression is available for excluding directories from the
4745 20. The autoreply transport checks the characters in options that define the
4746 message's headers; it allows continued headers, but it was checking with
4747 isspace() after an embedded newline instead of explicitly looking for a
4750 21. If all the "regular" hosts to which an address was routed had passed their
4751 expiry times, and had not reached their retry times, the address was
4752 bounced, even if fallback hosts were defined. Now Exim should go on to try
4755 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4756 equivalent code in the SMTP transport. Some hosts send humungous responses
4757 to HELO/EHLO, more than 1024 it seems.
4759 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4760 gives compiler warnings in some environments. I've now done it "properly",
4763 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4764 (because of problems with the built-in one) was declared to return uschar *
4765 instead of char *, causing compiler failure.
4767 25. Fixed a file descriptor leak when processing alias/forward files.
4769 26. Fixed a minor format string issue in dbfn.c.
4771 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4773 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4774 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4777 29. When a sender address is verified, it is cached, to save repeating the test
4778 when there is more than one recipient in a message. However, when the
4779 verification involves a callout, it is possible for different callout
4780 options to be set for different recipients. It is too complicated to keep
4781 track of this in the cache, so now Exim always runs a verification when a
4782 callout is required, relying on the callout cache for the optimization.
4783 The overhead is duplication of the address routing, but this should not be
4786 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4787 to be verified with callout=postmaster, and the main callout worked but the
4788 postmaster check failed, the verification correctly failed. However, if a
4789 subsequent RCPT command asked for sender verification *without* the
4790 postmaster check, incorrect caching caused this verification also to fail,
4793 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4794 it was not caching the DNS options (qualify_single, search_parents) that
4795 were used when the lookup failed. A subsequent lookup with different
4796 options therefore always gave the same answer, though there were cases
4797 where it should not have. (Example: a "domains = !$mx_any" option on a
4798 dnslookup router: the "domains" option is always processed without any
4799 widening, but the router might have qualify_single set.) Now Exim uses the
4800 cached value only when the same options are set.
4802 32. Added John Jetmore's "exipick" utility to the distribution.
4804 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4805 than a timeout (e.g. a certificate is required, and is not provided), an
4806 Exim server now closes the connection immediately. Previously it waited for
4807 the client to close - but if the client is SSL, it seems that they each
4808 wait for each other, leading to a delay before one of them times out.
4810 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4811 maintained 0.8.x compatibility because I don't think many are using it, and
4812 it is clearly obsolete.
4814 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4817 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4818 client certificate was expired. A simple patch fixes this, though I don't
4819 understand the full logic of why the verify callback is called multiple
4822 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4823 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4824 which causes problems with some clients (such as the Certicom SSL Plus
4825 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4826 disables the coutermeasure allowing Eudora to connect."
4828 38. Exim was not checking that a write() to a log file succeeded. This could
4829 lead to Bad Things if a log got too big, in particular if it hit a file
4830 size limit. Exim now panics and dies if it cannot write to a log file, just
4831 as it does if it cannot open a log file.
4833 39. Modified OS/Makefile-Linux so that it now contains
4835 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4837 The two -D definitions ensure that Exim is compiled with large file
4838 support, which makes it possible to handle log files that are bigger than
4841 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4842 instance) a domain was checked against a named list that involved a lookup,
4843 causing $domain_data to be set, then another domain was checked against the
4844 same list, then the first domain was re-checked, the value of $domain_data
4845 after the final check could be wrong. In particular, if the second check
4846 failed, it could be set empty. This bug probably also applied to
4849 41. The strip_trailing_dot option was not being applied to the address given
4850 with the -f command-line option.
4852 42. The code for reading a message's header from the spool was incrementing
4853 $received_count, but never initializing it. This meant that the value was
4854 incorrect (doubled) while delivering a message in the same process in which
4855 it was received. In the most common configuration of Exim, this never
4856 happens - a fresh exec is done - but it can happen when
4857 deliver_drop_privilege is set.
4859 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4860 it now includes up to 150 characters of the unexpected data in the log
4863 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4864 and building data strings. The size of both of these buffers was 10 000
4865 bytes - far larger than anybody would *ever* want, thought I. Needless to
4866 say, somebody hit the limit. I have increased the maximum line length to
4867 20 000 and the maximum data length of concatenated lines to 100 000. I have
4868 also fixed two bugs, because there was no checking on these buffers. Tsk,
4869 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4870 buffer is too small.
4872 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4873 lsearch lookups. Now it does.
4875 46. When parsing a route_list item in a manualroute router, a fixed-length
4876 buffer was used for the list of hosts. I made this 1024 bytes long,
4877 thinking that nobody would ever have a list of hosts that long. Wrong.
4878 Somebody had a whole pile of complicated expansion conditions, and the
4879 string was silently truncated, leading to an expansion error. It turns out
4880 that it is easier to change to an unlimited length (owing to other changes
4881 that have happened since this code was originally written) than to build
4882 structure for giving a limitation error. The length of the item that
4883 expands into the list of hosts is now unlimited.
4885 47. The lsearch lookup could not handle data where the length of text line was
4886 more than 4095 characters. Such lines were truncated, leading to shortened
4887 data being returned. It should now handle lines of any length.
4889 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4890 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4893 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4894 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4895 no longer happens in comment lines. A list of replacements is now placed
4896 at the head of all of the source files, except those whose only change is
4897 to replace PERL_COMMAND in the very first #! line.
4899 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4900 messages on the queue, with a bottom-up merge sort, using code contributed
4901 by Michael Haardt. This should make operations like -bp somewhat faster on
4902 large queues. It won't affect queue runners, except when queue_run_in_order
4905 51. Installed eximstats 1.31 in the distribution.
4907 52. Added support for SRV lookups to the dnslookup router.
4909 53. If an ACL referred to $message_body or $message_body_end, the value was not
4910 reset for any messages that followed in the same SMTP session.
4912 54. The store-handling optimization for building very long strings was not
4913 differentiating between the different store pools. I don't think this
4914 actually made any difference in practice, but I've tidied it.
4916 55. While running the routers to verify a sender address, $sender_address
4917 was still set to the sender address. This is wrong, because when routing to
4918 send a bounce to the sender, it would be empty. Therefore, I have changed
4919 it so that, while verifying a sender address, $sender_address is set to <>.
4920 (There is no change to what happens when verifying a recipient address.)
4922 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4923 target A or AAAA records (if not already returned) without resetting the
4924 qualify_single or search_parents options of the DNS resolver. These are
4925 inappropriate in this case because the targets of MX and SRV records must
4926 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4927 target that, when qualified, matched something in the local domain. These
4928 two options are now turned off when doing these lookups.
4930 57. It seems that at least some releases of Reiserfs (which does not have the
4931 concept of a fixed number of inodes) returns zero and not -1 for the
4932 number of available inodes. This interacted badly with check_spool_inodes,
4933 which assumed that -1 was the "no such thing" setting. What I have done is
4934 to check that the total number of inodes is greater than zero before doing
4935 the test of how many are available.
4937 58. When a "warn" ACL statement has a log_message modifier, the message is
4938 remembered, and not repeated. This is to avoid a lot of repetition when a
4939 message has many recipients that cause the same warning to be written.
4940 Howewer, Exim was preserving the list of already written lines for an
4941 entire SMTP session, which doesn't seem right. The memory is now reset if a
4942 new message is started.
4944 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4945 debugging output unless log_rewrite was also set.
4947 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4948 of (int)(handle) when we know that handle contains (void *)(-1).
4950 61. The Exim daemon panic-logs an error return when it closes the incoming
4951 connection. However "connection reset by peer" seems to be common, and
4952 isn't really an error worthy of noting specially, so that particular error
4955 62. When Exim is trying to find all the local interfaces, it used to panic and
4956 die if the ioctl to get the interface flags failed. However, it seems that
4957 on at least one OS (Solaris 9) it is possible to have an interface that is
4958 included in the list of interfaces, but for which you get a failure error
4959 for this call. This happens when the interface is not "plumbed" into a
4960 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4961 failure of the "get flags" call assumes that the interface is down.
4963 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4964 makes life easier for people who are doing arithmetic on fields extracted
4965 from dates, where you often get leading zeros that should not be
4966 interpreted as octal.
4968 64. Added qualify_domain to the redirect router, to override the global
4971 65. If a pathologically long header line contained very many addresses (the
4972 report of this problem mentioned 10 000) and each of them was rewritten,
4973 Exim could use up a very large amount of memory. (It kept on making new
4974 copies of the header line as it rewrote, and never released the old ones.)
4975 At the expense of a bit more processing, the header rewriting function has
4976 been changed so that it no longer eats memory in this way.
4978 66. The generation of the Received: header has been moved from the time that a
4979 message starts to be received, to the time that it finishes. The timestamp
4980 in the Received: header should now be very close to that of the <= log
4981 line. There are two side-effects of this change:
4983 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4984 logged header lines no longer include the local Received: line, because
4985 it has not yet been created. The same applies to a copy of the message
4986 that is returned to a non-SMTP sender when a message is rejected.
4988 (b) When a filter file is tested using -bf, no additional Received: header
4989 is added to the test message. After some thought, I decided that this
4992 This change does not affect the value of $received_for. It is still set
4993 after address rewriting, but before local_scan() is called.
4995 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4997 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4998 gave an unhelpful panic error message, and a defer error. I have managed to
4999 change this behaviour so that it now rejects any supplied certificate,
5000 which seems right, as the list of acceptable certificates is empty.
5002 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5003 gave an unhelpful defer error. I have not managed to make this reject any
5004 supplied certificates, but the error message it gives is "no certificate
5005 supplied", which is not helpful.
5007 70. exigrep's output now also includes lines that are not associated with any
5008 message, but which match the given pattern. Implemented by a patch from
5009 Martin Sluka, which also tidied up the Perl a bit.
5011 71. Recipient callout verification, like sender verification, was using <> in
5012 the MAIL FROM command. This isn't really the right thing, since the actual
5013 sender may affect whether the remote host accepts the recipient or not. I
5014 have changed it to use the actual sender in the callout; this means that
5015 the cache record is now keyed on a recipient/sender pair, not just the
5016 recipient address. There doesn't seem to be a real danger of callout loops,
5017 since a callout by the remote host to check the sender would use <>.
5018 [SEE ABOVE: changed after hitting problems.]
5020 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5021 temporary errors. However, in the case of such a code being given after
5022 the end of a data transmission (i.e. after ".") Exim was failing to write
5023 a retry record for the message. (Yes, there was some broken host that was
5024 actually sending 8xx at this point.)
5026 73. An unknown lookup type in a host list could cause Exim to panic-die when
5027 the list was checked. (An example that provoked this was putting <; in the
5028 middle of a list instead of at the start.) If this happened during a DATA
5029 ACL check, a -D file could be left lying around. This kind of configuration
5030 error no longer causes Exim to die; instead it causes a defer errror. The
5031 incident is still logged to the main and panic logs.
5033 74. Buglet left over from Exim 3 conversion. The message "too many messages
5034 in one connection" was written to the rejectlog but not the mainlog, except
5035 when address rewriting (yes!) was being logged.
5037 75. Added write_rejectlog option.
5039 76. When a system filter was run not as root (that is, when system_filter_user
5040 was set), the values of the $n variables were not being returned to the
5041 main process; thus, they were not subsequently available in the $sn
5044 77. Added +return_path_on_delivery log selector.
5046 78. A connection timeout was being treated differently from recipients deferred
5047 when testing hosts_max_try with a message that was older than the host's
5048 retry timeout. (The host should not be counted, thus allowing all hosts to
5049 be tried at least once before bouncing.) This may have been the cause of an
5050 occasionally reported bug whereby a message would remain on the queue
5051 longer than the retry timeout, but would be bounced if a delivery was
5052 forced. I say "may" because I never totally pinned down the problem;
5053 setting up timeout/retry tests is difficult. See also the next item.
5055 79. The ultimate address timeout was not being applied to errors that involved
5056 a combination of host plus message (for example, a timeout on a MAIL
5057 command). When an address resolved to a number of possible hosts, and they
5058 were not all tried for each delivery (e.g. because of hosts_max_try), a
5059 message could remain on the queue longer than the retry timeout.
5061 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5064 81. Fixed an obscure SMTP outgoing bug which required at least the following
5065 conditions: (a) there was another message waiting for the same server;
5066 (b) the server returned 5xx to all RCPT commands in the first message so
5067 that the message was not completed; (c) the server dropped the connection
5068 or gave a negative response to the RSET that Exim sends to abort the
5069 transaction. The observed case was a dropped connection after DATA that had
5070 been sent in pipelining mode. That is, the server had advertised PIPELINING
5071 but was not implementing it correctly. The effect of the bug was incorrect
5072 behaviour, such as trying another host, and this could lead to a crash.
5078 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5079 and daemon.c were passed as pointers to ints; they should have been
5080 pointers to socklen_t variables (which are typically unsigned ints).
5082 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5085 3. Fixed a really odd bug that affected only the testing scheme; patching a
5086 certain fixed string in the binary changed the value of another string that
5087 happened to be identical to the end of the original first string.
5089 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5090 name", it returns that address as the IP address. On some operating
5091 systems (e.g. Solaris), it also passes back the IP address string as the
5092 "host name". However, on others (e.g. Linux), it passes back an empty
5093 string. Exim wasn't checking for this, and was changing the host name to an
5094 empty string, assuming it had been canonicized.
5096 5. Although rare, it is permitted to have more than one PTR record for a given
5097 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5098 all the names associated with an address, because they do in Solaris.
5099 However, it seems that they do not in Linux for data that comes from the
5100 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5101 I found this out when I moved to a new Linux workstation and tried to run
5102 the Exim test suite.
5104 To get round this problem I have changed the code so that it now does its
5105 own call to the DNS to look up PTR records when searching for a host name.
5106 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5107 addresses that are only in /etc/hosts are still found.
5109 This behaviour is, however, controlled by an option called host_lookup_
5110 order, which defaults to "bydns:byaddr". If people want to use the other
5111 order, or indeed, just use one or the other means of lookup, they can
5112 specify it in this variable.
5114 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5115 some operating systems, this comes back from gethostbyaddr() as an empty
5116 string, and this is what Exim used to test for. However, it seems that in
5117 other systems, "." is yielded. Exim now tests for this case too.
5119 7. The values of check_spool_space and check_log_space are now held internally
5120 as a number of kilobytes instead of an absolute number of bytes. If a
5121 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5122 kilobyte. This means that much larger values can be stored.
5124 8. Exim monitor: an attempt to get the action menu when not actually pointing
5125 at a message produces an empty menu entitled "No message selected". This
5126 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5127 no entries in it ("Shell widget menu has zero width and/or height"). So I
5128 have added a single, blank menu entry in this case.
5130 9. Added ${quote_local_part.
5132 10. MIME decoding is now applied to the contents of Subject: header lines when
5135 11. Now that a reference to $sender_host_address automatically causes a reverse
5136 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5137 host lookup before query-style lookups in lists that might use this
5138 variable. This has therefore been abolished, and the "net-" prefix is no
5139 longer necessary for query-style lookups.
5141 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5142 have been a typo for LFLAGS, so it has been changed.
5144 13. The install script calls Exim with "-C /dev/null" in order to find the
5145 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5146 to be output. Howeve, since Exim outputs its version number before the
5147 error, it didn't break the script. It just looked ugly. I fixed this by
5148 always allowing "-C /dev/null" if the caller is root.
5150 14. Ignore overlarge ACL variable number when reading spool file - insurance
5151 against a later release with more variables having written the file.
5153 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5154 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5155 this, as well as the form without the "IPv6" on the front.
5157 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5158 OS/Makefile-Darwin file.
5160 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5161 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5163 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5164 variable. However, some OS use ints for inodes. I've added cast to long int
5165 to get rid of the compiler warning.
5167 19. I had forgotten to lock out "/../" in configuration file names when
5168 ALT_CONFIG_PREFIX was set.
5170 20. Routers used for verification do not need to specify transports. However,
5171 if such a router generated a host list, and callout was configured, Exim
5172 crashed, because it could not find a port number from the (non-existent)
5173 transport. It now assumes port 25 in this circumstance.
5175 21. Added the -t option to exigrep.
5177 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5178 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5179 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5180 them is set, LOOKUP_LSEARCH is forced.
5182 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5184 24. Added sender and host information to the "rejected by local_scan()" log
5185 line; previously there was no indication of these.
5187 25. Added .include_if_exists.
5189 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5190 Linux. It turns out that not all file systems support this. Apparently some
5191 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5192 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5193 not supported on the file descriptor, is now ignored when Exim is trying to
5194 sync a directory. This applies only to Linux.
5196 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5198 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5199 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5202 29. Added support for maildirsize files from supplied patch (modified a bit).
5204 30. The use of :fail: followed by an empty string could lead Exim to respond to
5205 sender verification failures with (e.g.):
5207 550 Verification failed for <xxx>
5208 550 Sender verify failed
5210 where the first response line was missing the '-' that indicates it is not
5211 the final line of the response.
5213 31. The loop for finding the name of the user that called Exim had a hardwired
5214 limit of 10; it now uses the value of finduser_retries, which is used for
5215 all other user lookups.
5217 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5220 33. Exim was neglecting to zero errno before one call of strtol() when
5221 expanding a string and expecting an integer value. On some systems this
5222 resulted in spurious "integer overflow" errors. Also, it was casting the
5223 result into an int without checking.
5225 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5226 did not work. The code looks as if it has *never* worked, though it appears
5227 to have been documented since at least releast 1.62. I have made it work.
5229 35. The "timeout_DNS" error in retry rules, also documented since at least
5230 1.62, also never worked. As it isn't clear exactly what this means, and
5231 clearly it isn't a major issue, I have abolished the feature by treating it
5232 as "timeout", and writing a warning to the main and panic logs.
5234 36. The display of retry rules for -brt wasn't always showing the error code
5237 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5238 timeout_connect_A, timeout_connect_MX.
5240 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5241 to the empty sender.
5243 39. The daemon was not analysing the content of -oX till after it had closed
5244 stderr and disconnected from the controlling terminal. This meant that any
5245 syntax errors were only noted on the panic log, and the return code from
5246 the command was 0. By re-arranging the code a little, I've made the
5247 decoding happen first, so such errors now appear on stderr, and the return
5248 code is 1. However, the actual setting up of the sockets still happens in
5249 the disconnected process, so errors there are still only recorded on the
5252 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5253 connections (as happens on some IP stacks) was logged at start up time as
5254 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5255 it from "IPv6 and IPv4", which means that two separate sockets are being
5258 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5259 says whether AF_INET or AF_INET6 was passed as an argument.
5261 42. Exiwhat output was messed up when time zones were included in log
5264 43. Exiwhat now gives more information about the daemon's listening ports,
5265 and whether -tls-on-connect was used.
5267 44. The "port" option of the smtp transport is now expanded.
5269 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5270 silently ignored. Now an error message is written to the main and panic
5273 46. There's a new ACL modifier called "logwrite" which writes to a log file
5274 as soon as it is encountered.
5276 47. Added $local_user_uid and $local_user_gid at routing time.
5278 48. Exim crashed when trying to verify a sender address that was being
5281 49. Exim was recognizing only a space character after ".include". It now also
5282 recognizes a tab character.
5284 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5285 extracting the relevant information from the specification. The man page no
5286 longer contains scrambled data for the -d option, and I've added a section
5287 at the front about calling Exim under different names.
5289 51. Added "extra_headers" argument to the "mail" command in filter files.
5291 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5294 53. Installed eximstats 1.29.
5296 54. Added transport_filter_timeout as a generic transport option.
5298 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5299 Cc: header lines. This was required by RFC 822, but it not required by RFC
5302 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5303 incoming messages that did not have them. Now it does so only if the
5304 message originates locally, that is, if there is no associated remote host
5305 address. When Resent- header lines are present, this applies to the Resent-
5306 lines rather than the non-Resent- lines.
5308 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5309 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5311 58. Messages for configuration errors now include the name of the main
5312 configuration file - useful now that there may be more than one file in a
5313 list (.included file names were always shown).
5315 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5316 for those rare installations that do not start the daemon as root or run it
5317 setuid root. I've cut out the call to initgroups() if the daemon is not
5320 60. The Exim user and group can now be bound into the binary as text strings
5321 that are looked up at the start of Exim's processing.
5323 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5325 62. Added $mailstore_basename variable.
5327 63. Installed patch to sieve.c from Michael Haardt.
5329 64. When Exim failed to open the panic log after failing to open the main log,
5330 the original message it was trying to log was written to stderr and debug
5331 output, but if they were not available (the usual case in production), it
5332 was lost. Now it is written to syslog before the two lines that record the
5333 failures to open the logs.
5335 65. Users' Exim filters run in subprocesses under the user's uid. It is
5336 possible for a "deliver" command or an alias in a "personal" command to
5337 provoke an address rewrite. If logging of address rewriting is configured,
5338 this fails because the process is not running as root or exim. There may be
5339 a better way of dealing with this, but for the moment (because 4.30 needs
5340 to be released), I have disabled address rewrite logging when running a
5341 filter in a non-root, non-exim process.
5347 1. The buildconfig auxiliary program wasn't quoting the value set for
5348 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5349 not defined. This bug was masked in 4.22 by the effect that was fixed in
5352 2. Some messages that were rejected after a message id was allocated were
5353 shown as "incomplete" by exigrep. It no longer does this for messages that
5354 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5356 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5357 have allow_domain_literals set, the ID did not get logged in the <= line.
5358 Domain literals are now always recognized in Message-ID: header lines.
5360 4. The first argument for a ${extract expansion item is the key name or field
5361 number. Leading and trailing spaces in this item were not being ignored,
5362 causing some misleading effects.
5364 5. When deliver_drop_privilege was set, single queue runner processes started
5365 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5366 same command in the process it spins off) were not dropping privilege.
5368 6. When the daemon running as "exim" started a queue runner, it always
5369 re-executed Exim in the spun-off process. This is a waste of effort when
5370 deliver_drop_privilege is set. The new process now just calls the
5371 queue-runner function directly.
5377 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5380 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5381 ignored. Though the use of -oP was forcing the writing of a pid file, it
5382 was always written to the default place.
5384 3. If the message "no IP address found for host xxxx" is generated during
5385 incoming verification, it is now followed by identification of the incoming
5386 connection (so you can more easily find what provoked it).
5388 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5390 5. Added some features to "harden" Exim a bit more against certain attacks:
5392 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5393 be put in Local/Makefile. This is like the never_users runtime option,
5394 but it cannot be overridden. The default setting is "root".
5396 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5397 prefix string with which any file named in a -C command line option
5400 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5401 is retained for -C and -D only if the caller of Exim is root. Without
5402 it, the exim user may also use -C and -D and retain privilege.
5404 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5405 command line option is disabled.
5407 6. Macro names set by the -D option must start with an upper case letter, just
5408 like macro names defined in the configuration file.
5410 7. Added "dereference=" facility to LDAP.
5412 8. Two instances of the typo "uknown" in the source files are fixed.
5414 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5415 the Configure-Makefile script screwed up while processing it.
5417 10. Incorporated PCRE 4.4.
5419 11. The SMTP synchronization check was not operating right at the start of an
5420 SMTP session. For example, it could not catch a HELO sent before the client
5421 waited for the greeting. There is now a check for outstanding input at the
5422 point when the greeting is written. Because of the duplex, asynchronous
5423 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5424 way, but not yet received, when the check is performed.
5426 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5427 on TCP/IP sockets, because this apparently causes some broken clients to
5430 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5431 unchanged) from the Cygwin maintainer.
5433 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5434 was supported instead of testing for mbx. Effectively a typo.
5436 15. The spa authenticator server code was not checking that the input it
5437 received was valid base64.
5439 16. The debug output line for the "set" modifier in ACLs was not showing the
5440 name of the variable that was being set.
5442 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5444 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5445 lookup on its own. Something else was needed to trigger the lookup. For
5446 example, a match in host_lookup or the need for a host name in a host list.
5447 Now, if $sender_host_name is referenced and the host name has not yet been
5448 looked up, a lookup is performed. If the lookup fails, the variable remains
5449 empty, and $host_lookup_failed is set to "1".
5451 19. Added "eqi" as a case-independent comparison operator.
5453 20. The saslauthd authentication condition could segfault if neither service
5454 nor realm was specified.
5456 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5457 error message that was logged was misleading, and incoming SMTP
5458 connections were dropped. The message is now more accurate, and temporary
5459 errors are given to SMTP connections.
5461 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5462 (except RSET) with a 421 error, until QUIT is received. However, it was
5463 failing to send a response to QUIT.
5465 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5466 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5467 ACL is now run after the helo_try_verify_hosts code.
5469 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5470 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5471 case-independent, so other case variants are also recognized). Apparently
5472 some systems use these upper case variants.
5474 25. If more than two messages were waiting for the same host, and a transport
5475 filter was specified for the transport, Exim sent two messages over the
5476 same TCP/IP connection, and then failed with "socket operation on non-
5477 socket" when it tried to send the third.
5479 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5481 27. The extern definition of crypt16() in expand.c was not being excluded when
5482 the OS had its own crypt16() function.
5484 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5485 as a preferred synonym for return_size_limit, both as an option and as an
5488 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5490 30. Changed the default configuration ACL to relax the local part checking rule
5491 for addresses that are not in any local domains. For these addresses,
5492 slashes and pipe symbols are allowed within local parts, but the sequence
5493 /../ is explicitly forbidden.
5495 31. SPA server authentication was not clearing the challenge buffer before
5498 32. log_message in a "warn" ACL statement was writing to the reject log as
5499 well as to the main log, which contradicts the documentation and doesn't
5500 seem right (because no rejection is happening). So I have stopped it.
5502 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5503 However, I am unable to do any testing of this.
5505 34. Fixed an infelicity in the appendfile transport. When checking directories
5506 for a mailbox, to see if any needed to be created, it was accidentally
5507 using path names with one or more superfluous leading slashes; tracing
5508 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5510 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5511 contents are added to the log line that is written for every discarded
5512 recipient. (Previously a log_message setting was ignored.)
5514 36. The ${quote: operator now quotes the string if it is empty.
5516 37. The install script runs exim in order to find its version number. If for
5517 some reason other than non-existence or emptiness, which it checks, it
5518 could not run './exim', it was installing it with an empty version number,
5519 i.e. as "exim-". This error state is now caught, and the installation is
5522 38. An argument was missing from the function that creates an error message
5523 when Exim fails to connect to the socket for saslauthd authentication.
5524 This could cause Exim to crash, or give a corrupted message.
5526 39. Added isip, isip4, and isip6 to ${if conditions.
5528 40. The ACL variables $acl_xx are now saved with the message, and can be
5529 accessed later in routers, transports, and filters.
5531 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5532 strings in the file are not string-expanded.
5534 42. If a MAIL command specified a SIZE value that was too large to fit into an
5535 int variable, the check against message_size_limit failed. Such values are
5536 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5537 day this will have to be increased, but I don't think I want to be around
5538 when emails are that large.
5545 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5546 iconv() is not standard in FreeBSD.
5548 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5549 IPv6 enabled. The observed symptom was a segmentation fault on return from
5550 the function os_common_find_running_interfaces() in src/os.c.
5552 3. In the check_special_case() function in daemon.c I had used "errno" as an
5553 argument name, which causes warnings on some systems. This was basically a
5554 typo, since it was named "eno" in the comments!
5556 4. The code that waits for the clock to tick (at a resolution of some fraction
5557 of a second) so as to ensure message-id uniqueness was always waiting for
5558 at least one whole tick, when it could have waited for less. [This is
5559 almost certainly not relevant at current processor speeds, where it is
5560 unlikely to ever wait at all. But we try to future-proof.]
5562 5. The function that sleeps for a time interval that includes fractions of a
5563 second contained a race. It did not block SIGALRM between setting the
5564 timer, and suspending (a couple of lines later). If the interval was short
5565 and the sigsuspend() was delayed until after it had expired, the suspension
5566 never ended. On busy systems this could lead to processes getting stuck for
5569 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5570 process, before it forks any delivery processes. The open lookup caching
5571 mechanism meant that the open file or database connection was passed into
5572 the delivery process. The problem was that delivery processes always tidy
5573 up cached lookup data. This could cause a problem for the next delivery
5574 process started by the queue runner, because the external queue runner
5575 process does not know about the closure. So the next delivery process
5576 still has data in the lookup cache. In the case of a file lookup, there was
5577 no problem because closing a file descriptor in a subprocess doesn't affect
5578 the parent. However, if the lookup was caching a connection to a database,
5579 the connection was closed, and the second delivery process was likely to
5580 see errors such as "PGSQL: query failed: server closed the connection
5581 unexpectedly". The problem has been fixed by closing all cached lookups
5582 in a queue runner before running a delivery process.
5584 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5585 seem to have the "const" qualifier which it has on other OS. I've
5588 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5589 *of the same type* (client or server) with the same public name that an
5590 error should be diagnosed.
5592 9. When Exim looked up a host name for an IP address, but failed to find the
5593 original IP address when looking up the host name (a safety check), it
5594 output the message "<ip address> does not match any IP for NULL", which was
5595 confusing, to say the least. The bug was that the host name should have
5596 appeared instead of "NULL".
5598 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5599 user that is built into the binary, and the -C or -D options is used, root
5600 privilege is dropped before the configuration file is read. In addition,
5601 logging is switched to stderr instead of the normal log files. If the
5602 configuration then re-defines the Exim user, the unprivileged environment
5603 is probably not what is expected, so Exim logs a panic warning message (but
5606 However, if deliver_drop_privilege is set, the unprivileged state may well
5607 be exactly what is intended, so the warning has been cut out in that case,
5608 and Exim is allowed to try to write to its normal log files.
5614 1. smtp_return_error_details was not giving details for temporary sender
5615 or receiver verification errors.
5617 2. Diagnose a configuration error if two authenticators have the same public
5620 3. Exim used not to create the message log file for a message until the first
5621 delivery attempt. This could be confusing when incoming messages were held
5622 for policy or load reasons. The message log file is now created at the time
5623 the message is received, and an initial "Received" line is written to it.
5625 4. The automatically generated man page for command line options had a minor
5626 bug that caused no ill effects; however, a more serious problem was that
5627 the procedure for building the man page automatically didn't always
5628 operate. Consequently, release 4.20 contains an out-of-date version. This
5629 shouldn't happen again.
5631 5. When building Exim with embedded Perl support, the script that builds the
5632 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5633 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5635 6. The freeze_tell option was not being used for messages that were frozen on
5636 arrival, either by an ACL or by local_scan().
5638 7. Added the smtp_incomplete_transaction log selector.
5640 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5641 was accepting AUTH without a new EHLO.
5643 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5644 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5647 10. Logging of TCP/IP connections (when configured) now happens in the main
5648 daemon process instead of the child process, so that the TCP/IP connection
5649 count is more accurate (but it can never be perfect).
5651 11. The use of "drop" in a nested ACL was not being handled correctly in the
5652 outer ACL. Now, if condition failure induced by the nested "drop" causes
5653 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5654 or "require"), the connection is dropped.
5656 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5657 that yield "discard" can now be used with an "accept" or a "discard" verb,
5658 but an error is generated for any others (because I can't see a useful way
5659 to define what should happen).
5661 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5662 are now processed in the same way as lines in the Exim configuration file.
5663 In particular, continuation lines are supported.
5665 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5667 15. Added -ti meaning -t -i.
5669 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5670 domains, and warn by logging if others are found.
5672 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5673 returned from GIFCONF when Exim is trying to find the list of interfaces on
5674 a host. The code in os.c has been modified to copy each ifreq to an aligned
5675 structure in all cases.
5677 Also, in some cases, the returned ifreq's were being copied to a 'struct
5678 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5679 means the last couple of bytes of an IPv6 address could be chopped if the
5680 ifreq contained only a normal sockaddr (14 bytes storage).
5682 18. Named domain lists were not supported in the hosts_treat_as_local option.
5683 An entry such as +xxxx was not recognized, and was treated as a literal
5686 19. Ensure that header lines added by a DATA ACL are included in the reject log
5687 if the ACL subsequently rejects the message.
5689 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5690 MD5 (which is deprecated).
5692 21. When testing a filter file using -bf, Exim was writing a message when it
5693 took the sender from a "From " line in the message, but it was not doing so
5694 when it took $return_path from a Return-Path: header line. It now does.
5696 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5697 with a valid header line field name (a series of printing characters
5698 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5700 23. Changed "disc" in the source to "disk" to conform to the documentation and
5701 the book and for uniformity.
5703 24. Ignore Sendmail's -Ooption=value command line item.
5705 25. When execve() failed while trying to run a command in a pipe transport,
5706 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5707 could be confused with a return value of 69 from the command itself. This
5708 has been changed to 127, the value the shell returns if it is asked to run
5709 a non-existent command. The wording for the related log line suggests a
5710 non-existent command as the problem.
5712 26. If received_header_text expands to an empty string, do not add a Received:
5713 header line to the message. (Well, it adds a token one on the spool, but
5714 marks it "old" so that it doesn't get used or transmitted.)
5716 27. Installed eximstats 1.28 (addition of -nt option).
5718 28. There was no check for failure on the call to getsockname() in the daemon
5719 code. This can fail if there is a shortage of resources on the system, with
5720 ENOMEM, for example. A temporary error is now given on failure.
5722 29. Contrary to the C standard, it seems that in some environments, the
5723 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5724 program. Exim now does this explicitly; it affects the formatting of
5725 timestamps using strftime().
5727 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5728 complaints. I've now initialized all the variables, to avoid this.
5730 32. Header lines added by a system filter were not being "seen" during
5731 transport-time rewrites.
5733 33. The info_callback() function passed to OpenSSL is set up with type void
5734 (*)(SSL *, int, int), as described somewhere. However, when calling the
5735 function (actually a macro) that sets it up, the type void(*)() is
5736 expected. I've put in a cast to prevent warnings from picky compilers.
5738 34. If a DNS black list lookup found a CNAME record, but there were no A
5739 records associated with the domain it pointed at, Exim crashed.
5741 35. If a DNS black list lookup returned more than one A record, Exim ignored
5742 all but the first. It now scans all returned addresses if a particular IP
5743 value is being sought. In this situation, the contents of the
5744 $dnslist_value variable are a list of all the addresses, separated by a
5747 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5748 to accept a host name and all its aliases if the forward lookup for any of
5749 them yielded the IP address of the incoming connection. Now it accepts only
5750 those names whose forward lookup yields the correct IP address. Any other
5751 names are discarded. This closes a loophole whereby a rogue DNS
5752 administrator could create reverse DNS records to break through a
5753 wildcarded host restriction in an ACL.
5755 37. If a user filter or a system filter that ran in a subprocess used any of
5756 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5757 the wrong values were passed to the pipe command ($thisaddress had the
5758 value of $0, $0 had the value of $1, etc). This bug was introduced by
5759 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5761 38. Improved the line breaking for long SMTP error messages from ACLs.
5762 Previously, if there was no break point between 40 and 75 characters, Exim
5763 left the rest of the message alone. Two changes have been made: (a) I've
5764 reduced the minimum length to 35 characters; (b) if it can't find a break
5765 point between 35 and 75 characters, it looks ahead and uses the first one
5766 that it finds. This may give the occasional overlong line, but at least the
5767 remaining text gets split now.
5769 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5770 file descriptors might be low, and that setting 1000 would always raise it.
5771 It turns out that in some environments, the limit is already over 1000 and
5772 that lowering it causes trouble. So now Exim takes care not to decrease it.
5774 40. When delivering a message, the value of $return_path is set to $sender_
5775 address at the start of routing (routers may change the value). By an
5776 oversight, this default was not being set up when an address was tested by
5777 -bt or -bv, which affected the outcome if any router or filter referred to
5780 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5781 log only when "message" or "log_message" are set. However, if one of the
5782 conditions was an address verification, or a call to a nested ACL, the
5783 messages generated by the underlying test were being passed through. This
5784 no longer happens. The underlying message is available in $acl_verify_
5785 message for both "message" and "log_message" expansions, so it can be
5786 passed through if needed.
5788 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5789 new expansion $bh_ to give the encoded byte string without charset
5790 translation. Translation happens only if iconv() is available; HAVE_ICONV
5791 indicates this at build time. HEADERS_CHARSET gives the charset to
5792 translate to; headers_charset can change it in the configuration, and
5793 "headers charset" can change it in an individual filter file.
5795 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5796 that creates RFC 2047 encoded "words" labels them as that charset instead
5797 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5798 expansion operator; (ii) when Exim creates a From: line for a local
5799 message; (iii) when a header line is rewritten to include a "phrase" part.
5801 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5802 buggy, causing it to skip the first lines of messages whose message ID
5803 ended in 'D'. This would not have bitten before Exim release 4.14, because
5804 message IDs were unlikely to end in 'D' before then. The effect was to have
5805 incorrect size information for certain domains.
5807 45. #include "config.h" was missing at the start of the crypt16.c module. This
5808 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5811 46. If there was a timeout during a "random" callout check, Exim treated it as
5812 a failure of the random address, and carried on sending RSET and the real
5813 address. If the delay was just some slowness somewhere, the response to the
5814 original RCPT would be taken as a response to RSET and so on, causing
5815 mayhem of various kinds.
5817 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5818 when I implemented it. It didn't allow for the fact that some option values
5819 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5820 the right test for positive values.
5822 48. Domain names in DNS records are case-independent. Exim always looks them up
5823 in lower case. Some resolvers return domain names in exactly the case they
5824 appear in the zone file, that is, they may contain uppercase letters. Not
5825 all resolvers do this - some return always lower case. Exim was treating a
5826 change of case by a resolver as a change of domain, similar to a widening
5827 of a domain abbreviation. This triggered its re-routing code and so it was
5828 trying to route what was effectively the same domain again. This normally
5829 caused routing to fail (because the router wouldn't handle the domain
5830 twice). Now Exim checks for this case specially, and just changes the
5831 casing of the domain that it ultimately uses when it transmits the message
5834 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5837 50. If a filter generated a file delivery with a non-absolute name (possible if
5838 no home directory exists for the router), the forbid_file option was not
5841 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5842 the existing equality matching.
5844 52. Exim was using ints instead of ino_t variables in some places where it was
5845 dealing with inode numbers.
5847 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5848 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5849 called TMPDIR, and if it finds it is different, it changes its value.
5851 54. The smtp_printf() function is now made available to local_scan() so
5852 additional output lines can be written before returning. There is also an
5853 smtp_fflush() function to enable the detection of a dropped connection.
5854 The variables smtp_input and smtp_batched_input are exported to
5857 55. Changed the default runtime configuration: the message "Unknown user"
5858 has been removed from the ACL, and instead placed on the localuser router,
5859 using the cannot_route_message feature. This means that any verification
5860 failures that generate their own messages won't get overridden. Similarly,
5861 the "Unrouteable address" message that was in the ACL for unverifiable
5862 relay addresses has also been removed.
5864 56. Added hosts_avoid_esmtp to the smtp transport.
5866 57. The exicyclog script was not checking for the esoteric option
5867 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5868 will work only if exicyclog is run under the appropriate euid.
5870 58. Following a discussion on the list, the rules by which Exim recognises line
5871 endings on incoming messages have been changed. The -dropcr and drop_cr
5872 options are now no-ops, retained only for backwards compatibility. The
5873 following line terminators are recognized: LF CRLF CR. However, special
5874 processing applies to CR:
5876 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5877 nor a local message in the state where . is a terminator.
5879 (ii) If a bare CR is encountered in a header line, an extra space is added
5880 after the line terminator so as not to end the header. The reasoning
5881 behind this is that bare CRs in header lines are most likely either
5882 to be mistakes, or people trying to play silly games.
5884 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5885 was being incorrectly given as 18 bytes larger than it should have been.
5886 This is a VOB (very old bug).
5888 60. This may never have affected anything current, but just in case it has:
5889 When the local host is found other than at the start of a list of hosts,
5890 the local host, those with the same MX, and any that follow, are discarded.
5891 When the list in question was part of a longer list of hosts, the following
5892 hosts (not currently being processed) were also being discarded. This no
5893 longer happens. I'm not sure if this situation could ever has previously
5896 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5899 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5900 header. This is something that is recommended in a new Internet Draft, and
5901 is something that is documented as being done by Sendmail. There are two
5902 possible values. For messages generated by the autoreply transport, Exim
5905 Auto-Submitted: auto-replied
5907 whereas for all other generated messages (e.g. bounces) it adds
5909 Auto-Submitted: auto-generated
5911 63. The "personal" condition in filters now includes a test for the
5912 Auto-Submitted: header. If it contains the string "auto-" the message it
5913 not considered personal.
5915 64. Added rcpt_include_affixes as a generic transport option.
5917 65. Added queue_only_override (default true).
5919 66. Added the syslog_duplication option.
5921 67. If what should have been the first header line of a message consisted of
5922 a space followed by a colon, Exim was mis-interpreting it as a header line.
5923 It isn't of course - it is syntactically invalid and should therefore be
5924 treated as the start of the message body. The misbehaviour could have
5925 caused a number of strange effects, including loss of data in subsequent
5926 header lines, and spool format errors.
5928 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5929 client host had authenticated. This control can now be exercised by an ACL
5930 for more flexibility.
5932 69. By default, callouts do not happen when testing with -bh. There is now a
5933 variant, -bhc, which does actually run the callout code, including
5934 consulting and updating the callout cache.
5936 70. Added support for saslauthd authentication, courtesy of Alexander
5939 71. If statvfs() failed on the spool or log directories while checking their
5940 size for availability, Exim confusingly gave the error "space shortage".
5941 Furthermore, in debugging mode it crashed with a floating point exception.
5942 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5943 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5944 serious problem, Exim now writes to the main and panic logs when this
5945 happens, with details of the failure. It then refuses to accept the
5946 incoming message, giving the message "spool directory problem" or "log
5947 directory problem" with a 421 code for SMTP messages.
5949 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5950 0, 1, and 2 exist, because some OS complain for execs without them (see
5951 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5952 descriptors. However, the code omitted to check that the open succeeded,
5953 causing mysterious errors if for some reason the permissions on /dev/null
5954 got screwed. Now Exim writes a message to the main and panic logs, and
5955 bombs out if it can't open /dev/null.
5957 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5958 interact so that it is all more flexible. It is supposed to remain
5959 backwards compatible. Also added extra_local_interfaces.
5961 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5962 to bomb out with an assertion failure - to the client this appears as a
5963 connection drop. This problem occurs in the part of the code that was taken
5964 from the Samba project. Fortunately, the assertion is in a very simple
5965 function, so I have fixed this by reproducing the function inline in the
5966 one place where it is called, and arranging for authentication to fail
5967 instead of killing the process with assert().
5969 75. The SPA client code was not working when the server requested OEM rather
5970 than Unicode encoding.
5972 76. Added code to make require_files with a specific uid setting more usable in
5973 the case where statting the file as root fails - usually a non-root-mounted
5974 NFS file system. When this happens and the failure is EACCES, Exim now
5975 forks a subprocess and does the per-uid checking as the relevant uid.
5977 77. Added process_log_path.
5979 78. If log_file_path was not explicitly set, a setting of check_log_space or
5980 check_log_inodes was ignored.
5982 79. If a space check for the spool or log partitions fails, the incident is now
5983 logged. Of course, in the latter case the data may get lost...
5985 80. Added the %p formatting code to string_format() so that it can be used to
5986 print addresses in debug_print(). Adjusted all the address printing in the
5987 debugging in store.c to use %p rather than %d.
5989 81. There was a concern that a line of code in smtp_in.c could overflow a
5990 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5991 initially expressed, the concern was not well-founded, because trailing
5992 spaces are removed early. However, if the trailing spaces were followed by
5993 a NULL, they did not get removed, so the overflow was possible. Two fixes
5996 (a) I re-wrote the offending code in a cleaner fashion.
5997 (b) If an incoming SMTP command contains a NULL character, it is rejected
6000 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6001 runs initgroups(), so that if the Exim user is in any additional groups,
6002 they will be used during message reception.
6008 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob