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. PPv2 support was
29 updated based on HAProxy spec change in May 2014.
31 JH/02 Add ${listextract {number}{list}{success}{fail}}.
33 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
34 Properly escape header and check for NULL return.
36 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
39 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
41 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
42 characters in header names, implemented as a verify condition.
43 Contributed by Michael Fischer v. Mollard.
45 TL/05 Rename SPF condition results err_perm and err_temp to standardized
46 results permerror and temperror. Previous values are deprecated but
47 still accepted. In a future release, err_perm and err_temp will be
48 completely removed, which will be a backward incompatibility if the
49 ACL tests for either of these two old results. Patch contributed by
50 user bes-internal on the mailing list.
52 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
54 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
55 selectors, in both main and reject logs.
57 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
60 JH/07 Add malware type "sock" for talking to simple daemon.
62 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
64 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
65 routers/transports under cutthrough routing.
67 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
68 numbers. Touch up "bool" conditional to keep the same definition.
70 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
72 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
74 JH/12 Expand items in router/transport headers_add or headers_remove lists
75 individually rather than the list as a whole. Bug 1452.
77 Required for reasonable handling of multiple headers_ options when
78 they may be empty; requires that headers_remove items with embedded
79 colons must have them doubled (or the list-separator changed).
81 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
82 view the policy declared in the DMARC record. Currently, $dmarc_status
83 is a combined value of both the record presence and the result of the
86 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
88 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
89 dnslookup router and the smtp transport (applying to the forward
92 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
93 of ldap servers used for a specific lookup. Patch provided by Heiko
96 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
97 New variable $lookup_dnssec_authenticated for observability.
99 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
100 Patch submitted by Lars Timman.
102 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
104 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
105 Requires trusted mode and valid format message id, aborts otherwise.
106 Patch contributed by Heiko Schlichting.
108 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
109 certextract with support for various fields. Bug 1358.
111 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
112 is requested by default, modifiable by smtp transport option
115 JH/22 Expansion operators ${md5:string} and ${sha1::string} can now
116 operate on certificate variables to give certificate fingerprints
117 Also new ${sha256:cert_variable}.
119 JH/23 The PRDR feature is moved from being Experimental into the mainline.
121 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
122 Christian Aistleitner.
124 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
126 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
127 file. Patch from Wolfgang Breyha.
129 JH/25 Expand the coverage of the delivery $host and $host_address to
130 client authenticators run in verify callout. Bug 1476.
136 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
138 PP/02 Make -n do something, by making it not do something.
139 When combined with -bP, the name of an option is not output.
141 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
144 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
145 $sender_host_name and config options to manage this, and basic check
148 PP/05 DSCP support for outbound connections and control modifier for inbound.
150 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
151 (Only plugin which currently uses this is kerberos4, which nobody should
152 be using, but we should make it available and other future plugins might
153 conceivably use it, even though it would break NAT; stuff *should* be
154 using channel bindings instead).
156 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
157 name; added for Sendmail compatibility; requires admin caller.
158 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
159 just ignore it); requires trusted caller.
160 Also parse but ignore: -Ac -Am -X<logfile>
163 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
165 TL/02 Add +smtp_confirmation as a default logging option.
167 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
168 Patch by Magnus Holmgren from 2007-02-20.
170 TL/04 Bugzilla 1281 - Spec typo.
171 Bugzilla 1283 - Spec typo.
172 Bugzilla 1290 - Spec grammar fixes.
174 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
176 TL/06 Add Experimental DMARC support using libopendmarc libraries.
178 TL/07 Fix an out of order global option causing a segfault. Reported to dev
179 mailing list by by Dmitry Isaikin.
181 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
183 JH/02 Support "G" suffix to numbers in ${if comparisons.
185 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
187 NM/01 Bugzilla 1197 - Spec typo
188 Bugzilla 1196 - Spec examples corrections
190 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
192 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
193 gnutls_enable_pkcs11, but renamed to more accurately indicate its
196 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
197 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
199 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
200 "acl {{name}{arg}...}", and optional args on acl condition
203 JH/05 Permit multiple router/transport headers_add/remove lines.
205 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
207 JH/07 Avoid using a waiting database for a single-message-only transport.
208 Performance patch from Paul Fisher. Bugzilla 1262.
210 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
213 JH/09 Add $headers_added variable, with content from use of ACL modifier
214 add_header (but not yet added to the message). Bugzilla 199.
216 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
217 Pulled from Bugzilla 817 by Wolfgang Breyha.
219 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
221 (nb: this is the same fix as in Exim 4.80.1)
223 JH/11 Add A= logging on delivery lines, and a client_set_id option on
226 JH/12 Add optional authenticated_sender logging to A= and a log_selector
229 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
231 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
232 advertise SMTP AUTH mechanism to us, instead of a generic
233 protocol violation error. Also, make Exim more robust to bad
234 data from the Dovecot auth socket.
236 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
238 When a queue runner is handling a message, Exim first routes the
239 recipient addresses, during which it prunes them based on the retry
240 hints database. After that it attempts to deliver the message to
241 any remaining recipients. It then updates the hints database using
244 So if a recipient address works intermittently, it can get repeatedly
245 deferred at routing time. The retry hints record remains fresh so the
246 address never reaches the final cutoff time.
248 This is a fairly common occurrence when a user is bumping up against
249 their storage quota. Exim had some logic in its local delivery code
250 to deal with this. However it did not apply to per-recipient defers
251 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
253 This change adds a proper retry rule check during routing so that the
254 final cutoff time is checked against the message's age. We only do
255 this check if there is an address retry record and there is not a
256 domain retry record; this implies that previous attempts to handle
257 the address had the retry_use_local_parts option turned on. We use
258 this as an approximation for the destination being like a local
259 delivery, as in LMTP.
261 I suspect this new check makes the old local delivery cutoff check
262 redundant, but I have not verified this so I left the code in place.
264 TF/02 Correct gecos expansion when From: is a prefix of the username.
266 Test 0254 submits a message to Exim with the header
270 When I ran the test suite under the user fanf2, Exim expanded
271 the header to contain my full name, whereas it should have added
272 a Resent-Sender: header. It erroneously treats any prefix of the
273 username as equal to the username.
275 This change corrects that bug.
277 GF/01 DCC debug and logging tidyup
278 Error conditions log to paniclog rather than rejectlog.
279 Debug lines prefixed by "DCC: " to remove any ambiguity.
281 TF/03 Avoid unnecessary rebuilds of lookup-related code.
283 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
284 Bug spotted by Jeremy Harris; was flawed since initial commit.
285 Would have resulted in OCSP responses post-SNI triggering an Exim
286 NULL dereference and crash.
288 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
290 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
291 Bug detection, analysis and fix by Samuel Thibault.
292 Bugzilla 1331, Debian bug #698092.
294 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
296 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
297 Server implementation by Todd Lyons, client by JH.
298 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
299 config variable "prdr_enable" controls whether the server
300 advertises the facility. If the client requests PRDR a new
301 acl_data_smtp_prdr ACL is called once for each recipient, after
302 the body content is received and before the acl_smtp_data ACL.
303 The client is controlled by bolth of: a hosts_try_prdr option
304 on the smtp transport, and the server advertisement.
305 Default client logging of deliveries and rejections involving
306 PRDR are flagged with the string "PRDR".
308 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
309 fclose(). Diagnosis by Todd Lyons.
311 PP/17 Update configure.default to handle IPv6 localhost better.
312 Patch by Alain Williams (plus minor tweaks).
315 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
316 This is now consistent with GnuTLS, and is now documented: the
317 previous undocumented portable approach to treating the option as
318 unset was to force an expansion failure. That still works, and
319 an empty string is now equivalent.
321 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
322 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
323 not performing validation itself.
325 PP/20 Added force_command boolean option to pipe transport.
326 Patch from Nick Koston, of cPanel Inc.
328 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
331 TF/04 Added udpsend ACL modifer and hexquote expansion operator
333 PP/21 Fix eximon continuous updating with timestamped log-files.
334 Broken in a format-string cleanup in 4.80, missed when I repaired the
335 other false fix of the same issue.
336 Report and fix from Heiko Schlichting.
339 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
340 Report from Prashanth Katuri.
342 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
343 It's SecureTransport, so affects any MacOS clients which use the
344 system-integrated TLS libraries, including email clients.
346 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
347 using a MIME ACL for non-SMTP local injection.
348 Report and assistance in diagnosis by Warren Baker.
350 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
352 JH/16 Fix comparisons for 64b. Bugzilla 1385.
354 TL/09 Add expansion variable $authenticated_fail_id to keep track of
355 last id that failed so it may be referenced in subsequent ACL's.
357 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
360 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
361 ldap library initialization, allowing self-signed CA's to be
362 used. Also properly sets require_cert option later in code by
363 using NULL (global ldap config) instead of ldap handle (per
364 session). Bug diagnosis and testing by alxgomz.
366 TL/12 Enhanced documentation in the ratelimit.pl script provided in
367 the src/util/ subdirectory.
369 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
370 renamed to Transport Post Delivery Action by Jeremy Harris, as
373 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
374 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
375 redis_servers = needs to be configured which will be used by the redis
376 lookup. Patch from Warren Baker, of The Packet Hub.
378 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
380 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
381 hostname or reverse DNS when processing a host list. Used suggestions
382 from multiple comments on this bug.
384 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
386 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
387 Missed a few lines, added it to make the runtest require no keyboard
390 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
391 contains upper case chars. Make router use caseful_local_part.
393 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
394 support when GnuTLS has been built with p11-kit.
400 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
402 This, or similar/improved, will also be change PP/11 of 4.82.
408 PP/01 Handle short writes when writing local log-files.
409 In practice, only affects FreeBSD (8 onwards).
410 Bugzilla 1053, with thanks to Dmitry Isaikin.
412 NM/01 Bugzilla 949 - Documentation tweak
414 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
417 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
419 PP/02 Implemented gsasl authenticator.
421 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
423 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
424 `pkg-config foo` for cflags/libs.
426 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
427 with rest of GSASL and with heimdal_gssapi.
429 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
430 `pkg-config foo` for cflags/libs for the TLS implementation.
432 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
433 properties get this fed in as external SSF. A number of robustness
434 and debugging improvements to the cyrus_sasl authenticator.
436 PP/08 cyrus_sasl server now expands the server_realm option.
438 PP/09 Bugzilla 1214 - Log authentication information in reject log.
439 Patch by Jeremy Harris.
441 PP/10 Added dbmjz lookup type.
443 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
445 PP/12 MAIL args handles TAB as well as SP, for better interop with
446 non-compliant senders.
447 Analysis and variant patch by Todd Lyons.
449 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
450 Bug report from Lars Müller <lars@samba.org> (via SUSE),
451 Patch from Dirk Mueller <dmueller@suse.com>
453 PP/13 tls_peerdn now print-escaped for spool files.
454 Observed some $tls_peerdn in wild which contained \n, which resulted
455 in spool file corruption.
457 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
458 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
459 or write after TLS renegotiation, which otherwise led to messages
462 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
463 as a tracking header (ie: a signed header comes before the signature).
464 Patch from Wolfgang Breyha.
466 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
467 comma-sep list; embedded commas doubled.
469 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
471 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
473 Report and patch from Dmitry Banschikov.
475 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
476 Removed SSL_clear() after SSL_new() which led to protocol negotiation
477 failures. We appear to now support TLS1.1+ with Exim.
479 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
480 lets Exim select keys and certificates based upon TLS SNI from client.
481 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
482 before an outbound SMTP session. New log_selector, +tls_sni.
484 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
485 NULL dereference. Report and patch from Alun Jones.
487 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
488 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
489 Not seeing resolver debug output on NetBSD, but suspect this is a
490 resolver implementation change.
492 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
493 Left warnings. Added "eximon gdb" invocation mode.
495 PP/21 Defaulting "accept_8bitmime" to true, not false.
497 PP/22 Added -bw for inetd wait mode support.
499 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
500 locate the relevant includes and libraries. Made this the default.
502 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
503 Bugzilla 1246, report and most of solution from Tomasz Kusy.
505 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
506 This may cause build issues on older platforms.
508 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
509 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
510 gnutls_require_mac & gnutls_require_protocols (no longer supported).
511 Added SNI support via GnuTLS too.
512 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
514 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
516 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
517 Applied second patch from Janne, implementing suggestion to default
518 multiple-strings-in-record handling to match SPF spec.
520 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
522 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
523 read-only, out of scope).
524 Patch from Wolfgang Breyha, report from Stuart Northfield.
526 PP/29 Fix three issues highlighted by clang analyser static analysis.
527 Only crash-plausible issue would require the Cambridge-specific
528 iplookup router and a misconfiguration.
529 Report from Marcin Mirosław.
531 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
533 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
534 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
535 As part of this, removing so much warning spew let me fix some minor
536 real issues in debug logging.
538 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
539 assignment on my part. Fixed.
541 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
542 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
543 Janne Snabb (who went above and beyond: thank you).
545 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
546 string otherwise requires a connection and a bunch more work and it's
547 relatively easy to get wrong. Should also expose TLS library linkage
550 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
551 64-bit ${eval} (JH/03).
553 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
554 GNU libc to support some of the 64-bit stuff, should not lead to
555 conflicts. Defined before os.h is pulled in, so if a given platform
556 needs to override this, it can.
558 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
559 protection layer was required, which is not implemented.
560 Bugzilla 1254, patch from Wolfgang Breyha.
562 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
563 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
564 tls_dhparam take prime identifiers. Also unbreak combination of
565 OpenSSL+DH_params+TLSSNI.
567 PP/39 Disable SSLv2 by default in OpenSSL support.
573 PP/01 Solaris build fix for Oracle's LDAP libraries.
574 Bugzilla 1109, patch from Stephen Usher.
576 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
578 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
581 TF/02 Fix a couple more cases where we did not log the error message
582 when unlink() failed. See also change 4.74-TF/03.
584 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
585 lock up or crash if it happened to be inside a call to libc when it
586 got a SIGUSR1 from exiwhat.
588 The SIGUSR1 handler appends the current process status to the process
589 log which is later printed by exiwhat. It used to use the general
590 purpose logging code to do this, but several functions it calls are
591 not safe for signals.
593 The new output code in the SIGUSR1 handler is specific to the process
594 log, and simple enough that it's easy to inspect for signal safety.
595 Removing some special cases also simplifies the general logging code.
596 Removing the spurious timestamps from the process log simplifies
599 TF/04 Improved ratelimit ACL condition.
601 The /noupdate option has been deprecated in favour of /readonly which
602 has clearer semantics. The /leaky, /strict, and /readonly update modes
603 are mutually exclusive. The update mode is no longer included in the
604 database key; it just determines when the database is updated. (This
605 means that when you upgrde Exim will forget old rate measurements.)
607 Exim now checks that the per_* options are used with an update mode that
608 makes sense for the current ACL. For example, when Exim is processing a
609 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
610 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
611 must specify per_mail/readonly. If you omit the update mode it defaults to
612 /leaky where that makes sense (as before) or /readonly where required.
614 The /noupdate option is now undocumented but still supported for
615 backwards compatibility. It is equivalent to /readonly except that in
616 ACLs where /readonly is required you may specify /leaky/noupdate or
617 /strict/noupdate which are treated the same as /readonly.
619 A useful new feature is the /count= option. This is a generalization
620 of the per_byte option, so that you can measure the throughput of other
621 aggregate values. For example, the per_byte option is now equivalent
622 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
624 The per_rcpt option has been generalized using the /count= mechanism
625 (though it's more complicated than the per_byte equivalence). When it is
626 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
627 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
628 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
629 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
630 Note that using per_rcpt with a non-readonly update mode in more than
631 one ACL will cause the recipients to be double-counted. (The per_mail
632 and per_byte options don't have this problem.)
634 The handling of very low rates has changed slightly. If the computed rate
635 is less than the event's count (usually one) then this event is the first
636 after a long gap. In this case the rate is set to the same as this event's
637 count, so that the first message of a spam run is counted properly.
639 The major new feature is a mechanism for counting the rate of unique
640 events. The new per_addr option counts the number of different
641 recipients that someone has sent messages to in the last time period. It
642 behaves like per_rcpt if all the recipient addresses are different, but
643 duplicate recipient addresses do not increase the measured rate. Like
644 the /count= option this is a general mechanism, so the per_addr option
645 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
646 example, measure the rate that a client uses different sender addresses
647 with the options per_mail/unique=$sender_address. There are further
648 details in the main documentation.
650 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
652 TF/06 Removed a few PCRE remnants.
654 TF/07 Automatically extract Exim's version number from tags in the git
655 repository when doing development or release builds.
657 PP/02 Raise smtp_cmd_buffer_size to 16kB.
658 Bugzilla 879. Patch from Paul Fisher.
660 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
661 Heavily based on revision 40f9a89a from Simon Arlott's tree.
664 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
666 PP/05 Variable $av_failed, true if the AV scanner deferred.
667 Bugzilla 1078. Patch from John Horne.
669 PP/06 Stop make process more reliably on build failure.
670 Bugzilla 1087. Patch from Heiko Schlittermann.
672 PP/07 Make maildir_use_size_file an _expandable_ boolean.
673 Bugzilla 1089. Patch from Heiko Schlittermann.
675 PP/08 Handle ${run} returning more data than OS pipe buffer size.
676 Bugzilla 1131. Patch from Holger Weiß.
678 PP/09 Handle IPv6 addresses with SPF.
679 Bugzilla 860. Patch from Wolfgang Breyha.
681 PP/10 GnuTLS: support TLS 1.2 & 1.1.
683 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
686 PP/11 match_* no longer expand right-hand-side by default.
687 New compile-time build option, EXPAND_LISTMATCH_RHS.
688 New expansion conditions, "inlist", "inlisti".
690 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
692 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
694 PP/14 fix log_write() format string regression from TF/03.
695 Bugzilla 1152. Patch from Dmitry Isaikin.
701 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
703 PP/02 Harmonised TLS library version reporting; only show if debugging.
704 Layout now matches that introduced for other libraries in 4.74 PP/03.
706 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
708 PP/04 New "dns_use_edns0" global option.
710 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
713 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
714 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
716 TK/01 Updated PolarSSL code to 0.14.2.
717 Bugzilla 1097. Patch from Andreas Metzler.
719 PP/07 Catch divide-by-zero in ${eval:...}.
722 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
725 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
726 format-string attack -- SECURITY: remote arbitrary code execution.
728 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
729 time unintentionally subject to list matching rules, letting the header
730 cause arbitrary Exim lookups (of items which can occur in lists, *not*
731 arbitrary string expansion). This allowed for information disclosure.
733 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
734 INT_MIN/-1 -- value coerced to INT_MAX.
740 NM/01 Workround for PCRE version dependency in version reporting
743 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
744 This fixes portability to compilers other than gcc, notably
745 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
747 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
748 makefiles for portability to HP-UX and POSIX correctness.
750 PP/01 Permit LOOKUP_foo enabling on the make command-line.
751 Also via indented variable definition in the Makefile.
752 (Debugging by Oliver Heesakkers).
754 PP/02 Restore caching of spamd results with expanded spamd_address.
755 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
757 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
758 Improves build reliability. Fix from: Frank Elsner
760 NM/02 Fix wide character breakage in the rfc2047 coding
761 Fixes bug 1064. Patch from Andrey N. Oktyabrski
763 NM/03 Allow underscore in dnslist lookups
764 Fixes bug 1026. Patch from Graeme Fowler
766 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
767 Code patches from Adam Ciarcinski of NetBSD.
769 NM/04 Fixed exiqgrep to cope with mailq missing size issue
772 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
773 is logged, to avoid truncation. Patch from John Horne.
775 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
776 Patch from Jakob Hirsch.
778 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
779 SQL string expansion failure details.
780 Patch from Andrey Oktyabrski.
782 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
783 Patch from Simon Arlott.
785 PP/09 New lookups functionality failed to compile on old gcc which rejects
786 extern declarations in function scope.
787 Patch from Oliver Fleischmann
789 PP/10 Use sig_atomic_t for flags set from signal handlers.
790 Check getgroups() return and improve debugging.
791 Fixed developed for diagnosis in bug 927 (which turned out to be
794 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
795 Patch from Mark Zealey.
797 PP/12 Bugzilla 1056: Improved spamd server selection.
798 Patch from Mark Zealey.
800 PP/13 Bugzilla 1086: Deal with maildir quota file races.
801 Based on patch from Heiko Schlittermann.
803 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
804 Patch from Uwe Doering, sign-off by Michael Haardt.
806 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
807 variable declaration deep within a block. Bug and patch from
810 PP/15 lookups-Makefile IRIX compatibilty coercion.
812 PP/16 Make DISABLE_DKIM build knob functional.
814 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
815 Patch by Simon Arlott
817 TF/03 Fix valgrind.h portability to C89 compilers that do not support
818 variable argument macros. Our copy now differs from upstream.
824 TF/01 Failure to get a lock on a hints database can have serious
825 consequences so log it to the panic log.
827 TF/02 Log LMTP confirmation messages in the same way as SMTP,
828 controlled using the smtp_confirmation log selector.
830 TF/03 Include the error message when we fail to unlink a spool file.
832 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
833 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
834 for maintaining out-of-tree patches for some time.
836 PP/01 Bugzilla 139: Documentation and portability issues.
837 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
838 Handle per-OS dynamic-module compilation flags.
840 PP/02 Let /dev/null have normal permissions.
841 The 4.73 fixes were a little too stringent and complained about the
842 permissions on /dev/null. Exempt it from some checks.
843 Reported by Andreas M. Kirchwitz.
845 PP/03 Report version information for many libraries, including
846 Exim version information for dynamically loaded libraries. Created
847 version.h, now support a version extension string for distributors
848 who patch heavily. Dynamic module ABI change.
850 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
851 privilege escalation vulnerability whereby the Exim run-time user
852 can cause root to append content of the attacker's choosing to
855 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
858 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
859 If dropping privileges for untrusted macros, we disabled normal logging
860 on the basis that it would fail; for the Exim run-time user, this is not
861 the case, and it resulted in successful deliveries going unlogged.
862 Fixed. Reported by Andreas Metzler.
868 PP/01 Date: & Message-Id: revert to normally being appended to a message,
869 only prepend for the Resent-* case. Fixes regression introduced in
870 Exim 4.70 by NM/22 for Bugzilla 607.
872 PP/02 Include check_rfc2047_length in configure.default because we're seeing
873 increasing numbers of administrators be bitten by this.
875 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
877 PP/03 Bugzilla 994: added openssl_options main configuration option.
879 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
881 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
883 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
885 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
886 without a peer certificate, leading to a segfault because of an
887 assumption that peers always have certificates. Be a little more
888 paranoid. Problem reported by Martin Tscholak.
890 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
891 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
892 NB: ClamAV planning to remove STREAM in "middle of 2010".
893 CL also introduces -bmalware, various -d+acl logging additions and
894 more caution in buffer sizes.
896 PP/09 Implemented reverse_ip expansion operator.
898 PP/10 Bugzilla 937: provide a "debug" ACL control.
900 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
902 PP/12 Bugzilla 973: Implement --version.
904 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
906 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
908 PP/15 Bugzilla 816: support multiple condition rules on Routers.
910 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
911 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
912 ignore trailing whitespace.
914 JJ/02 prevent non-panic DKIM error from being sent to paniclog
916 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
919 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
920 Notification from Dr Andrew Aitchison.
922 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
923 ExtendedDetectionInfo response format.
924 Notification from John Horne.
926 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
929 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
930 XSL and documented dependency on system catalogs, with examples of how
933 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
936 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
937 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
938 configuration file which is writeable by the Exim user or group.
940 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
941 of configuration files to cover files specified with the -C option if
942 they are going to be used with root privileges, not just the default
945 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
946 option (effectively making it always true).
948 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
949 files to be used while preserving root privileges.
951 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
952 that rogue child processes cannot use them.
954 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
955 run-time user, instead of root.
957 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
958 Exim run-time user without dropping privileges.
960 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
961 result string, instead of calling string_vformat() twice with the same
964 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
965 for other users. Others should always drop root privileges if they use
966 -C on the command line, even for a whitelisted configure file.
968 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
970 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
976 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
977 $data_path, and $header_path variables; fixed documentation bugs and
980 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
981 exipick to access non-standard spools, including the "frozen" queue
984 NM/01 Bugzilla 965: Support mysql stored procedures.
985 Patch from Alain Williams
987 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
989 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
990 Patch from Andreas Metzler
992 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
993 Patch from Kirill Miazine
995 NM/05 Bugzilla 671: Added umask to procmail example.
997 JJ/03 installed exipick 20100323.0, fixing doc bug
999 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1000 directory. Notification and patch from Dan Rosenberg.
1002 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1004 TK/02 Improve log output when DKIM signing operation fails.
1006 MH/01 Treat the transport option dkim_domain as a colon separated
1007 list, not as a single string, and sign the message with each element,
1008 omitting multiple occurences of the same signer.
1010 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1011 Bugzilla 985, 986. Patch by Simon Arlott
1013 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1014 Patch by Simon Arlott
1016 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1017 MBX locking. Notification from Dan Rosenberg.
1023 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1025 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1027 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1029 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1031 NM/04 Bugzilla 915: Flag broken perl installation during build.
1037 TK/01 Added patch by Johannes Berg that expands the main option
1038 "spamd_address" if it starts with a dollar sign.
1040 TK/02 Write list of recipients to X-Envelope-Sender header when building
1041 the mbox-format spool file for content scanning (suggested by Jakob
1044 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1045 (http://www.dcc-servers.net/) support via dccifd. Activated by
1046 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1048 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1049 by Mark Daniel Reidel <mr@df.eu>.
1051 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1052 When building exim an external PCRE library is now needed -
1053 PCRE is a system library on the majority of modern systems.
1054 See entry on PCRE_LIBS in EDITME file.
1056 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1057 conversation. Added nologin parameter to request.
1058 Patch contributed by Kirill Miazine.
1060 TF/01 Do not log submission mode rewrites if they do not change the address.
1062 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1064 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1065 log files in place. Contributed by Roberto Lima.
1067 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1069 TF/03 Bugzilla 615: When checking the local_parts router precondition
1070 after a local_part_suffix or local_part_prefix option, Exim now
1071 does not use the address's named list lookup cache, since this
1072 contains cached lookups for the whole local part.
1074 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1075 Robert Millan. Documentation is in experimental-spec.txt.
1077 TF/04 Bugzilla 668: Fix parallel build (make -j).
1079 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1081 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1082 Patch provided by Jan Srzednicki.
1084 TF/05 Leading white space used to be stripped from $spam_report which
1085 wrecked the formatting. Now it is preserved.
1087 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1088 that they are available at delivery time.
1090 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1092 TF/08 TLS error reporting now respects the incoming_interface and
1093 incoming_port log selectors.
1095 TF/09 Produce a more useful error message if an SMTP transport's hosts
1096 setting expands to an empty string.
1098 NM/06 Bugzilla 744: EXPN did not work under TLS.
1099 Patch provided by Phil Pennock.
1101 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1102 Patch provided by Richard Godbee.
1104 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1105 acl_smtp_notquit, added index entry.
1107 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1108 Patch provided by Eugene Bujak.
1110 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1111 accept(). Patch provided by Maxim Dounin.
1113 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1114 Patch provided by Phil Pennock.
1116 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1118 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1119 Patch provided by Brad "anomie" Jorsch.
1121 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1122 Patch provided by Dean Brooks.
1124 TK/05 Add native DKIM support (does not depend on external libraries).
1126 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1127 Patch provided by Graeme Fowler.
1129 NM/16 Bugzilla 851: Documentation example syntax fix.
1131 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1133 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1136 NM/19 Bugzilla 745: TLS version reporting.
1137 Patch provided by Phil Pennock.
1139 NM/20 Bugzilla 167: bool: condition support.
1140 Patch provided by Phil Pennock.
1142 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1143 clients. Patch provided by Phil Pennock.
1145 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1146 Patch provided by Brad "anomie" Jorsch.
1148 NM/23 Bugzilla 687: Fix misparses in eximstats.
1149 Patch provided by Heiko Schlittermann.
1151 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1152 Patch provided by Heiko Schlittermann.
1154 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1155 plus update to original patch.
1157 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1159 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1160 Patch provided by David Brownlee.
1162 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1164 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1166 NM/30 Bugzilla 888: TLS documentation bugfixes.
1168 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1170 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1171 Unlike the original bugzilla I have changed all shell scripts in src tree.
1173 NM/33 Bugzilla 898: Transport filter timeout fix.
1174 Patch by Todd Rinaldo.
1176 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1177 Patch by Serge Demonchaux.
1179 NM/35 Bugzilla 39: Base64 decode bug fixes.
1180 Patch by Jakob Hirsch.
1182 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1184 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1186 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1188 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1194 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1195 ALT-N's libdkim that I have put here:
1196 http://duncanthrax.net/exim-experimental/
1198 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1199 were called 'true' and it seems that C99 defines that as a reserved
1200 keyword to be used with 'bool' variable types. That means you could
1201 not include C99-style headers which use bools without triggering
1202 build errors in sieve.c.
1204 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1205 as mailq or other aliases. Changed the --help handling significantly
1206 to do whats expected. exim_usage() emits usage/help information.
1208 SC/01 Added the -bylocaldomain option to eximstats.
1210 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1212 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1214 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1220 PH/01 Another patch from the Sieve maintainer.
1222 PH/02 When an IPv6 address is converted to a string for single-key lookup
1223 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1224 dots are used instead of colons so that keys in lsearch files need not
1225 contain colons. This was done some time before quoting was made available
1226 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1227 (notated using the quote facility) so as to distinguish them from IPv4
1228 keys. This meant that lookups for IP addresses in host lists did not work
1229 for iplsearch lookups.
1231 This has been fixed by arranging for IPv6 addresses to be expressed with
1232 colons if the lookup type is iplsearch. This is not incompatible, because
1233 previously such lookups could never work.
1235 The situation is now rather anomolous, since one *can* have colons in
1236 ordinary lsearch keys. However, making the change in all cases is
1237 incompatible and would probably break a number of configurations.
1239 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1242 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1243 conversion specification without a maximum field width, thereby enabling
1244 a rogue spamd server to cause a buffer overflow. While nobody in their
1245 right mind would setup Exim to query an untrusted spamd server, an
1246 attacker that gains access to a server running spamd could potentially
1247 exploit this vulnerability to run arbitrary code as the Exim user.
1249 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1250 $primary_hostname instead of what libspf2 thinks the hosts name is.
1252 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1253 a directory entry by the name of the lookup key. Previously, if a
1254 symlink pointed to a non-existing file or a file in a directory that
1255 Exim lacked permissions to read, a lookup for a key matching that
1256 symlink would fail. Now it is enough that a matching directory entry
1257 exists, symlink or not. (Bugzilla 503.)
1259 PH/03 The body_linecount and body_zerocount variables are now exported in the
1262 PH/04 Added the $dnslist_matched variable.
1264 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1265 This means they are set thereafter only if the connection becomes
1268 PH/06 Added the client_condition to authenticators so that some can be skipped
1269 by clients under certain conditions.
1271 PH/07 The error message for a badly-placed control=no_multiline_responses left
1272 "_responses" off the end of the name.
1274 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1276 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1277 (without spaces) instead of just copying the configuration text.
1279 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1281 PH/11 Added $max_received_linelength.
1283 PH/12 Added +ignore_defer and +include_defer to host lists.
1285 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1286 way in which PCRE > 7.0 is built.
1288 PH/14 Implemented queue_only_load_latch.
1290 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1291 MAIL command. The effect was to mangle the value on 64-bit systems.
1293 PH/16 Another patch from the Sieve maintainer.
1295 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1297 PH/18 If a system quota error occurred while trying to create the file for
1298 a maildir delivery, the message "Mailbox is full" was not appended to the
1299 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1300 only to a quota excession during the actual writing of the file.
1302 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1303 characters?) which causes problems in log lines. The DN values are now
1304 passed through string_printing() before being added to log lines.
1306 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1307 and InterBase are left for another time.)
1309 PH/21 Added message_body_newlines option.
1311 PH/22 Guard against possible overflow in moan_check_errorcopy().
1313 PH/23 POSIX allows open() to be a macro; guard against that.
1315 PH/24 If the recipient of an error message contained an @ in the local part
1316 (suitably quoted, of course), incorrect values were put in $domain and
1317 $local_part during the evaluation of errors_copy.
1323 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1324 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1327 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1328 issue a MAIL command.
1330 PH/02 In an ACL statement such as
1332 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1334 if a client was not listed at all, or was listed with a value other than
1335 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1336 the condition was not true (as it should be), so access was not denied.
1337 The bug was that the ! inversion was incorrectly passed on to the second
1338 item. This has been fixed.
1340 PH/03 Added additional dnslists conditions == and =& which are different from
1341 = and & when the dns lookup returns more than one IP address.
1343 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1344 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1346 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1347 FSYNC, which compiles an option called disable_fsync that allows for
1348 bypassing fsync(). The documentation is heavily laced with warnings.
1350 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1352 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1353 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1354 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1355 including adding "make clean"; (3) Added -fPIC when compiling the test
1356 dynamically loaded module, to get rid of a warning.
1358 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1359 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1360 = 0s. The bug is otherwise harmless.
1362 PH/07 There was a bug in the dovecot authenticator such that the value of
1363 $auth1 could be overwritten, and so not correctly preserved, after a
1364 successful authentication. This usually meant that the value preserved by
1365 the server_setid option was incorrect.
1367 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1369 PH/09 Installed PCRE release 7.0.
1371 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1372 run for batched SMTP input. It is now run at the start of every message
1373 in the batch. While fixing this I discovered that the process information
1374 (output by running exiwhat) was not always getting set for -bs and -bS
1375 input. This is fixed, and it now also says "batched" for BSMTP.
1377 PH/11 Added control=no_pipelining.
1379 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1380 patch, slightly modified), and move the expansion of helo_data till after
1381 the connection is made in the smtp transport (so it can use these
1384 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1386 PH/14 Added log_selector = +pid.
1388 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1390 PH/16 Add ${if forany and ${if forall.
1392 PH/17 Added dsn_from option to vary the From: line in DSNs.
1394 PH/18 Flush SMTP output before performing a callout, unless control =
1395 no_callout_flush is set.
1397 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1398 was true (the default) a successful delivery failed to delete the retry
1399 item, thus causing premature timeout of the address. The bug is now
1402 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1404 PH/21 Long custom messages for fakedefer and fakereject are now split up
1405 into multiline reponses in the same way that messages for "deny" and
1406 other ACL rejections are.
1408 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1409 with slight modification.
1411 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1412 draft, changing the syntax and factoring some duplicate code".
1414 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1415 for deliveries of the second and subsequent messages over the same SMTP
1418 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1419 ${reduce, with only minor "tidies".
1421 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1423 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1424 expansion side effects.
1426 PH/27 When a message times out after an over-quota error from an Exim-imposed
1427 quota, the bounce message says "mailbox is full". This message was not
1428 being given when it was a system quota that was exceeded. It now should
1431 MH/03 Made $recipients available in local_scan(). local_scan() already has
1432 better access to the recipient list through recipients_list[], but
1433 $recipients can be useful in postmaster-provided expansion strings.
1435 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1436 in the case of a MAIL command with additional options following the
1437 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1438 were accidentally chopped off.
1440 PH/29 SMTP synchronization checks are implemented when a command is read -
1441 there is a check that no more input is waiting when there shouldn't be
1442 any. However, for some commands, a delay in an ACL can mean that it is
1443 some time before the response is written. In this time, more input might
1444 arrive, invalidly. So now there are extra checks after an ACL has run for
1445 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1446 pipelining has not been advertised.
1448 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1450 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1451 returned in the SMTP response when smtp_return_error_details was false.
1452 This has been fixed.
1454 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1455 instead of the C I/O that was originally supplied, because problems were
1456 reported on Solaris.
1458 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1459 Exim which did not show up earlier: it was assuming that a call to
1460 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1461 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1462 was a macro that became an assignment, so it seemed to work. This has
1463 changed to a proper function call with a void return, hence the compile
1464 error. Exim's code has been fixed.
1466 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1469 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1471 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1473 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1474 because it turns out to be too big or there is a timeout) there is no
1475 "Completed" line in the log. When some messages of this type were
1476 selected by exigrep, they were listed as "not completed". Others were
1477 picked up by some special patterns. I have improved the selection
1478 criteria to be more general.
1480 PH/38 The host_find_failed option in the manualroute router can now be set
1481 to "ignore", to completely ignore a host whose IP address cannot be
1482 found. If all hosts are ignored, the behaviour is controlled by the new
1483 host_all_ignored option.
1485 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1486 homing or because of multiple MX records with /mx) generated more than
1487 one IP address, and the following item turned out to be the local host,
1488 all the secondary addresses of the first item were incorrectly removed
1489 from the list, along with the local host and any following hosts (which
1490 is what is supposed to happen).
1492 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1493 whoever called Exim into the -H file. In the case of the daemon it was
1494 behaving confusingly. When first started, it used values for whoever
1495 started the daemon, but after a SIGHUP it used the Exim user (because it
1496 calls itself on a restart). I have changed the code so that it now always
1499 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1500 message are rejected with the same error (e.g. no authentication or bad
1501 sender address), and a DATA command is nevertheless sent (as can happen
1502 with PIPELINING or a stupid MUA), the error message that was given to the
1503 RCPT commands is included in the rejection of the DATA command. This is
1504 intended to be helpful for MUAs that show only the final error to their
1507 PH/42 Another patch from the Sieve maintainer.
1509 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1510 Eximstats - Fixed some broken HTML links and added missing column headers
1512 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1513 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1515 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1517 PH/43 Yet another patch from the Sieve maintainer.
1519 PH/44 I found a way to check for a TCP/IP connection going away before sending
1520 the response to the final '.' that terminates a message, but only in the
1521 case where the client has not sent further data following the '.'
1522 (unfortunately, this is allowed). However, in many cases there won't be
1523 any further data because there won't be any more messages to send. A call
1524 to select() can be used: if it shows that the input is "ready", there is
1525 either input waiting, or the socket has been closed. An attempt to read
1526 the next input character can distinguish the two cases. Previously, Exim
1527 would have sent an OK response which the client would never have see.
1528 This could lead to message repetition. This fix should cure that, at
1529 least in a lot of common cases.
1531 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1532 advertised in response to EHLO.
1538 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1539 fixed by 4.65/MH/01 (is this a record?) are fixed:
1541 (i) An empty string was always treated as zero by the numeric comparison
1542 operators. This behaviour has been restored.
1544 (ii) It is documented that the numeric comparison operators always treat
1545 their arguments as decimal numbers. This was broken in that numbers
1546 starting with 0 were being interpreted as octal.
1548 While fixing these problems I realized that there was another issue that
1549 hadn't been noticed. Values of message_size_limit (both the global option
1550 and the transport option) were treated as octal if they started with 0.
1551 The documentation was vague. These values are now always treated as
1552 decimal, and I will make that clear in the documentation.
1558 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1559 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1562 MH/01 Don't check that the operands of numeric comparison operators are
1563 integers when their expansion is in "skipping" mode (fixes bug
1564 introduced by 4.64-PH/07).
1566 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1567 child addresses, Exim now panics and dies. Previously, because the count
1568 is held in a short int, deliveries were likely to be lost. As such a
1569 large number of recipients for a single message is ridiculous
1570 (performance will be very, very poor), I have chosen to impose a limit
1571 rather than extend the field.
1577 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1578 leftover -K file (the existence of which was triggered by #402).
1579 While we were at it, introduced process PID as part of the -K
1580 filename. This should rule out race conditions when creating
1583 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1584 processing considerably. Previous code took too long for large mails,
1585 triggering a timeout which in turn triggers #401.
1587 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1588 in the DK code in transports.c. sendfile() is not really portable,
1589 hence the _LINUX specificness.
1591 TF/01 In the add_headers option to the mail command in an Exim filter,
1592 there was a bug that Exim would claim a syntax error in any
1593 header after the first one which had an odd number of characters
1596 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1597 callout verification, Exim cached a "reject" for the entire domain. This
1598 is correct for most verifications, but it is not correct for a recipient
1599 verification with use_sender or use_postmaster set, because in that case
1600 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1601 case of MAIL FROM:<> rejection from other early rejections (e.g.
1602 rejection of HELO). When verifying a recipient using a non-null MAIL
1603 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1604 Whatever the result of the callout, the value of the domain cache is
1605 left unchanged (for any other kind of callout, getting as far as trying
1606 RCPT means that the domain itself is ok).
1608 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1611 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1612 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1613 session, the daemon ignores ECONNECT errors and logs others; it now
1614 ignores EPIPE as well.
1616 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1617 (quoted-printable decoding).
1619 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1620 later the small subsequent patch to fix an introduced bug.
1622 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1624 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1626 PH/08 An error is now given if message_size_limit is specified negative.
1628 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1629 to be given (somewhat) arbitrary names.
1631 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1634 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1635 miscellaneous code fixes
1637 PH/10 Added the log_reject_target ACL modifier to specify where to log
1640 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1641 hostname. This is wrong, because it relates to the incoming message (and
1642 probably the interface on which it is arriving) and not to the outgoing
1643 callout (which could be using a different interface). This has been
1644 changed to use the value of the helo_data option from the smtp transport
1645 instead - this is what is used when a message is actually being sent. If
1646 there is no remote transport (possible with a router that sets up host
1647 addresses), $smtp_active_hostname is used.
1649 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1650 tweaks were necessary in order to get it to work (see also 21 below):
1651 (a) The code assumed that strncpy() returns a negative number on buffer
1652 overflow, which isn't the case. Replaced with Exim's string_format()
1654 (b) There were several signed/unsigned issues. I just did the minimum
1655 hacking in of casts. There is scope for a larger refactoring.
1656 (c) The code used strcasecmp() which is not a standard C function.
1657 Replaced with Exim's strcmpic() function.
1658 (d) The code set only $1; it now sets $auth1 as well.
1659 (e) A simple test gave the error "authentication client didn't specify
1660 service in request". It would seem that Dovecot has changed its
1661 interface. Fortunately there's a specification; I followed it and
1662 changed what the client sends and it appears to be working now.
1664 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1667 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1668 address A is aliased to B and C, where B exists and C does not. Without
1669 -v the output is "A verified" because verification stops after a
1670 successful redirection if more than one address is generated. However,
1671 with -v the child addresses are also verified. Exim was outputting "A
1672 failed to verify" and then showing the successful verification for C,
1673 with its parentage. It now outputs "B failed to verify", showing B's
1674 parentage before showing the successful verification of C.
1676 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1677 look up a TXT record in a specific list after matching in a combined
1680 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1681 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1682 they consult the DNS. I had assumed they would set it the way they
1683 wanted; and indeed my experiments on Linux seem to show that in some
1684 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1685 To be on the safe side, however, I have now made the interface to
1686 host_find_byname() similar to host_find_bydns(), with an argument
1687 containing the DNS resolver options. The host_find_byname() function now
1688 sets these options at its start, just as host_find_bydns() does. The smtp
1689 transport options dns_qualify_single and dns_search_parents are passed to
1690 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1691 of host_find_byname() use the default settings of RES_DEFNAMES
1692 (qualify_single) but not RES_DNSRCH (search_parents).
1694 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1695 spool_read_header() do less string testing, by means of a preliminary
1696 switch on the second character of optional "-foo" lines. (This is
1697 overdue, caused by the large number of possibilities that now exist.
1698 Originally there were few.) While I was there, I also converted the
1699 str(n)cmp tests so they don't re-test the leading "-" and the first
1700 character, in the hope this might squeeze out yet more improvement.
1702 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1703 flag allowing group syntax was set by the header_syntax check but not
1704 turned off, possible causing trouble later; (2) The flag was not being
1705 set at all for the header_verify test, causing "group"-style headers to
1706 be rejected. I have now set it in this case, and also caused header_
1707 verify to ignore an empty address taken from a group. While doing this, I
1708 came across some other cases where the code for allowing group syntax
1709 while scanning a header line wasn't quite right (mostly, not resetting
1710 the flag correctly in the right place). These bugs could have caused
1711 trouble for malformed header lines. I hope it is now all correct.
1713 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1714 with the "reply" argument non-NULL. The code, however (which originally
1715 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1716 but it didn't always do it. This confused somebody who was copying the
1717 code for some other use. I have removed all the tests.
1719 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1720 feature that was used to support insecure browsers during the U.S. crypto
1721 embargo. It requires special client support, and Exim is probably the
1722 only MTA that supported it -- and would never use it because real RSA is
1723 always available. This code has been removed, because it had the bad
1724 effect of slowing Exim down by computing (never used) parameters for the
1725 RSA_EXPORT functionality.
1727 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1728 authenticator to fail if there's a tab character in the incoming data
1729 (there should never be unless someone is messing about, as it's supposed
1730 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1731 if the connection is using TLS or if the remote IP is the same as the
1732 local IP, and the "valid-client-cert option" if a client certificate has
1735 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1736 authenticators. This can be used for authorization after authentication
1737 succeeds. (In the case of plaintext, it servers for both authentication
1740 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1741 if any retry times were supplied.
1743 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1744 connection, where there is no client IP address to check. In this
1745 situation, the verify now always succeeds.
1747 PH/25 Applied John Jetmore's -Mset patch.
1749 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1751 PH/27 In a string expansion for a processed (not raw) header when multiple
1752 headers of the same name were present, leading whitespace was being
1753 removed from all of them, but trailing whitespace was being removed only
1754 from the last one. Now trailing whitespace is removed from each header
1755 before concatenation. Completely empty headers in a concatenation (as
1756 before) are ignored.
1758 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1759 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1761 PH/29 [Removed. This was a change that I later backed out, and forgot to
1762 correct the ChangeLog entry (that I had efficiently created) before
1763 committing the later change.]
1765 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1766 address errors (4xx response to RCPT) over the same connection as other
1767 messages routed to the same hosts. Such deliveries are always "forced",
1768 so retry times are not inspected. This resulted in far too many retries
1769 for the affected addresses. The effect occurred only when there were more
1770 hosts than the hosts_max_try setting in the smtp transport when it had
1771 the 4xx errors. Those hosts that it had tried were not added to the list
1772 of hosts for which the message was waiting, so if all were tried, there
1773 was no problem. Two fixes have been applied:
1775 (i) If there are any address or message errors in an SMTP delivery, none
1776 of the hosts (tried or untried) are now added to the list of hosts
1777 for which the message is waiting, so the message should not be a
1778 candidate for sending over the same connection that was used for a
1779 successful delivery of some other message. This seems entirely
1780 reasonable: after all the message is NOT "waiting for some host".
1781 This is so "obvious" that I'm not sure why it wasn't done
1782 previously. Hope I haven't missed anything, but it can't do any
1783 harm, as the worst effect is to miss an optimization.
1785 (ii) If, despite (i), such a delivery is accidentally attempted, the
1786 routing retry time is respected, so at least it doesn't keep
1787 hammering the server.
1789 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1790 in ${readsocket because some servers need this prod.
1792 PH/32 Added some extra debug output when updating a wait-xxx database.
1794 PH/33 The hint "could be header name not terminated by colon", which has been
1795 given for certain expansion errors for a long time, was not being given
1796 for the ${if def:h_colon_omitted{... case.
1798 PH/34 The spec says: "With one important exception, whenever a domain list is
1799 being scanned, $domain contains the subject domain." There was at least
1800 one case where this was not true.
1802 PH/35 The error "getsockname() failed: connection reset by peer" was being
1803 written to the panic log as well as the main log, but it isn't really
1804 panic-worthy as it just means the connection died rather early on. I have
1805 removed the panic log writing for the ECONNRESET error when getsockname()
1808 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1809 runs only) independently of the message's sender address. This meant
1810 that, if the 4xx error was in fact related to the sender, a different
1811 message to the same recipient with a different sender could confuse
1812 things. In particualar, this can happen when sending to a greylisting
1813 server, but other circumstances could also provoke similar problems.
1814 I have changed the default so that the retry time for these errors is now
1815 based a combination of the sender and recipient addresses. This change
1816 can be overridden by setting address_retry_include_sender=false in the
1819 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1820 remote server are returned as part of bounce messages. This was not
1821 happening for LMTP over a pipe (the lmtp transport), but now it is the
1822 same for both kinds of LMTP.
1824 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1825 in header lines that were in fact CNAMEs. This is no longer the case.
1827 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1828 and queue runs started by the daemon processed all messages. This has
1829 been fixed so that -R and -S can now usefully be given with -q<time>.
1831 PH/40 Import PCRE release 6.7 (fixes some bugs).
1833 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1835 PH/42 Give an error if -q is specified more than once.
1837 PH/43 Renamed the variables $interface_address and $interface_port as
1838 $received_ip_address and $received_port, to make it clear that these
1839 values apply to message reception, and not to the outgoing interface when
1840 a message is delivered. (The old names remain recognized, of course.)
1842 PH/44 There was no timeout on the connect() call when using a Unix domain
1843 socket in the ${readsocket expansion. There now is.
1845 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1846 be meaningful with "accept".
1848 SC/01 Eximstats V1.43
1849 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1851 SC/02 Eximstats V1.44
1852 Use a glob alias rather than an array ref in the generated
1853 parser. This improves both readability and performance.
1855 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1856 Collect SpamAssassin and rejection statistics.
1857 Don't display local sender or destination tables unless
1858 there is data to show.
1859 Added average volumes into the top table text output.
1861 SC/04 Eximstats V1.46
1862 Collect data on the number of addresses (recipients)
1863 as well as the number of messages.
1865 SC/05 Eximstats V1.47
1866 Added 'Message too big' to the list of mail rejection
1867 reasons (thanks to Marco Gaiarin).
1869 SC/06 Eximstats V1.48
1870 Mainlog lines which have GMT offsets and are too short to
1871 have a flag are now skipped.
1873 SC/07 Eximstats V1.49 (Alain Williams)
1874 Added the -emptyok flag.
1876 SC/08 Eximstats V1.50
1877 Fixes for obtaining the IP address from reject messages.
1879 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1880 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1881 whitesspace changes from 4.64-PH/27
1883 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1886 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1887 are found, allow negative numbers in numeric criteria)
1889 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1891 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1892 to match changes made in 4.64-PH/43
1894 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1896 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1897 "domain" is actually an IP address, and force a failure. This locks out
1898 those revolvers/nameservers that support "A-for-A" lookups, in
1899 contravention of the specifications.
1901 PH/48 When a host name was looked up from an IP address, and the subsequent
1902 forward lookup of the name timed out, the host name was left in
1903 $sender_host_name, contrary to the specification.
1905 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1906 restricted to single-key lookups, Exim was not diagnosing an error if
1907 * or *@ was used with a query-style lookup.
1909 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1911 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1912 long ago, but noone interested enough thought of it. Let's just say that
1913 the "1.1" means that there are some new functions that weren't there at
1914 some point in the past.
1916 PH/51 Error processing for expansion failure of helo_data from an smtp
1917 transport during callout processing was broken.
1919 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1920 tested/used via the -bh/-bhc/-bs options.
1922 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1923 bug, fixed in subsequent PCRE releases).
1925 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1926 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1928 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1934 SC/01 Use a glob alias rather than an array ref in eximstats generated
1935 parser. This improves both readability and performance.
1937 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1938 Don't display local sender or destination tables in eximstats unless
1939 there is data to show.
1940 Added average volumes into the eximstats top table text output.
1942 SC/03 Collect data on the number of addresses (recipients) as well
1943 as the number of messages in eximstats.
1945 TF/01 Correct an error in the documentation for the redirect router. Exim
1946 does (usually) call initgroups() when daemonizing.
1948 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1949 with consistent privilege compared to when running as a daemon.
1951 TF/03 Note in the spec that $authenticated_id is not set for local
1952 submissions from trusted users.
1954 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1955 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1957 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1958 by adding some example configuration directives to the default
1959 configuration file. A little bit of work is required to uncomment the
1960 directives and define how usernames and passwords are checked, but
1961 there is now a framework to start from.
1963 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1964 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1965 without this. I don't know how relevant this is to other LDAP libraries.
1967 PH/02 Add the verb name to the "unknown ACL verb" error.
1969 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1971 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1973 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1974 directory not expanded when it should be if an expanded home directory
1975 was set for the address (which is overridden by the transport).
1977 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1980 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1981 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1982 because it is too late at that time, and has no effect.
1984 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1985 security issue with \' (bugzilla #107). I could not use the
1986 PQescapeStringConn() function, because it needs a PGconn value as one of
1989 PH/08 When testing addresses using -bt, indicate those final addresses that
1990 are duplicates that would not cause an additional delivery. At least one
1991 person was confused, thinking that -bt output corresponded to deliveries.
1992 (Suppressing duplicates isn't a good idea as you lose the information
1993 about possibly different redirections that led to the duplicates.)
1995 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1996 systems where poll() doesn't work, in particular OS X.
1998 PH/10 Added more information to debugging output for retry time not reached.
2000 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2001 operations in malware.c.
2003 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2006 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2007 syslog_duplication set false, log lines that would normally be written
2008 both the the main log and to the reject log were not written to syslog at
2011 PH/14 In the default configuration, change the use of "message" in ACL warn
2012 statements to "add_header".
2014 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2015 not followed by a command (e.g. "seen endif").
2017 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2018 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2021 PH/17 Added extra conditions to the default value of delay_warning_condition
2025 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2026 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2027 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2030 The Auto-Submitted: and various List- headers are standardised, whereas I
2031 don't think Precedence: ever was.
2033 PH/18 Refactored debugging code in route_finduser() to show more information,
2034 in particular, the error code if getpwnam() issues one.
2036 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2037 This is apparently needed in addition to the PH/07 change above to avoid
2038 any possible encoding problems.
2040 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2041 but not after initializing Perl.
2043 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2044 output them only if debugging. By default they are written stderr,
2045 apparently, which is not desirable.
2047 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2050 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2053 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2055 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2056 authenticated or an ident call has been made. Suppress the default
2057 values for $authenticated_id and $authenticated_sender (but permit -oMai
2058 and -oMas) when testing with -bh.
2060 PH/24 Re-jigged the order of the tests in the default configuration so that the
2061 tests for valid domains and recipients precede the DNS black list and CSA
2062 tests, on the grounds that those ones are more expensive.
2064 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2065 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2066 This bug exists in every version of Exim that I still have, right back to
2069 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2070 However, an attempt to turn on case-sensitivity in a regex key by
2071 including (?-i) didn't work because the subject string was already
2072 lowercased, and the effects were non-intuitive. It turns out that a
2073 one-line patch can be used to allow (?-i) to work as expected.
2079 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2080 other effects) broke the use of negated acl sub-conditions.
2082 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2085 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2086 "Deny" causes Exim to reject the incoming connection with a 554 error.
2087 Unfortunately, if there is a major crisis, such as a disk failure,
2088 tcp-wrappers gives "deny", whereas what one would like would be some
2089 kind of temporary error. A kludge has been added to help with this.
2090 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2091 554 error is used if errno is still zero or contains ENOENT (which occurs
2092 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2095 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2097 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2098 errors. Otherwise a message that provokes a temporary error (when other
2099 messages do not) can cause a whole host to time out.
2101 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2102 addresses were routed directly to files or pipes from a redirect router.
2103 File deliveries just didn't batch; pipe deliveries might have suffered
2106 PH/06 A failure to get a lock for a hints database would erroneously always say
2107 "Failed to get write lock", even when it was really a read lock.
2109 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2110 of 0600. This has been changed to use the value of the lockfile_mode
2111 option (which defaults to 0600).
2113 PH/08 Applied small patch from the Sieve maintainer.
2115 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2116 folder from quota calculations, a direct delivery into this folder messed
2117 up the contents of the maildirsize file. This was because the regex was
2118 used only to exclude .Trash (or whatever) when the size of the mailbox
2119 was calculated. There was no check that a delivery was happening into an
2120 excluded directory. This bug has been fixed by ignoring all quota
2121 processing for deliveries into excluded directories.
2123 PH/10 Added the maildirfolder_create_regex option to appendfile.
2129 PH/01 The code for finding all the local interface addresses on a FreeBSD
2130 system running IPv6 was broken. This may well have applied to all BSD
2131 systems, as well as to others that have similar system calls. The broken
2132 code found IPv4 interfaces correctly, but gave incorrect values for the
2133 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2134 that it would not match correctly against @[] and not recognize the IPv6
2137 PH/02 The ipliteral router was not recognizing addresses of the form user@
2138 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2140 PH/03 Added disable_ipv6.
2142 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2143 decoded form, because it is most often used to construct To: headers
2144 lines in autoreplies, and the decoded form may well be syntactically
2145 invalid. However, $reply_address has leading white space removed, and all
2146 newlines turned into spaces so that the autoreply transport does not
2149 PH/05 If group was specified without a user on a router, and no group or user
2150 was specified on a transport, the group from the router was ignored.
2152 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2153 for visible compile-time settings that can be used to change these
2154 numbers, for those that want even more. Backwards compatibility with old
2155 spool files has been maintained. However, going back to a previous Exim
2156 release will lost any variables that are in spool files.
2158 PH/07 Two small changes when running in the test harness: increase delay when
2159 passing a TCP/IP connection to a new process, in case the original
2160 process has to generate a bounce, and remove special handling of
2161 127.0.0.2 (sic), which is no longer necessary.
2163 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2164 be the same on different OS.
2166 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2169 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2170 whether --show-vars was specified or not
2172 JJ/02 exipick: Added support for new ACL variable spool format introduced
2175 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2176 syntactically invalid From: or Reply-to: line, and a filter used this to
2177 generate an autoreply, and therefore failed to obtain an address for the
2178 autoreply, Exim could try to deliver to a non-existent relative file
2179 name, causing unrelated and misleading errors. What now happens is that
2180 it logs this as a hard delivery error, but does not attempt to create a
2183 PH/11 The exinext utility has a -C option for testing purposes, but although
2184 the given file was scanned by exinext itself; it wasn't being passed on
2185 when Exim was called.
2187 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2188 an end-of-file indication when reading a command response.
2190 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2191 compiled. In many other places in Exim, IPv6 addresses are always
2192 recognized, so I have changed this. It also means that IPv4 domain
2193 literals of the form [IPV4:n.n.n.n] are now always recognized.
2195 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2196 used if the router is not running as root, for example, when verifying at
2197 ACL time, or when using -bh. The debugging output from this situation was
2198 non-existent - all you got was a failure to exec. I have made two
2201 (a) Failures to set uid/gid, the current directory, or a process leader
2202 in a subprocess such as that created by queryprogram now generate
2203 suitable debugging ouput when -d is set.
2205 (b) The queryprogram router detects when it is not running as root,
2206 outputs suitable debugging information if -d is set, and then runs
2207 the subprocess without attempting to change uid/gid.
2209 PH/15 Minor change to Makefile for building test_host (undocumented testing
2212 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2213 additional section of a DNS packet that returns MX or SRV records.
2214 Instead, it always explicitly searches for A/AAAA records. This avoids
2215 major problems that occur when a DNS server includes only records of one
2216 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2217 fixed another bug: if SRV records were looked up and the corresponding
2218 address records were *not* found in the additional section, the port
2219 values from the SRV records were lost.
2221 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2222 using the correct key (the original address) when searching the retry
2223 rules in order to find which one to use for generating the retry hint.
2225 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2226 adding the default one. Similarly, if it contains a Reply-To: header, the
2227 errors_reply_to option, if set, is not used.
2229 PH/19 When calculating a retry time, Exim used to measure the "time since
2230 failure" by looking at the "first failed" field in the retry record. Now
2231 it does not use this if it is later than than the arrival time of the
2232 message. Instead it uses the arrival time. This makes for better
2233 behaviour in cases where some deliveries succeed, thus re-setting the
2234 "first failed" field. An example is a quota failure for a huge message
2235 when small messages continue to be delivered. Without this change, the
2236 "time since failure" will always be short, possible causing more frequent
2237 delivery attempts for the huge message than are intended.
2238 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2240 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2241 $1, $2, $3) because the numerical variables can be reset during some
2242 expansion items (e.g. "match"), thereby losing the authentication data.
2244 PH/21 Make -bV show the size of off_t variables so that the test suite can
2245 decide whether to run tests for quotas > 2G.
2247 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2248 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2249 filecount value is greater than 2G or if a quota value is greater than 2G
2250 on a system where the size of off_t is not greater than 4, a panic error
2253 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2254 never match. The debug and -bh output now contains an explicit error
2255 message indicating a malformed IPv4 address or mask.
2257 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2258 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2259 PH/23 above applies.
2261 PH/25 Do not write to syslog when running in the test harness. The only
2262 occasion when this arises is a failure to open the main or panic logs
2263 (for which there is an explicit test).
2265 PH/26 Added the /no_tell option to "control=freeze".
2267 PH/27 If a host name lookup failed very early in a connection, for example, if
2268 the IP address matched host_lookup and the reverse lookup yielded a name
2269 that did not have a forward lookup, an error message of the form "no IP
2270 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2271 could be logged. Now it outputs the IP address instead of "NULL".
2273 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2274 allows the sender and the authenticated sender to be set when
2275 submitting a message from within Exim. Since child_open_exim() is
2276 documented for local_scan(), the new function should be too.
2278 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2279 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2280 results in an empty string is now treated as unset.
2282 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2284 PH/31 Added sender_verify_fail logging option.
2286 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2287 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2288 tidied the source and removed it altogether.
2290 PH/33 When a queue run was abandoned because the load average was too high, a
2291 log line was always written; now it is written only if the queue_run log
2292 selector is set. In addition, the log line for abandonment now contains
2293 information about the queue run such as the pid. This is always present
2294 in "start" and "stop" lines but was omitted from the "abandon" line.
2296 PH/34 Omit spaces between a header name and the colon in the error message that
2297 is given when verify = headers_syntax fails (if there are lots of them,
2298 the message gets confusing).
2300 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2301 names, as there are now some PTR records that contain slashes. This check
2302 is only to protect against broken name servers that fall over on strange
2303 characters, so the fact that it applies to all lookups doesn't matter.
2305 PH/36 Now that the new test suite is complete, we can remove some of the
2306 special code in Exim that was needed for the old test suite. For example,
2307 sorting DNS records because real resolvers return them in an arbitrary
2308 order. The new test suite's fake resolver always returns records in the
2311 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2312 bounces) except when queue_only is set, to avoid logging races between
2313 the different processes.
2315 PH/38 Panic-die if .include specifies a non-absolute path.
2317 PH/39 A tweak to the "H" retry rule from its user.
2319 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2320 a label. They prevented compilation on older perls.
2322 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2323 a warning to be raised on newish perls.
2325 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2326 on queue. Changes to match documented behaviour of showing count of
2327 messages matching specified criteria.
2329 PH/40 Changed the default ident timeout from 30s to 5s.
2331 PH/41 Added support for the use of login_cap features, on those BSD systems
2332 that have them, for controlling the resources used by pipe deliveries.
2334 PH/42 The content-scanning code uses fopen() to create files in which to put
2335 message data. Previously it was not paying any attention to the mode of
2336 the files. Exim runs with umask(0) because the rest of the code creates
2337 files with open(), and sets the required mode explicitly. Thus, these
2338 files were ending up world-writeable. This was not a big issue, because,
2339 being within the spool directory, they were not world-accessible. I have
2340 created a function called modefopen, which takes an additional mode
2341 argument. It sets umask(777), creates the file, chmods it to the required
2342 mode, then resets the umask. All the relevant calls to fopen() in the
2343 content scanning code have been changed to use this function.
2345 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2346 to 24 hours. This avoids potential overflow problems when processing G
2347 and H retry rules. I suspect nobody ever tinkers with this value.
2349 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2351 PH/45 When the plaintext authenticator is running as a client, the server's
2352 challenges are checked to ensure they are valid base64 strings. By
2353 default, the authentication attempt is cancelled if an invalid string is
2354 received. Setting client_ignore_invalid_base64 true ignores these errors.
2355 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2356 they are received. Thus, the responses can be made to depend on the
2357 challenges. If an invalid string is ignored, an empty string is placed in
2360 PH/46 Messages that are created by the autoreply transport now contains a
2361 References: header, in accordance with RFCs 2822 and 3834.
2363 PH/47 Added authenticated_sender_force to the smtp transport.
2365 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2367 PH/49 Installed latest patch from the Sieve maintainer.
2369 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2370 was also set, the appendfile transport was unnecessarily scanning a
2371 directory of message files (e.g. for maildir delivery) to find the count
2372 of files (along with the size), even though it did not need this
2373 information. It now does the scan only if it needs to find either the
2374 size of the count of files.
2376 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2378 PH/52 Two bugs concerned with error handling when the smtp transport is
2381 (i) Exim was not creating retry information for temporary errors given
2382 for individual recipients after the DATA command when the smtp transport
2383 was used in LMTP mode. This meant that they could be retried too
2384 frequently, and not timed out correctly.
2386 (ii) Exim was setting the flag that allows error details to be returned
2387 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2388 recipients that were returned after the DATA command.
2390 PH/53 This is related to PH/52, but is more general: for any failing address,
2391 when detailed error information was permitted to be returned to the
2392 sender, but the error was temporary, then after the final timeout, only
2393 "retry timeout exceeded" was returned. Now it returns the full error as
2394 well as "retry timeout exceeded".
2396 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2397 do this, and (what is worse) MTAs that accept it.
2399 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2400 will now be deprecated.
2402 PH/56 New os.c-cygwin from the Cygwin maintainer.
2404 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2405 formats (previously only available in exim formats via -bpr, -bpru,
2406 and -bpra. Now also available in native and exiqgrep formats)
2408 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2409 with very large, slow to parse queues
2411 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2413 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2415 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2416 responses to authentication challenges, though it was showing the
2417 challenges; (ii) I've removed the CR characters from the debug output for
2420 PH/58 Allow for the insertion of a newline as well as a space when a string
2421 is turned into more than one encoded-word during RFC 2047 encoding. The
2422 Sieve code now uses this.
2424 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2425 data_4xx, lost_connection, tls_required.
2427 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2428 message was being sent as an SMTP response.
2430 PH/61 Add -l and -k options to exicyclog.
2432 PH/62 When verifying, if an address was redirected to one new address, so that
2433 verification continued, and the new address failed or deferred after
2434 having set something in $address_data, the value of $address_data was not
2435 passed back to the ACL. This was different to the case when no
2436 redirection occurred. The value is now passed back in both cases.
2438 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2439 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2440 use login_cap.h, so on its own it isn't the distinguishing feature. The
2441 new name refers directly to the setclassresources() function.
2443 PH/65 Added configuration files for NetBSD3.
2445 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2447 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2448 is preferred over IPv4.
2450 PH/68 The bounce_return_message and bounce_return_body options were not being
2451 honoured for bounces generated during the reception of non-SMTP messages.
2452 In particular, this applied to messages rejected by the ACL. This bug has
2453 been fixed. However, if bounce_return_message is true and bounce_return_
2454 body is false, the headers that are returned for a non-SMTP message
2455 include only those that have been read before the error was detected.
2456 (In the case of an ACL rejection, they have all been read.)
2458 PH/69 The HTML version of the specification is now built in a directory called
2459 spec_html instead of spec.html, because the latter looks like a path with
2460 a MIME-type, and this confuses some software.
2462 PH/70 Catch two compiler warnings in sieve.c.
2464 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2465 function verify_get_ident() calls ip_connect() to connect a socket, but
2466 if the "connect()" function timed out, ip_connect() used to close the
2467 socket. However, verify_get_ident() also closes the socket later, and in
2468 between Exim writes to the log, which may get opened at this point. When
2469 the socket was closed in ip_connect(), the log could get the same file
2470 descriptor number as the socket. This naturally causes chaos. The fix is
2471 not to close the socket in ip_connect(); the socket should be closed by
2472 the function that creates it. There was only one place in the code where
2473 this was missing, in the iplookup router, which I don't think anybody now
2474 uses, but I've fixed it anyway.
2476 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2477 well as to direct DNS lookups. Otherwise the handling of names in host
2478 lists is inconsistent and therefore confusing.
2484 PH/01 Two changes to the default runtime configuration:
2486 (1) Move the checks for relay_from_hosts and authenticated clients from
2487 after to before the (commented out) DNS black list checks.
2489 (2) Add control=submission to the relay_from_hosts and authenticated
2490 clients checks, on the grounds that messages accepted by these
2491 statements are most likely to be submissions.
2493 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2495 (1) Generate an error if the third argument for the ${prvs expansion is
2498 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2501 (3) Reset the variables that are obtained from the first argument of
2502 ${prvscheck and used in the second argument before leaving the code,
2503 because their memory is reclaimed, so using them afterwards may do
2506 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2507 one (it's much easier than Tom thought :-).
2509 (5) Because of (4), we can now allow for the use of $prvscheck_result
2510 inside the third argument.
2512 PH/03 For some reason, the default setting of PATH when running a command from
2513 a pipe transport was just "/usr/bin". I have changed it to
2516 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2517 anything to be listed in the output from -bV.
2519 PH/05 When a filter generated an autoreply, the entire To: header line was
2520 quoted in the delivery log line, like this:
2522 => >A.N.Other <ano@some.domain> <original@ddress> ...
2524 This has been changed so that it extracts the operative address. There
2525 may be more than one such address. If so, they are comma-separated, like
2528 => >ano@some.domain,ona@other.domain <original@ddress> ...
2530 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2531 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2532 not being looked up in the rDNS to get a host name, Exim was showing the
2533 IP address twice in Received: lines, even though the IP addresses were
2534 identical. For example:
2536 Received: from [1.2.3.4] (helo=[1.2.3.4])
2538 However, if the real host name was known, it was omitting the HELO data
2539 if it matched the actual IP address. This has been tidied up so that it
2540 doesn't show the same IP address twice.
2542 PH/07 When both +timestamp and +memory debugging was on, the value given by
2543 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2544 called by the debug printing, thereby overwriting the timestamp buffer.
2545 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2547 PH/08 When the original message was included in an autoreply transport, it
2548 always said "this is a copy of the message, including all the headers",
2549 even if body_only or headers_only was set. It now gives an appropriate
2552 PH/09 Applied a patch from the Sieve maintainer which:
2554 o fixes some comments
2555 o adds the (disabled) notify extension core
2556 o adds some debug output for the result of if/elsif tests
2557 o points to the current vacation draft in the documentation
2558 and documents the missing references header update
2562 o fixes a bug in processing the envelope test (when testing
2563 multiple envelope elements, the last element determinted the
2566 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2567 Electronic Mail") by including:
2569 Auto-submitted: auto-generated
2571 in the messages that it generates (bounce messages and others, such as
2572 warnings). In the case of bounce messages for non-SMTP mesages, there was
2573 also a typo: it was using "Auto_submitted" (underscore instead of
2574 hyphen). Since every message generated by Exim is necessarily in response
2575 to another message, thes have all been changed to:
2577 Auto-Submitted: auto-replied
2579 in accordance with these statements in the RFC:
2581 The auto-replied keyword:
2583 - SHOULD be used on messages sent in direct response to another
2584 message by an automatic process,
2586 - MUST NOT be used on manually-generated messages,
2588 - MAY be used on Delivery Status Notifications (DSNs) and Message
2589 Disposition Notifications (MDNs),
2591 - MUST NOT be used on messages generated by automatic or periodic
2592 processes, except for messages which are automatic responses to
2595 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2596 to the default Received: header definition.
2598 PH/12 Added log selector acl_warn_skipped (default on).
2600 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2601 variables because (a) they are in the wrong storage pool and (b) even if
2602 they were copied, it wouldn't work properly because of the caching.
2604 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2605 checking when decoding. Apparently there are clients that generate
2606 overlong encoded strings. Why am I not surprised?
2608 PH/15 If the first argument of "${if match_address" was not empty, but did not
2609 contain an "@" character, Exim crashed. Now it writes a panic log message
2610 and treats the condition as false.
2612 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2614 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2615 extension "envelope-auth". The code is finished and in agreement with
2616 other implementations, but there is no documentation so far and in fact,
2617 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2618 not changing the active code.
2620 Print executed "if" and "elsif" statements when debugging is used. This
2621 helps a great deal to understand what a filter does.
2623 Document more things not specified clearly in RFC3028. I had all this
2624 sorted out, when out of a sudden new issues came to my mind. Oops."
2626 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2629 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2630 canonical form (as documented). However, after a host name lookup from
2631 the IP address, check_host() was doing a simple string comparison with
2632 addresses acquired from the DNS when checking that the found name did
2633 have the original IP as one of its addresses. Since any found IPv6
2634 addresses are likely to be in abbreviated form, the comparison could
2635 fail. Luckily, there already exists a function for doing the comparison
2636 by converting both addresses to binary, so now that is used instead of
2637 the text comparison.
2639 PH/20 There was another similar case to PH/19, when a complete host name was
2640 given in a host list; looking up its IP address could give an abbreviated
2641 form, whereas the current host's name might or might not be abbreviated.
2642 The same fix has been applied.
2648 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2649 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2652 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2653 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2655 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2657 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2658 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2659 Also, for "Auto-Submitted", treat anything other than "no" as
2660 non-personal, in accordance with RFC 3834. (Previously it treated
2661 anything starting "auto-" as non-personal.)
2663 TF/01 The control=submission/name=... option had a problem with syntax
2664 errors if the name included a slash character. The /name= option
2665 now slurps the rest of the string, so it can include any characters
2666 but it must come last in the list of options (after /sender_retain
2669 PH/05 Some modifications to the interface to the fake nameserver for the new
2677 TK/01 Added the "success_on_redirect" address verification option. See
2678 NewStuff for rationale and an example.
2680 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2682 PH/02 Patch to exigrep to allow it to work on syslog lines.
2684 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2685 fread() to skip over the body file's header line, because in Cygwin the
2686 header line is locked and is inaccessible.
2688 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2689 co-exist for some time) to make it clear that it is the Exim ID that is
2690 referenced, not the Message-ID: header line.
2692 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2693 string_format() function, because snprintf() does not exist on all
2696 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2697 ${stat: expansion item.
2699 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2700 protocol synchronization error", to keep the pedants happy.
2702 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2703 well as for IRIX systems, when gcc is being used. See the host.c source
2706 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2708 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2711 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2712 transport and to the smtp transport in LMTP mode.
2714 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2716 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2717 run a filter in a subprocess. This could lead to confusion in subsequent
2718 lookups in the parent process. There should also be a search_tidyup() at
2719 the end of the subprocess.
2721 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2722 only if the host matched helo_try_verify_hosts, which caused the
2723 verification to occur when the EHLO/HELO command was issued. The ACL just
2724 tested the remembered result. Now, if a previous verification attempt has
2725 not happened, "verify = helo" does it there and then.
2727 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2729 TK/03 Fix log output including CR from clamd.
2731 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2732 exist provoked a memory error which could cause a segfault.
2734 PH/15 Installed PCRE 6.2
2736 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2738 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2739 of the problem. Specifically, suggested +O2 rather than +O1 for the
2742 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2744 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2745 record, the port was not correctly logged when the outgoing_port log
2746 selector was set (it logged the transort's default port).
2748 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2749 fallback_hosts, and "hosts" in the smtp transport.
2751 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2752 host errors such as "Connection refused".
2754 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2755 authentication with radiusclient 0.4.9:
2757 - Error returned from rc_read_config was caught wrongly
2758 - Username/password not passed on to radius server due to wrong length.
2760 The presumption is that some radiusclient API changes for 4.51/PH/17
2761 were not taken care of correctly. The code is still untested by me (my
2762 Linux distribution still has 0.3.2 of radiusclient), but it was
2763 contributed by a Radius user.
2765 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2766 expanding the "port" option of the smtp transport.
2768 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2769 while reading a MIME header. Thanks to Tom Hughes for a patch.
2771 PH/24 Include config.h inside local_scan.h so that configuration settings are
2774 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2775 that in an ACL for RCPT (for example), you can examine exactly what was
2778 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2779 commands, but it was not correctly comparing the address with the actual
2780 client host address. Thus, it would show the EHLO address in Received:
2781 header lines when this was not necessary.
2783 PH/27 Added the % operator to ${eval:}.
2785 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2786 it should be ignoring failures (because with -C, for example, it has lost
2787 privilege). It wasn't ignoring creation failures other than "already
2790 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2793 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2794 because an input line was too long, either on its own, or by virtue of
2795 too many continuations, the temporary file was not being removed, and the
2796 return code was incorrect.
2798 PH/31 Missing "BOOL" in function definition in filtertest.c.
2800 PH/32 Applied Sieve patches from the maintainer.
2802 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2804 PH/33 Added "verify = not_blind".
2806 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2807 Local/Makefile (with some defaults set). These are used in built scripts
2808 such as exicyclog, but they have never been used in the exim_install
2809 script (though there are many overriding facilities there). I have
2810 arranged that the exim_install script now takes note of these two
2813 PH/35 Installed configuration files for Dragonfly.
2815 PH/36 When a locally submitted message by a trusted user did not contain a
2816 From: header, and the sender address was obtained from -f or from an SMTP
2817 MAIL command, and the trusted user did not use -F to supply a sender
2818 name, $originator_name was incorrectly used when constructing a From:
2819 header. Furthermore, $originator_name was used for submission mode
2820 messages from external hosts without From: headers in a similar way,
2821 which is clearly wrong.
2823 PH/37 Added control=suppress_local_fixups.
2825 PH/38 When log_selector = +received_sender was set, and the addition of the
2826 sender made the log line's construction buffer exactly full, or one byte
2827 less than full, an overflow happened when the terminating "\n" was
2830 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2831 when the result of a list match is failure because a DNS lookup failed.
2833 PH/40 RM_COMMAND is now used in the building process.
2835 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2836 the "build-* directories that it finds.
2838 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2839 address in a domain literal was a prefix of an interface address.
2841 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2842 when verifying a sender address, unless rewrite_headers is false.
2844 PH/44 Wrote a long comment about why errors_to addresses are verified as
2845 recipients, not senders.
2847 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2848 the ratelimit ACL was added.
2850 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2852 PH/46 Added extra information about PostgreSQL errors to the error string.
2854 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2855 avoiding the need to install special zones in a real server. This is
2856 backwards compatible; if it can't find the fake resolver, it drops back.
2857 Thus, both old and new test suites can be run.
2859 TF/02 Added util/ratelimit.pl
2861 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2862 clock is set back in time.
2864 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2865 Candler <B.Candler@pobox.com>.
2867 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2868 OK for addresses that are the result of redirections.
2870 PH/48 A number of further additions for the benefit of the new test suite,
2871 including a fake gethostbyname() that interfaces to the fake DNS resolver
2874 TF/06 The fix for widen_domains has also been applied to qualify_single and
2875 search_parents which are the other dnslookup options that can cause
2878 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2881 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2883 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2884 service" if it encountered an MX record with an empty target hostname.
2885 The message is now "an MX or SRV record indicated no SMTP service".
2887 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2888 if the DNS of the sending site is misconfigured. This is quite a
2889 common situation. This change restores the behaviour of treating a
2890 helo verification defer as a failure.
2892 PH/51 If self=fail was set on a router, the bounce message did not include the
2893 actual error message.
2899 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2901 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2902 command itself ended in error, the underlying message about the transport
2903 filter timeout was being overwritten with the pipe command error. Now the
2904 underlying error message should be appended to the second error message.
2906 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2908 PH/02 Reduce the amount of output that "make" produces by default. Full output
2909 can still be requested.
2911 PH/03 The warning log line about a condition test deferring for a "warn" verb
2912 was being output only once per connection, rather than after each
2913 occurrence (because it was using the same function as for successful
2914 "warn" verbs). This seems wrong, so I have changed it.
2916 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2917 it should not have, which might have caused a crash in the right
2918 circumstances, but probably never did.
2920 PH/04 Installed a modified version of Tony Finch's patch to make submission
2921 mode fix the return path as well as the Sender: header line, and to
2922 add a /name= option so that you can make the user's friendly name appear
2925 TF/03 Added the control = fakedefer ACL modifier.
2927 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2928 Mark Lowes for thorough testing.
2930 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2932 TK/03 Merged latest SRS patch from Miles Wilton.
2934 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2935 with the definition in sysexits.h (which is #included earlier).
2936 Fortunately, Exim does not actually use EX_OK. The code used to try to
2937 preserve the sysexits.h value, by assumimg that macro definitions were
2938 scanned for macro replacements. I have been disabused of this notion,
2939 so now the code just undefines EX_OK before #including unistd.h.
2941 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2942 in the smtp transport. When a block could not be written in a single
2943 write() function, the timeout was being re-applied to each part-write.
2944 This seems wrong - if the receiver was accepting one byte at a time it
2945 would take for ever. The timeout is now adjusted when this happens. It
2946 doesn't have to be particularly precise.
2948 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2949 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2951 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2952 without a domain if the check to <postmaster@domain> fails.
2954 SC/01 Eximstats: added -xls and the ability to specify output files
2955 (patch written by Frank Heydlauf).
2957 SC/02 Eximstats: use FileHandles for outputing results.
2959 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2961 SC/04 Eximstats: fixed display of large numbers with -nvr option
2963 SC/05 Eximstats: fixed merging of reports with empty tables.
2965 SC/06 Eximstats: added the -include_original_destination flag
2967 SC/07 Eximstats: removed tabs and trailing whitespace.
2969 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2971 TK/06 MBOX spool code: Add real "From " MBOX separator line
2972 so the .eml file is really in mbox format (even though
2973 most programs do not really care). Patch from Alex Miller.
2975 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2976 The latter is generated from $received_to and is only set if the
2977 message has one envelope recipient. SA can use these headers,
2978 obviously out-of-the-box. Patch from Alex Miller.
2980 PH/08 The ${def test on a variable was returning false if the variable's
2981 value was "0", contrary to what the specification has always said!
2982 The result should be true unless the variable is empty.
2984 PH/09 The syntax error of a character other than { following "${if
2985 def:variable_name" (after optional whitespace) was not being diagnosed.
2986 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2987 accidental colon was present, for example, could give incorrect results.
2989 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2990 result is used (not including appendfile, where other changes are about
2993 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2994 This involved changing a lot of size variables from int to off_t. It
2995 should work with maildirs and everything.
2997 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2998 spamd dying while we are connected to it.
3000 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3003 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3004 function for BDB 4.3.
3006 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3008 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3009 the sources for PCRE have been re-arranged and more files are now
3012 PH/15 The code I had for printing potentially long long variables in PH/11
3013 above was not the best (it lost precision). The length of off_t variables
3014 is now inspected at build time, and an appropriate printing format (%ld
3015 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3016 to be "long long int" or "long int". This is needed for the internal
3017 formatting function string_vformat().
3019 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3020 the configuration file to be ":syslog", then the script "guesses" where
3021 the logs files are, rather than using the compiled in default. In our
3022 case the guess is not the same as the compiled default, so the script
3023 suddenly stopped working when I started to use syslog. The patch checks
3024 to see if log_file_path is "". If so, it attempts to read it from exim
3025 with no configuration file to get the compiled in version, before it
3026 falls back to the previous guessing code."
3028 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3029 implementing BATV in an Exim configuration. See NewStuff for the gory
3032 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3033 Makefile that are specific to HP-UX.
3035 PH/18 If the "use_postmaster" option was set for a recipient callout together
3036 with the "random" option, the postmaster address was used as the MAIL
3037 FROM address for the random test, but not for the subsequent recipient
3038 test. It is now used for both.
3040 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3041 patch removes a few documentation additions to RFC 3028, because the
3042 latest draft now contains them. It adds the new en;ascii-case comparator
3043 and a new error check for 8bit text in MIME parts. Comparator and
3044 require names are now matched exactly. I enabled the subaddress
3045 extension, but it is not well tested yet (read: it works for me)."
3047 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3048 rework some of the code of TK/09 above to avoid the hardwired use of
3049 "%lld" and "long long". Replaced the call to snprintf() with a call to
3052 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3053 records point to non-existent hosts", "retry timeout exceeded", and
3054 "retry time not reached for any host after a long failure period".
3056 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3057 experimental DomainKeys support:
3059 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3060 (2) On an error such as an illegally used "control", the wrong name for
3061 the control was given.
3063 These problems did NOT occur unless DomainKeys support was compiled.
3065 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3067 PH/24 Added ${if match_ip condition.
3069 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3070 databases so that it will be absolutely obvious if a crash occurs in the
3071 DB library. This is a regular occurrence (often caused by mis-matched
3074 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3075 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3076 picked up on a user's system that detects such things. There doesn't seem
3077 to be a gcc warning option for this - only an attribute that has to be
3078 put on the function's prototype. It seems that in Fedora Core 4 they have
3079 set this on a number of new functions. No doubt there will be more in due
3082 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3083 specify a transport. However, if an address that was verified by such a
3084 router was the subject of a callout, Exim crashed because it tried to
3085 read the rcpt_include_affixes from the non-existent transport. Now it
3086 just assumes that the setting of that option is false. This bug was
3087 introduced by 4.51/PH/31.
3089 PH/28 Changed -d+all to exclude +memory, because that information is very
3090 rarely of interest, but it makes the output a lot bigger. People tend to
3091 do -d+all out of habit.
3093 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3094 code in os-type was giving problems when libc.so lives in lib64, like on
3097 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3098 aren't the modern standard, and it seems that some systems' include files
3099 don't always have them. Exim was already checking for some of the newer
3100 ones like T_AAAA, and defining it itself. I've added checks for all the
3101 record types that Exim uses.
3103 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3104 not automatically generating a new one, as it is supposed to. This
3105 prevented TLS from working. If the file did exist, but contained invalid
3106 data, a new version was generated, as expected. It was only the case of a
3107 non-existent file that was broken.
3109 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3110 with a change in libdomainkeys > 0.64.
3112 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3113 from DNS. If the selector record carries the flag, it now has
3114 precedence over the domain-wide flag.
3116 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3118 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3119 the use of a port name that isn't defined in /etc/services) occurred, the
3120 message was deferred as in a normal delivery, and thus remained on the
3121 spool, instead of being failed because of the mua_wrapper setting. This
3122 is now fixed, and I tidied up some of the mua_wrapper messages at the
3125 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3126 the messages in a hash of arrays rather than using individual hashes.
3127 This is a bit cleaner and results in dramatic memory savings, albeit
3128 at a slight CPU cost.
3130 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3131 as requested by Marc Sherman.
3133 SC/10 Eximstats: added histograms for user specified patterns as requested
3136 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3138 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3139 fopen() in the content-scanning modules that did not already have it.
3145 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3146 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3148 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3150 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3152 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3153 to test Sieve filters that use "vacation".
3155 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3156 that changes the way the GnuTLS parameters are stored in the cache file.
3157 The new format can be generated externally. For backward compatibility,
3158 if the data in the cache doesn't make sense, Exim assumes it has read an
3159 old-format file, and it generates new data and writes a new file. This
3160 means that you can't go back to an older release without removing the
3163 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3164 work if there are any delivery delays because "one_time" forces the
3165 parent to be marked "delivered", so its unseen clone is never tried
3166 again. For this reason, Exim now forbids the simultaneous setting of
3169 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3170 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3171 entry if you want to know the details. The fix, however, neglected to
3172 consider the case when local delivery batching is involved. The test for
3173 "previously delivered" was not happening when checking to see if an
3174 address could be batched with a previous (undelivered) one; under
3175 certain circumstances this could lead to multiple deliveries to the same
3178 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3179 in its include files, and this causes problems building Exim.
3181 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3182 header_syntax) but Exim was just ignoring anything given after a slash.
3183 In particular, this caused confusion with an attempt to use "verify =
3184 reverse_host_lookup/defer_ok". An error is now given when options are
3185 supplied for verify items that do not have them. (Maybe reverse_host_
3186 lookup should have a defer_ok option, but that's a different point.)
3188 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3189 defined by RFC 821) to 2048, because there were problems with some AUTH
3190 commands, and RFC 1869 says the size should be increased for extended
3191 SMTP commands that take arguments.
3193 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3196 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3197 "unknown" error; now it says that the functionality isn't in the binary.
3199 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3200 an address' error message when a string expansion fails (syntax or
3201 whatever). Otherwise the password may appear in the log. Following change
3202 PH/42 below, there is no longer a chance of it appearing in a bounce
3205 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3207 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3208 of its addresses was ever tried. (Bugzilla bug #2.)
3210 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3211 the result incorrectly in the debug output. (It correctly added a newline
3212 to what was transported.)
3214 TF/01 Added $received_time.
3216 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3217 commented out examples of how to interface to a virus scanner and to
3218 SpamAssassin. Also added commented examples of av_scanner and
3219 spamd_address settings.
3221 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3222 and controls are allowed in which ACLs. There were a couple of minor
3223 errors. Some of the entries in the conditions table (which is a table of
3224 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3225 negation of where the condition IS allowed.
3227 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3229 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3230 header file does not have a version number, so I've had to invent a new
3231 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3232 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3233 radiusclient), but it was contributed by a Radius user.
3235 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3236 files or directories, for OpenSSL.
3238 PH/19 When an Exim process that is running as root has to create an Exim log
3239 file, it does so in a subprocess that runs as exim:exim so as to get the
3240 ownership right at creation (otherwise, other Exim processes might see
3241 the file with the wrong ownership). There was no test for failure of this
3242 fork() call, which would lead to the process getting stuck as it waited
3243 for a non-existent subprocess. Forks do occasionally fail when resources
3244 run out. I reviewed all the other calls to fork(); they all seem to check
3247 PH/20 When checking for unexpected SMTP input at connect time (before writing
3248 the banner), Exim was not dealing correctly with a non-positive return
3249 from the read() function. If the client had disconnected by this time,
3250 the result was a log entry for a synchronization error with an empty
3251 string after "input=" when read() returned zero. If read() returned -1
3252 (an event I could not check), uninitialized data bytes were printed.
3253 There were reports of junk text (parts of files, etc) appearing after
3256 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3258 PH/22 Added support for macro redefinition, and (re)definition in between
3259 driver and ACL definitions.
3261 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3262 forgetting to use the resulting value; it was using the unexpanded value.
3264 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3265 hadn't been configured. The fix is from Juergen Kreileder, who
3266 understands it better than I do:
3268 "Here's what I see happening with three configured cyrus_sasl
3269 authenticators configured (plain, login, cram-md5):
3271 On startup auth_cyrus_sasl_init() gets called for each of these.
3272 This means three calls to sasl_listmech() without a specified mech_list.
3273 => SASL tests which mechs of all available mechs actually work
3274 => three warnings about OTP not working
3275 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3277 With the patch, sasl_listmech() also gets called three times. But now
3278 SASL's mech_list option is set to the server_mech specified in the the
3279 authenticator. Or in other words, the answer from sasl_listmech()
3280 gets limited to just the mech you're testing for (which is different
3282 => the return list contains just 'plain' or 'login', 'cram-md5' or
3283 nothing depending on the value of ob->server_mech.
3285 I've just tested the patch: Authentication still works fine,
3286 unavailable mechs specified in the exim configuration are still
3287 caught, and the auth.log warnings about OTP are gone."
3289 PH/25 When debugging is enabled, the contents of the command line are added
3290 to the debugging output, even when log_selector=+arguments is not
3293 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3294 answer is "GNU", and only if the return is "GNU/something" is the answer
3297 PH/27 $acl_verify_message is now set immediately after the failure of a
3298 verification in an ACL, and so is available in subsequent modifiers. In
3299 particular, the message can be preserved by coding like this:
3301 warn !verify = sender
3302 set acl_m0 = $acl_verify_message
3304 Previously, $acl_verify_message was set only while expanding "message"
3305 and "log_message" when a very denied access.
3307 PH/28 Modified OS/os.c-Linux with
3309 -#ifndef OS_LOAD_AVERAGE
3310 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3312 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3313 nomenclature these days.)
3315 PH/29 Installed patch from the Sieve maintainer that adds the options
3316 sieve_useraddress and sieve_subaddress to the redirect router.
3318 PH/30 In these circumstances:
3319 . Two addresses routed to the same list of hosts;
3320 . First host does not offer TLS;
3321 . First host accepts first address;
3322 . First host gives temporary error to second address;
3323 . Second host offers TLS and a TLS session is established;
3324 . Second host accepts second address.
3325 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3326 and peerdn, if requested) that were in fact used only for the second
3329 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3330 attention to any local part prefix or suffix that was matched by the
3331 router that accepted the address. It now behaves in the same way as it
3332 does for delivery: the affixes are removed from the local part unless
3333 rcpt_include_affixes is set on the transport.
3335 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3336 timeout during the DATA phase of an incoming message.
3338 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3339 applied a patch sanctioned by the Sieve maintainer.
3341 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3342 the uid or gid is negative. A case of a negative gid caused this to be
3343 noticed. The fix allows for either to be negative.
3345 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3346 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3349 PH/36 The change PH/12 above was broken. Fixed it.
3351 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3352 the grounds that routing the same address twice would always produce the
3353 same answer. This might have been true once, but it is certainly no
3354 longer true now. Routing a child address may depend on the previous
3355 routing that produced that child. Some complicated redirection strategies
3356 went wrong when messages had multiple recipients, and made Exim's
3357 behaviour dependent on the order in which the addresses were given.
3359 I have moved the duplicate checking until after the routing is complete.
3360 Exim scans the addresses that are assigned to local and remote
3361 transports, and removes any duplicates. This means that more work will be
3362 done, as duplicates will always all be routed, but duplicates are
3363 presumably rare, so I don't expect this is of any significance.
3365 For deliveries to pipes, files, and autoreplies, the duplicate checking
3366 still happens during the routing process, since they are not going to be
3369 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3370 It corrects a timeout issue with spamd. This is Ian's comment: "The
3371 background is that sometimes spamd either never reads data from a
3372 connection it has accepted, or it never writes response data. The exiscan
3373 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3374 blindly assumes that writes won't block so it may never time out."
3376 PH/39 Allow G after quota size as well as K and M.
3378 PH/40 The value set for $authenticated_id in an authenticator may not contain
3379 binary zeroes or newlines because the value is written to log lines and
3380 to spool files. There was no check on this. Now the value is run through
3381 the string_printing() function so that such characters are converted to
3382 printable escape sequences.
3384 PH/41 $message_linecount is a new variable that contains the total number of
3385 lines in the message. Compare $body_linecount, which is the count for the
3388 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3389 bounce and delay warning messages, except in certain special cases, which
3392 (a) An SMTP error message from a remote host;
3393 (b) A message specified in a :fail: redirection;
3394 (c) A message specified in a "fail" command in a system filter;
3395 (d) A message specified in a FAIL return from the queryprogram router;
3396 (e) A message specified by the cannot_route_message router option.
3398 In these cases only, Exim does include the error details in bounce and
3399 warning messages. There are also a few cases where bland messages such
3400 as "unrouteable address" or "local delivery error" are given.
3402 PH/43 $value is now also set for the "else" part of a ${run expansion.
3404 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3405 being worked on, but at least Exim now implements the latest version to
3408 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3409 process to complete was treated as a delivery failure, a timeout while
3410 writing the message to the pipe was logged, but erroneously treated as a
3411 successful delivery. Such timeouts include transport filter timeouts. For
3412 consistency with the overall process timeout, these timeouts are now
3413 treated as errors, giving rise to delivery failures by default. However,
3414 there is now a new Boolean option for the pipe transport called
3415 timeout_defer, which, if set TRUE, converts the failures into defers for
3416 both kinds of timeout. A transport filter timeout is now identified in
3419 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3420 systems where "make" and "gmake" are different, calling "gmake" at top
3421 level broke things. I've arranged for the value of $(MAKE) to be passed
3422 from the Makefile to this script so that it can call the same version of
3426 A note about Exim versions 4.44 and 4.50
3427 ----------------------------------------
3429 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3430 changes of various kinds. As a consequence, a big documentation update was
3431 needed. This delayed the release for rather longer than seemed good, especially
3432 in the light of a couple of (minor) security issues. Therefore, the changes
3433 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3434 release. So 4.44 and 4.50 are in effect two different branches that both start
3437 I have left the 4.50 change log unchanged; it contains all the changes since
3438 4.43. The change log for 4.44 is below; many of its items are identical to
3439 those for 4.50. This seems to be the most sensible way to preserve the
3440 historical information.
3446 1. Minor wording change to the doc/README.SIEVE file.
3448 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3449 computation of the current number of files was incorrect.
3451 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3452 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3455 4. Give more explanation in the error message when the command for a transport
3456 filter fails to execute.
3458 5. There are several places where Exim runs a non-Exim command in a
3459 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3460 was being done only for the command run by the queryprogram router. It is
3461 now done for all such subprocesses. The other cases are: ${run, transport
3462 filters, and the commands run by the lmtp and pipe transports.
3464 6. Added CONFIGURE_GROUP build-time option.
3466 7. Some older OS have a limit of 256 on the maximum number of file
3467 descriptors. Exim was using setrlimit() to set 1000 as a large value
3468 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3469 systems. I've change it so that if it can't get 1000, it tries for 256.
3471 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3472 was an oversight, and furthermore, ever since the addition of extra
3473 controls (e.g. 4.43/32), the checks on when to allow different forms of
3474 "control" were broken. There should now be diagnostics for all cases when a
3475 control that does not make sense is encountered.
3477 9. Added the /retain_sender option to "control=submission".
3479 10. $recipients is now available in the predata ACL (oversight).
3481 11. Tidy the search cache before the fork to do a delivery from a message
3482 received from the command line. Otherwise the child will trigger a lookup
3483 failure and thereby defer the delivery if it tries to use (for example) a
3484 cached ldap connection that the parent has called unbind on.
3486 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3487 of $address_data from the recipient verification was clobbered by the
3488 sender verification.
3490 13. The value of address_data from a sender verification is now available in
3491 $sender_address_data in subsequent conditions in the ACL statement.
3493 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3495 15. Added a new option "connect=<time>" to callout options, to set a different
3498 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3499 was its contents. (It was OK if the option was not defined at all.)
3501 17. A "Completed" log line is now written for messages that are removed from
3502 the spool by the -Mrm option.
3504 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3505 information about exactly what failed.
3507 19. Added -dd to debug only the daemon process.
3509 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3510 handles timeouts, both on the server side and network timeouts. Renamed the
3511 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3513 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3514 It is now set to "smtps".
3516 22. $host_address is now set to the target address during the checking of
3517 ignore_target_hosts.
3519 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3520 being passed; this would have caused $sender_host_name to have been used if
3521 matching the list had actually called for a host name (not very likely,
3522 since this list is usually IP addresses). A host name is now passed as
3525 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3526 code that specifies a non-restarting handler (typically sigaction() in
3527 modern systems) in an attempt to fix a rare and obscure crash bug.
3529 25. Narrowed the window for a race in the daemon that could cause it to ignore
3530 SIGCHLD signals. This is not a major problem, because they are used only to
3531 wake it up if nothing else does.
3533 26. A malformed maildirsize file could cause Exim to calculate negative values
3534 for the mailbox size or file count. Odd effects could occur as a result.
3535 The maildirsize information is now recalculated if the size or filecount
3538 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3539 support for a long time. Removed HAVE_SYS_VFS_H.
3541 28. Installed the latest version of exipick from John Jetmore.
3543 29. In an address list, if the pattern was not a regular expression, an empty
3544 subject address (from a bounce message) matched only if the pattern was an
3545 empty string. Non-empty patterns were not even tested. This was the wrong
3546 because it is perfectly reasonable to use an empty address as part of a
3547 database query. An empty address is now tested by patterns that are
3548 lookups. However, all the other forms of pattern expect the subject to
3549 contain a local part and a domain, and therefore, for them, an empty
3550 address still always fails if the pattern is not itself empty.
3552 30. Exim went into a mad DNS loop when attempting to do a callout where the
3553 host was specified on an smtp transport, and looking it up yielded more
3554 than one IP address.
3556 31. Re-factored the code for checking spool and log partition space into a
3557 function that finds that data and another that does the check. The former
3558 is then used to implement four new variables: $spool_space, $log_space,
3559 $spool_inodes, and $log_inodes.
3561 32. The RFC2047 encoding function was originally intended for short strings
3562 such as real names; it was not keeping to the 75-character limit for
3563 encoded words that the RFC imposes. It now respects the limit, and
3564 generates multiple encoded words if necessary. To be on the safe side, I
3565 have increased the buffer size for the ${rfc2047: expansion operator from
3568 33. It is now permitted to omit both strings after an "if" condition; if the
3569 condition is true, the result is "true". As before, when the second string
3570 is omitted, a false condition yields an empty string. This makes it less
3571 cumbersome to write custom ACL and router conditions.
3573 34. Failure to deliver a bounce message always caused it to be frozen, even if
3574 there was an errors_to setting on the router. The errors_to setting is now
3577 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3578 canonical form (fully expanded) before being placed in
3579 $sender_host_address.
3581 36. The table in the code that translates DNS record types into text (T_A to
3582 "A" for instance) was missing entries for NS and CNAME. It is just possible
3583 that this could have caused confusion if both these types were looked up
3584 for the same domain, because the text type is used as part of Exim's
3585 per-process caching. But the chance of anyone hitting this buglet seems
3588 37. The dnsdb lookup has been extended in a number of ways.
3590 (1) There is a new type, "zns", which walks up the domain tree until it
3591 finds some nameserver records. It should be used with care.
3593 (2) There is a new type, "mxh", which is like "mx" except that it returns
3594 just the host names, not the priorities.
3596 (3) It is now possible to give a list of domains (or IP addresses) to be
3597 looked up. The behaviour when one of the lookups defers can be
3598 controlled by a keyword.
3600 (4) It is now possible to specify the separator character for use when
3601 multiple records are returned.
3603 38. The dnslists ACL condition has been extended: it is now possible to supply
3604 a list of IP addresses and/or domains to be looked up in a particular DNS
3607 39. Added log_selector=+queue_time_overall.
3609 40. When running the queue in the test harness, wait just a tad after forking a
3610 delivery process, to get repeatability of debugging output.
3612 41. Include certificate and key file names in error message when GnuTLS fails
3613 to set them up, because the GnuTLS error message doesn't include the name
3614 of the failing file when there is a problem reading it.
3616 42. Allow both -bf and -bF in the same test run.
3618 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3620 44. The "Exiscan patch" is now merged into the mainline Exim source.
3622 45. Sometimes the final signoff response after QUIT could fail to get
3623 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3624 < 0 before doing a fflush(). This bug looks as though it goes back to the
3625 introduction of TLS in release 3.20, but "sometimes" must have been rare
3626 because the tests only now provoked it.
3628 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3629 (this can affect the format of dates).
3631 47. exim_tidydb, when checking for the continued existence of a message for
3632 which it has found a message-specific retry record, was not finding
3633 messages that were in split spool directories. Consequently, it was
3634 deleting retry records that should have stayed in existence.
3636 48. Steve fixed some bugs in eximstats.
3638 49. The SPA authentication driver was not abandoning authentication and moving
3639 on to the next authenticator when an expansion was forced to fail,
3640 contradicting the general specification for all authenticators. Instead it
3641 was generating a temporary error. It now behaves as specified.
3643 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3644 (the order specifies the preference for clients). The order is now AES256,
3645 AES128, 3DES, ARCFOUR128.
3647 51. Small patch to Sieve code - explicitly set From: when generating an
3650 52. Exim crashed if a remote delivery caused a very long error message to be
3651 recorded - for instance if somebody sent an entire SpamAssassin report back
3652 as a large number of 550 error lines. This bug was coincidentally fixed by
3653 increasing the size of one of Exim's internal buffers (big_buffer) that
3654 happened as part of the Exiscan merge. However, to be on the safe side, I
3655 have made the code more robust (and fixed the comments that describe what
3658 53. Now that there can be additional text after "Completed" in log lines (if
3659 the queue_time_overall log selector is set), a one-byte patch to exigrep
3660 was needed to allow it to recognize "Completed" as not the last thing in
3663 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3664 patch that reportedly fixes this has been added. I am not expert enough to
3665 create a test for it. This is what the patch creator wrote:
3667 "I found a little strange behaviour of ldap code when working with
3668 Windows 2003 AD Domain, where users was placed in more than one
3669 Organization Units. When I tried to give exim partial DN, the exit code
3670 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3671 But simultaneously result of request was absolutely normal ldap result,
3672 so I produce this patch..."
3674 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3675 REFERENCE, so I have modified the code to exclude the patch when that macro
3678 55. Some experimental protocols are using DNS PTR records for new purposes. The
3679 keys for these records are domain names, not reversed IP addresses. The
3680 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3681 leaves it alone. Component reversal etc. now happens only for IP addresses.
3684 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3686 57. Double the size of the debug message buffer (to 2048) so that more of very
3687 long debug lines gets shown.
3689 58. The exicyclog utility now does better if the number of log files to keep
3690 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3692 59. Two changes related to the smtp_active_hostname option:
3694 (1) $smtp_active_hostname is now available as a variable.
3695 (2) The default for smtp_banner uses $smtp_active_hostname instead
3696 of $primary_hostname.
3698 60. The host_aton() function is supposed to be passed a string that is known
3699 to be a valid IP address. However, in the case of IPv6 addresses, it was
3700 not checking this. This is a hostage to fortune. Exim now panics and dies
3701 if the condition is not met. A case was found where this could be provoked
3702 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3703 components; fortuitously, this particular loophole had already been fixed
3704 by change 4.50/55 above.
3706 If there are any other similar loopholes, the new check in host_aton()
3707 itself should stop them being exploited. The report I received stated that
3708 data on the command line could provoke the exploit when Exim was running as
3709 exim, but did not say which command line option was involved. All I could
3710 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3711 running as the user.
3714 61. There was a buffer overflow vulnerability in the SPA authentication code
3715 (which came originally from the Samba project). I have added a test to the
3716 spa_base64_to_bits() function which I hope fixes it.
3719 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3720 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3722 63. The daemon start-up calls getloadavg() while still root for those OS that
3723 need the first call to be done as root, but it missed one case: when
3724 deliver_queue_load_max is set with deliver_drop_privilege. This is
3725 necessary for the benefit of the queue runner, because there is no re-exec
3726 when deliver_drop_privilege is set.
3728 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3729 This has been fixed.
3731 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3732 was in use, was not putting the data itself into the right store pool;
3733 consequently, it could be overwritten for a subsequent message in the same
3734 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3737 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3739 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3740 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3741 the function were treating the return as a boolean value, which happened to
3742 work because 0=false and not-0=true, but is not correct code.
3744 68. The host_aton() function was not handling scoped IPv6 addresses (those
3745 with, for example, "%eth0" on the end) correctly.
3747 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3748 negated items (that is, ~something) in unsigned ints. Some compilers
3749 apparently mutter when there is no cast.
3751 70. If an address verification called from an ACL failed, and did not produce a
3752 user-specific message (i.e. there was only a "system" message), nothing was
3753 put in $acl_verify_message. In this situation, it now puts the system
3756 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3757 session; change 4.31/43 added the unwanted input to the log line - except
3758 that it did not do this in the start of session case. It now does.
3760 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3761 This is wrong and can cause the other end to generate a synchronization
3762 error if it is another Exim or anything else that does the synchronization
3763 check. A QUIT command is no longer sent after a timeout.
3765 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3766 during host lookups.
3768 74. The defer_ok option of callout verification was not working if it was used
3769 when verifying addresses in header lines, that is, for this case:
3771 verify = header_sender/callout=defer_ok
3773 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3774 those file descriptors could be used for SMTP connections. If anything
3775 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3776 could be sent to the SMTP client, causing chaos. The daemon now opens
3777 stdin, stdout, and stderr to /dev/null when it puts itself into the
3780 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3781 log by default. The user can override this with suitable Perl magic.
3783 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3784 the log message when discard triggers, was not working for the DATA ACL or
3785 for the non-SMTP ACL.
3787 78. Error message wording change in sieve.c.
3789 79. If smtp_accept_max_per_host was set, the number of connections could be
3790 restricted to fewer than expected, because the daemon was trying to set up
3791 a new connection before checking whether the processes handling previous
3792 connections had finished. The check for completed processes is now done
3793 earlier. On busy systems, this bug wouldn't be noticed because something
3794 else would have woken the daemon, and it would have reaped the completed
3797 80. If a message was submitted locally by a user whose login name contained one
3798 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3799 It caused a spool format error. I have fixed the spool reading code. A
3800 related problem was that the "from" clause in the Received: line became
3801 illegal because of the space(s). It is now covered by ${quote_local_part.
3803 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3806 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3808 83. Patch from Sieve maintainer to fix unterminated string problem in
3809 "vacation" handling.
3811 84. Some minor changes to the Linux configuration files to help with other
3812 OS variants using glibc.
3814 85. One more patch for Sieve to update vacation handling to latest spec.
3817 ----------------------------------------------------
3818 See the note above about the 4.44 and 4.50 releases.
3819 ----------------------------------------------------
3825 1. Change 4.43/35 introduced a bug that caused file counts to be
3826 incorrectly computed when quota_filecount was set in an appendfile
3829 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3830 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3833 3. Give more explanation in the error message when the command for a transport
3834 filter fails to execute.
3836 4. There are several places where Exim runs a non-Exim command in a
3837 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3838 was being done only for the command run by the queryprogram router. It is
3839 now done for all such subprocesses. The other cases are: ${run, transport
3840 filters, and the commands run by the lmtp and pipe transports.
3842 5. Some older OS have a limit of 256 on the maximum number of file
3843 descriptors. Exim was using setrlimit() to set 1000 as a large value
3844 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3845 systems. I've change it so that if it can't get 1000, it tries for 256.
3847 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3848 was an oversight, and furthermore, ever since the addition of extra
3849 controls (e.g. 4.43/32), the checks on when to allow different forms of
3850 "control" were broken. There should now be diagnostics for all cases when a
3851 control that does not make sense is encountered.
3853 7. $recipients is now available in the predata ACL (oversight).
3855 8. Tidy the search cache before the fork to do a delivery from a message
3856 received from the command line. Otherwise the child will trigger a lookup
3857 failure and thereby defer the delivery if it tries to use (for example) a
3858 cached ldap connection that the parent has called unbind on.
3860 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3861 of $address_data from the recipient verification was clobbered by the
3862 sender verification.
3864 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3865 was its contents. (It was OK if the option was not defined at all.)
3867 11. A "Completed" log line is now written for messages that are removed from
3868 the spool by the -Mrm option.
3870 12. $host_address is now set to the target address during the checking of
3871 ignore_target_hosts.
3873 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3874 being passed; this would have caused $sender_host_name to have been used if
3875 matching the list had actually called for a host name (not very likely,
3876 since this list is usually IP addresses). A host name is now passed as
3879 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3880 code that specifies a non-restarting handler (typically sigaction() in
3881 modern systems) in an attempt to fix a rare and obscure crash bug.
3883 15. Narrowed the window for a race in the daemon that could cause it to ignore
3884 SIGCHLD signals. This is not a major problem, because they are used only to
3885 wake it up if nothing else does.
3887 16. A malformed maildirsize file could cause Exim to calculate negative values
3888 for the mailbox size or file count. Odd effects could occur as a result.
3889 The maildirsize information is now recalculated if the size or filecount
3892 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3893 support for a long time. Removed HAVE_SYS_VFS_H.
3895 18. Updated exipick to current release from John Jetmore.
3897 19. Allow an empty sender to be matched against a lookup in an address list.
3898 Previously the only cases considered were a regular expression, or an
3901 20. Exim went into a mad DNS lookup loop when doing a callout where the
3902 host was specified on the transport, if the DNS lookup yielded more than
3905 21. The RFC2047 encoding function was originally intended for short strings
3906 such as real names; it was not keeping to the 75-character limit for
3907 encoded words that the RFC imposes. It now respects the limit, and
3908 generates multiple encoded words if necessary. To be on the safe side, I
3909 have increased the buffer size for the ${rfc2047: expansion operator from
3912 22. Failure to deliver a bounce message always caused it to be frozen, even if
3913 there was an errors_to setting on the router. The errors_to setting is now
3916 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3917 canonical form (fully expanded) before being placed in
3918 $sender_host_address.
3920 24. Updated eximstats to version 1.33
3922 25. Include certificate and key file names in error message when GnuTLS fails
3923 to set them up, because the GnuTLS error message doesn't include the name
3924 of the failing file when there is a problem reading it.
3926 26. Expand error message when OpenSSL has problems setting up cert/key files.
3929 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3930 (this can affect the format of dates).
3932 28. exim_tidydb, when checking for the continued existence of a message for
3933 which it has found a message-specific retry record, was not finding
3934 messages that were in split spool directories. Consequently, it was
3935 deleting retry records that should have stayed in existence.
3937 29. eximstats updated to version 1.35
3938 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3939 1.35 - bugfix such that pie charts by volume are generated correctly
3941 30. The SPA authentication driver was not abandoning authentication and moving
3942 on to the next authenticator when an expansion was forced to fail,
3943 contradicting the general specification for all authenticators. Instead it
3944 was generating a temporary error. It now behaves as specified.
3946 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3947 (the order specifies the preference for clients). The order is now AES256,
3948 AES128, 3DES, ARCFOUR128.
3950 31. Small patch to Sieve code - explicitly set From: when generating an
3953 32. Exim crashed if a remote delivery caused a very long error message to be
3954 recorded - for instance if somebody sent an entire SpamAssassin report back
3955 as a large number of 550 error lines. This bug was coincidentally fixed by
3956 increasing the size of one of Exim's internal buffers (big_buffer) that
3957 happened as part of the Exiscan merge. However, to be on the safe side, I
3958 have made the code more robust (and fixed the comments that describe what
3961 33. Some experimental protocols are using DNS PTR records for new purposes. The
3962 keys for these records are domain names, not reversed IP addresses. The
3963 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3964 leaves it alone. Component reversal etc. now happens only for IP addresses.
3967 34. The host_aton() function is supposed to be passed a string that is known
3968 to be a valid IP address. However, in the case of IPv6 addresses, it was
3969 not checking this. This is a hostage to fortune. Exim now panics and dies
3970 if the condition is not met. A case was found where this could be provoked
3971 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3972 components; fortuitously, this particular loophole had already been fixed
3973 by change 4.50/55 or 4.44/33 above.
3975 If there are any other similar loopholes, the new check in host_aton()
3976 itself should stop them being exploited. The report I received stated that
3977 data on the command line could provoke the exploit when Exim was running as
3978 exim, but did not say which command line option was involved. All I could
3979 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3980 running as the user.
3983 35. There was a buffer overflow vulnerability in the SPA authentication code
3984 (which came originally from the Samba project). I have added a test to the
3985 spa_base64_to_bits() function which I hope fixes it.
3988 36. The daemon start-up calls getloadavg() while still root for those OS that
3989 need the first call to be done as root, but it missed one case: when
3990 deliver_queue_load_max is set with deliver_drop_privilege. This is
3991 necessary for the benefit of the queue runner, because there is no re-exec
3992 when deliver_drop_privilege is set.
3994 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3995 was in use, was not putting the data itself into the right store pool;
3996 consequently, it could be overwritten for a subsequent message in the same
3997 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4000 38. Sometimes the final signoff response after QUIT could fail to get
4001 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4002 < 0 before doing a fflush(). This bug looks as though it goes back to the
4003 introduction of TLS in release 3.20, but "sometimes" must have been rare
4004 because the tests only now provoked it.
4010 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4011 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4012 Now it returns an integer. A number of calls were still expecting a T/F
4013 return. Fortuitously, in all cases, the tests worked in OK situations,
4014 which is the norm. However, things would have gone wrong on any write
4015 failures on the smtp file descriptor. This function is used when sending
4016 messages over SMTP and also when doing verify callouts.
4018 2. When Exim is called to do synchronous delivery of a locally submitted
4019 message (the -odf or -odi options), it no longer closes stderr before doing
4022 3. Implemented the mua_wrapper option.
4024 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4026 5. Implemented the functions header_remove(), header_testname(),
4027 header_add_at_position(), and receive_remove_recipient(), and exported them
4030 6. If an ACL "warn" statement specified the addition of headers, Exim already
4031 inserted X-ACL-Warn: at the start if there was no header name. However, it
4032 was not making this test for the second and subsequent header lines if
4033 there were newlines in the string. This meant that an invalid header could
4034 be inserted if Exim was badly configured.
4036 7. Allow an ACL "warn" statement to add header lines at the start or after all
4037 the Received: headers, as well as at the end.
4039 8. Added the rcpt_4xx retry error code.
4041 9. Added postmaster_mailfrom=xxx to callout verification option.
4043 10. Added mailfrom=xxxx to the callout verification option, for verify=
4046 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4047 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4049 12. Inserted some casts to stop certain compilers warning when using pointer
4050 differences as field lengths or precisions in printf-type calls (mostly
4051 affecting debugging statements).
4053 13. Added optional readline() support for -be (dynamically loaded).
4055 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4056 same clock tick as a message's arrival, so that its received time was the
4057 same as the "first fail" time on the retry record, and that message
4058 remained on the queue past the ultimate address timeout, every queue runner
4059 would try a delivery (because it was past the ultimate address timeout) but
4060 after another failure, the ultimate address timeout, which should have then
4061 bounced the address, did not kick in. This was a "< instead of <=" error;
4062 in most cases the first failure would have been in the next clock tick
4063 after the received time, and all would be well.
4065 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4066 being recognized when the domain list was tested by the match_domain
4067 condition in an expansion string.
4069 16. Added the ${str2b64: operator.
4071 17. Exim was always calling setrlimit() to set a large limit for the number of
4072 processes, without checking whether the existing limit was already
4073 adequate. (It did check for the limit on file descriptors.) Furthermore,
4074 errors from getrlimit() and setrlimit() were being ignored. Now they are
4075 logged to the main and panic logs, but Exim does carry on, to try to do its
4076 job under whatever limits there are.
4078 18. Imported PCRE 5.0.
4080 19. Trivial typo in log message " temporarily refused connection" (the leading
4083 20. If the log selector return_path_on_delivery was set and an address was
4084 redirected to /dev/null, the delivery process crashed because it assumed
4085 that a return path would always be set for a "successful" delivery. In this
4086 case, the whole delivery is bypassed as an optimization, and therefore no
4089 21. Internal re-arrangement: the function for sending a challenge and reading
4090 a response while authentication was assuming a zero-terminated challenge
4091 string. It's now changed to take a pointer and a length, to allow for
4092 binary data in such strings.
4094 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4096 23. Exim was not respecting finduser_retries when seeking the login of the
4097 uid under which it was called; it was always trying 10 times. (The default
4098 setting of finduser_retries is zero.) Also, it was sleeping after the final
4099 failure, which is pointless.
4101 24. Implemented tls_on_connect_ports.
4103 25. Implemented acl_smtp_predata.
4105 26. If the domain in control=submission is set empty, Exim assumes that the
4106 authenticated id is a complete email address when it generates From: or
4107 Sender: header lines.
4109 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4110 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4111 chown and chgrp in /bin and hostname in /usr/bin.
4113 28. Exim was keeping the "process log" file open after each use, just as it
4114 does for the main log. This opens the possibility of it remaining open for
4115 long periods when the USR1 signal hits a daemon. Occasional processlog
4116 errors were reported, that could have been caused by this. Anyway, it seems
4117 much more sensible not to leave this file open at all, so that is what now
4120 29. The long-running daemon process does not normally write to the log once it
4121 has entered its main loop, and it closes the log before doing so. This is
4122 so that log files can straightforwardly be renamed and moved. However,
4123 there are a couple of unusual error situations where the daemon does write
4124 log entries, and I had neglected to close the log afterwards.
4126 30. The text of an SMTP error response that was received during a remote
4127 delivery was being truncated at 512 bytes. This is too short for some of
4128 the long messages that one sometimes sees. I've increased the limit to
4131 31. It is now possible to make retry rules that apply only when a message has a
4132 specific sender, in particular, an empty sender.
4134 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4135 it possible to be selective about when SMTP synchronization is enforced.
4137 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4139 32. Implemented hosts_connection_nolog.
4141 33. Added an ACL for QUIT.
4143 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4146 35. Added mailbox_size and mailbox_filecount to appendfile.
4148 36. Added control = no_multiline_responses to ACLs.
4150 37. There was a bug in the logic of the code that waits for the clock to tick
4151 in the case where the clock went backwards by a substantial amount such
4152 that the microsecond fraction of "now" was more than the microsecond
4153 fraction of "then" (but the whole seconds number was less).
4155 38. Added support for the libradius Radius client library this is found on
4156 FreeBSD (previously only the radiusclient library was supported).
4162 1. When certain lookups returned multiple values in the form name=value, the
4163 quoting of the values was not always being done properly. Specifically:
4164 (a) If the value started with a double quote, but contained no whitespace,
4166 (b) If the value contained whitespace other than a space character (i.e.
4167 tabs or newlines or carriage returns) it was not quoted.
4168 This fix has been applied to the mysql and pgsql lookups by writing a
4169 separate quoting function and calling it from the lookup code. The fix
4170 should probably also be applied to nisplus, ibase and oracle lookups, but
4171 since I cannot test any of those, I have not disturbed their existing code.
4173 2. A hit in the callout cache for a specific address caused a log line with no
4174 reason for rejecting RCPT. Now it says "Previous (cached) callout
4175 verification failure".
4177 3. There was an off-by-one bug in the queryprogram router. An over-long
4178 return line was truncated at 256 instead of 255 characters, thereby
4179 overflowing its buffer with the terminating zero. As well as fixing this, I
4180 have increased the buffer size to 1024 (and made a note to document this).
4182 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4183 when Exim is waiting for an SMTP response from a remote server, Exim
4184 restarts its select() call on the socket, thereby resetting its timeout.
4185 This is not a problem when such interrupts are rare. Somebody set up a cron
4186 job to run exiwhat every 2 minutes, which is less than the normal select()
4187 timeout (5 or 10 minutes). This meant that the select() timeout never
4188 kicked in because it was always reset. I have fixed this by comparing the
4189 time when an interrupt arrives with the time at the start of the first call
4190 to select(). If more time than the timeout has elapsed, the interrupt is
4191 treated as a timeout.
4193 5. Some internal re-factoring in preparation for the addition of Sieve
4194 extensions (by MH). In particular, the "personal" test is moved to a
4195 separate function, and given an option for scanning Cc: and Bcc: (which is
4196 not set for Exim filters).
4198 6. When Exim created an email address using the login of the caller as the
4199 local part (e.g. when creating a From: or Sender: header line), it was not
4200 quoting the local part when it contained special characters such as @.
4202 7. Installed new OpenBSD configuration files.
4204 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4205 try to make them clearer.
4207 9. Callout options, other than the timeout value, were being ignored when
4208 verifying sender addresses in header lines. For example, when using
4210 verify = header_sender/callout=no_cache
4212 the cache was (incorrectly) being used.
4214 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4215 only the Cygwin environment.
4217 11. When return_path_on_delivery was set as a log selector, if different remote
4218 addresses in the same message used different return paths and parallel
4219 remote delivery occurred, the wrong values would sometimes be logged.
4220 (Whenever a remote delivery process finished, the return path value from
4221 the most recently started remote delivery process was logged.)
4223 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4224 lines when AUTH and/or TLS are in use. This is the "received protocol"
4225 field. Exim used to use "asmtp" for authenticated SMTP, without any
4226 indication (in the protocol name) for TLS use. Now it follows the RFC and
4227 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4228 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4229 names appear in log lines as well as in Received: header lines.
4231 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4232 extensions, and comparison tests, and to fix some bugs.
4234 14. Changes to the "personal" filter test:
4236 (1) The test was buggy in that it was just doing the equivalent of
4237 "contains" tests on header lines. For example, if a user's address was
4238 anne@some.where, the "personal" test would incorrectly be true for
4240 To: susanne@some.where
4242 This test is now done by extracting each address from the header in turn,
4243 and checking the entire address. Other tests that are part of "personal"
4244 are now done using regular expressions (for example, to check local parts
4245 of addresses in From: header lines).
4247 (2) The list of non-personal local parts in From: addresses has been
4248 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4249 taken from the Sieve specification recommendations.
4251 (3) If the message contains any header line starting with "List-" it is
4252 treated as non-personal.
4254 (4) The test for "circular" in the Subject: header line has been removed
4255 because it now seems ill-conceived.
4257 15. Minor typos in src/EDITME comments corrected.
4259 16. Installed latest exipick from John Jetmore.
4261 17. If headers_add on a router specified a text string that was too long for
4262 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4263 of string_sprintf() is now avoided.
4265 18. $message_body_size was not set (it was always zero) when running the DATA
4266 ACL and the local_scan() function.
4268 19. For the "mail" command in an Exim filter, no default was being set for
4269 the once_repeat time, causing a random time value to be used if "once" was
4270 specified. (If the value happened to be <= 0, no repeat happened.) The
4271 default is now 0s, meaning "never repeat". The "vacation" command was OK
4272 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4273 (I found it when inspecting the code).
4275 20. There is now an overall timeout for performing a callout verification. It
4276 defaults to 4 times the callout timeout, which applies to individual SMTP
4277 commands during the callout. The overall timeout applies when there is more
4278 than one host that can be tried. The timeout is checked before trying the
4279 next host. This prevents very long delays if there are a large number of
4280 hosts and all are timing out (e.g. when the network connections are timing
4281 out). The value of the overall timeout can be changed by specifying an
4282 additional sub-option for "callout", called "maxwait". For example:
4284 verify = sender/callout=5s,maxwait=20s
4286 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4287 to the end before writing, but this should make it even safer).
4289 22. Exim was forgetting that it had advertised PIPELINING for the second and
4290 subsequent messages on an SMTP connection. It was also not resetting its
4291 memory on STARTTLS and an internal HELO.
4293 23. When Exim logs an SMTP synchronization error within a session, it now
4294 records whether PIPELINING has been advertised or not.
4296 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4297 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4298 rather than long int.
4300 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4302 26. Added the never_mail option to autoreply.
4308 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4309 crash if the getsockname() call failed; this can happen if a connection is
4310 closed very soon after it is established. The problem was simply in the
4311 order in which certain operations were done, causing Exim to try to write
4312 to the SMTP stream before it had set up the file descriptor. The bug has
4313 been fixed by making things happen in the correct order.
4319 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4320 before the connection was closed, thus losing the rejection response.
4322 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4323 some early Solaris releases, but causes trouble in current releases where
4324 socklen_t is defined.
4326 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4329 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4332 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4333 that releasing the top of it at the end releases what was used for sub-
4334 expansions (unless the block got too big). However, discard this block if
4335 the first thing is a variable or header, so that we can use its block when
4336 it is dynamic (useful for very large $message_headers, for example).
4338 6. Lookups now cache *every* query, not just the most recent. A new, separate
4339 store pool is used for this. It can be recovered when all lookup caches are
4340 flushed. Lookups now release memory at the end of their result strings.
4341 This has involved some general refactoring of the lookup sources.
4343 7. Some code has been added to the store_xxx() functions to reduce the amount
4344 of flapping under certain conditions.
4346 8. log_incoming_interface used to affect only the <= reception log lines. Now
4347 it causes the local interface and port to be added to several more SMTP log
4348 lines, for example "SMTP connection from", and rejection lines.
4350 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4352 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4354 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4355 could be overwritten at the end of the current message (or the start of a
4356 new message if it was set in a HELO ACL). The value is now preserved for
4357 the duration of the SMTP connection.
4359 12. If a transport had a headers_rewrite setting, and a matching header line
4360 contained an unqualified address, that address was qualified, even if it
4361 did not match any rewriting rules. The underlying bug was that the values
4362 of the flags that permit the existence of unqualified sender and recipient
4363 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4364 non-local messages, and by -bnq for local messages) were not being
4365 preserved with the message after it was received.
4367 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4368 "next input=" as part of the text comprising the host identity instead of
4369 the correct text. The code was using the same buffer for two different
4370 strings. However, depending on which order the printing function evaluated
4371 its arguments, the bug did not always show up. Under Linux, for example, my
4372 test suite worked just fine.
4374 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4375 doesn't work with some older versions of Perl. It has been changed to "my",
4376 which in any case is probably the better facility to use.
4378 15. A really picky compiler found some instances of statements for creating
4379 error messages that either had too many or two few arguments for the format
4382 16. The size of the buffer for calls to the DNS resolver has been increased
4383 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4384 for addresses that have a lot of PTR records. This alleviates a problem; it
4385 does not fully solve it.
4387 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4388 buffer now truncates the list and logs the incident, which is the same
4389 action as happens when Exim is looking up a host name and its aliases.
4390 Previously in this situation something unpredictable would happen;
4391 sometimes it was "internal error: store_reset failed".
4393 18. If a server dropped the connection unexpectedly when an Exim client was
4394 using GnuTLS and trying to read a response, the client delivery process
4395 crashed while trying to generate an error log message.
4397 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4398 string, for example:
4400 warn message = H1: something\nH2: something
4402 the text was added as a single header line from Exim's point of view
4403 though it ended up OK in the delivered message. However, searching for the
4404 second and subsequent header lines using $h_h2: did not work. This has been
4405 fixed. Similarly, if a system filter added multiple headers in this way,
4406 the routers could not see them.
4408 20. Expanded the error message when iplsearch is called with an invalid key to
4409 suggest using net-iplsearch in a host list.
4411 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4412 ACLs are no longer actually imposed (and a message to that effect is
4415 22. If a "gecos" field in a passwd entry contained escaped characters, in
4416 particular, if it contained a \" sequence, Exim got it wrong when building
4417 a From: or a Sender: header from that name. A second bug also caused
4418 incorrect handling when an unquoted " was present following a character
4419 that needed quoting.
4421 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4422 was not being matched caselessly.
4424 24. Arranged for all hyphens in the exim.8 source to be escaped with
4427 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4428 properly. Recipient callout cache records were still being keyed to include
4429 the sender, even when use_sender was set false. This led to far more
4430 callouts that were necessary. The sender is no longer included in the key
4431 when use_sender is false.
4433 26. Added "control = submission" modifier to ACLs.
4435 27. Added the ${base62d: operator to decode base 62 numbers.
4437 28. dnsdb lookups can now access SRV records.
4439 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4440 the configuration file.
4442 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4443 (-v) mode. This makes the output for a verbose queue run more intelligible.
4445 31. Added a use_postmaster feature to recipient callouts.
4447 32. Added the $body_zerocount variable, containing the number of binary zero
4448 bytes in the message body.
4450 33. The time of last modification of the "new" subdirectory is now used as the
4451 "mailbox time last read" when there is a quota error for a maildir
4454 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4456 35. Added +ignore_unknown as a special item in host lists.
4458 36. Code for decoding IPv6 addresses in host lists is now included, even if
4459 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4460 address was recognized as an IP address, but was then not correctly decoded
4461 into binary, causing unexpected and incorrect effects when compared with
4468 1. Very minor rewording of debugging text in manualroute to say "list of
4469 hosts" instead of "hostlist".
4471 2. If verify=header_syntax was set, and a header line with an unqualified
4472 address (no domain) and a large number of spaces between the end of the
4473 name and the colon was received, the reception process suffered a buffer
4474 overflow, and (when I tested it) crashed. This was caused by some obsolete
4475 code that should have been removed. The fix is to remove it!
4477 3. When running in the test harness, delay a bit after writing a bounce
4478 message to get a bit more predictability in the log output.
4480 4. Added a call to search_tidyup() just before forking a reception process. In
4481 theory, someone could use a lookup in the expansion of smtp_accept_max_
4482 per_host which, without the tidyup, could leave open a database connection.
4484 5. Added the variables $recipient_data and $sender_data which get set from a
4485 lookup success in an ACL "recipients" or "senders" condition, or a router
4486 "senders" option, similar to $domain_data and $local_part_data.
4488 6. Moved the writing of debug_print from before to after the "senders" test
4491 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4492 problems for message scanning, either using a data ACL, or using
4493 local_scan() because the Received: header was not generated till after they
4494 were called (in order to set the time as the time of reception completion).
4495 I have revised the way this works. The header is now generated after the
4496 body is received, but before the ACL or local_scan() are called. After they
4497 are run, the timestamp in the header is updated.
4503 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4504 before starting a queue runner without re-exec. This happened only when
4505 deliver_drop_privilege was set or when the Exim user was set to root. The
4506 effect of the bug was that timeouts during subsequent deliveries caused
4507 crashes instead of being properly handled. The handler is now left at its
4508 default (and expected) setting.
4510 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4511 message, again when deliver_drop_privilege is set or Exim is run as root.
4512 The bug described in (1) was not present in this case, but the tidying up
4513 of the other signals was missing. I have made the two cases consistent.
4515 3. The ignore_target_hosts setting on a manualroute router was being ignored
4516 for hosts that were looked up using the /MX notation.
4518 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4521 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4522 operated on the sender address. After changing the $sender_address to <>
4523 for the sender address verify, Exim was re-instated it as the original
4524 (before rewriting) address, but remembering that it had rewritten it, so it
4525 wasn't rewriting it again. This bug also had the effect of breaking the
4526 sender address verification caching when the sender address was rewritten.
4528 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4529 This has been changed so that if the ip literal address matches
4530 ignore_target_hosts, the router declines.
4532 7. Added expansion conditions match_domain, match_address, and match_local_
4533 part (NOT match_host).
4535 8. The placeholder for the Received: header didn't have a length field set.
4537 9. Added code to Exim itself and to exim_lock to test for a specific race
4538 condition that could lead to file corruption when using MBX delivery. The
4539 issue is with the lockfile that is created in /tmp. If this file is removed
4540 after a process has opened it but before that process has acquired a lock,
4541 there is the potential for a second process to recreate the file and also
4542 acquire a lock. This could lead to two Exim processes writing to the file
4543 at the same time. The added code performs the same test as UW imapd; it
4544 checks after acquiring the lock that its file descriptor still refers to
4545 the same named file.
4547 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4548 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4551 11. Added the smtp_active_hostname option. If used, this will typically be made
4552 to depend on the incoming interface address. Because $interface_address is
4553 not set up until the daemon has forked a reception process, error responses
4554 that can happen earlier (such as "too many connections") no longer contain
4557 12. If an expansion in a condition on a "warn" statement fails because a lookup
4558 defers, the "warn" statement is abandoned, and the next ACL statement is
4559 processed. Previously this caused the whole ACL to be aborted.
4561 13. Added the iplsearch lookup type.
4563 14. Added ident_timeout as a log selector.
4565 15. Added tls_certificate_verified as a log selector.
4567 16. Added a global option tls_require_ciphers (compare the smtp transport
4568 option of the same name). This controls incoming TLS connections.
4570 17. I finally figured out how to make tls_require_ciphers do a similar thing
4571 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4572 before starting the TLS session.
4574 18. Tabs are now shown as \t in -bP output.
4576 19. If the log selector return_path_on_delivery was set, Exim crashed when
4577 bouncing a message because it had too many Received: header lines.
4579 20. If two routers both had headers_remove settings, and the first one included
4580 a superfluous trailing colon, the final name in the first list and the
4581 first name in the second list were incorrectly joined into one item (with a
4582 colon in the middle).
4588 1. Added -C and -D options to the exinext utility, mainly to make it easier
4589 to include in the automated testing, but these could be helpful when
4590 multiple configurations are in use.
4592 2. The exinext utility was not formatting the output nicely when there was
4593 an alternate port involved in the retry record key, nor when there was a
4594 message id as well (for retries that were specific to a specific message
4595 and a specific host). It was also confused by IPv6 addresses, because of
4596 the additional colons they contain. I have fixed the IPv4 problem, and
4597 patched it up to do a reasonable job for IPv6.
4599 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4600 delivery, the log line now contains "pipelined" if PIPELINING was used.
4602 4. An SMTP transport process used to panic and die if the bind() call to set
4603 an explicit outgoing interface failed. This has been changed; it is now
4604 treated in the same way as a connect() failure.
4606 5. A reference to $sender_host_name in the part of a conditional expansion
4607 that was being skipped was still causing a DNS lookup. This no longer
4610 6. The def: expansion condition was not recognizing references to header lines
4611 that used bh_ and bheader_.
4613 7. Added the _cache feature to named lists.
4615 8. The code for checking quota_filecount in the appendfile transport was
4616 allowing one more file than it should have been.
4618 9. For compatibility with Sendmail, the command line option
4626 and sets the incoming protocol and host name (for trusted callers). The
4627 host name and its colon can be omitted when only the protocol is to be set.
4628 Note the Exim already has two private options, -pd and -ps, that refer to
4629 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4630 "s", but I don't think that's a major issue.
4632 10. A number of refactoring changes to the code, none of which should affect
4635 (a) The number of logging options was getting close to filling up the
4636 32-bit word that was used as a bit map. I have split them into two classes:
4637 those that are passed in the argument to log_write(), and those that are
4638 only ever tested independently outside of that function. These are now in
4639 separate 32-bit words, so there is plenty of room for expansion again.
4640 There is no change in the user interface or the logging behaviour.
4642 (b) When building, for example, log lines, the code previously used a
4643 macro that called string_cat() twice, in order to add two strings. This is
4644 not really sufficiently general. Furthermore, there was one instance where
4645 it was actually wrong because one of the argument was used twice, and in
4646 one call a function was used. (As it happened, calling the function twice
4647 did not affect the overall behaviour.) The macro has been replaced by a
4648 function that can join an arbitrary number of extra strings onto a growing
4651 (c) The code for expansion conditions now uses a table and a binary chop
4652 instead of a serial search (which was left over from when there were very
4653 few conditions). Also, it now recognizes conditions like "pam" even when
4654 the relevant support is not compiled in: a suitably worded error message is
4655 given if an attempt is made to use such a condition.
4657 11. Added ${time_interval:xxxxx}.
4659 12. A bug was causing one of the ddress fields not to be passed back correctly
4660 from remote delivery subprocesses. The field in question was not being
4661 subsequently used, so this caused to problems in practice.
4663 13. Added new log selectors queue_time and deliver_time.
4665 14. Might have fixed a bug in maildirsizefile handling that threw up
4666 "unexpected character" debug warnings, and recalculated the data
4667 unnecessarily. In any case, I expanded the warning message to give more
4670 15. Added the message "Restricted characters in address" to the statements in
4671 the default ACL that block characters like @ and % in local parts.
4673 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4674 Three changes have been made:
4676 (a) There was a serious bug; a negative response to MAIL caused the whole
4677 recipient domain to be cached as invalid, thereby blocking all messages
4678 to all local parts at the same domain, from all senders. This bug has
4679 been fixed. The domain is no longer cached after a negative response to
4680 MAIL if the sender used is not empty.
4682 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4685 (c) A new callout option, "use_sender" has been added for people who want
4686 the modified behaviour.
4692 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4695 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4696 indeed breaks things for older releases.
4698 3. Added additional logging to the case where there is a problem reading data
4699 from a filter that is running in a subprocess using a pipe, in order to
4700 try to track down a specific problem.
4702 4. Testing facility fudge: when running in the test harness and attempting
4703 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4704 getting "No route to host". Convert this to a timeout.
4706 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4709 6. Some OS don't have socklen_t but use size_t instead. This affects the
4710 fifth argument of getsockopt() amongst other things. This is now
4711 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4712 can be set for individual OS. I have set it for SunOS5, OSF1, and
4713 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4714 some earlier ones do not.
4716 7. Change 4.30/15 was not doing the test caselessly.
4718 8. The standard form for an IPv6 address literal was being rejected by address
4719 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4720 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4721 this, as well as the form without the "IPv6" on the front (but only when
4722 address literals are enabled, of course).
4724 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4726 10. Exim crashed if a message with an empty sender address specified by -f
4727 encountered a router with an errors_to setting. This could be provoked only
4728 by a command such as
4732 where an empty string was supplied; "<>" did not hit this bug.
4734 11. Installed PCRE release 4.5.
4736 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4737 remained set. It is now erased.
4739 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4740 times from message ids (which are base 36 rather than the normal 62).
4742 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4743 were being counted as actual protocol errors, and logged if the log
4744 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4745 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4746 and DATA following a set of rejected RCPTs do not count as protocol errors.
4747 In other words, Exim assumes they were pipelined, though this may not
4748 actually be the case. Of course, in all cases the client gets an
4749 appropriate error code.
4751 15. If a lookup fails in an ACL condition, a message about the failure may
4752 be available; it is used if testing the ACL cannot continue, because most
4753 such messages specify what the cause of the deferral is. However, some
4754 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4755 that caused an old message to be retained and used if a later statement
4756 caused a defer, replacing the real cause of the deferral.
4758 16. If an IP address had so many PTR records that the DNS lookup buffer
4759 was not large enough to hold them, Exim could crash while trying to process
4760 the truncated data. It now detects and logs this case.
4762 17. Further to 4.21/58, another change has been made: if (and only if) the
4763 first line of a message (the first header line) ends with CRLF, a bare LF
4764 in a subsequent header line has a space inserted after it, so as not to
4765 terminate the header.
4767 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4768 unnecessarily, used atoi() instead of strtol(), and didn't check the
4769 termination when getting file sizes from file names by regex.
4771 19. Completely re-implemented the support for maildirsize files, in the light
4772 of a number of problems with the previous contributed implementation
4773 (4.30/29). In particular:
4775 . If the quota is zero, the maildirsize file is maintained, but no quota is
4778 . If the maildir directory does not exist, it is created before any attempt
4779 to write a maildirsize file.
4781 . The quota value in the file is just a cache; if the quota is changed in
4782 the transport, the new value overrides.
4784 . A regular expression is available for excluding directories from the
4787 20. The autoreply transport checks the characters in options that define the
4788 message's headers; it allows continued headers, but it was checking with
4789 isspace() after an embedded newline instead of explicitly looking for a
4792 21. If all the "regular" hosts to which an address was routed had passed their
4793 expiry times, and had not reached their retry times, the address was
4794 bounced, even if fallback hosts were defined. Now Exim should go on to try
4797 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4798 equivalent code in the SMTP transport. Some hosts send humungous responses
4799 to HELO/EHLO, more than 1024 it seems.
4801 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4802 gives compiler warnings in some environments. I've now done it "properly",
4805 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4806 (because of problems with the built-in one) was declared to return uschar *
4807 instead of char *, causing compiler failure.
4809 25. Fixed a file descriptor leak when processing alias/forward files.
4811 26. Fixed a minor format string issue in dbfn.c.
4813 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4815 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4816 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4819 29. When a sender address is verified, it is cached, to save repeating the test
4820 when there is more than one recipient in a message. However, when the
4821 verification involves a callout, it is possible for different callout
4822 options to be set for different recipients. It is too complicated to keep
4823 track of this in the cache, so now Exim always runs a verification when a
4824 callout is required, relying on the callout cache for the optimization.
4825 The overhead is duplication of the address routing, but this should not be
4828 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4829 to be verified with callout=postmaster, and the main callout worked but the
4830 postmaster check failed, the verification correctly failed. However, if a
4831 subsequent RCPT command asked for sender verification *without* the
4832 postmaster check, incorrect caching caused this verification also to fail,
4835 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4836 it was not caching the DNS options (qualify_single, search_parents) that
4837 were used when the lookup failed. A subsequent lookup with different
4838 options therefore always gave the same answer, though there were cases
4839 where it should not have. (Example: a "domains = !$mx_any" option on a
4840 dnslookup router: the "domains" option is always processed without any
4841 widening, but the router might have qualify_single set.) Now Exim uses the
4842 cached value only when the same options are set.
4844 32. Added John Jetmore's "exipick" utility to the distribution.
4846 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4847 than a timeout (e.g. a certificate is required, and is not provided), an
4848 Exim server now closes the connection immediately. Previously it waited for
4849 the client to close - but if the client is SSL, it seems that they each
4850 wait for each other, leading to a delay before one of them times out.
4852 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4853 maintained 0.8.x compatibility because I don't think many are using it, and
4854 it is clearly obsolete.
4856 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4859 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4860 client certificate was expired. A simple patch fixes this, though I don't
4861 understand the full logic of why the verify callback is called multiple
4864 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4865 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4866 which causes problems with some clients (such as the Certicom SSL Plus
4867 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4868 disables the coutermeasure allowing Eudora to connect."
4870 38. Exim was not checking that a write() to a log file succeeded. This could
4871 lead to Bad Things if a log got too big, in particular if it hit a file
4872 size limit. Exim now panics and dies if it cannot write to a log file, just
4873 as it does if it cannot open a log file.
4875 39. Modified OS/Makefile-Linux so that it now contains
4877 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4879 The two -D definitions ensure that Exim is compiled with large file
4880 support, which makes it possible to handle log files that are bigger than
4883 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4884 instance) a domain was checked against a named list that involved a lookup,
4885 causing $domain_data to be set, then another domain was checked against the
4886 same list, then the first domain was re-checked, the value of $domain_data
4887 after the final check could be wrong. In particular, if the second check
4888 failed, it could be set empty. This bug probably also applied to
4891 41. The strip_trailing_dot option was not being applied to the address given
4892 with the -f command-line option.
4894 42. The code for reading a message's header from the spool was incrementing
4895 $received_count, but never initializing it. This meant that the value was
4896 incorrect (doubled) while delivering a message in the same process in which
4897 it was received. In the most common configuration of Exim, this never
4898 happens - a fresh exec is done - but it can happen when
4899 deliver_drop_privilege is set.
4901 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4902 it now includes up to 150 characters of the unexpected data in the log
4905 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4906 and building data strings. The size of both of these buffers was 10 000
4907 bytes - far larger than anybody would *ever* want, thought I. Needless to
4908 say, somebody hit the limit. I have increased the maximum line length to
4909 20 000 and the maximum data length of concatenated lines to 100 000. I have
4910 also fixed two bugs, because there was no checking on these buffers. Tsk,
4911 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4912 buffer is too small.
4914 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4915 lsearch lookups. Now it does.
4917 46. When parsing a route_list item in a manualroute router, a fixed-length
4918 buffer was used for the list of hosts. I made this 1024 bytes long,
4919 thinking that nobody would ever have a list of hosts that long. Wrong.
4920 Somebody had a whole pile of complicated expansion conditions, and the
4921 string was silently truncated, leading to an expansion error. It turns out
4922 that it is easier to change to an unlimited length (owing to other changes
4923 that have happened since this code was originally written) than to build
4924 structure for giving a limitation error. The length of the item that
4925 expands into the list of hosts is now unlimited.
4927 47. The lsearch lookup could not handle data where the length of text line was
4928 more than 4095 characters. Such lines were truncated, leading to shortened
4929 data being returned. It should now handle lines of any length.
4931 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4932 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4935 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4936 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4937 no longer happens in comment lines. A list of replacements is now placed
4938 at the head of all of the source files, except those whose only change is
4939 to replace PERL_COMMAND in the very first #! line.
4941 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4942 messages on the queue, with a bottom-up merge sort, using code contributed
4943 by Michael Haardt. This should make operations like -bp somewhat faster on
4944 large queues. It won't affect queue runners, except when queue_run_in_order
4947 51. Installed eximstats 1.31 in the distribution.
4949 52. Added support for SRV lookups to the dnslookup router.
4951 53. If an ACL referred to $message_body or $message_body_end, the value was not
4952 reset for any messages that followed in the same SMTP session.
4954 54. The store-handling optimization for building very long strings was not
4955 differentiating between the different store pools. I don't think this
4956 actually made any difference in practice, but I've tidied it.
4958 55. While running the routers to verify a sender address, $sender_address
4959 was still set to the sender address. This is wrong, because when routing to
4960 send a bounce to the sender, it would be empty. Therefore, I have changed
4961 it so that, while verifying a sender address, $sender_address is set to <>.
4962 (There is no change to what happens when verifying a recipient address.)
4964 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4965 target A or AAAA records (if not already returned) without resetting the
4966 qualify_single or search_parents options of the DNS resolver. These are
4967 inappropriate in this case because the targets of MX and SRV records must
4968 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4969 target that, when qualified, matched something in the local domain. These
4970 two options are now turned off when doing these lookups.
4972 57. It seems that at least some releases of Reiserfs (which does not have the
4973 concept of a fixed number of inodes) returns zero and not -1 for the
4974 number of available inodes. This interacted badly with check_spool_inodes,
4975 which assumed that -1 was the "no such thing" setting. What I have done is
4976 to check that the total number of inodes is greater than zero before doing
4977 the test of how many are available.
4979 58. When a "warn" ACL statement has a log_message modifier, the message is
4980 remembered, and not repeated. This is to avoid a lot of repetition when a
4981 message has many recipients that cause the same warning to be written.
4982 Howewer, Exim was preserving the list of already written lines for an
4983 entire SMTP session, which doesn't seem right. The memory is now reset if a
4984 new message is started.
4986 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4987 debugging output unless log_rewrite was also set.
4989 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4990 of (int)(handle) when we know that handle contains (void *)(-1).
4992 61. The Exim daemon panic-logs an error return when it closes the incoming
4993 connection. However "connection reset by peer" seems to be common, and
4994 isn't really an error worthy of noting specially, so that particular error
4997 62. When Exim is trying to find all the local interfaces, it used to panic and
4998 die if the ioctl to get the interface flags failed. However, it seems that
4999 on at least one OS (Solaris 9) it is possible to have an interface that is
5000 included in the list of interfaces, but for which you get a failure error
5001 for this call. This happens when the interface is not "plumbed" into a
5002 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5003 failure of the "get flags" call assumes that the interface is down.
5005 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5006 makes life easier for people who are doing arithmetic on fields extracted
5007 from dates, where you often get leading zeros that should not be
5008 interpreted as octal.
5010 64. Added qualify_domain to the redirect router, to override the global
5013 65. If a pathologically long header line contained very many addresses (the
5014 report of this problem mentioned 10 000) and each of them was rewritten,
5015 Exim could use up a very large amount of memory. (It kept on making new
5016 copies of the header line as it rewrote, and never released the old ones.)
5017 At the expense of a bit more processing, the header rewriting function has
5018 been changed so that it no longer eats memory in this way.
5020 66. The generation of the Received: header has been moved from the time that a
5021 message starts to be received, to the time that it finishes. The timestamp
5022 in the Received: header should now be very close to that of the <= log
5023 line. There are two side-effects of this change:
5025 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5026 logged header lines no longer include the local Received: line, because
5027 it has not yet been created. The same applies to a copy of the message
5028 that is returned to a non-SMTP sender when a message is rejected.
5030 (b) When a filter file is tested using -bf, no additional Received: header
5031 is added to the test message. After some thought, I decided that this
5034 This change does not affect the value of $received_for. It is still set
5035 after address rewriting, but before local_scan() is called.
5037 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5039 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5040 gave an unhelpful panic error message, and a defer error. I have managed to
5041 change this behaviour so that it now rejects any supplied certificate,
5042 which seems right, as the list of acceptable certificates is empty.
5044 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5045 gave an unhelpful defer error. I have not managed to make this reject any
5046 supplied certificates, but the error message it gives is "no certificate
5047 supplied", which is not helpful.
5049 70. exigrep's output now also includes lines that are not associated with any
5050 message, but which match the given pattern. Implemented by a patch from
5051 Martin Sluka, which also tidied up the Perl a bit.
5053 71. Recipient callout verification, like sender verification, was using <> in
5054 the MAIL FROM command. This isn't really the right thing, since the actual
5055 sender may affect whether the remote host accepts the recipient or not. I
5056 have changed it to use the actual sender in the callout; this means that
5057 the cache record is now keyed on a recipient/sender pair, not just the
5058 recipient address. There doesn't seem to be a real danger of callout loops,
5059 since a callout by the remote host to check the sender would use <>.
5060 [SEE ABOVE: changed after hitting problems.]
5062 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5063 temporary errors. However, in the case of such a code being given after
5064 the end of a data transmission (i.e. after ".") Exim was failing to write
5065 a retry record for the message. (Yes, there was some broken host that was
5066 actually sending 8xx at this point.)
5068 73. An unknown lookup type in a host list could cause Exim to panic-die when
5069 the list was checked. (An example that provoked this was putting <; in the
5070 middle of a list instead of at the start.) If this happened during a DATA
5071 ACL check, a -D file could be left lying around. This kind of configuration
5072 error no longer causes Exim to die; instead it causes a defer errror. The
5073 incident is still logged to the main and panic logs.
5075 74. Buglet left over from Exim 3 conversion. The message "too many messages
5076 in one connection" was written to the rejectlog but not the mainlog, except
5077 when address rewriting (yes!) was being logged.
5079 75. Added write_rejectlog option.
5081 76. When a system filter was run not as root (that is, when system_filter_user
5082 was set), the values of the $n variables were not being returned to the
5083 main process; thus, they were not subsequently available in the $sn
5086 77. Added +return_path_on_delivery log selector.
5088 78. A connection timeout was being treated differently from recipients deferred
5089 when testing hosts_max_try with a message that was older than the host's
5090 retry timeout. (The host should not be counted, thus allowing all hosts to
5091 be tried at least once before bouncing.) This may have been the cause of an
5092 occasionally reported bug whereby a message would remain on the queue
5093 longer than the retry timeout, but would be bounced if a delivery was
5094 forced. I say "may" because I never totally pinned down the problem;
5095 setting up timeout/retry tests is difficult. See also the next item.
5097 79. The ultimate address timeout was not being applied to errors that involved
5098 a combination of host plus message (for example, a timeout on a MAIL
5099 command). When an address resolved to a number of possible hosts, and they
5100 were not all tried for each delivery (e.g. because of hosts_max_try), a
5101 message could remain on the queue longer than the retry timeout.
5103 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5106 81. Fixed an obscure SMTP outgoing bug which required at least the following
5107 conditions: (a) there was another message waiting for the same server;
5108 (b) the server returned 5xx to all RCPT commands in the first message so
5109 that the message was not completed; (c) the server dropped the connection
5110 or gave a negative response to the RSET that Exim sends to abort the
5111 transaction. The observed case was a dropped connection after DATA that had
5112 been sent in pipelining mode. That is, the server had advertised PIPELINING
5113 but was not implementing it correctly. The effect of the bug was incorrect
5114 behaviour, such as trying another host, and this could lead to a crash.
5120 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5121 and daemon.c were passed as pointers to ints; they should have been
5122 pointers to socklen_t variables (which are typically unsigned ints).
5124 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5127 3. Fixed a really odd bug that affected only the testing scheme; patching a
5128 certain fixed string in the binary changed the value of another string that
5129 happened to be identical to the end of the original first string.
5131 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5132 name", it returns that address as the IP address. On some operating
5133 systems (e.g. Solaris), it also passes back the IP address string as the
5134 "host name". However, on others (e.g. Linux), it passes back an empty
5135 string. Exim wasn't checking for this, and was changing the host name to an
5136 empty string, assuming it had been canonicized.
5138 5. Although rare, it is permitted to have more than one PTR record for a given
5139 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5140 all the names associated with an address, because they do in Solaris.
5141 However, it seems that they do not in Linux for data that comes from the
5142 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5143 I found this out when I moved to a new Linux workstation and tried to run
5144 the Exim test suite.
5146 To get round this problem I have changed the code so that it now does its
5147 own call to the DNS to look up PTR records when searching for a host name.
5148 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5149 addresses that are only in /etc/hosts are still found.
5151 This behaviour is, however, controlled by an option called host_lookup_
5152 order, which defaults to "bydns:byaddr". If people want to use the other
5153 order, or indeed, just use one or the other means of lookup, they can
5154 specify it in this variable.
5156 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5157 some operating systems, this comes back from gethostbyaddr() as an empty
5158 string, and this is what Exim used to test for. However, it seems that in
5159 other systems, "." is yielded. Exim now tests for this case too.
5161 7. The values of check_spool_space and check_log_space are now held internally
5162 as a number of kilobytes instead of an absolute number of bytes. If a
5163 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5164 kilobyte. This means that much larger values can be stored.
5166 8. Exim monitor: an attempt to get the action menu when not actually pointing
5167 at a message produces an empty menu entitled "No message selected". This
5168 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5169 no entries in it ("Shell widget menu has zero width and/or height"). So I
5170 have added a single, blank menu entry in this case.
5172 9. Added ${quote_local_part.
5174 10. MIME decoding is now applied to the contents of Subject: header lines when
5177 11. Now that a reference to $sender_host_address automatically causes a reverse
5178 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5179 host lookup before query-style lookups in lists that might use this
5180 variable. This has therefore been abolished, and the "net-" prefix is no
5181 longer necessary for query-style lookups.
5183 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5184 have been a typo for LFLAGS, so it has been changed.
5186 13. The install script calls Exim with "-C /dev/null" in order to find the
5187 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5188 to be output. Howeve, since Exim outputs its version number before the
5189 error, it didn't break the script. It just looked ugly. I fixed this by
5190 always allowing "-C /dev/null" if the caller is root.
5192 14. Ignore overlarge ACL variable number when reading spool file - insurance
5193 against a later release with more variables having written the file.
5195 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5196 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5197 this, as well as the form without the "IPv6" on the front.
5199 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5200 OS/Makefile-Darwin file.
5202 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5203 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5205 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5206 variable. However, some OS use ints for inodes. I've added cast to long int
5207 to get rid of the compiler warning.
5209 19. I had forgotten to lock out "/../" in configuration file names when
5210 ALT_CONFIG_PREFIX was set.
5212 20. Routers used for verification do not need to specify transports. However,
5213 if such a router generated a host list, and callout was configured, Exim
5214 crashed, because it could not find a port number from the (non-existent)
5215 transport. It now assumes port 25 in this circumstance.
5217 21. Added the -t option to exigrep.
5219 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5220 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5221 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5222 them is set, LOOKUP_LSEARCH is forced.
5224 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5226 24. Added sender and host information to the "rejected by local_scan()" log
5227 line; previously there was no indication of these.
5229 25. Added .include_if_exists.
5231 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5232 Linux. It turns out that not all file systems support this. Apparently some
5233 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5234 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5235 not supported on the file descriptor, is now ignored when Exim is trying to
5236 sync a directory. This applies only to Linux.
5238 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5240 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5241 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5244 29. Added support for maildirsize files from supplied patch (modified a bit).
5246 30. The use of :fail: followed by an empty string could lead Exim to respond to
5247 sender verification failures with (e.g.):
5249 550 Verification failed for <xxx>
5250 550 Sender verify failed
5252 where the first response line was missing the '-' that indicates it is not
5253 the final line of the response.
5255 31. The loop for finding the name of the user that called Exim had a hardwired
5256 limit of 10; it now uses the value of finduser_retries, which is used for
5257 all other user lookups.
5259 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5262 33. Exim was neglecting to zero errno before one call of strtol() when
5263 expanding a string and expecting an integer value. On some systems this
5264 resulted in spurious "integer overflow" errors. Also, it was casting the
5265 result into an int without checking.
5267 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5268 did not work. The code looks as if it has *never* worked, though it appears
5269 to have been documented since at least releast 1.62. I have made it work.
5271 35. The "timeout_DNS" error in retry rules, also documented since at least
5272 1.62, also never worked. As it isn't clear exactly what this means, and
5273 clearly it isn't a major issue, I have abolished the feature by treating it
5274 as "timeout", and writing a warning to the main and panic logs.
5276 36. The display of retry rules for -brt wasn't always showing the error code
5279 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5280 timeout_connect_A, timeout_connect_MX.
5282 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5283 to the empty sender.
5285 39. The daemon was not analysing the content of -oX till after it had closed
5286 stderr and disconnected from the controlling terminal. This meant that any
5287 syntax errors were only noted on the panic log, and the return code from
5288 the command was 0. By re-arranging the code a little, I've made the
5289 decoding happen first, so such errors now appear on stderr, and the return
5290 code is 1. However, the actual setting up of the sockets still happens in
5291 the disconnected process, so errors there are still only recorded on the
5294 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5295 connections (as happens on some IP stacks) was logged at start up time as
5296 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5297 it from "IPv6 and IPv4", which means that two separate sockets are being
5300 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5301 says whether AF_INET or AF_INET6 was passed as an argument.
5303 42. Exiwhat output was messed up when time zones were included in log
5306 43. Exiwhat now gives more information about the daemon's listening ports,
5307 and whether -tls-on-connect was used.
5309 44. The "port" option of the smtp transport is now expanded.
5311 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5312 silently ignored. Now an error message is written to the main and panic
5315 46. There's a new ACL modifier called "logwrite" which writes to a log file
5316 as soon as it is encountered.
5318 47. Added $local_user_uid and $local_user_gid at routing time.
5320 48. Exim crashed when trying to verify a sender address that was being
5323 49. Exim was recognizing only a space character after ".include". It now also
5324 recognizes a tab character.
5326 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5327 extracting the relevant information from the specification. The man page no
5328 longer contains scrambled data for the -d option, and I've added a section
5329 at the front about calling Exim under different names.
5331 51. Added "extra_headers" argument to the "mail" command in filter files.
5333 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5336 53. Installed eximstats 1.29.
5338 54. Added transport_filter_timeout as a generic transport option.
5340 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5341 Cc: header lines. This was required by RFC 822, but it not required by RFC
5344 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5345 incoming messages that did not have them. Now it does so only if the
5346 message originates locally, that is, if there is no associated remote host
5347 address. When Resent- header lines are present, this applies to the Resent-
5348 lines rather than the non-Resent- lines.
5350 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5351 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5353 58. Messages for configuration errors now include the name of the main
5354 configuration file - useful now that there may be more than one file in a
5355 list (.included file names were always shown).
5357 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5358 for those rare installations that do not start the daemon as root or run it
5359 setuid root. I've cut out the call to initgroups() if the daemon is not
5362 60. The Exim user and group can now be bound into the binary as text strings
5363 that are looked up at the start of Exim's processing.
5365 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5367 62. Added $mailstore_basename variable.
5369 63. Installed patch to sieve.c from Michael Haardt.
5371 64. When Exim failed to open the panic log after failing to open the main log,
5372 the original message it was trying to log was written to stderr and debug
5373 output, but if they were not available (the usual case in production), it
5374 was lost. Now it is written to syslog before the two lines that record the
5375 failures to open the logs.
5377 65. Users' Exim filters run in subprocesses under the user's uid. It is
5378 possible for a "deliver" command or an alias in a "personal" command to
5379 provoke an address rewrite. If logging of address rewriting is configured,
5380 this fails because the process is not running as root or exim. There may be
5381 a better way of dealing with this, but for the moment (because 4.30 needs
5382 to be released), I have disabled address rewrite logging when running a
5383 filter in a non-root, non-exim process.
5389 1. The buildconfig auxiliary program wasn't quoting the value set for
5390 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5391 not defined. This bug was masked in 4.22 by the effect that was fixed in
5394 2. Some messages that were rejected after a message id was allocated were
5395 shown as "incomplete" by exigrep. It no longer does this for messages that
5396 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5398 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5399 have allow_domain_literals set, the ID did not get logged in the <= line.
5400 Domain literals are now always recognized in Message-ID: header lines.
5402 4. The first argument for a ${extract expansion item is the key name or field
5403 number. Leading and trailing spaces in this item were not being ignored,
5404 causing some misleading effects.
5406 5. When deliver_drop_privilege was set, single queue runner processes started
5407 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5408 same command in the process it spins off) were not dropping privilege.
5410 6. When the daemon running as "exim" started a queue runner, it always
5411 re-executed Exim in the spun-off process. This is a waste of effort when
5412 deliver_drop_privilege is set. The new process now just calls the
5413 queue-runner function directly.
5419 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5422 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5423 ignored. Though the use of -oP was forcing the writing of a pid file, it
5424 was always written to the default place.
5426 3. If the message "no IP address found for host xxxx" is generated during
5427 incoming verification, it is now followed by identification of the incoming
5428 connection (so you can more easily find what provoked it).
5430 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5432 5. Added some features to "harden" Exim a bit more against certain attacks:
5434 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5435 be put in Local/Makefile. This is like the never_users runtime option,
5436 but it cannot be overridden. The default setting is "root".
5438 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5439 prefix string with which any file named in a -C command line option
5442 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5443 is retained for -C and -D only if the caller of Exim is root. Without
5444 it, the exim user may also use -C and -D and retain privilege.
5446 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5447 command line option is disabled.
5449 6. Macro names set by the -D option must start with an upper case letter, just
5450 like macro names defined in the configuration file.
5452 7. Added "dereference=" facility to LDAP.
5454 8. Two instances of the typo "uknown" in the source files are fixed.
5456 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5457 the Configure-Makefile script screwed up while processing it.
5459 10. Incorporated PCRE 4.4.
5461 11. The SMTP synchronization check was not operating right at the start of an
5462 SMTP session. For example, it could not catch a HELO sent before the client
5463 waited for the greeting. There is now a check for outstanding input at the
5464 point when the greeting is written. Because of the duplex, asynchronous
5465 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5466 way, but not yet received, when the check is performed.
5468 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5469 on TCP/IP sockets, because this apparently causes some broken clients to
5472 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5473 unchanged) from the Cygwin maintainer.
5475 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5476 was supported instead of testing for mbx. Effectively a typo.
5478 15. The spa authenticator server code was not checking that the input it
5479 received was valid base64.
5481 16. The debug output line for the "set" modifier in ACLs was not showing the
5482 name of the variable that was being set.
5484 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5486 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5487 lookup on its own. Something else was needed to trigger the lookup. For
5488 example, a match in host_lookup or the need for a host name in a host list.
5489 Now, if $sender_host_name is referenced and the host name has not yet been
5490 looked up, a lookup is performed. If the lookup fails, the variable remains
5491 empty, and $host_lookup_failed is set to "1".
5493 19. Added "eqi" as a case-independent comparison operator.
5495 20. The saslauthd authentication condition could segfault if neither service
5496 nor realm was specified.
5498 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5499 error message that was logged was misleading, and incoming SMTP
5500 connections were dropped. The message is now more accurate, and temporary
5501 errors are given to SMTP connections.
5503 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5504 (except RSET) with a 421 error, until QUIT is received. However, it was
5505 failing to send a response to QUIT.
5507 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5508 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5509 ACL is now run after the helo_try_verify_hosts code.
5511 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5512 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5513 case-independent, so other case variants are also recognized). Apparently
5514 some systems use these upper case variants.
5516 25. If more than two messages were waiting for the same host, and a transport
5517 filter was specified for the transport, Exim sent two messages over the
5518 same TCP/IP connection, and then failed with "socket operation on non-
5519 socket" when it tried to send the third.
5521 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5523 27. The extern definition of crypt16() in expand.c was not being excluded when
5524 the OS had its own crypt16() function.
5526 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5527 as a preferred synonym for return_size_limit, both as an option and as an
5530 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5532 30. Changed the default configuration ACL to relax the local part checking rule
5533 for addresses that are not in any local domains. For these addresses,
5534 slashes and pipe symbols are allowed within local parts, but the sequence
5535 /../ is explicitly forbidden.
5537 31. SPA server authentication was not clearing the challenge buffer before
5540 32. log_message in a "warn" ACL statement was writing to the reject log as
5541 well as to the main log, which contradicts the documentation and doesn't
5542 seem right (because no rejection is happening). So I have stopped it.
5544 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5545 However, I am unable to do any testing of this.
5547 34. Fixed an infelicity in the appendfile transport. When checking directories
5548 for a mailbox, to see if any needed to be created, it was accidentally
5549 using path names with one or more superfluous leading slashes; tracing
5550 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5552 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5553 contents are added to the log line that is written for every discarded
5554 recipient. (Previously a log_message setting was ignored.)
5556 36. The ${quote: operator now quotes the string if it is empty.
5558 37. The install script runs exim in order to find its version number. If for
5559 some reason other than non-existence or emptiness, which it checks, it
5560 could not run './exim', it was installing it with an empty version number,
5561 i.e. as "exim-". This error state is now caught, and the installation is
5564 38. An argument was missing from the function that creates an error message
5565 when Exim fails to connect to the socket for saslauthd authentication.
5566 This could cause Exim to crash, or give a corrupted message.
5568 39. Added isip, isip4, and isip6 to ${if conditions.
5570 40. The ACL variables $acl_xx are now saved with the message, and can be
5571 accessed later in routers, transports, and filters.
5573 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5574 strings in the file are not string-expanded.
5576 42. If a MAIL command specified a SIZE value that was too large to fit into an
5577 int variable, the check against message_size_limit failed. Such values are
5578 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5579 day this will have to be increased, but I don't think I want to be around
5580 when emails are that large.
5587 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5588 iconv() is not standard in FreeBSD.
5590 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5591 IPv6 enabled. The observed symptom was a segmentation fault on return from
5592 the function os_common_find_running_interfaces() in src/os.c.
5594 3. In the check_special_case() function in daemon.c I had used "errno" as an
5595 argument name, which causes warnings on some systems. This was basically a
5596 typo, since it was named "eno" in the comments!
5598 4. The code that waits for the clock to tick (at a resolution of some fraction
5599 of a second) so as to ensure message-id uniqueness was always waiting for
5600 at least one whole tick, when it could have waited for less. [This is
5601 almost certainly not relevant at current processor speeds, where it is
5602 unlikely to ever wait at all. But we try to future-proof.]
5604 5. The function that sleeps for a time interval that includes fractions of a
5605 second contained a race. It did not block SIGALRM between setting the
5606 timer, and suspending (a couple of lines later). If the interval was short
5607 and the sigsuspend() was delayed until after it had expired, the suspension
5608 never ended. On busy systems this could lead to processes getting stuck for
5611 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5612 process, before it forks any delivery processes. The open lookup caching
5613 mechanism meant that the open file or database connection was passed into
5614 the delivery process. The problem was that delivery processes always tidy
5615 up cached lookup data. This could cause a problem for the next delivery
5616 process started by the queue runner, because the external queue runner
5617 process does not know about the closure. So the next delivery process
5618 still has data in the lookup cache. In the case of a file lookup, there was
5619 no problem because closing a file descriptor in a subprocess doesn't affect
5620 the parent. However, if the lookup was caching a connection to a database,
5621 the connection was closed, and the second delivery process was likely to
5622 see errors such as "PGSQL: query failed: server closed the connection
5623 unexpectedly". The problem has been fixed by closing all cached lookups
5624 in a queue runner before running a delivery process.
5626 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5627 seem to have the "const" qualifier which it has on other OS. I've
5630 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5631 *of the same type* (client or server) with the same public name that an
5632 error should be diagnosed.
5634 9. When Exim looked up a host name for an IP address, but failed to find the
5635 original IP address when looking up the host name (a safety check), it
5636 output the message "<ip address> does not match any IP for NULL", which was
5637 confusing, to say the least. The bug was that the host name should have
5638 appeared instead of "NULL".
5640 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5641 user that is built into the binary, and the -C or -D options is used, root
5642 privilege is dropped before the configuration file is read. In addition,
5643 logging is switched to stderr instead of the normal log files. If the
5644 configuration then re-defines the Exim user, the unprivileged environment
5645 is probably not what is expected, so Exim logs a panic warning message (but
5648 However, if deliver_drop_privilege is set, the unprivileged state may well
5649 be exactly what is intended, so the warning has been cut out in that case,
5650 and Exim is allowed to try to write to its normal log files.
5656 1. smtp_return_error_details was not giving details for temporary sender
5657 or receiver verification errors.
5659 2. Diagnose a configuration error if two authenticators have the same public
5662 3. Exim used not to create the message log file for a message until the first
5663 delivery attempt. This could be confusing when incoming messages were held
5664 for policy or load reasons. The message log file is now created at the time
5665 the message is received, and an initial "Received" line is written to it.
5667 4. The automatically generated man page for command line options had a minor
5668 bug that caused no ill effects; however, a more serious problem was that
5669 the procedure for building the man page automatically didn't always
5670 operate. Consequently, release 4.20 contains an out-of-date version. This
5671 shouldn't happen again.
5673 5. When building Exim with embedded Perl support, the script that builds the
5674 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5675 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5677 6. The freeze_tell option was not being used for messages that were frozen on
5678 arrival, either by an ACL or by local_scan().
5680 7. Added the smtp_incomplete_transaction log selector.
5682 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5683 was accepting AUTH without a new EHLO.
5685 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5686 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5689 10. Logging of TCP/IP connections (when configured) now happens in the main
5690 daemon process instead of the child process, so that the TCP/IP connection
5691 count is more accurate (but it can never be perfect).
5693 11. The use of "drop" in a nested ACL was not being handled correctly in the
5694 outer ACL. Now, if condition failure induced by the nested "drop" causes
5695 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5696 or "require"), the connection is dropped.
5698 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5699 that yield "discard" can now be used with an "accept" or a "discard" verb,
5700 but an error is generated for any others (because I can't see a useful way
5701 to define what should happen).
5703 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5704 are now processed in the same way as lines in the Exim configuration file.
5705 In particular, continuation lines are supported.
5707 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5709 15. Added -ti meaning -t -i.
5711 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5712 domains, and warn by logging if others are found.
5714 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5715 returned from GIFCONF when Exim is trying to find the list of interfaces on
5716 a host. The code in os.c has been modified to copy each ifreq to an aligned
5717 structure in all cases.
5719 Also, in some cases, the returned ifreq's were being copied to a 'struct
5720 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5721 means the last couple of bytes of an IPv6 address could be chopped if the
5722 ifreq contained only a normal sockaddr (14 bytes storage).
5724 18. Named domain lists were not supported in the hosts_treat_as_local option.
5725 An entry such as +xxxx was not recognized, and was treated as a literal
5728 19. Ensure that header lines added by a DATA ACL are included in the reject log
5729 if the ACL subsequently rejects the message.
5731 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5732 MD5 (which is deprecated).
5734 21. When testing a filter file using -bf, Exim was writing a message when it
5735 took the sender from a "From " line in the message, but it was not doing so
5736 when it took $return_path from a Return-Path: header line. It now does.
5738 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5739 with a valid header line field name (a series of printing characters
5740 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5742 23. Changed "disc" in the source to "disk" to conform to the documentation and
5743 the book and for uniformity.
5745 24. Ignore Sendmail's -Ooption=value command line item.
5747 25. When execve() failed while trying to run a command in a pipe transport,
5748 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5749 could be confused with a return value of 69 from the command itself. This
5750 has been changed to 127, the value the shell returns if it is asked to run
5751 a non-existent command. The wording for the related log line suggests a
5752 non-existent command as the problem.
5754 26. If received_header_text expands to an empty string, do not add a Received:
5755 header line to the message. (Well, it adds a token one on the spool, but
5756 marks it "old" so that it doesn't get used or transmitted.)
5758 27. Installed eximstats 1.28 (addition of -nt option).
5760 28. There was no check for failure on the call to getsockname() in the daemon
5761 code. This can fail if there is a shortage of resources on the system, with
5762 ENOMEM, for example. A temporary error is now given on failure.
5764 29. Contrary to the C standard, it seems that in some environments, the
5765 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5766 program. Exim now does this explicitly; it affects the formatting of
5767 timestamps using strftime().
5769 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5770 complaints. I've now initialized all the variables, to avoid this.
5772 32. Header lines added by a system filter were not being "seen" during
5773 transport-time rewrites.
5775 33. The info_callback() function passed to OpenSSL is set up with type void
5776 (*)(SSL *, int, int), as described somewhere. However, when calling the
5777 function (actually a macro) that sets it up, the type void(*)() is
5778 expected. I've put in a cast to prevent warnings from picky compilers.
5780 34. If a DNS black list lookup found a CNAME record, but there were no A
5781 records associated with the domain it pointed at, Exim crashed.
5783 35. If a DNS black list lookup returned more than one A record, Exim ignored
5784 all but the first. It now scans all returned addresses if a particular IP
5785 value is being sought. In this situation, the contents of the
5786 $dnslist_value variable are a list of all the addresses, separated by a
5789 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5790 to accept a host name and all its aliases if the forward lookup for any of
5791 them yielded the IP address of the incoming connection. Now it accepts only
5792 those names whose forward lookup yields the correct IP address. Any other
5793 names are discarded. This closes a loophole whereby a rogue DNS
5794 administrator could create reverse DNS records to break through a
5795 wildcarded host restriction in an ACL.
5797 37. If a user filter or a system filter that ran in a subprocess used any of
5798 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5799 the wrong values were passed to the pipe command ($thisaddress had the
5800 value of $0, $0 had the value of $1, etc). This bug was introduced by
5801 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5803 38. Improved the line breaking for long SMTP error messages from ACLs.
5804 Previously, if there was no break point between 40 and 75 characters, Exim
5805 left the rest of the message alone. Two changes have been made: (a) I've
5806 reduced the minimum length to 35 characters; (b) if it can't find a break
5807 point between 35 and 75 characters, it looks ahead and uses the first one
5808 that it finds. This may give the occasional overlong line, but at least the
5809 remaining text gets split now.
5811 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5812 file descriptors might be low, and that setting 1000 would always raise it.
5813 It turns out that in some environments, the limit is already over 1000 and
5814 that lowering it causes trouble. So now Exim takes care not to decrease it.
5816 40. When delivering a message, the value of $return_path is set to $sender_
5817 address at the start of routing (routers may change the value). By an
5818 oversight, this default was not being set up when an address was tested by
5819 -bt or -bv, which affected the outcome if any router or filter referred to
5822 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5823 log only when "message" or "log_message" are set. However, if one of the
5824 conditions was an address verification, or a call to a nested ACL, the
5825 messages generated by the underlying test were being passed through. This
5826 no longer happens. The underlying message is available in $acl_verify_
5827 message for both "message" and "log_message" expansions, so it can be
5828 passed through if needed.
5830 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5831 new expansion $bh_ to give the encoded byte string without charset
5832 translation. Translation happens only if iconv() is available; HAVE_ICONV
5833 indicates this at build time. HEADERS_CHARSET gives the charset to
5834 translate to; headers_charset can change it in the configuration, and
5835 "headers charset" can change it in an individual filter file.
5837 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5838 that creates RFC 2047 encoded "words" labels them as that charset instead
5839 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5840 expansion operator; (ii) when Exim creates a From: line for a local
5841 message; (iii) when a header line is rewritten to include a "phrase" part.
5843 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5844 buggy, causing it to skip the first lines of messages whose message ID
5845 ended in 'D'. This would not have bitten before Exim release 4.14, because
5846 message IDs were unlikely to end in 'D' before then. The effect was to have
5847 incorrect size information for certain domains.
5849 45. #include "config.h" was missing at the start of the crypt16.c module. This
5850 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5853 46. If there was a timeout during a "random" callout check, Exim treated it as
5854 a failure of the random address, and carried on sending RSET and the real
5855 address. If the delay was just some slowness somewhere, the response to the
5856 original RCPT would be taken as a response to RSET and so on, causing
5857 mayhem of various kinds.
5859 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5860 when I implemented it. It didn't allow for the fact that some option values
5861 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5862 the right test for positive values.
5864 48. Domain names in DNS records are case-independent. Exim always looks them up
5865 in lower case. Some resolvers return domain names in exactly the case they
5866 appear in the zone file, that is, they may contain uppercase letters. Not
5867 all resolvers do this - some return always lower case. Exim was treating a
5868 change of case by a resolver as a change of domain, similar to a widening
5869 of a domain abbreviation. This triggered its re-routing code and so it was
5870 trying to route what was effectively the same domain again. This normally
5871 caused routing to fail (because the router wouldn't handle the domain
5872 twice). Now Exim checks for this case specially, and just changes the
5873 casing of the domain that it ultimately uses when it transmits the message
5876 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5879 50. If a filter generated a file delivery with a non-absolute name (possible if
5880 no home directory exists for the router), the forbid_file option was not
5883 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5884 the existing equality matching.
5886 52. Exim was using ints instead of ino_t variables in some places where it was
5887 dealing with inode numbers.
5889 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5890 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5891 called TMPDIR, and if it finds it is different, it changes its value.
5893 54. The smtp_printf() function is now made available to local_scan() so
5894 additional output lines can be written before returning. There is also an
5895 smtp_fflush() function to enable the detection of a dropped connection.
5896 The variables smtp_input and smtp_batched_input are exported to
5899 55. Changed the default runtime configuration: the message "Unknown user"
5900 has been removed from the ACL, and instead placed on the localuser router,
5901 using the cannot_route_message feature. This means that any verification
5902 failures that generate their own messages won't get overridden. Similarly,
5903 the "Unrouteable address" message that was in the ACL for unverifiable
5904 relay addresses has also been removed.
5906 56. Added hosts_avoid_esmtp to the smtp transport.
5908 57. The exicyclog script was not checking for the esoteric option
5909 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5910 will work only if exicyclog is run under the appropriate euid.
5912 58. Following a discussion on the list, the rules by which Exim recognises line
5913 endings on incoming messages have been changed. The -dropcr and drop_cr
5914 options are now no-ops, retained only for backwards compatibility. The
5915 following line terminators are recognized: LF CRLF CR. However, special
5916 processing applies to CR:
5918 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5919 nor a local message in the state where . is a terminator.
5921 (ii) If a bare CR is encountered in a header line, an extra space is added
5922 after the line terminator so as not to end the header. The reasoning
5923 behind this is that bare CRs in header lines are most likely either
5924 to be mistakes, or people trying to play silly games.
5926 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5927 was being incorrectly given as 18 bytes larger than it should have been.
5928 This is a VOB (very old bug).
5930 60. This may never have affected anything current, but just in case it has:
5931 When the local host is found other than at the start of a list of hosts,
5932 the local host, those with the same MX, and any that follow, are discarded.
5933 When the list in question was part of a longer list of hosts, the following
5934 hosts (not currently being processed) were also being discarded. This no
5935 longer happens. I'm not sure if this situation could ever has previously
5938 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5941 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5942 header. This is something that is recommended in a new Internet Draft, and
5943 is something that is documented as being done by Sendmail. There are two
5944 possible values. For messages generated by the autoreply transport, Exim
5947 Auto-Submitted: auto-replied
5949 whereas for all other generated messages (e.g. bounces) it adds
5951 Auto-Submitted: auto-generated
5953 63. The "personal" condition in filters now includes a test for the
5954 Auto-Submitted: header. If it contains the string "auto-" the message it
5955 not considered personal.
5957 64. Added rcpt_include_affixes as a generic transport option.
5959 65. Added queue_only_override (default true).
5961 66. Added the syslog_duplication option.
5963 67. If what should have been the first header line of a message consisted of
5964 a space followed by a colon, Exim was mis-interpreting it as a header line.
5965 It isn't of course - it is syntactically invalid and should therefore be
5966 treated as the start of the message body. The misbehaviour could have
5967 caused a number of strange effects, including loss of data in subsequent
5968 header lines, and spool format errors.
5970 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5971 client host had authenticated. This control can now be exercised by an ACL
5972 for more flexibility.
5974 69. By default, callouts do not happen when testing with -bh. There is now a
5975 variant, -bhc, which does actually run the callout code, including
5976 consulting and updating the callout cache.
5978 70. Added support for saslauthd authentication, courtesy of Alexander
5981 71. If statvfs() failed on the spool or log directories while checking their
5982 size for availability, Exim confusingly gave the error "space shortage".
5983 Furthermore, in debugging mode it crashed with a floating point exception.
5984 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5985 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5986 serious problem, Exim now writes to the main and panic logs when this
5987 happens, with details of the failure. It then refuses to accept the
5988 incoming message, giving the message "spool directory problem" or "log
5989 directory problem" with a 421 code for SMTP messages.
5991 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5992 0, 1, and 2 exist, because some OS complain for execs without them (see
5993 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5994 descriptors. However, the code omitted to check that the open succeeded,
5995 causing mysterious errors if for some reason the permissions on /dev/null
5996 got screwed. Now Exim writes a message to the main and panic logs, and
5997 bombs out if it can't open /dev/null.
5999 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6000 interact so that it is all more flexible. It is supposed to remain
6001 backwards compatible. Also added extra_local_interfaces.
6003 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6004 to bomb out with an assertion failure - to the client this appears as a
6005 connection drop. This problem occurs in the part of the code that was taken
6006 from the Samba project. Fortunately, the assertion is in a very simple
6007 function, so I have fixed this by reproducing the function inline in the
6008 one place where it is called, and arranging for authentication to fail
6009 instead of killing the process with assert().
6011 75. The SPA client code was not working when the server requested OEM rather
6012 than Unicode encoding.
6014 76. Added code to make require_files with a specific uid setting more usable in
6015 the case where statting the file as root fails - usually a non-root-mounted
6016 NFS file system. When this happens and the failure is EACCES, Exim now
6017 forks a subprocess and does the per-uid checking as the relevant uid.
6019 77. Added process_log_path.
6021 78. If log_file_path was not explicitly set, a setting of check_log_space or
6022 check_log_inodes was ignored.
6024 79. If a space check for the spool or log partitions fails, the incident is now
6025 logged. Of course, in the latter case the data may get lost...
6027 80. Added the %p formatting code to string_format() so that it can be used to
6028 print addresses in debug_print(). Adjusted all the address printing in the
6029 debugging in store.c to use %p rather than %d.
6031 81. There was a concern that a line of code in smtp_in.c could overflow a
6032 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6033 initially expressed, the concern was not well-founded, because trailing
6034 spaces are removed early. However, if the trailing spaces were followed by
6035 a NULL, they did not get removed, so the overflow was possible. Two fixes
6038 (a) I re-wrote the offending code in a cleaner fashion.
6039 (b) If an incoming SMTP command contains a NULL character, it is rejected
6042 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6043 runs initgroups(), so that if the Exim user is in any additional groups,
6044 they will be used during message reception.
6050 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob