1 Change log file for Exim from version 4.21
2 -------------------------------------------
8 TF/01 Correctly close the server side of TLS when forking for delivery.
10 When a message was received over SMTP with TLS, Exim failed to clear up
11 the incoming connection properly after forking off the child process to
12 deliver the message. In some situations the subsequent outgoing
13 delivery connection happened to have the same fd number as the incoming
14 connection previously had. Exim would try to use TLS and fail, logging
15 a "Bad file descriptor" error.
17 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
18 utilities have not been installed.
20 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
21 temporary space as the ACL may create new global variables.
23 TL/01 LDAP support uses per connection or global context settings, depending
24 upon the detected version of the libraries at build time.
26 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
27 to extract and use the src ip:port in logging and expansions as if it
28 were a direct connection from the outside internet.
30 JH/02 Add ${listextract {number}{list}{success}{fail}}.
32 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
33 Properly escape header and check for NULL return.
35 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
38 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
40 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
41 characters in header names, implemented as a verify condition.
42 Contributed by Michael Fischer v. Mollard.
44 TL/05 Rename SPF condition results err_perm and err_temp to standardized
45 results permerror and temperror. Previous values are deprecated but
46 still accepted. In a future release, err_perm and err_temp will be
47 completely removed, which will be a backward incompatibility if the
48 ACL tests for either of these two old results. Patch contributed by
49 user bes-internal on the mailing list.
51 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
53 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
54 selectors, in both main and reject logs.
56 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
59 JH/07 Add malware type "sock" for talking to simple daemon.
61 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
63 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
64 routers/transports under cutthrough routing.
66 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
67 numbers. Touch up "bool" conditional to keep the same definition.
69 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
71 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
73 JH/12 Expand items in router/transport headers_add or headers_remove lists
74 individually rather than the list as a whole. Bug 1452.
76 Required for reasonable handling of multiple headers_ options when
77 they may be empty; requires that headers_remove items with embedded
78 colons must have them doubled (or the list-separator changed).
80 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
81 view the policy declared in the DMARC record. Currently, $dmarc_status
82 is a combined value of both the record presence and the result of the
85 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
87 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
88 dnslookup router and the smtp transport (applying to the forward
91 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
92 of ldap servers used for a specific lookup. Patch provided by Heiko
95 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
96 New variable $lookup_dnssec_authenticated for observability.
98 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
99 Patch submitted by Lars Timman.
101 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
103 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
104 Requires trusted mode and valid format message id, aborts otherwise.
105 Patch contributed by Heiko Schlichting.
107 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
108 certextract with support for various fields. Bug 1358.
110 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
111 is requested by default, modifiable by smtp transport option
114 JH/22 Expansion operators ${md5:string} and ${sha1::string} can now
115 operate on certificate variables to give certificate fingerprints
121 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
123 PP/02 Make -n do something, by making it not do something.
124 When combined with -bP, the name of an option is not output.
126 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
129 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
130 $sender_host_name and config options to manage this, and basic check
133 PP/05 DSCP support for outbound connections and control modifier for inbound.
135 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
136 (Only plugin which currently uses this is kerberos4, which nobody should
137 be using, but we should make it available and other future plugins might
138 conceivably use it, even though it would break NAT; stuff *should* be
139 using channel bindings instead).
141 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
142 name; added for Sendmail compatibility; requires admin caller.
143 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
144 just ignore it); requires trusted caller.
145 Also parse but ignore: -Ac -Am -X<logfile>
148 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
150 TL/02 Add +smtp_confirmation as a default logging option.
152 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
153 Patch by Magnus Holmgren from 2007-02-20.
155 TL/04 Bugzilla 1281 - Spec typo.
156 Bugzilla 1283 - Spec typo.
157 Bugzilla 1290 - Spec grammar fixes.
159 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
161 TL/06 Add Experimental DMARC support using libopendmarc libraries.
163 TL/07 Fix an out of order global option causing a segfault. Reported to dev
164 mailing list by by Dmitry Isaikin.
166 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
168 JH/02 Support "G" suffix to numbers in ${if comparisons.
170 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
172 NM/01 Bugzilla 1197 - Spec typo
173 Bugzilla 1196 - Spec examples corrections
175 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
177 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
178 gnutls_enable_pkcs11, but renamed to more accurately indicate its
181 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
182 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
184 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
185 "acl {{name}{arg}...}", and optional args on acl condition
188 JH/05 Permit multiple router/transport headers_add/remove lines.
190 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
192 JH/07 Avoid using a waiting database for a single-message-only transport.
193 Performance patch from Paul Fisher. Bugzilla 1262.
195 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
198 JH/09 Add $headers_added variable, with content from use of ACL modifier
199 add_header (but not yet added to the message). Bugzilla 199.
201 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
202 Pulled from Bugzilla 817 by Wolfgang Breyha.
204 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
206 (nb: this is the same fix as in Exim 4.80.1)
208 JH/11 Add A= logging on delivery lines, and a client_set_id option on
211 JH/12 Add optional authenticated_sender logging to A= and a log_selector
214 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
216 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
217 advertise SMTP AUTH mechanism to us, instead of a generic
218 protocol violation error. Also, make Exim more robust to bad
219 data from the Dovecot auth socket.
221 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
223 When a queue runner is handling a message, Exim first routes the
224 recipient addresses, during which it prunes them based on the retry
225 hints database. After that it attempts to deliver the message to
226 any remaining recipients. It then updates the hints database using
229 So if a recipient address works intermittently, it can get repeatedly
230 deferred at routing time. The retry hints record remains fresh so the
231 address never reaches the final cutoff time.
233 This is a fairly common occurrence when a user is bumping up against
234 their storage quota. Exim had some logic in its local delivery code
235 to deal with this. However it did not apply to per-recipient defers
236 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
238 This change adds a proper retry rule check during routing so that the
239 final cutoff time is checked against the message's age. We only do
240 this check if there is an address retry record and there is not a
241 domain retry record; this implies that previous attempts to handle
242 the address had the retry_use_local_parts option turned on. We use
243 this as an approximation for the destination being like a local
244 delivery, as in LMTP.
246 I suspect this new check makes the old local delivery cutoff check
247 redundant, but I have not verified this so I left the code in place.
249 TF/02 Correct gecos expansion when From: is a prefix of the username.
251 Test 0254 submits a message to Exim with the header
255 When I ran the test suite under the user fanf2, Exim expanded
256 the header to contain my full name, whereas it should have added
257 a Resent-Sender: header. It erroneously treats any prefix of the
258 username as equal to the username.
260 This change corrects that bug.
262 GF/01 DCC debug and logging tidyup
263 Error conditions log to paniclog rather than rejectlog.
264 Debug lines prefixed by "DCC: " to remove any ambiguity.
266 TF/03 Avoid unnecessary rebuilds of lookup-related code.
268 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
269 Bug spotted by Jeremy Harris; was flawed since initial commit.
270 Would have resulted in OCSP responses post-SNI triggering an Exim
271 NULL dereference and crash.
273 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
275 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
276 Bug detection, analysis and fix by Samuel Thibault.
277 Bugzilla 1331, Debian bug #698092.
279 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
281 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
282 Server implementation by Todd Lyons, client by JH.
283 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
284 config variable "prdr_enable" controls whether the server
285 advertises the facility. If the client requests PRDR a new
286 acl_data_smtp_prdr ACL is called once for each recipient, after
287 the body content is received and before the acl_smtp_data ACL.
288 The client is controlled by bolth of: a hosts_try_prdr option
289 on the smtp transport, and the server advertisement.
290 Default client logging of deliveries and rejections involving
291 PRDR are flagged with the string "PRDR".
293 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
294 fclose(). Diagnosis by Todd Lyons.
296 PP/17 Update configure.default to handle IPv6 localhost better.
297 Patch by Alain Williams (plus minor tweaks).
300 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
301 This is now consistent with GnuTLS, and is now documented: the
302 previous undocumented portable approach to treating the option as
303 unset was to force an expansion failure. That still works, and
304 an empty string is now equivalent.
306 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
307 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
308 not performing validation itself.
310 PP/20 Added force_command boolean option to pipe transport.
311 Patch from Nick Koston, of cPanel Inc.
313 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
316 TF/04 Added udpsend ACL modifer and hexquote expansion operator
318 PP/21 Fix eximon continuous updating with timestamped log-files.
319 Broken in a format-string cleanup in 4.80, missed when I repaired the
320 other false fix of the same issue.
321 Report and fix from Heiko Schlichting.
324 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
325 Report from Prashanth Katuri.
327 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
328 It's SecureTransport, so affects any MacOS clients which use the
329 system-integrated TLS libraries, including email clients.
331 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
332 using a MIME ACL for non-SMTP local injection.
333 Report and assistance in diagnosis by Warren Baker.
335 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
337 JH/16 Fix comparisons for 64b. Bugzilla 1385.
339 TL/09 Add expansion variable $authenticated_fail_id to keep track of
340 last id that failed so it may be referenced in subsequent ACL's.
342 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
345 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
346 ldap library initialization, allowing self-signed CA's to be
347 used. Also properly sets require_cert option later in code by
348 using NULL (global ldap config) instead of ldap handle (per
349 session). Bug diagnosis and testing by alxgomz.
351 TL/12 Enhanced documentation in the ratelimit.pl script provided in
352 the src/util/ subdirectory.
354 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
355 renamed to Transport Post Delivery Action by Jeremy Harris, as
358 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
359 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
360 redis_servers = needs to be configured which will be used by the redis
361 lookup. Patch from Warren Baker, of The Packet Hub.
363 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
365 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
366 hostname or reverse DNS when processing a host list. Used suggestions
367 from multiple comments on this bug.
369 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
371 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
372 Missed a few lines, added it to make the runtest require no keyboard
375 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
376 contains upper case chars. Make router use caseful_local_part.
378 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
379 support when GnuTLS has been built with p11-kit.
385 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
387 This, or similar/improved, will also be change PP/11 of 4.82.
393 PP/01 Handle short writes when writing local log-files.
394 In practice, only affects FreeBSD (8 onwards).
395 Bugzilla 1053, with thanks to Dmitry Isaikin.
397 NM/01 Bugzilla 949 - Documentation tweak
399 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
402 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
404 PP/02 Implemented gsasl authenticator.
406 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
408 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
409 `pkg-config foo` for cflags/libs.
411 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
412 with rest of GSASL and with heimdal_gssapi.
414 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
415 `pkg-config foo` for cflags/libs for the TLS implementation.
417 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
418 properties get this fed in as external SSF. A number of robustness
419 and debugging improvements to the cyrus_sasl authenticator.
421 PP/08 cyrus_sasl server now expands the server_realm option.
423 PP/09 Bugzilla 1214 - Log authentication information in reject log.
424 Patch by Jeremy Harris.
426 PP/10 Added dbmjz lookup type.
428 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
430 PP/12 MAIL args handles TAB as well as SP, for better interop with
431 non-compliant senders.
432 Analysis and variant patch by Todd Lyons.
434 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
435 Bug report from Lars Müller <lars@samba.org> (via SUSE),
436 Patch from Dirk Mueller <dmueller@suse.com>
438 PP/13 tls_peerdn now print-escaped for spool files.
439 Observed some $tls_peerdn in wild which contained \n, which resulted
440 in spool file corruption.
442 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
443 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
444 or write after TLS renegotiation, which otherwise led to messages
447 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
448 as a tracking header (ie: a signed header comes before the signature).
449 Patch from Wolfgang Breyha.
451 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
452 comma-sep list; embedded commas doubled.
454 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
456 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
458 Report and patch from Dmitry Banschikov.
460 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
461 Removed SSL_clear() after SSL_new() which led to protocol negotiation
462 failures. We appear to now support TLS1.1+ with Exim.
464 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
465 lets Exim select keys and certificates based upon TLS SNI from client.
466 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
467 before an outbound SMTP session. New log_selector, +tls_sni.
469 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
470 NULL dereference. Report and patch from Alun Jones.
472 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
473 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
474 Not seeing resolver debug output on NetBSD, but suspect this is a
475 resolver implementation change.
477 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
478 Left warnings. Added "eximon gdb" invocation mode.
480 PP/21 Defaulting "accept_8bitmime" to true, not false.
482 PP/22 Added -bw for inetd wait mode support.
484 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
485 locate the relevant includes and libraries. Made this the default.
487 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
488 Bugzilla 1246, report and most of solution from Tomasz Kusy.
490 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
491 This may cause build issues on older platforms.
493 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
494 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
495 gnutls_require_mac & gnutls_require_protocols (no longer supported).
496 Added SNI support via GnuTLS too.
497 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
499 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
501 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
502 Applied second patch from Janne, implementing suggestion to default
503 multiple-strings-in-record handling to match SPF spec.
505 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
507 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
508 read-only, out of scope).
509 Patch from Wolfgang Breyha, report from Stuart Northfield.
511 PP/29 Fix three issues highlighted by clang analyser static analysis.
512 Only crash-plausible issue would require the Cambridge-specific
513 iplookup router and a misconfiguration.
514 Report from Marcin Mirosław.
516 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
518 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
519 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
520 As part of this, removing so much warning spew let me fix some minor
521 real issues in debug logging.
523 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
524 assignment on my part. Fixed.
526 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
527 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
528 Janne Snabb (who went above and beyond: thank you).
530 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
531 string otherwise requires a connection and a bunch more work and it's
532 relatively easy to get wrong. Should also expose TLS library linkage
535 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
536 64-bit ${eval} (JH/03).
538 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
539 GNU libc to support some of the 64-bit stuff, should not lead to
540 conflicts. Defined before os.h is pulled in, so if a given platform
541 needs to override this, it can.
543 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
544 protection layer was required, which is not implemented.
545 Bugzilla 1254, patch from Wolfgang Breyha.
547 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
548 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
549 tls_dhparam take prime identifiers. Also unbreak combination of
550 OpenSSL+DH_params+TLSSNI.
552 PP/39 Disable SSLv2 by default in OpenSSL support.
558 PP/01 Solaris build fix for Oracle's LDAP libraries.
559 Bugzilla 1109, patch from Stephen Usher.
561 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
563 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
566 TF/02 Fix a couple more cases where we did not log the error message
567 when unlink() failed. See also change 4.74-TF/03.
569 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
570 lock up or crash if it happened to be inside a call to libc when it
571 got a SIGUSR1 from exiwhat.
573 The SIGUSR1 handler appends the current process status to the process
574 log which is later printed by exiwhat. It used to use the general
575 purpose logging code to do this, but several functions it calls are
576 not safe for signals.
578 The new output code in the SIGUSR1 handler is specific to the process
579 log, and simple enough that it's easy to inspect for signal safety.
580 Removing some special cases also simplifies the general logging code.
581 Removing the spurious timestamps from the process log simplifies
584 TF/04 Improved ratelimit ACL condition.
586 The /noupdate option has been deprecated in favour of /readonly which
587 has clearer semantics. The /leaky, /strict, and /readonly update modes
588 are mutually exclusive. The update mode is no longer included in the
589 database key; it just determines when the database is updated. (This
590 means that when you upgrde Exim will forget old rate measurements.)
592 Exim now checks that the per_* options are used with an update mode that
593 makes sense for the current ACL. For example, when Exim is processing a
594 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
595 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
596 must specify per_mail/readonly. If you omit the update mode it defaults to
597 /leaky where that makes sense (as before) or /readonly where required.
599 The /noupdate option is now undocumented but still supported for
600 backwards compatibility. It is equivalent to /readonly except that in
601 ACLs where /readonly is required you may specify /leaky/noupdate or
602 /strict/noupdate which are treated the same as /readonly.
604 A useful new feature is the /count= option. This is a generalization
605 of the per_byte option, so that you can measure the throughput of other
606 aggregate values. For example, the per_byte option is now equivalent
607 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
609 The per_rcpt option has been generalized using the /count= mechanism
610 (though it's more complicated than the per_byte equivalence). When it is
611 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
612 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
613 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
614 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
615 Note that using per_rcpt with a non-readonly update mode in more than
616 one ACL will cause the recipients to be double-counted. (The per_mail
617 and per_byte options don't have this problem.)
619 The handling of very low rates has changed slightly. If the computed rate
620 is less than the event's count (usually one) then this event is the first
621 after a long gap. In this case the rate is set to the same as this event's
622 count, so that the first message of a spam run is counted properly.
624 The major new feature is a mechanism for counting the rate of unique
625 events. The new per_addr option counts the number of different
626 recipients that someone has sent messages to in the last time period. It
627 behaves like per_rcpt if all the recipient addresses are different, but
628 duplicate recipient addresses do not increase the measured rate. Like
629 the /count= option this is a general mechanism, so the per_addr option
630 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
631 example, measure the rate that a client uses different sender addresses
632 with the options per_mail/unique=$sender_address. There are further
633 details in the main documentation.
635 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
637 TF/06 Removed a few PCRE remnants.
639 TF/07 Automatically extract Exim's version number from tags in the git
640 repository when doing development or release builds.
642 PP/02 Raise smtp_cmd_buffer_size to 16kB.
643 Bugzilla 879. Patch from Paul Fisher.
645 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
646 Heavily based on revision 40f9a89a from Simon Arlott's tree.
649 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
651 PP/05 Variable $av_failed, true if the AV scanner deferred.
652 Bugzilla 1078. Patch from John Horne.
654 PP/06 Stop make process more reliably on build failure.
655 Bugzilla 1087. Patch from Heiko Schlittermann.
657 PP/07 Make maildir_use_size_file an _expandable_ boolean.
658 Bugzilla 1089. Patch from Heiko Schlittermann.
660 PP/08 Handle ${run} returning more data than OS pipe buffer size.
661 Bugzilla 1131. Patch from Holger Weiß.
663 PP/09 Handle IPv6 addresses with SPF.
664 Bugzilla 860. Patch from Wolfgang Breyha.
666 PP/10 GnuTLS: support TLS 1.2 & 1.1.
668 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
671 PP/11 match_* no longer expand right-hand-side by default.
672 New compile-time build option, EXPAND_LISTMATCH_RHS.
673 New expansion conditions, "inlist", "inlisti".
675 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
677 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
679 PP/14 fix log_write() format string regression from TF/03.
680 Bugzilla 1152. Patch from Dmitry Isaikin.
686 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
688 PP/02 Harmonised TLS library version reporting; only show if debugging.
689 Layout now matches that introduced for other libraries in 4.74 PP/03.
691 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
693 PP/04 New "dns_use_edns0" global option.
695 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
698 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
699 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
701 TK/01 Updated PolarSSL code to 0.14.2.
702 Bugzilla 1097. Patch from Andreas Metzler.
704 PP/07 Catch divide-by-zero in ${eval:...}.
707 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
710 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
711 format-string attack -- SECURITY: remote arbitrary code execution.
713 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
714 time unintentionally subject to list matching rules, letting the header
715 cause arbitrary Exim lookups (of items which can occur in lists, *not*
716 arbitrary string expansion). This allowed for information disclosure.
718 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
719 INT_MIN/-1 -- value coerced to INT_MAX.
725 NM/01 Workround for PCRE version dependency in version reporting
728 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
729 This fixes portability to compilers other than gcc, notably
730 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
732 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
733 makefiles for portability to HP-UX and POSIX correctness.
735 PP/01 Permit LOOKUP_foo enabling on the make command-line.
736 Also via indented variable definition in the Makefile.
737 (Debugging by Oliver Heesakkers).
739 PP/02 Restore caching of spamd results with expanded spamd_address.
740 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
742 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
743 Improves build reliability. Fix from: Frank Elsner
745 NM/02 Fix wide character breakage in the rfc2047 coding
746 Fixes bug 1064. Patch from Andrey N. Oktyabrski
748 NM/03 Allow underscore in dnslist lookups
749 Fixes bug 1026. Patch from Graeme Fowler
751 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
752 Code patches from Adam Ciarcinski of NetBSD.
754 NM/04 Fixed exiqgrep to cope with mailq missing size issue
757 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
758 is logged, to avoid truncation. Patch from John Horne.
760 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
761 Patch from Jakob Hirsch.
763 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
764 SQL string expansion failure details.
765 Patch from Andrey Oktyabrski.
767 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
768 Patch from Simon Arlott.
770 PP/09 New lookups functionality failed to compile on old gcc which rejects
771 extern declarations in function scope.
772 Patch from Oliver Fleischmann
774 PP/10 Use sig_atomic_t for flags set from signal handlers.
775 Check getgroups() return and improve debugging.
776 Fixed developed for diagnosis in bug 927 (which turned out to be
779 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
780 Patch from Mark Zealey.
782 PP/12 Bugzilla 1056: Improved spamd server selection.
783 Patch from Mark Zealey.
785 PP/13 Bugzilla 1086: Deal with maildir quota file races.
786 Based on patch from Heiko Schlittermann.
788 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
789 Patch from Uwe Doering, sign-off by Michael Haardt.
791 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
792 variable declaration deep within a block. Bug and patch from
795 PP/15 lookups-Makefile IRIX compatibilty coercion.
797 PP/16 Make DISABLE_DKIM build knob functional.
799 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
800 Patch by Simon Arlott
802 TF/03 Fix valgrind.h portability to C89 compilers that do not support
803 variable argument macros. Our copy now differs from upstream.
809 TF/01 Failure to get a lock on a hints database can have serious
810 consequences so log it to the panic log.
812 TF/02 Log LMTP confirmation messages in the same way as SMTP,
813 controlled using the smtp_confirmation log selector.
815 TF/03 Include the error message when we fail to unlink a spool file.
817 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
818 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
819 for maintaining out-of-tree patches for some time.
821 PP/01 Bugzilla 139: Documentation and portability issues.
822 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
823 Handle per-OS dynamic-module compilation flags.
825 PP/02 Let /dev/null have normal permissions.
826 The 4.73 fixes were a little too stringent and complained about the
827 permissions on /dev/null. Exempt it from some checks.
828 Reported by Andreas M. Kirchwitz.
830 PP/03 Report version information for many libraries, including
831 Exim version information for dynamically loaded libraries. Created
832 version.h, now support a version extension string for distributors
833 who patch heavily. Dynamic module ABI change.
835 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
836 privilege escalation vulnerability whereby the Exim run-time user
837 can cause root to append content of the attacker's choosing to
840 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
843 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
844 If dropping privileges for untrusted macros, we disabled normal logging
845 on the basis that it would fail; for the Exim run-time user, this is not
846 the case, and it resulted in successful deliveries going unlogged.
847 Fixed. Reported by Andreas Metzler.
853 PP/01 Date: & Message-Id: revert to normally being appended to a message,
854 only prepend for the Resent-* case. Fixes regression introduced in
855 Exim 4.70 by NM/22 for Bugzilla 607.
857 PP/02 Include check_rfc2047_length in configure.default because we're seeing
858 increasing numbers of administrators be bitten by this.
860 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
862 PP/03 Bugzilla 994: added openssl_options main configuration option.
864 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
866 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
868 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
870 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
871 without a peer certificate, leading to a segfault because of an
872 assumption that peers always have certificates. Be a little more
873 paranoid. Problem reported by Martin Tscholak.
875 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
876 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
877 NB: ClamAV planning to remove STREAM in "middle of 2010".
878 CL also introduces -bmalware, various -d+acl logging additions and
879 more caution in buffer sizes.
881 PP/09 Implemented reverse_ip expansion operator.
883 PP/10 Bugzilla 937: provide a "debug" ACL control.
885 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
887 PP/12 Bugzilla 973: Implement --version.
889 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
891 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
893 PP/15 Bugzilla 816: support multiple condition rules on Routers.
895 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
896 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
897 ignore trailing whitespace.
899 JJ/02 prevent non-panic DKIM error from being sent to paniclog
901 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
904 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
905 Notification from Dr Andrew Aitchison.
907 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
908 ExtendedDetectionInfo response format.
909 Notification from John Horne.
911 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
914 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
915 XSL and documented dependency on system catalogs, with examples of how
918 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
921 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
922 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
923 configuration file which is writeable by the Exim user or group.
925 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
926 of configuration files to cover files specified with the -C option if
927 they are going to be used with root privileges, not just the default
930 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
931 option (effectively making it always true).
933 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
934 files to be used while preserving root privileges.
936 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
937 that rogue child processes cannot use them.
939 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
940 run-time user, instead of root.
942 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
943 Exim run-time user without dropping privileges.
945 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
946 result string, instead of calling string_vformat() twice with the same
949 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
950 for other users. Others should always drop root privileges if they use
951 -C on the command line, even for a whitelisted configure file.
953 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
955 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
961 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
962 $data_path, and $header_path variables; fixed documentation bugs and
965 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
966 exipick to access non-standard spools, including the "frozen" queue
969 NM/01 Bugzilla 965: Support mysql stored procedures.
970 Patch from Alain Williams
972 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
974 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
975 Patch from Andreas Metzler
977 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
978 Patch from Kirill Miazine
980 NM/05 Bugzilla 671: Added umask to procmail example.
982 JJ/03 installed exipick 20100323.0, fixing doc bug
984 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
985 directory. Notification and patch from Dan Rosenberg.
987 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
989 TK/02 Improve log output when DKIM signing operation fails.
991 MH/01 Treat the transport option dkim_domain as a colon separated
992 list, not as a single string, and sign the message with each element,
993 omitting multiple occurences of the same signer.
995 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
996 Bugzilla 985, 986. Patch by Simon Arlott
998 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
999 Patch by Simon Arlott
1001 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1002 MBX locking. Notification from Dan Rosenberg.
1008 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1010 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1012 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1014 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1016 NM/04 Bugzilla 915: Flag broken perl installation during build.
1022 TK/01 Added patch by Johannes Berg that expands the main option
1023 "spamd_address" if it starts with a dollar sign.
1025 TK/02 Write list of recipients to X-Envelope-Sender header when building
1026 the mbox-format spool file for content scanning (suggested by Jakob
1029 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1030 (http://www.dcc-servers.net/) support via dccifd. Activated by
1031 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1033 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1034 by Mark Daniel Reidel <mr@df.eu>.
1036 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1037 When building exim an external PCRE library is now needed -
1038 PCRE is a system library on the majority of modern systems.
1039 See entry on PCRE_LIBS in EDITME file.
1041 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1042 conversation. Added nologin parameter to request.
1043 Patch contributed by Kirill Miazine.
1045 TF/01 Do not log submission mode rewrites if they do not change the address.
1047 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1049 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1050 log files in place. Contributed by Roberto Lima.
1052 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1054 TF/03 Bugzilla 615: When checking the local_parts router precondition
1055 after a local_part_suffix or local_part_prefix option, Exim now
1056 does not use the address's named list lookup cache, since this
1057 contains cached lookups for the whole local part.
1059 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1060 Robert Millan. Documentation is in experimental-spec.txt.
1062 TF/04 Bugzilla 668: Fix parallel build (make -j).
1064 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1066 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1067 Patch provided by Jan Srzednicki.
1069 TF/05 Leading white space used to be stripped from $spam_report which
1070 wrecked the formatting. Now it is preserved.
1072 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1073 that they are available at delivery time.
1075 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1077 TF/08 TLS error reporting now respects the incoming_interface and
1078 incoming_port log selectors.
1080 TF/09 Produce a more useful error message if an SMTP transport's hosts
1081 setting expands to an empty string.
1083 NM/06 Bugzilla 744: EXPN did not work under TLS.
1084 Patch provided by Phil Pennock.
1086 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1087 Patch provided by Richard Godbee.
1089 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1090 acl_smtp_notquit, added index entry.
1092 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1093 Patch provided by Eugene Bujak.
1095 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1096 accept(). Patch provided by Maxim Dounin.
1098 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1099 Patch provided by Phil Pennock.
1101 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1103 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1104 Patch provided by Brad "anomie" Jorsch.
1106 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1107 Patch provided by Dean Brooks.
1109 TK/05 Add native DKIM support (does not depend on external libraries).
1111 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1112 Patch provided by Graeme Fowler.
1114 NM/16 Bugzilla 851: Documentation example syntax fix.
1116 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1118 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1121 NM/19 Bugzilla 745: TLS version reporting.
1122 Patch provided by Phil Pennock.
1124 NM/20 Bugzilla 167: bool: condition support.
1125 Patch provided by Phil Pennock.
1127 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1128 clients. Patch provided by Phil Pennock.
1130 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1131 Patch provided by Brad "anomie" Jorsch.
1133 NM/23 Bugzilla 687: Fix misparses in eximstats.
1134 Patch provided by Heiko Schlittermann.
1136 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1137 Patch provided by Heiko Schlittermann.
1139 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1140 plus update to original patch.
1142 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1144 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1145 Patch provided by David Brownlee.
1147 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1149 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1151 NM/30 Bugzilla 888: TLS documentation bugfixes.
1153 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1155 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1156 Unlike the original bugzilla I have changed all shell scripts in src tree.
1158 NM/33 Bugzilla 898: Transport filter timeout fix.
1159 Patch by Todd Rinaldo.
1161 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1162 Patch by Serge Demonchaux.
1164 NM/35 Bugzilla 39: Base64 decode bug fixes.
1165 Patch by Jakob Hirsch.
1167 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1169 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1171 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1173 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1179 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1180 ALT-N's libdkim that I have put here:
1181 http://duncanthrax.net/exim-experimental/
1183 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1184 were called 'true' and it seems that C99 defines that as a reserved
1185 keyword to be used with 'bool' variable types. That means you could
1186 not include C99-style headers which use bools without triggering
1187 build errors in sieve.c.
1189 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1190 as mailq or other aliases. Changed the --help handling significantly
1191 to do whats expected. exim_usage() emits usage/help information.
1193 SC/01 Added the -bylocaldomain option to eximstats.
1195 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1197 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1199 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1205 PH/01 Another patch from the Sieve maintainer.
1207 PH/02 When an IPv6 address is converted to a string for single-key lookup
1208 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1209 dots are used instead of colons so that keys in lsearch files need not
1210 contain colons. This was done some time before quoting was made available
1211 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1212 (notated using the quote facility) so as to distinguish them from IPv4
1213 keys. This meant that lookups for IP addresses in host lists did not work
1214 for iplsearch lookups.
1216 This has been fixed by arranging for IPv6 addresses to be expressed with
1217 colons if the lookup type is iplsearch. This is not incompatible, because
1218 previously such lookups could never work.
1220 The situation is now rather anomolous, since one *can* have colons in
1221 ordinary lsearch keys. However, making the change in all cases is
1222 incompatible and would probably break a number of configurations.
1224 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1227 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1228 conversion specification without a maximum field width, thereby enabling
1229 a rogue spamd server to cause a buffer overflow. While nobody in their
1230 right mind would setup Exim to query an untrusted spamd server, an
1231 attacker that gains access to a server running spamd could potentially
1232 exploit this vulnerability to run arbitrary code as the Exim user.
1234 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1235 $primary_hostname instead of what libspf2 thinks the hosts name is.
1237 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1238 a directory entry by the name of the lookup key. Previously, if a
1239 symlink pointed to a non-existing file or a file in a directory that
1240 Exim lacked permissions to read, a lookup for a key matching that
1241 symlink would fail. Now it is enough that a matching directory entry
1242 exists, symlink or not. (Bugzilla 503.)
1244 PH/03 The body_linecount and body_zerocount variables are now exported in the
1247 PH/04 Added the $dnslist_matched variable.
1249 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1250 This means they are set thereafter only if the connection becomes
1253 PH/06 Added the client_condition to authenticators so that some can be skipped
1254 by clients under certain conditions.
1256 PH/07 The error message for a badly-placed control=no_multiline_responses left
1257 "_responses" off the end of the name.
1259 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1261 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1262 (without spaces) instead of just copying the configuration text.
1264 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1266 PH/11 Added $max_received_linelength.
1268 PH/12 Added +ignore_defer and +include_defer to host lists.
1270 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1271 way in which PCRE > 7.0 is built.
1273 PH/14 Implemented queue_only_load_latch.
1275 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1276 MAIL command. The effect was to mangle the value on 64-bit systems.
1278 PH/16 Another patch from the Sieve maintainer.
1280 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1282 PH/18 If a system quota error occurred while trying to create the file for
1283 a maildir delivery, the message "Mailbox is full" was not appended to the
1284 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1285 only to a quota excession during the actual writing of the file.
1287 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1288 characters?) which causes problems in log lines. The DN values are now
1289 passed through string_printing() before being added to log lines.
1291 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1292 and InterBase are left for another time.)
1294 PH/21 Added message_body_newlines option.
1296 PH/22 Guard against possible overflow in moan_check_errorcopy().
1298 PH/23 POSIX allows open() to be a macro; guard against that.
1300 PH/24 If the recipient of an error message contained an @ in the local part
1301 (suitably quoted, of course), incorrect values were put in $domain and
1302 $local_part during the evaluation of errors_copy.
1308 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1309 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1312 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1313 issue a MAIL command.
1315 PH/02 In an ACL statement such as
1317 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1319 if a client was not listed at all, or was listed with a value other than
1320 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1321 the condition was not true (as it should be), so access was not denied.
1322 The bug was that the ! inversion was incorrectly passed on to the second
1323 item. This has been fixed.
1325 PH/03 Added additional dnslists conditions == and =& which are different from
1326 = and & when the dns lookup returns more than one IP address.
1328 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1329 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1331 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1332 FSYNC, which compiles an option called disable_fsync that allows for
1333 bypassing fsync(). The documentation is heavily laced with warnings.
1335 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1337 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1338 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1339 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1340 including adding "make clean"; (3) Added -fPIC when compiling the test
1341 dynamically loaded module, to get rid of a warning.
1343 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1344 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1345 = 0s. The bug is otherwise harmless.
1347 PH/07 There was a bug in the dovecot authenticator such that the value of
1348 $auth1 could be overwritten, and so not correctly preserved, after a
1349 successful authentication. This usually meant that the value preserved by
1350 the server_setid option was incorrect.
1352 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1354 PH/09 Installed PCRE release 7.0.
1356 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1357 run for batched SMTP input. It is now run at the start of every message
1358 in the batch. While fixing this I discovered that the process information
1359 (output by running exiwhat) was not always getting set for -bs and -bS
1360 input. This is fixed, and it now also says "batched" for BSMTP.
1362 PH/11 Added control=no_pipelining.
1364 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1365 patch, slightly modified), and move the expansion of helo_data till after
1366 the connection is made in the smtp transport (so it can use these
1369 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1371 PH/14 Added log_selector = +pid.
1373 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1375 PH/16 Add ${if forany and ${if forall.
1377 PH/17 Added dsn_from option to vary the From: line in DSNs.
1379 PH/18 Flush SMTP output before performing a callout, unless control =
1380 no_callout_flush is set.
1382 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1383 was true (the default) a successful delivery failed to delete the retry
1384 item, thus causing premature timeout of the address. The bug is now
1387 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1389 PH/21 Long custom messages for fakedefer and fakereject are now split up
1390 into multiline reponses in the same way that messages for "deny" and
1391 other ACL rejections are.
1393 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1394 with slight modification.
1396 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1397 draft, changing the syntax and factoring some duplicate code".
1399 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1400 for deliveries of the second and subsequent messages over the same SMTP
1403 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1404 ${reduce, with only minor "tidies".
1406 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1408 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1409 expansion side effects.
1411 PH/27 When a message times out after an over-quota error from an Exim-imposed
1412 quota, the bounce message says "mailbox is full". This message was not
1413 being given when it was a system quota that was exceeded. It now should
1416 MH/03 Made $recipients available in local_scan(). local_scan() already has
1417 better access to the recipient list through recipients_list[], but
1418 $recipients can be useful in postmaster-provided expansion strings.
1420 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1421 in the case of a MAIL command with additional options following the
1422 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1423 were accidentally chopped off.
1425 PH/29 SMTP synchronization checks are implemented when a command is read -
1426 there is a check that no more input is waiting when there shouldn't be
1427 any. However, for some commands, a delay in an ACL can mean that it is
1428 some time before the response is written. In this time, more input might
1429 arrive, invalidly. So now there are extra checks after an ACL has run for
1430 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1431 pipelining has not been advertised.
1433 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1435 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1436 returned in the SMTP response when smtp_return_error_details was false.
1437 This has been fixed.
1439 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1440 instead of the C I/O that was originally supplied, because problems were
1441 reported on Solaris.
1443 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1444 Exim which did not show up earlier: it was assuming that a call to
1445 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1446 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1447 was a macro that became an assignment, so it seemed to work. This has
1448 changed to a proper function call with a void return, hence the compile
1449 error. Exim's code has been fixed.
1451 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1454 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1456 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1458 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1459 because it turns out to be too big or there is a timeout) there is no
1460 "Completed" line in the log. When some messages of this type were
1461 selected by exigrep, they were listed as "not completed". Others were
1462 picked up by some special patterns. I have improved the selection
1463 criteria to be more general.
1465 PH/38 The host_find_failed option in the manualroute router can now be set
1466 to "ignore", to completely ignore a host whose IP address cannot be
1467 found. If all hosts are ignored, the behaviour is controlled by the new
1468 host_all_ignored option.
1470 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1471 homing or because of multiple MX records with /mx) generated more than
1472 one IP address, and the following item turned out to be the local host,
1473 all the secondary addresses of the first item were incorrectly removed
1474 from the list, along with the local host and any following hosts (which
1475 is what is supposed to happen).
1477 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1478 whoever called Exim into the -H file. In the case of the daemon it was
1479 behaving confusingly. When first started, it used values for whoever
1480 started the daemon, but after a SIGHUP it used the Exim user (because it
1481 calls itself on a restart). I have changed the code so that it now always
1484 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1485 message are rejected with the same error (e.g. no authentication or bad
1486 sender address), and a DATA command is nevertheless sent (as can happen
1487 with PIPELINING or a stupid MUA), the error message that was given to the
1488 RCPT commands is included in the rejection of the DATA command. This is
1489 intended to be helpful for MUAs that show only the final error to their
1492 PH/42 Another patch from the Sieve maintainer.
1494 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1495 Eximstats - Fixed some broken HTML links and added missing column headers
1497 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1498 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1500 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1502 PH/43 Yet another patch from the Sieve maintainer.
1504 PH/44 I found a way to check for a TCP/IP connection going away before sending
1505 the response to the final '.' that terminates a message, but only in the
1506 case where the client has not sent further data following the '.'
1507 (unfortunately, this is allowed). However, in many cases there won't be
1508 any further data because there won't be any more messages to send. A call
1509 to select() can be used: if it shows that the input is "ready", there is
1510 either input waiting, or the socket has been closed. An attempt to read
1511 the next input character can distinguish the two cases. Previously, Exim
1512 would have sent an OK response which the client would never have see.
1513 This could lead to message repetition. This fix should cure that, at
1514 least in a lot of common cases.
1516 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1517 advertised in response to EHLO.
1523 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1524 fixed by 4.65/MH/01 (is this a record?) are fixed:
1526 (i) An empty string was always treated as zero by the numeric comparison
1527 operators. This behaviour has been restored.
1529 (ii) It is documented that the numeric comparison operators always treat
1530 their arguments as decimal numbers. This was broken in that numbers
1531 starting with 0 were being interpreted as octal.
1533 While fixing these problems I realized that there was another issue that
1534 hadn't been noticed. Values of message_size_limit (both the global option
1535 and the transport option) were treated as octal if they started with 0.
1536 The documentation was vague. These values are now always treated as
1537 decimal, and I will make that clear in the documentation.
1543 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1544 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1547 MH/01 Don't check that the operands of numeric comparison operators are
1548 integers when their expansion is in "skipping" mode (fixes bug
1549 introduced by 4.64-PH/07).
1551 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1552 child addresses, Exim now panics and dies. Previously, because the count
1553 is held in a short int, deliveries were likely to be lost. As such a
1554 large number of recipients for a single message is ridiculous
1555 (performance will be very, very poor), I have chosen to impose a limit
1556 rather than extend the field.
1562 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1563 leftover -K file (the existence of which was triggered by #402).
1564 While we were at it, introduced process PID as part of the -K
1565 filename. This should rule out race conditions when creating
1568 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1569 processing considerably. Previous code took too long for large mails,
1570 triggering a timeout which in turn triggers #401.
1572 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1573 in the DK code in transports.c. sendfile() is not really portable,
1574 hence the _LINUX specificness.
1576 TF/01 In the add_headers option to the mail command in an Exim filter,
1577 there was a bug that Exim would claim a syntax error in any
1578 header after the first one which had an odd number of characters
1581 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1582 callout verification, Exim cached a "reject" for the entire domain. This
1583 is correct for most verifications, but it is not correct for a recipient
1584 verification with use_sender or use_postmaster set, because in that case
1585 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1586 case of MAIL FROM:<> rejection from other early rejections (e.g.
1587 rejection of HELO). When verifying a recipient using a non-null MAIL
1588 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1589 Whatever the result of the callout, the value of the domain cache is
1590 left unchanged (for any other kind of callout, getting as far as trying
1591 RCPT means that the domain itself is ok).
1593 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1596 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1597 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1598 session, the daemon ignores ECONNECT errors and logs others; it now
1599 ignores EPIPE as well.
1601 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1602 (quoted-printable decoding).
1604 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1605 later the small subsequent patch to fix an introduced bug.
1607 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1609 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1611 PH/08 An error is now given if message_size_limit is specified negative.
1613 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1614 to be given (somewhat) arbitrary names.
1616 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1619 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1620 miscellaneous code fixes
1622 PH/10 Added the log_reject_target ACL modifier to specify where to log
1625 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1626 hostname. This is wrong, because it relates to the incoming message (and
1627 probably the interface on which it is arriving) and not to the outgoing
1628 callout (which could be using a different interface). This has been
1629 changed to use the value of the helo_data option from the smtp transport
1630 instead - this is what is used when a message is actually being sent. If
1631 there is no remote transport (possible with a router that sets up host
1632 addresses), $smtp_active_hostname is used.
1634 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1635 tweaks were necessary in order to get it to work (see also 21 below):
1636 (a) The code assumed that strncpy() returns a negative number on buffer
1637 overflow, which isn't the case. Replaced with Exim's string_format()
1639 (b) There were several signed/unsigned issues. I just did the minimum
1640 hacking in of casts. There is scope for a larger refactoring.
1641 (c) The code used strcasecmp() which is not a standard C function.
1642 Replaced with Exim's strcmpic() function.
1643 (d) The code set only $1; it now sets $auth1 as well.
1644 (e) A simple test gave the error "authentication client didn't specify
1645 service in request". It would seem that Dovecot has changed its
1646 interface. Fortunately there's a specification; I followed it and
1647 changed what the client sends and it appears to be working now.
1649 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1652 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1653 address A is aliased to B and C, where B exists and C does not. Without
1654 -v the output is "A verified" because verification stops after a
1655 successful redirection if more than one address is generated. However,
1656 with -v the child addresses are also verified. Exim was outputting "A
1657 failed to verify" and then showing the successful verification for C,
1658 with its parentage. It now outputs "B failed to verify", showing B's
1659 parentage before showing the successful verification of C.
1661 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1662 look up a TXT record in a specific list after matching in a combined
1665 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1666 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1667 they consult the DNS. I had assumed they would set it the way they
1668 wanted; and indeed my experiments on Linux seem to show that in some
1669 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1670 To be on the safe side, however, I have now made the interface to
1671 host_find_byname() similar to host_find_bydns(), with an argument
1672 containing the DNS resolver options. The host_find_byname() function now
1673 sets these options at its start, just as host_find_bydns() does. The smtp
1674 transport options dns_qualify_single and dns_search_parents are passed to
1675 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1676 of host_find_byname() use the default settings of RES_DEFNAMES
1677 (qualify_single) but not RES_DNSRCH (search_parents).
1679 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1680 spool_read_header() do less string testing, by means of a preliminary
1681 switch on the second character of optional "-foo" lines. (This is
1682 overdue, caused by the large number of possibilities that now exist.
1683 Originally there were few.) While I was there, I also converted the
1684 str(n)cmp tests so they don't re-test the leading "-" and the first
1685 character, in the hope this might squeeze out yet more improvement.
1687 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1688 flag allowing group syntax was set by the header_syntax check but not
1689 turned off, possible causing trouble later; (2) The flag was not being
1690 set at all for the header_verify test, causing "group"-style headers to
1691 be rejected. I have now set it in this case, and also caused header_
1692 verify to ignore an empty address taken from a group. While doing this, I
1693 came across some other cases where the code for allowing group syntax
1694 while scanning a header line wasn't quite right (mostly, not resetting
1695 the flag correctly in the right place). These bugs could have caused
1696 trouble for malformed header lines. I hope it is now all correct.
1698 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1699 with the "reply" argument non-NULL. The code, however (which originally
1700 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1701 but it didn't always do it. This confused somebody who was copying the
1702 code for some other use. I have removed all the tests.
1704 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1705 feature that was used to support insecure browsers during the U.S. crypto
1706 embargo. It requires special client support, and Exim is probably the
1707 only MTA that supported it -- and would never use it because real RSA is
1708 always available. This code has been removed, because it had the bad
1709 effect of slowing Exim down by computing (never used) parameters for the
1710 RSA_EXPORT functionality.
1712 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1713 authenticator to fail if there's a tab character in the incoming data
1714 (there should never be unless someone is messing about, as it's supposed
1715 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1716 if the connection is using TLS or if the remote IP is the same as the
1717 local IP, and the "valid-client-cert option" if a client certificate has
1720 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1721 authenticators. This can be used for authorization after authentication
1722 succeeds. (In the case of plaintext, it servers for both authentication
1725 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1726 if any retry times were supplied.
1728 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1729 connection, where there is no client IP address to check. In this
1730 situation, the verify now always succeeds.
1732 PH/25 Applied John Jetmore's -Mset patch.
1734 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1736 PH/27 In a string expansion for a processed (not raw) header when multiple
1737 headers of the same name were present, leading whitespace was being
1738 removed from all of them, but trailing whitespace was being removed only
1739 from the last one. Now trailing whitespace is removed from each header
1740 before concatenation. Completely empty headers in a concatenation (as
1741 before) are ignored.
1743 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1744 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1746 PH/29 [Removed. This was a change that I later backed out, and forgot to
1747 correct the ChangeLog entry (that I had efficiently created) before
1748 committing the later change.]
1750 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1751 address errors (4xx response to RCPT) over the same connection as other
1752 messages routed to the same hosts. Such deliveries are always "forced",
1753 so retry times are not inspected. This resulted in far too many retries
1754 for the affected addresses. The effect occurred only when there were more
1755 hosts than the hosts_max_try setting in the smtp transport when it had
1756 the 4xx errors. Those hosts that it had tried were not added to the list
1757 of hosts for which the message was waiting, so if all were tried, there
1758 was no problem. Two fixes have been applied:
1760 (i) If there are any address or message errors in an SMTP delivery, none
1761 of the hosts (tried or untried) are now added to the list of hosts
1762 for which the message is waiting, so the message should not be a
1763 candidate for sending over the same connection that was used for a
1764 successful delivery of some other message. This seems entirely
1765 reasonable: after all the message is NOT "waiting for some host".
1766 This is so "obvious" that I'm not sure why it wasn't done
1767 previously. Hope I haven't missed anything, but it can't do any
1768 harm, as the worst effect is to miss an optimization.
1770 (ii) If, despite (i), such a delivery is accidentally attempted, the
1771 routing retry time is respected, so at least it doesn't keep
1772 hammering the server.
1774 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1775 in ${readsocket because some servers need this prod.
1777 PH/32 Added some extra debug output when updating a wait-xxx database.
1779 PH/33 The hint "could be header name not terminated by colon", which has been
1780 given for certain expansion errors for a long time, was not being given
1781 for the ${if def:h_colon_omitted{... case.
1783 PH/34 The spec says: "With one important exception, whenever a domain list is
1784 being scanned, $domain contains the subject domain." There was at least
1785 one case where this was not true.
1787 PH/35 The error "getsockname() failed: connection reset by peer" was being
1788 written to the panic log as well as the main log, but it isn't really
1789 panic-worthy as it just means the connection died rather early on. I have
1790 removed the panic log writing for the ECONNRESET error when getsockname()
1793 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1794 runs only) independently of the message's sender address. This meant
1795 that, if the 4xx error was in fact related to the sender, a different
1796 message to the same recipient with a different sender could confuse
1797 things. In particualar, this can happen when sending to a greylisting
1798 server, but other circumstances could also provoke similar problems.
1799 I have changed the default so that the retry time for these errors is now
1800 based a combination of the sender and recipient addresses. This change
1801 can be overridden by setting address_retry_include_sender=false in the
1804 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1805 remote server are returned as part of bounce messages. This was not
1806 happening for LMTP over a pipe (the lmtp transport), but now it is the
1807 same for both kinds of LMTP.
1809 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1810 in header lines that were in fact CNAMEs. This is no longer the case.
1812 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1813 and queue runs started by the daemon processed all messages. This has
1814 been fixed so that -R and -S can now usefully be given with -q<time>.
1816 PH/40 Import PCRE release 6.7 (fixes some bugs).
1818 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1820 PH/42 Give an error if -q is specified more than once.
1822 PH/43 Renamed the variables $interface_address and $interface_port as
1823 $received_ip_address and $received_port, to make it clear that these
1824 values apply to message reception, and not to the outgoing interface when
1825 a message is delivered. (The old names remain recognized, of course.)
1827 PH/44 There was no timeout on the connect() call when using a Unix domain
1828 socket in the ${readsocket expansion. There now is.
1830 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1831 be meaningful with "accept".
1833 SC/01 Eximstats V1.43
1834 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1836 SC/02 Eximstats V1.44
1837 Use a glob alias rather than an array ref in the generated
1838 parser. This improves both readability and performance.
1840 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1841 Collect SpamAssassin and rejection statistics.
1842 Don't display local sender or destination tables unless
1843 there is data to show.
1844 Added average volumes into the top table text output.
1846 SC/04 Eximstats V1.46
1847 Collect data on the number of addresses (recipients)
1848 as well as the number of messages.
1850 SC/05 Eximstats V1.47
1851 Added 'Message too big' to the list of mail rejection
1852 reasons (thanks to Marco Gaiarin).
1854 SC/06 Eximstats V1.48
1855 Mainlog lines which have GMT offsets and are too short to
1856 have a flag are now skipped.
1858 SC/07 Eximstats V1.49 (Alain Williams)
1859 Added the -emptyok flag.
1861 SC/08 Eximstats V1.50
1862 Fixes for obtaining the IP address from reject messages.
1864 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1865 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1866 whitesspace changes from 4.64-PH/27
1868 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1871 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1872 are found, allow negative numbers in numeric criteria)
1874 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1876 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1877 to match changes made in 4.64-PH/43
1879 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1881 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1882 "domain" is actually an IP address, and force a failure. This locks out
1883 those revolvers/nameservers that support "A-for-A" lookups, in
1884 contravention of the specifications.
1886 PH/48 When a host name was looked up from an IP address, and the subsequent
1887 forward lookup of the name timed out, the host name was left in
1888 $sender_host_name, contrary to the specification.
1890 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1891 restricted to single-key lookups, Exim was not diagnosing an error if
1892 * or *@ was used with a query-style lookup.
1894 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1896 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1897 long ago, but noone interested enough thought of it. Let's just say that
1898 the "1.1" means that there are some new functions that weren't there at
1899 some point in the past.
1901 PH/51 Error processing for expansion failure of helo_data from an smtp
1902 transport during callout processing was broken.
1904 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1905 tested/used via the -bh/-bhc/-bs options.
1907 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1908 bug, fixed in subsequent PCRE releases).
1910 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1911 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1913 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1919 SC/01 Use a glob alias rather than an array ref in eximstats generated
1920 parser. This improves both readability and performance.
1922 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1923 Don't display local sender or destination tables in eximstats unless
1924 there is data to show.
1925 Added average volumes into the eximstats top table text output.
1927 SC/03 Collect data on the number of addresses (recipients) as well
1928 as the number of messages in eximstats.
1930 TF/01 Correct an error in the documentation for the redirect router. Exim
1931 does (usually) call initgroups() when daemonizing.
1933 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1934 with consistent privilege compared to when running as a daemon.
1936 TF/03 Note in the spec that $authenticated_id is not set for local
1937 submissions from trusted users.
1939 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1940 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1942 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1943 by adding some example configuration directives to the default
1944 configuration file. A little bit of work is required to uncomment the
1945 directives and define how usernames and passwords are checked, but
1946 there is now a framework to start from.
1948 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1949 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1950 without this. I don't know how relevant this is to other LDAP libraries.
1952 PH/02 Add the verb name to the "unknown ACL verb" error.
1954 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1956 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1958 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1959 directory not expanded when it should be if an expanded home directory
1960 was set for the address (which is overridden by the transport).
1962 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1965 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1966 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1967 because it is too late at that time, and has no effect.
1969 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1970 security issue with \' (bugzilla #107). I could not use the
1971 PQescapeStringConn() function, because it needs a PGconn value as one of
1974 PH/08 When testing addresses using -bt, indicate those final addresses that
1975 are duplicates that would not cause an additional delivery. At least one
1976 person was confused, thinking that -bt output corresponded to deliveries.
1977 (Suppressing duplicates isn't a good idea as you lose the information
1978 about possibly different redirections that led to the duplicates.)
1980 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1981 systems where poll() doesn't work, in particular OS X.
1983 PH/10 Added more information to debugging output for retry time not reached.
1985 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1986 operations in malware.c.
1988 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1991 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1992 syslog_duplication set false, log lines that would normally be written
1993 both the the main log and to the reject log were not written to syslog at
1996 PH/14 In the default configuration, change the use of "message" in ACL warn
1997 statements to "add_header".
1999 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2000 not followed by a command (e.g. "seen endif").
2002 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2003 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2006 PH/17 Added extra conditions to the default value of delay_warning_condition
2010 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2011 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2012 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2015 The Auto-Submitted: and various List- headers are standardised, whereas I
2016 don't think Precedence: ever was.
2018 PH/18 Refactored debugging code in route_finduser() to show more information,
2019 in particular, the error code if getpwnam() issues one.
2021 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2022 This is apparently needed in addition to the PH/07 change above to avoid
2023 any possible encoding problems.
2025 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2026 but not after initializing Perl.
2028 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2029 output them only if debugging. By default they are written stderr,
2030 apparently, which is not desirable.
2032 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2035 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2038 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2040 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2041 authenticated or an ident call has been made. Suppress the default
2042 values for $authenticated_id and $authenticated_sender (but permit -oMai
2043 and -oMas) when testing with -bh.
2045 PH/24 Re-jigged the order of the tests in the default configuration so that the
2046 tests for valid domains and recipients precede the DNS black list and CSA
2047 tests, on the grounds that those ones are more expensive.
2049 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2050 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2051 This bug exists in every version of Exim that I still have, right back to
2054 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2055 However, an attempt to turn on case-sensitivity in a regex key by
2056 including (?-i) didn't work because the subject string was already
2057 lowercased, and the effects were non-intuitive. It turns out that a
2058 one-line patch can be used to allow (?-i) to work as expected.
2064 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2065 other effects) broke the use of negated acl sub-conditions.
2067 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2070 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2071 "Deny" causes Exim to reject the incoming connection with a 554 error.
2072 Unfortunately, if there is a major crisis, such as a disk failure,
2073 tcp-wrappers gives "deny", whereas what one would like would be some
2074 kind of temporary error. A kludge has been added to help with this.
2075 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2076 554 error is used if errno is still zero or contains ENOENT (which occurs
2077 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2080 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2082 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2083 errors. Otherwise a message that provokes a temporary error (when other
2084 messages do not) can cause a whole host to time out.
2086 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2087 addresses were routed directly to files or pipes from a redirect router.
2088 File deliveries just didn't batch; pipe deliveries might have suffered
2091 PH/06 A failure to get a lock for a hints database would erroneously always say
2092 "Failed to get write lock", even when it was really a read lock.
2094 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2095 of 0600. This has been changed to use the value of the lockfile_mode
2096 option (which defaults to 0600).
2098 PH/08 Applied small patch from the Sieve maintainer.
2100 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2101 folder from quota calculations, a direct delivery into this folder messed
2102 up the contents of the maildirsize file. This was because the regex was
2103 used only to exclude .Trash (or whatever) when the size of the mailbox
2104 was calculated. There was no check that a delivery was happening into an
2105 excluded directory. This bug has been fixed by ignoring all quota
2106 processing for deliveries into excluded directories.
2108 PH/10 Added the maildirfolder_create_regex option to appendfile.
2114 PH/01 The code for finding all the local interface addresses on a FreeBSD
2115 system running IPv6 was broken. This may well have applied to all BSD
2116 systems, as well as to others that have similar system calls. The broken
2117 code found IPv4 interfaces correctly, but gave incorrect values for the
2118 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2119 that it would not match correctly against @[] and not recognize the IPv6
2122 PH/02 The ipliteral router was not recognizing addresses of the form user@
2123 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2125 PH/03 Added disable_ipv6.
2127 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2128 decoded form, because it is most often used to construct To: headers
2129 lines in autoreplies, and the decoded form may well be syntactically
2130 invalid. However, $reply_address has leading white space removed, and all
2131 newlines turned into spaces so that the autoreply transport does not
2134 PH/05 If group was specified without a user on a router, and no group or user
2135 was specified on a transport, the group from the router was ignored.
2137 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2138 for visible compile-time settings that can be used to change these
2139 numbers, for those that want even more. Backwards compatibility with old
2140 spool files has been maintained. However, going back to a previous Exim
2141 release will lost any variables that are in spool files.
2143 PH/07 Two small changes when running in the test harness: increase delay when
2144 passing a TCP/IP connection to a new process, in case the original
2145 process has to generate a bounce, and remove special handling of
2146 127.0.0.2 (sic), which is no longer necessary.
2148 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2149 be the same on different OS.
2151 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2154 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2155 whether --show-vars was specified or not
2157 JJ/02 exipick: Added support for new ACL variable spool format introduced
2160 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2161 syntactically invalid From: or Reply-to: line, and a filter used this to
2162 generate an autoreply, and therefore failed to obtain an address for the
2163 autoreply, Exim could try to deliver to a non-existent relative file
2164 name, causing unrelated and misleading errors. What now happens is that
2165 it logs this as a hard delivery error, but does not attempt to create a
2168 PH/11 The exinext utility has a -C option for testing purposes, but although
2169 the given file was scanned by exinext itself; it wasn't being passed on
2170 when Exim was called.
2172 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2173 an end-of-file indication when reading a command response.
2175 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2176 compiled. In many other places in Exim, IPv6 addresses are always
2177 recognized, so I have changed this. It also means that IPv4 domain
2178 literals of the form [IPV4:n.n.n.n] are now always recognized.
2180 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2181 used if the router is not running as root, for example, when verifying at
2182 ACL time, or when using -bh. The debugging output from this situation was
2183 non-existent - all you got was a failure to exec. I have made two
2186 (a) Failures to set uid/gid, the current directory, or a process leader
2187 in a subprocess such as that created by queryprogram now generate
2188 suitable debugging ouput when -d is set.
2190 (b) The queryprogram router detects when it is not running as root,
2191 outputs suitable debugging information if -d is set, and then runs
2192 the subprocess without attempting to change uid/gid.
2194 PH/15 Minor change to Makefile for building test_host (undocumented testing
2197 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2198 additional section of a DNS packet that returns MX or SRV records.
2199 Instead, it always explicitly searches for A/AAAA records. This avoids
2200 major problems that occur when a DNS server includes only records of one
2201 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2202 fixed another bug: if SRV records were looked up and the corresponding
2203 address records were *not* found in the additional section, the port
2204 values from the SRV records were lost.
2206 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2207 using the correct key (the original address) when searching the retry
2208 rules in order to find which one to use for generating the retry hint.
2210 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2211 adding the default one. Similarly, if it contains a Reply-To: header, the
2212 errors_reply_to option, if set, is not used.
2214 PH/19 When calculating a retry time, Exim used to measure the "time since
2215 failure" by looking at the "first failed" field in the retry record. Now
2216 it does not use this if it is later than than the arrival time of the
2217 message. Instead it uses the arrival time. This makes for better
2218 behaviour in cases where some deliveries succeed, thus re-setting the
2219 "first failed" field. An example is a quota failure for a huge message
2220 when small messages continue to be delivered. Without this change, the
2221 "time since failure" will always be short, possible causing more frequent
2222 delivery attempts for the huge message than are intended.
2223 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2225 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2226 $1, $2, $3) because the numerical variables can be reset during some
2227 expansion items (e.g. "match"), thereby losing the authentication data.
2229 PH/21 Make -bV show the size of off_t variables so that the test suite can
2230 decide whether to run tests for quotas > 2G.
2232 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2233 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2234 filecount value is greater than 2G or if a quota value is greater than 2G
2235 on a system where the size of off_t is not greater than 4, a panic error
2238 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2239 never match. The debug and -bh output now contains an explicit error
2240 message indicating a malformed IPv4 address or mask.
2242 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2243 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2244 PH/23 above applies.
2246 PH/25 Do not write to syslog when running in the test harness. The only
2247 occasion when this arises is a failure to open the main or panic logs
2248 (for which there is an explicit test).
2250 PH/26 Added the /no_tell option to "control=freeze".
2252 PH/27 If a host name lookup failed very early in a connection, for example, if
2253 the IP address matched host_lookup and the reverse lookup yielded a name
2254 that did not have a forward lookup, an error message of the form "no IP
2255 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2256 could be logged. Now it outputs the IP address instead of "NULL".
2258 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2259 allows the sender and the authenticated sender to be set when
2260 submitting a message from within Exim. Since child_open_exim() is
2261 documented for local_scan(), the new function should be too.
2263 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2264 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2265 results in an empty string is now treated as unset.
2267 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2269 PH/31 Added sender_verify_fail logging option.
2271 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2272 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2273 tidied the source and removed it altogether.
2275 PH/33 When a queue run was abandoned because the load average was too high, a
2276 log line was always written; now it is written only if the queue_run log
2277 selector is set. In addition, the log line for abandonment now contains
2278 information about the queue run such as the pid. This is always present
2279 in "start" and "stop" lines but was omitted from the "abandon" line.
2281 PH/34 Omit spaces between a header name and the colon in the error message that
2282 is given when verify = headers_syntax fails (if there are lots of them,
2283 the message gets confusing).
2285 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2286 names, as there are now some PTR records that contain slashes. This check
2287 is only to protect against broken name servers that fall over on strange
2288 characters, so the fact that it applies to all lookups doesn't matter.
2290 PH/36 Now that the new test suite is complete, we can remove some of the
2291 special code in Exim that was needed for the old test suite. For example,
2292 sorting DNS records because real resolvers return them in an arbitrary
2293 order. The new test suite's fake resolver always returns records in the
2296 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2297 bounces) except when queue_only is set, to avoid logging races between
2298 the different processes.
2300 PH/38 Panic-die if .include specifies a non-absolute path.
2302 PH/39 A tweak to the "H" retry rule from its user.
2304 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2305 a label. They prevented compilation on older perls.
2307 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2308 a warning to be raised on newish perls.
2310 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2311 on queue. Changes to match documented behaviour of showing count of
2312 messages matching specified criteria.
2314 PH/40 Changed the default ident timeout from 30s to 5s.
2316 PH/41 Added support for the use of login_cap features, on those BSD systems
2317 that have them, for controlling the resources used by pipe deliveries.
2319 PH/42 The content-scanning code uses fopen() to create files in which to put
2320 message data. Previously it was not paying any attention to the mode of
2321 the files. Exim runs with umask(0) because the rest of the code creates
2322 files with open(), and sets the required mode explicitly. Thus, these
2323 files were ending up world-writeable. This was not a big issue, because,
2324 being within the spool directory, they were not world-accessible. I have
2325 created a function called modefopen, which takes an additional mode
2326 argument. It sets umask(777), creates the file, chmods it to the required
2327 mode, then resets the umask. All the relevant calls to fopen() in the
2328 content scanning code have been changed to use this function.
2330 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2331 to 24 hours. This avoids potential overflow problems when processing G
2332 and H retry rules. I suspect nobody ever tinkers with this value.
2334 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2336 PH/45 When the plaintext authenticator is running as a client, the server's
2337 challenges are checked to ensure they are valid base64 strings. By
2338 default, the authentication attempt is cancelled if an invalid string is
2339 received. Setting client_ignore_invalid_base64 true ignores these errors.
2340 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2341 they are received. Thus, the responses can be made to depend on the
2342 challenges. If an invalid string is ignored, an empty string is placed in
2345 PH/46 Messages that are created by the autoreply transport now contains a
2346 References: header, in accordance with RFCs 2822 and 3834.
2348 PH/47 Added authenticated_sender_force to the smtp transport.
2350 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2352 PH/49 Installed latest patch from the Sieve maintainer.
2354 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2355 was also set, the appendfile transport was unnecessarily scanning a
2356 directory of message files (e.g. for maildir delivery) to find the count
2357 of files (along with the size), even though it did not need this
2358 information. It now does the scan only if it needs to find either the
2359 size of the count of files.
2361 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2363 PH/52 Two bugs concerned with error handling when the smtp transport is
2366 (i) Exim was not creating retry information for temporary errors given
2367 for individual recipients after the DATA command when the smtp transport
2368 was used in LMTP mode. This meant that they could be retried too
2369 frequently, and not timed out correctly.
2371 (ii) Exim was setting the flag that allows error details to be returned
2372 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2373 recipients that were returned after the DATA command.
2375 PH/53 This is related to PH/52, but is more general: for any failing address,
2376 when detailed error information was permitted to be returned to the
2377 sender, but the error was temporary, then after the final timeout, only
2378 "retry timeout exceeded" was returned. Now it returns the full error as
2379 well as "retry timeout exceeded".
2381 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2382 do this, and (what is worse) MTAs that accept it.
2384 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2385 will now be deprecated.
2387 PH/56 New os.c-cygwin from the Cygwin maintainer.
2389 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2390 formats (previously only available in exim formats via -bpr, -bpru,
2391 and -bpra. Now also available in native and exiqgrep formats)
2393 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2394 with very large, slow to parse queues
2396 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2398 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2400 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2401 responses to authentication challenges, though it was showing the
2402 challenges; (ii) I've removed the CR characters from the debug output for
2405 PH/58 Allow for the insertion of a newline as well as a space when a string
2406 is turned into more than one encoded-word during RFC 2047 encoding. The
2407 Sieve code now uses this.
2409 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2410 data_4xx, lost_connection, tls_required.
2412 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2413 message was being sent as an SMTP response.
2415 PH/61 Add -l and -k options to exicyclog.
2417 PH/62 When verifying, if an address was redirected to one new address, so that
2418 verification continued, and the new address failed or deferred after
2419 having set something in $address_data, the value of $address_data was not
2420 passed back to the ACL. This was different to the case when no
2421 redirection occurred. The value is now passed back in both cases.
2423 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2424 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2425 use login_cap.h, so on its own it isn't the distinguishing feature. The
2426 new name refers directly to the setclassresources() function.
2428 PH/65 Added configuration files for NetBSD3.
2430 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2432 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2433 is preferred over IPv4.
2435 PH/68 The bounce_return_message and bounce_return_body options were not being
2436 honoured for bounces generated during the reception of non-SMTP messages.
2437 In particular, this applied to messages rejected by the ACL. This bug has
2438 been fixed. However, if bounce_return_message is true and bounce_return_
2439 body is false, the headers that are returned for a non-SMTP message
2440 include only those that have been read before the error was detected.
2441 (In the case of an ACL rejection, they have all been read.)
2443 PH/69 The HTML version of the specification is now built in a directory called
2444 spec_html instead of spec.html, because the latter looks like a path with
2445 a MIME-type, and this confuses some software.
2447 PH/70 Catch two compiler warnings in sieve.c.
2449 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2450 function verify_get_ident() calls ip_connect() to connect a socket, but
2451 if the "connect()" function timed out, ip_connect() used to close the
2452 socket. However, verify_get_ident() also closes the socket later, and in
2453 between Exim writes to the log, which may get opened at this point. When
2454 the socket was closed in ip_connect(), the log could get the same file
2455 descriptor number as the socket. This naturally causes chaos. The fix is
2456 not to close the socket in ip_connect(); the socket should be closed by
2457 the function that creates it. There was only one place in the code where
2458 this was missing, in the iplookup router, which I don't think anybody now
2459 uses, but I've fixed it anyway.
2461 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2462 well as to direct DNS lookups. Otherwise the handling of names in host
2463 lists is inconsistent and therefore confusing.
2469 PH/01 Two changes to the default runtime configuration:
2471 (1) Move the checks for relay_from_hosts and authenticated clients from
2472 after to before the (commented out) DNS black list checks.
2474 (2) Add control=submission to the relay_from_hosts and authenticated
2475 clients checks, on the grounds that messages accepted by these
2476 statements are most likely to be submissions.
2478 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2480 (1) Generate an error if the third argument for the ${prvs expansion is
2483 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2486 (3) Reset the variables that are obtained from the first argument of
2487 ${prvscheck and used in the second argument before leaving the code,
2488 because their memory is reclaimed, so using them afterwards may do
2491 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2492 one (it's much easier than Tom thought :-).
2494 (5) Because of (4), we can now allow for the use of $prvscheck_result
2495 inside the third argument.
2497 PH/03 For some reason, the default setting of PATH when running a command from
2498 a pipe transport was just "/usr/bin". I have changed it to
2501 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2502 anything to be listed in the output from -bV.
2504 PH/05 When a filter generated an autoreply, the entire To: header line was
2505 quoted in the delivery log line, like this:
2507 => >A.N.Other <ano@some.domain> <original@ddress> ...
2509 This has been changed so that it extracts the operative address. There
2510 may be more than one such address. If so, they are comma-separated, like
2513 => >ano@some.domain,ona@other.domain <original@ddress> ...
2515 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2516 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2517 not being looked up in the rDNS to get a host name, Exim was showing the
2518 IP address twice in Received: lines, even though the IP addresses were
2519 identical. For example:
2521 Received: from [1.2.3.4] (helo=[1.2.3.4])
2523 However, if the real host name was known, it was omitting the HELO data
2524 if it matched the actual IP address. This has been tidied up so that it
2525 doesn't show the same IP address twice.
2527 PH/07 When both +timestamp and +memory debugging was on, the value given by
2528 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2529 called by the debug printing, thereby overwriting the timestamp buffer.
2530 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2532 PH/08 When the original message was included in an autoreply transport, it
2533 always said "this is a copy of the message, including all the headers",
2534 even if body_only or headers_only was set. It now gives an appropriate
2537 PH/09 Applied a patch from the Sieve maintainer which:
2539 o fixes some comments
2540 o adds the (disabled) notify extension core
2541 o adds some debug output for the result of if/elsif tests
2542 o points to the current vacation draft in the documentation
2543 and documents the missing references header update
2547 o fixes a bug in processing the envelope test (when testing
2548 multiple envelope elements, the last element determinted the
2551 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2552 Electronic Mail") by including:
2554 Auto-submitted: auto-generated
2556 in the messages that it generates (bounce messages and others, such as
2557 warnings). In the case of bounce messages for non-SMTP mesages, there was
2558 also a typo: it was using "Auto_submitted" (underscore instead of
2559 hyphen). Since every message generated by Exim is necessarily in response
2560 to another message, thes have all been changed to:
2562 Auto-Submitted: auto-replied
2564 in accordance with these statements in the RFC:
2566 The auto-replied keyword:
2568 - SHOULD be used on messages sent in direct response to another
2569 message by an automatic process,
2571 - MUST NOT be used on manually-generated messages,
2573 - MAY be used on Delivery Status Notifications (DSNs) and Message
2574 Disposition Notifications (MDNs),
2576 - MUST NOT be used on messages generated by automatic or periodic
2577 processes, except for messages which are automatic responses to
2580 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2581 to the default Received: header definition.
2583 PH/12 Added log selector acl_warn_skipped (default on).
2585 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2586 variables because (a) they are in the wrong storage pool and (b) even if
2587 they were copied, it wouldn't work properly because of the caching.
2589 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2590 checking when decoding. Apparently there are clients that generate
2591 overlong encoded strings. Why am I not surprised?
2593 PH/15 If the first argument of "${if match_address" was not empty, but did not
2594 contain an "@" character, Exim crashed. Now it writes a panic log message
2595 and treats the condition as false.
2597 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2599 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2600 extension "envelope-auth". The code is finished and in agreement with
2601 other implementations, but there is no documentation so far and in fact,
2602 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2603 not changing the active code.
2605 Print executed "if" and "elsif" statements when debugging is used. This
2606 helps a great deal to understand what a filter does.
2608 Document more things not specified clearly in RFC3028. I had all this
2609 sorted out, when out of a sudden new issues came to my mind. Oops."
2611 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2614 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2615 canonical form (as documented). However, after a host name lookup from
2616 the IP address, check_host() was doing a simple string comparison with
2617 addresses acquired from the DNS when checking that the found name did
2618 have the original IP as one of its addresses. Since any found IPv6
2619 addresses are likely to be in abbreviated form, the comparison could
2620 fail. Luckily, there already exists a function for doing the comparison
2621 by converting both addresses to binary, so now that is used instead of
2622 the text comparison.
2624 PH/20 There was another similar case to PH/19, when a complete host name was
2625 given in a host list; looking up its IP address could give an abbreviated
2626 form, whereas the current host's name might or might not be abbreviated.
2627 The same fix has been applied.
2633 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2634 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2637 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2638 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2640 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2642 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2643 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2644 Also, for "Auto-Submitted", treat anything other than "no" as
2645 non-personal, in accordance with RFC 3834. (Previously it treated
2646 anything starting "auto-" as non-personal.)
2648 TF/01 The control=submission/name=... option had a problem with syntax
2649 errors if the name included a slash character. The /name= option
2650 now slurps the rest of the string, so it can include any characters
2651 but it must come last in the list of options (after /sender_retain
2654 PH/05 Some modifications to the interface to the fake nameserver for the new
2662 TK/01 Added the "success_on_redirect" address verification option. See
2663 NewStuff for rationale and an example.
2665 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2667 PH/02 Patch to exigrep to allow it to work on syslog lines.
2669 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2670 fread() to skip over the body file's header line, because in Cygwin the
2671 header line is locked and is inaccessible.
2673 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2674 co-exist for some time) to make it clear that it is the Exim ID that is
2675 referenced, not the Message-ID: header line.
2677 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2678 string_format() function, because snprintf() does not exist on all
2681 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2682 ${stat: expansion item.
2684 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2685 protocol synchronization error", to keep the pedants happy.
2687 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2688 well as for IRIX systems, when gcc is being used. See the host.c source
2691 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2693 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2696 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2697 transport and to the smtp transport in LMTP mode.
2699 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2701 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2702 run a filter in a subprocess. This could lead to confusion in subsequent
2703 lookups in the parent process. There should also be a search_tidyup() at
2704 the end of the subprocess.
2706 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2707 only if the host matched helo_try_verify_hosts, which caused the
2708 verification to occur when the EHLO/HELO command was issued. The ACL just
2709 tested the remembered result. Now, if a previous verification attempt has
2710 not happened, "verify = helo" does it there and then.
2712 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2714 TK/03 Fix log output including CR from clamd.
2716 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2717 exist provoked a memory error which could cause a segfault.
2719 PH/15 Installed PCRE 6.2
2721 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2723 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2724 of the problem. Specifically, suggested +O2 rather than +O1 for the
2727 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2729 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2730 record, the port was not correctly logged when the outgoing_port log
2731 selector was set (it logged the transort's default port).
2733 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2734 fallback_hosts, and "hosts" in the smtp transport.
2736 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2737 host errors such as "Connection refused".
2739 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2740 authentication with radiusclient 0.4.9:
2742 - Error returned from rc_read_config was caught wrongly
2743 - Username/password not passed on to radius server due to wrong length.
2745 The presumption is that some radiusclient API changes for 4.51/PH/17
2746 were not taken care of correctly. The code is still untested by me (my
2747 Linux distribution still has 0.3.2 of radiusclient), but it was
2748 contributed by a Radius user.
2750 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2751 expanding the "port" option of the smtp transport.
2753 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2754 while reading a MIME header. Thanks to Tom Hughes for a patch.
2756 PH/24 Include config.h inside local_scan.h so that configuration settings are
2759 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2760 that in an ACL for RCPT (for example), you can examine exactly what was
2763 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2764 commands, but it was not correctly comparing the address with the actual
2765 client host address. Thus, it would show the EHLO address in Received:
2766 header lines when this was not necessary.
2768 PH/27 Added the % operator to ${eval:}.
2770 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2771 it should be ignoring failures (because with -C, for example, it has lost
2772 privilege). It wasn't ignoring creation failures other than "already
2775 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2778 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2779 because an input line was too long, either on its own, or by virtue of
2780 too many continuations, the temporary file was not being removed, and the
2781 return code was incorrect.
2783 PH/31 Missing "BOOL" in function definition in filtertest.c.
2785 PH/32 Applied Sieve patches from the maintainer.
2787 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2789 PH/33 Added "verify = not_blind".
2791 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2792 Local/Makefile (with some defaults set). These are used in built scripts
2793 such as exicyclog, but they have never been used in the exim_install
2794 script (though there are many overriding facilities there). I have
2795 arranged that the exim_install script now takes note of these two
2798 PH/35 Installed configuration files for Dragonfly.
2800 PH/36 When a locally submitted message by a trusted user did not contain a
2801 From: header, and the sender address was obtained from -f or from an SMTP
2802 MAIL command, and the trusted user did not use -F to supply a sender
2803 name, $originator_name was incorrectly used when constructing a From:
2804 header. Furthermore, $originator_name was used for submission mode
2805 messages from external hosts without From: headers in a similar way,
2806 which is clearly wrong.
2808 PH/37 Added control=suppress_local_fixups.
2810 PH/38 When log_selector = +received_sender was set, and the addition of the
2811 sender made the log line's construction buffer exactly full, or one byte
2812 less than full, an overflow happened when the terminating "\n" was
2815 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2816 when the result of a list match is failure because a DNS lookup failed.
2818 PH/40 RM_COMMAND is now used in the building process.
2820 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2821 the "build-* directories that it finds.
2823 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2824 address in a domain literal was a prefix of an interface address.
2826 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2827 when verifying a sender address, unless rewrite_headers is false.
2829 PH/44 Wrote a long comment about why errors_to addresses are verified as
2830 recipients, not senders.
2832 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2833 the ratelimit ACL was added.
2835 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2837 PH/46 Added extra information about PostgreSQL errors to the error string.
2839 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2840 avoiding the need to install special zones in a real server. This is
2841 backwards compatible; if it can't find the fake resolver, it drops back.
2842 Thus, both old and new test suites can be run.
2844 TF/02 Added util/ratelimit.pl
2846 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2847 clock is set back in time.
2849 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2850 Candler <B.Candler@pobox.com>.
2852 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2853 OK for addresses that are the result of redirections.
2855 PH/48 A number of further additions for the benefit of the new test suite,
2856 including a fake gethostbyname() that interfaces to the fake DNS resolver
2859 TF/06 The fix for widen_domains has also been applied to qualify_single and
2860 search_parents which are the other dnslookup options that can cause
2863 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2866 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2868 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2869 service" if it encountered an MX record with an empty target hostname.
2870 The message is now "an MX or SRV record indicated no SMTP service".
2872 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2873 if the DNS of the sending site is misconfigured. This is quite a
2874 common situation. This change restores the behaviour of treating a
2875 helo verification defer as a failure.
2877 PH/51 If self=fail was set on a router, the bounce message did not include the
2878 actual error message.
2884 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2886 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2887 command itself ended in error, the underlying message about the transport
2888 filter timeout was being overwritten with the pipe command error. Now the
2889 underlying error message should be appended to the second error message.
2891 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2893 PH/02 Reduce the amount of output that "make" produces by default. Full output
2894 can still be requested.
2896 PH/03 The warning log line about a condition test deferring for a "warn" verb
2897 was being output only once per connection, rather than after each
2898 occurrence (because it was using the same function as for successful
2899 "warn" verbs). This seems wrong, so I have changed it.
2901 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2902 it should not have, which might have caused a crash in the right
2903 circumstances, but probably never did.
2905 PH/04 Installed a modified version of Tony Finch's patch to make submission
2906 mode fix the return path as well as the Sender: header line, and to
2907 add a /name= option so that you can make the user's friendly name appear
2910 TF/03 Added the control = fakedefer ACL modifier.
2912 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2913 Mark Lowes for thorough testing.
2915 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2917 TK/03 Merged latest SRS patch from Miles Wilton.
2919 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2920 with the definition in sysexits.h (which is #included earlier).
2921 Fortunately, Exim does not actually use EX_OK. The code used to try to
2922 preserve the sysexits.h value, by assumimg that macro definitions were
2923 scanned for macro replacements. I have been disabused of this notion,
2924 so now the code just undefines EX_OK before #including unistd.h.
2926 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2927 in the smtp transport. When a block could not be written in a single
2928 write() function, the timeout was being re-applied to each part-write.
2929 This seems wrong - if the receiver was accepting one byte at a time it
2930 would take for ever. The timeout is now adjusted when this happens. It
2931 doesn't have to be particularly precise.
2933 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2934 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2936 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2937 without a domain if the check to <postmaster@domain> fails.
2939 SC/01 Eximstats: added -xls and the ability to specify output files
2940 (patch written by Frank Heydlauf).
2942 SC/02 Eximstats: use FileHandles for outputing results.
2944 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2946 SC/04 Eximstats: fixed display of large numbers with -nvr option
2948 SC/05 Eximstats: fixed merging of reports with empty tables.
2950 SC/06 Eximstats: added the -include_original_destination flag
2952 SC/07 Eximstats: removed tabs and trailing whitespace.
2954 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2956 TK/06 MBOX spool code: Add real "From " MBOX separator line
2957 so the .eml file is really in mbox format (even though
2958 most programs do not really care). Patch from Alex Miller.
2960 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2961 The latter is generated from $received_to and is only set if the
2962 message has one envelope recipient. SA can use these headers,
2963 obviously out-of-the-box. Patch from Alex Miller.
2965 PH/08 The ${def test on a variable was returning false if the variable's
2966 value was "0", contrary to what the specification has always said!
2967 The result should be true unless the variable is empty.
2969 PH/09 The syntax error of a character other than { following "${if
2970 def:variable_name" (after optional whitespace) was not being diagnosed.
2971 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2972 accidental colon was present, for example, could give incorrect results.
2974 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2975 result is used (not including appendfile, where other changes are about
2978 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2979 This involved changing a lot of size variables from int to off_t. It
2980 should work with maildirs and everything.
2982 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2983 spamd dying while we are connected to it.
2985 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2988 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2989 function for BDB 4.3.
2991 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2993 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2994 the sources for PCRE have been re-arranged and more files are now
2997 PH/15 The code I had for printing potentially long long variables in PH/11
2998 above was not the best (it lost precision). The length of off_t variables
2999 is now inspected at build time, and an appropriate printing format (%ld
3000 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3001 to be "long long int" or "long int". This is needed for the internal
3002 formatting function string_vformat().
3004 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3005 the configuration file to be ":syslog", then the script "guesses" where
3006 the logs files are, rather than using the compiled in default. In our
3007 case the guess is not the same as the compiled default, so the script
3008 suddenly stopped working when I started to use syslog. The patch checks
3009 to see if log_file_path is "". If so, it attempts to read it from exim
3010 with no configuration file to get the compiled in version, before it
3011 falls back to the previous guessing code."
3013 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3014 implementing BATV in an Exim configuration. See NewStuff for the gory
3017 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3018 Makefile that are specific to HP-UX.
3020 PH/18 If the "use_postmaster" option was set for a recipient callout together
3021 with the "random" option, the postmaster address was used as the MAIL
3022 FROM address for the random test, but not for the subsequent recipient
3023 test. It is now used for both.
3025 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3026 patch removes a few documentation additions to RFC 3028, because the
3027 latest draft now contains them. It adds the new en;ascii-case comparator
3028 and a new error check for 8bit text in MIME parts. Comparator and
3029 require names are now matched exactly. I enabled the subaddress
3030 extension, but it is not well tested yet (read: it works for me)."
3032 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3033 rework some of the code of TK/09 above to avoid the hardwired use of
3034 "%lld" and "long long". Replaced the call to snprintf() with a call to
3037 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3038 records point to non-existent hosts", "retry timeout exceeded", and
3039 "retry time not reached for any host after a long failure period".
3041 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3042 experimental DomainKeys support:
3044 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3045 (2) On an error such as an illegally used "control", the wrong name for
3046 the control was given.
3048 These problems did NOT occur unless DomainKeys support was compiled.
3050 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3052 PH/24 Added ${if match_ip condition.
3054 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3055 databases so that it will be absolutely obvious if a crash occurs in the
3056 DB library. This is a regular occurrence (often caused by mis-matched
3059 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3060 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3061 picked up on a user's system that detects such things. There doesn't seem
3062 to be a gcc warning option for this - only an attribute that has to be
3063 put on the function's prototype. It seems that in Fedora Core 4 they have
3064 set this on a number of new functions. No doubt there will be more in due
3067 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3068 specify a transport. However, if an address that was verified by such a
3069 router was the subject of a callout, Exim crashed because it tried to
3070 read the rcpt_include_affixes from the non-existent transport. Now it
3071 just assumes that the setting of that option is false. This bug was
3072 introduced by 4.51/PH/31.
3074 PH/28 Changed -d+all to exclude +memory, because that information is very
3075 rarely of interest, but it makes the output a lot bigger. People tend to
3076 do -d+all out of habit.
3078 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3079 code in os-type was giving problems when libc.so lives in lib64, like on
3082 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3083 aren't the modern standard, and it seems that some systems' include files
3084 don't always have them. Exim was already checking for some of the newer
3085 ones like T_AAAA, and defining it itself. I've added checks for all the
3086 record types that Exim uses.
3088 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3089 not automatically generating a new one, as it is supposed to. This
3090 prevented TLS from working. If the file did exist, but contained invalid
3091 data, a new version was generated, as expected. It was only the case of a
3092 non-existent file that was broken.
3094 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3095 with a change in libdomainkeys > 0.64.
3097 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3098 from DNS. If the selector record carries the flag, it now has
3099 precedence over the domain-wide flag.
3101 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3103 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3104 the use of a port name that isn't defined in /etc/services) occurred, the
3105 message was deferred as in a normal delivery, and thus remained on the
3106 spool, instead of being failed because of the mua_wrapper setting. This
3107 is now fixed, and I tidied up some of the mua_wrapper messages at the
3110 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3111 the messages in a hash of arrays rather than using individual hashes.
3112 This is a bit cleaner and results in dramatic memory savings, albeit
3113 at a slight CPU cost.
3115 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3116 as requested by Marc Sherman.
3118 SC/10 Eximstats: added histograms for user specified patterns as requested
3121 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3123 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3124 fopen() in the content-scanning modules that did not already have it.
3130 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3131 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3133 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3135 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3137 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3138 to test Sieve filters that use "vacation".
3140 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3141 that changes the way the GnuTLS parameters are stored in the cache file.
3142 The new format can be generated externally. For backward compatibility,
3143 if the data in the cache doesn't make sense, Exim assumes it has read an
3144 old-format file, and it generates new data and writes a new file. This
3145 means that you can't go back to an older release without removing the
3148 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3149 work if there are any delivery delays because "one_time" forces the
3150 parent to be marked "delivered", so its unseen clone is never tried
3151 again. For this reason, Exim now forbids the simultaneous setting of
3154 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3155 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3156 entry if you want to know the details. The fix, however, neglected to
3157 consider the case when local delivery batching is involved. The test for
3158 "previously delivered" was not happening when checking to see if an
3159 address could be batched with a previous (undelivered) one; under
3160 certain circumstances this could lead to multiple deliveries to the same
3163 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3164 in its include files, and this causes problems building Exim.
3166 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3167 header_syntax) but Exim was just ignoring anything given after a slash.
3168 In particular, this caused confusion with an attempt to use "verify =
3169 reverse_host_lookup/defer_ok". An error is now given when options are
3170 supplied for verify items that do not have them. (Maybe reverse_host_
3171 lookup should have a defer_ok option, but that's a different point.)
3173 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3174 defined by RFC 821) to 2048, because there were problems with some AUTH
3175 commands, and RFC 1869 says the size should be increased for extended
3176 SMTP commands that take arguments.
3178 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3181 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3182 "unknown" error; now it says that the functionality isn't in the binary.
3184 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3185 an address' error message when a string expansion fails (syntax or
3186 whatever). Otherwise the password may appear in the log. Following change
3187 PH/42 below, there is no longer a chance of it appearing in a bounce
3190 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3192 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3193 of its addresses was ever tried. (Bugzilla bug #2.)
3195 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3196 the result incorrectly in the debug output. (It correctly added a newline
3197 to what was transported.)
3199 TF/01 Added $received_time.
3201 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3202 commented out examples of how to interface to a virus scanner and to
3203 SpamAssassin. Also added commented examples of av_scanner and
3204 spamd_address settings.
3206 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3207 and controls are allowed in which ACLs. There were a couple of minor
3208 errors. Some of the entries in the conditions table (which is a table of
3209 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3210 negation of where the condition IS allowed.
3212 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3214 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3215 header file does not have a version number, so I've had to invent a new
3216 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3217 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3218 radiusclient), but it was contributed by a Radius user.
3220 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3221 files or directories, for OpenSSL.
3223 PH/19 When an Exim process that is running as root has to create an Exim log
3224 file, it does so in a subprocess that runs as exim:exim so as to get the
3225 ownership right at creation (otherwise, other Exim processes might see
3226 the file with the wrong ownership). There was no test for failure of this
3227 fork() call, which would lead to the process getting stuck as it waited
3228 for a non-existent subprocess. Forks do occasionally fail when resources
3229 run out. I reviewed all the other calls to fork(); they all seem to check
3232 PH/20 When checking for unexpected SMTP input at connect time (before writing
3233 the banner), Exim was not dealing correctly with a non-positive return
3234 from the read() function. If the client had disconnected by this time,
3235 the result was a log entry for a synchronization error with an empty
3236 string after "input=" when read() returned zero. If read() returned -1
3237 (an event I could not check), uninitialized data bytes were printed.
3238 There were reports of junk text (parts of files, etc) appearing after
3241 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3243 PH/22 Added support for macro redefinition, and (re)definition in between
3244 driver and ACL definitions.
3246 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3247 forgetting to use the resulting value; it was using the unexpanded value.
3249 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3250 hadn't been configured. The fix is from Juergen Kreileder, who
3251 understands it better than I do:
3253 "Here's what I see happening with three configured cyrus_sasl
3254 authenticators configured (plain, login, cram-md5):
3256 On startup auth_cyrus_sasl_init() gets called for each of these.
3257 This means three calls to sasl_listmech() without a specified mech_list.
3258 => SASL tests which mechs of all available mechs actually work
3259 => three warnings about OTP not working
3260 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3262 With the patch, sasl_listmech() also gets called three times. But now
3263 SASL's mech_list option is set to the server_mech specified in the the
3264 authenticator. Or in other words, the answer from sasl_listmech()
3265 gets limited to just the mech you're testing for (which is different
3267 => the return list contains just 'plain' or 'login', 'cram-md5' or
3268 nothing depending on the value of ob->server_mech.
3270 I've just tested the patch: Authentication still works fine,
3271 unavailable mechs specified in the exim configuration are still
3272 caught, and the auth.log warnings about OTP are gone."
3274 PH/25 When debugging is enabled, the contents of the command line are added
3275 to the debugging output, even when log_selector=+arguments is not
3278 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3279 answer is "GNU", and only if the return is "GNU/something" is the answer
3282 PH/27 $acl_verify_message is now set immediately after the failure of a
3283 verification in an ACL, and so is available in subsequent modifiers. In
3284 particular, the message can be preserved by coding like this:
3286 warn !verify = sender
3287 set acl_m0 = $acl_verify_message
3289 Previously, $acl_verify_message was set only while expanding "message"
3290 and "log_message" when a very denied access.
3292 PH/28 Modified OS/os.c-Linux with
3294 -#ifndef OS_LOAD_AVERAGE
3295 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3297 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3298 nomenclature these days.)
3300 PH/29 Installed patch from the Sieve maintainer that adds the options
3301 sieve_useraddress and sieve_subaddress to the redirect router.
3303 PH/30 In these circumstances:
3304 . Two addresses routed to the same list of hosts;
3305 . First host does not offer TLS;
3306 . First host accepts first address;
3307 . First host gives temporary error to second address;
3308 . Second host offers TLS and a TLS session is established;
3309 . Second host accepts second address.
3310 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3311 and peerdn, if requested) that were in fact used only for the second
3314 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3315 attention to any local part prefix or suffix that was matched by the
3316 router that accepted the address. It now behaves in the same way as it
3317 does for delivery: the affixes are removed from the local part unless
3318 rcpt_include_affixes is set on the transport.
3320 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3321 timeout during the DATA phase of an incoming message.
3323 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3324 applied a patch sanctioned by the Sieve maintainer.
3326 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3327 the uid or gid is negative. A case of a negative gid caused this to be
3328 noticed. The fix allows for either to be negative.
3330 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3331 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3334 PH/36 The change PH/12 above was broken. Fixed it.
3336 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3337 the grounds that routing the same address twice would always produce the
3338 same answer. This might have been true once, but it is certainly no
3339 longer true now. Routing a child address may depend on the previous
3340 routing that produced that child. Some complicated redirection strategies
3341 went wrong when messages had multiple recipients, and made Exim's
3342 behaviour dependent on the order in which the addresses were given.
3344 I have moved the duplicate checking until after the routing is complete.
3345 Exim scans the addresses that are assigned to local and remote
3346 transports, and removes any duplicates. This means that more work will be
3347 done, as duplicates will always all be routed, but duplicates are
3348 presumably rare, so I don't expect this is of any significance.
3350 For deliveries to pipes, files, and autoreplies, the duplicate checking
3351 still happens during the routing process, since they are not going to be
3354 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3355 It corrects a timeout issue with spamd. This is Ian's comment: "The
3356 background is that sometimes spamd either never reads data from a
3357 connection it has accepted, or it never writes response data. The exiscan
3358 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3359 blindly assumes that writes won't block so it may never time out."
3361 PH/39 Allow G after quota size as well as K and M.
3363 PH/40 The value set for $authenticated_id in an authenticator may not contain
3364 binary zeroes or newlines because the value is written to log lines and
3365 to spool files. There was no check on this. Now the value is run through
3366 the string_printing() function so that such characters are converted to
3367 printable escape sequences.
3369 PH/41 $message_linecount is a new variable that contains the total number of
3370 lines in the message. Compare $body_linecount, which is the count for the
3373 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3374 bounce and delay warning messages, except in certain special cases, which
3377 (a) An SMTP error message from a remote host;
3378 (b) A message specified in a :fail: redirection;
3379 (c) A message specified in a "fail" command in a system filter;
3380 (d) A message specified in a FAIL return from the queryprogram router;
3381 (e) A message specified by the cannot_route_message router option.
3383 In these cases only, Exim does include the error details in bounce and
3384 warning messages. There are also a few cases where bland messages such
3385 as "unrouteable address" or "local delivery error" are given.
3387 PH/43 $value is now also set for the "else" part of a ${run expansion.
3389 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3390 being worked on, but at least Exim now implements the latest version to
3393 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3394 process to complete was treated as a delivery failure, a timeout while
3395 writing the message to the pipe was logged, but erroneously treated as a
3396 successful delivery. Such timeouts include transport filter timeouts. For
3397 consistency with the overall process timeout, these timeouts are now
3398 treated as errors, giving rise to delivery failures by default. However,
3399 there is now a new Boolean option for the pipe transport called
3400 timeout_defer, which, if set TRUE, converts the failures into defers for
3401 both kinds of timeout. A transport filter timeout is now identified in
3404 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3405 systems where "make" and "gmake" are different, calling "gmake" at top
3406 level broke things. I've arranged for the value of $(MAKE) to be passed
3407 from the Makefile to this script so that it can call the same version of
3411 A note about Exim versions 4.44 and 4.50
3412 ----------------------------------------
3414 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3415 changes of various kinds. As a consequence, a big documentation update was
3416 needed. This delayed the release for rather longer than seemed good, especially
3417 in the light of a couple of (minor) security issues. Therefore, the changes
3418 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3419 release. So 4.44 and 4.50 are in effect two different branches that both start
3422 I have left the 4.50 change log unchanged; it contains all the changes since
3423 4.43. The change log for 4.44 is below; many of its items are identical to
3424 those for 4.50. This seems to be the most sensible way to preserve the
3425 historical information.
3431 1. Minor wording change to the doc/README.SIEVE file.
3433 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3434 computation of the current number of files was incorrect.
3436 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3437 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3440 4. Give more explanation in the error message when the command for a transport
3441 filter fails to execute.
3443 5. There are several places where Exim runs a non-Exim command in a
3444 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3445 was being done only for the command run by the queryprogram router. It is
3446 now done for all such subprocesses. The other cases are: ${run, transport
3447 filters, and the commands run by the lmtp and pipe transports.
3449 6. Added CONFIGURE_GROUP build-time option.
3451 7. Some older OS have a limit of 256 on the maximum number of file
3452 descriptors. Exim was using setrlimit() to set 1000 as a large value
3453 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3454 systems. I've change it so that if it can't get 1000, it tries for 256.
3456 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3457 was an oversight, and furthermore, ever since the addition of extra
3458 controls (e.g. 4.43/32), the checks on when to allow different forms of
3459 "control" were broken. There should now be diagnostics for all cases when a
3460 control that does not make sense is encountered.
3462 9. Added the /retain_sender option to "control=submission".
3464 10. $recipients is now available in the predata ACL (oversight).
3466 11. Tidy the search cache before the fork to do a delivery from a message
3467 received from the command line. Otherwise the child will trigger a lookup
3468 failure and thereby defer the delivery if it tries to use (for example) a
3469 cached ldap connection that the parent has called unbind on.
3471 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3472 of $address_data from the recipient verification was clobbered by the
3473 sender verification.
3475 13. The value of address_data from a sender verification is now available in
3476 $sender_address_data in subsequent conditions in the ACL statement.
3478 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3480 15. Added a new option "connect=<time>" to callout options, to set a different
3483 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3484 was its contents. (It was OK if the option was not defined at all.)
3486 17. A "Completed" log line is now written for messages that are removed from
3487 the spool by the -Mrm option.
3489 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3490 information about exactly what failed.
3492 19. Added -dd to debug only the daemon process.
3494 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3495 handles timeouts, both on the server side and network timeouts. Renamed the
3496 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3498 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3499 It is now set to "smtps".
3501 22. $host_address is now set to the target address during the checking of
3502 ignore_target_hosts.
3504 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3505 being passed; this would have caused $sender_host_name to have been used if
3506 matching the list had actually called for a host name (not very likely,
3507 since this list is usually IP addresses). A host name is now passed as
3510 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3511 code that specifies a non-restarting handler (typically sigaction() in
3512 modern systems) in an attempt to fix a rare and obscure crash bug.
3514 25. Narrowed the window for a race in the daemon that could cause it to ignore
3515 SIGCHLD signals. This is not a major problem, because they are used only to
3516 wake it up if nothing else does.
3518 26. A malformed maildirsize file could cause Exim to calculate negative values
3519 for the mailbox size or file count. Odd effects could occur as a result.
3520 The maildirsize information is now recalculated if the size or filecount
3523 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3524 support for a long time. Removed HAVE_SYS_VFS_H.
3526 28. Installed the latest version of exipick from John Jetmore.
3528 29. In an address list, if the pattern was not a regular expression, an empty
3529 subject address (from a bounce message) matched only if the pattern was an
3530 empty string. Non-empty patterns were not even tested. This was the wrong
3531 because it is perfectly reasonable to use an empty address as part of a
3532 database query. An empty address is now tested by patterns that are
3533 lookups. However, all the other forms of pattern expect the subject to
3534 contain a local part and a domain, and therefore, for them, an empty
3535 address still always fails if the pattern is not itself empty.
3537 30. Exim went into a mad DNS loop when attempting to do a callout where the
3538 host was specified on an smtp transport, and looking it up yielded more
3539 than one IP address.
3541 31. Re-factored the code for checking spool and log partition space into a
3542 function that finds that data and another that does the check. The former
3543 is then used to implement four new variables: $spool_space, $log_space,
3544 $spool_inodes, and $log_inodes.
3546 32. The RFC2047 encoding function was originally intended for short strings
3547 such as real names; it was not keeping to the 75-character limit for
3548 encoded words that the RFC imposes. It now respects the limit, and
3549 generates multiple encoded words if necessary. To be on the safe side, I
3550 have increased the buffer size for the ${rfc2047: expansion operator from
3553 33. It is now permitted to omit both strings after an "if" condition; if the
3554 condition is true, the result is "true". As before, when the second string
3555 is omitted, a false condition yields an empty string. This makes it less
3556 cumbersome to write custom ACL and router conditions.
3558 34. Failure to deliver a bounce message always caused it to be frozen, even if
3559 there was an errors_to setting on the router. The errors_to setting is now
3562 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3563 canonical form (fully expanded) before being placed in
3564 $sender_host_address.
3566 36. The table in the code that translates DNS record types into text (T_A to
3567 "A" for instance) was missing entries for NS and CNAME. It is just possible
3568 that this could have caused confusion if both these types were looked up
3569 for the same domain, because the text type is used as part of Exim's
3570 per-process caching. But the chance of anyone hitting this buglet seems
3573 37. The dnsdb lookup has been extended in a number of ways.
3575 (1) There is a new type, "zns", which walks up the domain tree until it
3576 finds some nameserver records. It should be used with care.
3578 (2) There is a new type, "mxh", which is like "mx" except that it returns
3579 just the host names, not the priorities.
3581 (3) It is now possible to give a list of domains (or IP addresses) to be
3582 looked up. The behaviour when one of the lookups defers can be
3583 controlled by a keyword.
3585 (4) It is now possible to specify the separator character for use when
3586 multiple records are returned.
3588 38. The dnslists ACL condition has been extended: it is now possible to supply
3589 a list of IP addresses and/or domains to be looked up in a particular DNS
3592 39. Added log_selector=+queue_time_overall.
3594 40. When running the queue in the test harness, wait just a tad after forking a
3595 delivery process, to get repeatability of debugging output.
3597 41. Include certificate and key file names in error message when GnuTLS fails
3598 to set them up, because the GnuTLS error message doesn't include the name
3599 of the failing file when there is a problem reading it.
3601 42. Allow both -bf and -bF in the same test run.
3603 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3605 44. The "Exiscan patch" is now merged into the mainline Exim source.
3607 45. Sometimes the final signoff response after QUIT could fail to get
3608 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3609 < 0 before doing a fflush(). This bug looks as though it goes back to the
3610 introduction of TLS in release 3.20, but "sometimes" must have been rare
3611 because the tests only now provoked it.
3613 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3614 (this can affect the format of dates).
3616 47. exim_tidydb, when checking for the continued existence of a message for
3617 which it has found a message-specific retry record, was not finding
3618 messages that were in split spool directories. Consequently, it was
3619 deleting retry records that should have stayed in existence.
3621 48. Steve fixed some bugs in eximstats.
3623 49. The SPA authentication driver was not abandoning authentication and moving
3624 on to the next authenticator when an expansion was forced to fail,
3625 contradicting the general specification for all authenticators. Instead it
3626 was generating a temporary error. It now behaves as specified.
3628 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3629 (the order specifies the preference for clients). The order is now AES256,
3630 AES128, 3DES, ARCFOUR128.
3632 51. Small patch to Sieve code - explicitly set From: when generating an
3635 52. Exim crashed if a remote delivery caused a very long error message to be
3636 recorded - for instance if somebody sent an entire SpamAssassin report back
3637 as a large number of 550 error lines. This bug was coincidentally fixed by
3638 increasing the size of one of Exim's internal buffers (big_buffer) that
3639 happened as part of the Exiscan merge. However, to be on the safe side, I
3640 have made the code more robust (and fixed the comments that describe what
3643 53. Now that there can be additional text after "Completed" in log lines (if
3644 the queue_time_overall log selector is set), a one-byte patch to exigrep
3645 was needed to allow it to recognize "Completed" as not the last thing in
3648 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3649 patch that reportedly fixes this has been added. I am not expert enough to
3650 create a test for it. This is what the patch creator wrote:
3652 "I found a little strange behaviour of ldap code when working with
3653 Windows 2003 AD Domain, where users was placed in more than one
3654 Organization Units. When I tried to give exim partial DN, the exit code
3655 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3656 But simultaneously result of request was absolutely normal ldap result,
3657 so I produce this patch..."
3659 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3660 REFERENCE, so I have modified the code to exclude the patch when that macro
3663 55. Some experimental protocols are using DNS PTR records for new purposes. The
3664 keys for these records are domain names, not reversed IP addresses. The
3665 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3666 leaves it alone. Component reversal etc. now happens only for IP addresses.
3669 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3671 57. Double the size of the debug message buffer (to 2048) so that more of very
3672 long debug lines gets shown.
3674 58. The exicyclog utility now does better if the number of log files to keep
3675 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3677 59. Two changes related to the smtp_active_hostname option:
3679 (1) $smtp_active_hostname is now available as a variable.
3680 (2) The default for smtp_banner uses $smtp_active_hostname instead
3681 of $primary_hostname.
3683 60. The host_aton() function is supposed to be passed a string that is known
3684 to be a valid IP address. However, in the case of IPv6 addresses, it was
3685 not checking this. This is a hostage to fortune. Exim now panics and dies
3686 if the condition is not met. A case was found where this could be provoked
3687 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3688 components; fortuitously, this particular loophole had already been fixed
3689 by change 4.50/55 above.
3691 If there are any other similar loopholes, the new check in host_aton()
3692 itself should stop them being exploited. The report I received stated that
3693 data on the command line could provoke the exploit when Exim was running as
3694 exim, but did not say which command line option was involved. All I could
3695 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3696 running as the user.
3699 61. There was a buffer overflow vulnerability in the SPA authentication code
3700 (which came originally from the Samba project). I have added a test to the
3701 spa_base64_to_bits() function which I hope fixes it.
3704 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3705 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3707 63. The daemon start-up calls getloadavg() while still root for those OS that
3708 need the first call to be done as root, but it missed one case: when
3709 deliver_queue_load_max is set with deliver_drop_privilege. This is
3710 necessary for the benefit of the queue runner, because there is no re-exec
3711 when deliver_drop_privilege is set.
3713 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3714 This has been fixed.
3716 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3717 was in use, was not putting the data itself into the right store pool;
3718 consequently, it could be overwritten for a subsequent message in the same
3719 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3722 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3724 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3725 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3726 the function were treating the return as a boolean value, which happened to
3727 work because 0=false and not-0=true, but is not correct code.
3729 68. The host_aton() function was not handling scoped IPv6 addresses (those
3730 with, for example, "%eth0" on the end) correctly.
3732 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3733 negated items (that is, ~something) in unsigned ints. Some compilers
3734 apparently mutter when there is no cast.
3736 70. If an address verification called from an ACL failed, and did not produce a
3737 user-specific message (i.e. there was only a "system" message), nothing was
3738 put in $acl_verify_message. In this situation, it now puts the system
3741 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3742 session; change 4.31/43 added the unwanted input to the log line - except
3743 that it did not do this in the start of session case. It now does.
3745 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3746 This is wrong and can cause the other end to generate a synchronization
3747 error if it is another Exim or anything else that does the synchronization
3748 check. A QUIT command is no longer sent after a timeout.
3750 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3751 during host lookups.
3753 74. The defer_ok option of callout verification was not working if it was used
3754 when verifying addresses in header lines, that is, for this case:
3756 verify = header_sender/callout=defer_ok
3758 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3759 those file descriptors could be used for SMTP connections. If anything
3760 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3761 could be sent to the SMTP client, causing chaos. The daemon now opens
3762 stdin, stdout, and stderr to /dev/null when it puts itself into the
3765 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3766 log by default. The user can override this with suitable Perl magic.
3768 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3769 the log message when discard triggers, was not working for the DATA ACL or
3770 for the non-SMTP ACL.
3772 78. Error message wording change in sieve.c.
3774 79. If smtp_accept_max_per_host was set, the number of connections could be
3775 restricted to fewer than expected, because the daemon was trying to set up
3776 a new connection before checking whether the processes handling previous
3777 connections had finished. The check for completed processes is now done
3778 earlier. On busy systems, this bug wouldn't be noticed because something
3779 else would have woken the daemon, and it would have reaped the completed
3782 80. If a message was submitted locally by a user whose login name contained one
3783 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3784 It caused a spool format error. I have fixed the spool reading code. A
3785 related problem was that the "from" clause in the Received: line became
3786 illegal because of the space(s). It is now covered by ${quote_local_part.
3788 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3791 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3793 83. Patch from Sieve maintainer to fix unterminated string problem in
3794 "vacation" handling.
3796 84. Some minor changes to the Linux configuration files to help with other
3797 OS variants using glibc.
3799 85. One more patch for Sieve to update vacation handling to latest spec.
3802 ----------------------------------------------------
3803 See the note above about the 4.44 and 4.50 releases.
3804 ----------------------------------------------------
3810 1. Change 4.43/35 introduced a bug that caused file counts to be
3811 incorrectly computed when quota_filecount was set in an appendfile
3814 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3815 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3818 3. Give more explanation in the error message when the command for a transport
3819 filter fails to execute.
3821 4. There are several places where Exim runs a non-Exim command in a
3822 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3823 was being done only for the command run by the queryprogram router. It is
3824 now done for all such subprocesses. The other cases are: ${run, transport
3825 filters, and the commands run by the lmtp and pipe transports.
3827 5. Some older OS have a limit of 256 on the maximum number of file
3828 descriptors. Exim was using setrlimit() to set 1000 as a large value
3829 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3830 systems. I've change it so that if it can't get 1000, it tries for 256.
3832 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3833 was an oversight, and furthermore, ever since the addition of extra
3834 controls (e.g. 4.43/32), the checks on when to allow different forms of
3835 "control" were broken. There should now be diagnostics for all cases when a
3836 control that does not make sense is encountered.
3838 7. $recipients is now available in the predata ACL (oversight).
3840 8. Tidy the search cache before the fork to do a delivery from a message
3841 received from the command line. Otherwise the child will trigger a lookup
3842 failure and thereby defer the delivery if it tries to use (for example) a
3843 cached ldap connection that the parent has called unbind on.
3845 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3846 of $address_data from the recipient verification was clobbered by the
3847 sender verification.
3849 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3850 was its contents. (It was OK if the option was not defined at all.)
3852 11. A "Completed" log line is now written for messages that are removed from
3853 the spool by the -Mrm option.
3855 12. $host_address is now set to the target address during the checking of
3856 ignore_target_hosts.
3858 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3859 being passed; this would have caused $sender_host_name to have been used if
3860 matching the list had actually called for a host name (not very likely,
3861 since this list is usually IP addresses). A host name is now passed as
3864 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3865 code that specifies a non-restarting handler (typically sigaction() in
3866 modern systems) in an attempt to fix a rare and obscure crash bug.
3868 15. Narrowed the window for a race in the daemon that could cause it to ignore
3869 SIGCHLD signals. This is not a major problem, because they are used only to
3870 wake it up if nothing else does.
3872 16. A malformed maildirsize file could cause Exim to calculate negative values
3873 for the mailbox size or file count. Odd effects could occur as a result.
3874 The maildirsize information is now recalculated if the size or filecount
3877 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3878 support for a long time. Removed HAVE_SYS_VFS_H.
3880 18. Updated exipick to current release from John Jetmore.
3882 19. Allow an empty sender to be matched against a lookup in an address list.
3883 Previously the only cases considered were a regular expression, or an
3886 20. Exim went into a mad DNS lookup loop when doing a callout where the
3887 host was specified on the transport, if the DNS lookup yielded more than
3890 21. The RFC2047 encoding function was originally intended for short strings
3891 such as real names; it was not keeping to the 75-character limit for
3892 encoded words that the RFC imposes. It now respects the limit, and
3893 generates multiple encoded words if necessary. To be on the safe side, I
3894 have increased the buffer size for the ${rfc2047: expansion operator from
3897 22. Failure to deliver a bounce message always caused it to be frozen, even if
3898 there was an errors_to setting on the router. The errors_to setting is now
3901 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3902 canonical form (fully expanded) before being placed in
3903 $sender_host_address.
3905 24. Updated eximstats to version 1.33
3907 25. Include certificate and key file names in error message when GnuTLS fails
3908 to set them up, because the GnuTLS error message doesn't include the name
3909 of the failing file when there is a problem reading it.
3911 26. Expand error message when OpenSSL has problems setting up cert/key files.
3914 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3915 (this can affect the format of dates).
3917 28. exim_tidydb, when checking for the continued existence of a message for
3918 which it has found a message-specific retry record, was not finding
3919 messages that were in split spool directories. Consequently, it was
3920 deleting retry records that should have stayed in existence.
3922 29. eximstats updated to version 1.35
3923 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3924 1.35 - bugfix such that pie charts by volume are generated correctly
3926 30. The SPA authentication driver was not abandoning authentication and moving
3927 on to the next authenticator when an expansion was forced to fail,
3928 contradicting the general specification for all authenticators. Instead it
3929 was generating a temporary error. It now behaves as specified.
3931 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3932 (the order specifies the preference for clients). The order is now AES256,
3933 AES128, 3DES, ARCFOUR128.
3935 31. Small patch to Sieve code - explicitly set From: when generating an
3938 32. Exim crashed if a remote delivery caused a very long error message to be
3939 recorded - for instance if somebody sent an entire SpamAssassin report back
3940 as a large number of 550 error lines. This bug was coincidentally fixed by
3941 increasing the size of one of Exim's internal buffers (big_buffer) that
3942 happened as part of the Exiscan merge. However, to be on the safe side, I
3943 have made the code more robust (and fixed the comments that describe what
3946 33. Some experimental protocols are using DNS PTR records for new purposes. The
3947 keys for these records are domain names, not reversed IP addresses. The
3948 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3949 leaves it alone. Component reversal etc. now happens only for IP addresses.
3952 34. The host_aton() function is supposed to be passed a string that is known
3953 to be a valid IP address. However, in the case of IPv6 addresses, it was
3954 not checking this. This is a hostage to fortune. Exim now panics and dies
3955 if the condition is not met. A case was found where this could be provoked
3956 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3957 components; fortuitously, this particular loophole had already been fixed
3958 by change 4.50/55 or 4.44/33 above.
3960 If there are any other similar loopholes, the new check in host_aton()
3961 itself should stop them being exploited. The report I received stated that
3962 data on the command line could provoke the exploit when Exim was running as
3963 exim, but did not say which command line option was involved. All I could
3964 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3965 running as the user.
3968 35. There was a buffer overflow vulnerability in the SPA authentication code
3969 (which came originally from the Samba project). I have added a test to the
3970 spa_base64_to_bits() function which I hope fixes it.
3973 36. The daemon start-up calls getloadavg() while still root for those OS that
3974 need the first call to be done as root, but it missed one case: when
3975 deliver_queue_load_max is set with deliver_drop_privilege. This is
3976 necessary for the benefit of the queue runner, because there is no re-exec
3977 when deliver_drop_privilege is set.
3979 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3980 was in use, was not putting the data itself into the right store pool;
3981 consequently, it could be overwritten for a subsequent message in the same
3982 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3985 38. Sometimes the final signoff response after QUIT could fail to get
3986 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3987 < 0 before doing a fflush(). This bug looks as though it goes back to the
3988 introduction of TLS in release 3.20, but "sometimes" must have been rare
3989 because the tests only now provoked it.
3995 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3996 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3997 Now it returns an integer. A number of calls were still expecting a T/F
3998 return. Fortuitously, in all cases, the tests worked in OK situations,
3999 which is the norm. However, things would have gone wrong on any write
4000 failures on the smtp file descriptor. This function is used when sending
4001 messages over SMTP and also when doing verify callouts.
4003 2. When Exim is called to do synchronous delivery of a locally submitted
4004 message (the -odf or -odi options), it no longer closes stderr before doing
4007 3. Implemented the mua_wrapper option.
4009 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4011 5. Implemented the functions header_remove(), header_testname(),
4012 header_add_at_position(), and receive_remove_recipient(), and exported them
4015 6. If an ACL "warn" statement specified the addition of headers, Exim already
4016 inserted X-ACL-Warn: at the start if there was no header name. However, it
4017 was not making this test for the second and subsequent header lines if
4018 there were newlines in the string. This meant that an invalid header could
4019 be inserted if Exim was badly configured.
4021 7. Allow an ACL "warn" statement to add header lines at the start or after all
4022 the Received: headers, as well as at the end.
4024 8. Added the rcpt_4xx retry error code.
4026 9. Added postmaster_mailfrom=xxx to callout verification option.
4028 10. Added mailfrom=xxxx to the callout verification option, for verify=
4031 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4032 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4034 12. Inserted some casts to stop certain compilers warning when using pointer
4035 differences as field lengths or precisions in printf-type calls (mostly
4036 affecting debugging statements).
4038 13. Added optional readline() support for -be (dynamically loaded).
4040 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4041 same clock tick as a message's arrival, so that its received time was the
4042 same as the "first fail" time on the retry record, and that message
4043 remained on the queue past the ultimate address timeout, every queue runner
4044 would try a delivery (because it was past the ultimate address timeout) but
4045 after another failure, the ultimate address timeout, which should have then
4046 bounced the address, did not kick in. This was a "< instead of <=" error;
4047 in most cases the first failure would have been in the next clock tick
4048 after the received time, and all would be well.
4050 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4051 being recognized when the domain list was tested by the match_domain
4052 condition in an expansion string.
4054 16. Added the ${str2b64: operator.
4056 17. Exim was always calling setrlimit() to set a large limit for the number of
4057 processes, without checking whether the existing limit was already
4058 adequate. (It did check for the limit on file descriptors.) Furthermore,
4059 errors from getrlimit() and setrlimit() were being ignored. Now they are
4060 logged to the main and panic logs, but Exim does carry on, to try to do its
4061 job under whatever limits there are.
4063 18. Imported PCRE 5.0.
4065 19. Trivial typo in log message " temporarily refused connection" (the leading
4068 20. If the log selector return_path_on_delivery was set and an address was
4069 redirected to /dev/null, the delivery process crashed because it assumed
4070 that a return path would always be set for a "successful" delivery. In this
4071 case, the whole delivery is bypassed as an optimization, and therefore no
4074 21. Internal re-arrangement: the function for sending a challenge and reading
4075 a response while authentication was assuming a zero-terminated challenge
4076 string. It's now changed to take a pointer and a length, to allow for
4077 binary data in such strings.
4079 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4081 23. Exim was not respecting finduser_retries when seeking the login of the
4082 uid under which it was called; it was always trying 10 times. (The default
4083 setting of finduser_retries is zero.) Also, it was sleeping after the final
4084 failure, which is pointless.
4086 24. Implemented tls_on_connect_ports.
4088 25. Implemented acl_smtp_predata.
4090 26. If the domain in control=submission is set empty, Exim assumes that the
4091 authenticated id is a complete email address when it generates From: or
4092 Sender: header lines.
4094 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4095 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4096 chown and chgrp in /bin and hostname in /usr/bin.
4098 28. Exim was keeping the "process log" file open after each use, just as it
4099 does for the main log. This opens the possibility of it remaining open for
4100 long periods when the USR1 signal hits a daemon. Occasional processlog
4101 errors were reported, that could have been caused by this. Anyway, it seems
4102 much more sensible not to leave this file open at all, so that is what now
4105 29. The long-running daemon process does not normally write to the log once it
4106 has entered its main loop, and it closes the log before doing so. This is
4107 so that log files can straightforwardly be renamed and moved. However,
4108 there are a couple of unusual error situations where the daemon does write
4109 log entries, and I had neglected to close the log afterwards.
4111 30. The text of an SMTP error response that was received during a remote
4112 delivery was being truncated at 512 bytes. This is too short for some of
4113 the long messages that one sometimes sees. I've increased the limit to
4116 31. It is now possible to make retry rules that apply only when a message has a
4117 specific sender, in particular, an empty sender.
4119 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4120 it possible to be selective about when SMTP synchronization is enforced.
4122 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4124 32. Implemented hosts_connection_nolog.
4126 33. Added an ACL for QUIT.
4128 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4131 35. Added mailbox_size and mailbox_filecount to appendfile.
4133 36. Added control = no_multiline_responses to ACLs.
4135 37. There was a bug in the logic of the code that waits for the clock to tick
4136 in the case where the clock went backwards by a substantial amount such
4137 that the microsecond fraction of "now" was more than the microsecond
4138 fraction of "then" (but the whole seconds number was less).
4140 38. Added support for the libradius Radius client library this is found on
4141 FreeBSD (previously only the radiusclient library was supported).
4147 1. When certain lookups returned multiple values in the form name=value, the
4148 quoting of the values was not always being done properly. Specifically:
4149 (a) If the value started with a double quote, but contained no whitespace,
4151 (b) If the value contained whitespace other than a space character (i.e.
4152 tabs or newlines or carriage returns) it was not quoted.
4153 This fix has been applied to the mysql and pgsql lookups by writing a
4154 separate quoting function and calling it from the lookup code. The fix
4155 should probably also be applied to nisplus, ibase and oracle lookups, but
4156 since I cannot test any of those, I have not disturbed their existing code.
4158 2. A hit in the callout cache for a specific address caused a log line with no
4159 reason for rejecting RCPT. Now it says "Previous (cached) callout
4160 verification failure".
4162 3. There was an off-by-one bug in the queryprogram router. An over-long
4163 return line was truncated at 256 instead of 255 characters, thereby
4164 overflowing its buffer with the terminating zero. As well as fixing this, I
4165 have increased the buffer size to 1024 (and made a note to document this).
4167 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4168 when Exim is waiting for an SMTP response from a remote server, Exim
4169 restarts its select() call on the socket, thereby resetting its timeout.
4170 This is not a problem when such interrupts are rare. Somebody set up a cron
4171 job to run exiwhat every 2 minutes, which is less than the normal select()
4172 timeout (5 or 10 minutes). This meant that the select() timeout never
4173 kicked in because it was always reset. I have fixed this by comparing the
4174 time when an interrupt arrives with the time at the start of the first call
4175 to select(). If more time than the timeout has elapsed, the interrupt is
4176 treated as a timeout.
4178 5. Some internal re-factoring in preparation for the addition of Sieve
4179 extensions (by MH). In particular, the "personal" test is moved to a
4180 separate function, and given an option for scanning Cc: and Bcc: (which is
4181 not set for Exim filters).
4183 6. When Exim created an email address using the login of the caller as the
4184 local part (e.g. when creating a From: or Sender: header line), it was not
4185 quoting the local part when it contained special characters such as @.
4187 7. Installed new OpenBSD configuration files.
4189 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4190 try to make them clearer.
4192 9. Callout options, other than the timeout value, were being ignored when
4193 verifying sender addresses in header lines. For example, when using
4195 verify = header_sender/callout=no_cache
4197 the cache was (incorrectly) being used.
4199 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4200 only the Cygwin environment.
4202 11. When return_path_on_delivery was set as a log selector, if different remote
4203 addresses in the same message used different return paths and parallel
4204 remote delivery occurred, the wrong values would sometimes be logged.
4205 (Whenever a remote delivery process finished, the return path value from
4206 the most recently started remote delivery process was logged.)
4208 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4209 lines when AUTH and/or TLS are in use. This is the "received protocol"
4210 field. Exim used to use "asmtp" for authenticated SMTP, without any
4211 indication (in the protocol name) for TLS use. Now it follows the RFC and
4212 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4213 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4214 names appear in log lines as well as in Received: header lines.
4216 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4217 extensions, and comparison tests, and to fix some bugs.
4219 14. Changes to the "personal" filter test:
4221 (1) The test was buggy in that it was just doing the equivalent of
4222 "contains" tests on header lines. For example, if a user's address was
4223 anne@some.where, the "personal" test would incorrectly be true for
4225 To: susanne@some.where
4227 This test is now done by extracting each address from the header in turn,
4228 and checking the entire address. Other tests that are part of "personal"
4229 are now done using regular expressions (for example, to check local parts
4230 of addresses in From: header lines).
4232 (2) The list of non-personal local parts in From: addresses has been
4233 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4234 taken from the Sieve specification recommendations.
4236 (3) If the message contains any header line starting with "List-" it is
4237 treated as non-personal.
4239 (4) The test for "circular" in the Subject: header line has been removed
4240 because it now seems ill-conceived.
4242 15. Minor typos in src/EDITME comments corrected.
4244 16. Installed latest exipick from John Jetmore.
4246 17. If headers_add on a router specified a text string that was too long for
4247 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4248 of string_sprintf() is now avoided.
4250 18. $message_body_size was not set (it was always zero) when running the DATA
4251 ACL and the local_scan() function.
4253 19. For the "mail" command in an Exim filter, no default was being set for
4254 the once_repeat time, causing a random time value to be used if "once" was
4255 specified. (If the value happened to be <= 0, no repeat happened.) The
4256 default is now 0s, meaning "never repeat". The "vacation" command was OK
4257 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4258 (I found it when inspecting the code).
4260 20. There is now an overall timeout for performing a callout verification. It
4261 defaults to 4 times the callout timeout, which applies to individual SMTP
4262 commands during the callout. The overall timeout applies when there is more
4263 than one host that can be tried. The timeout is checked before trying the
4264 next host. This prevents very long delays if there are a large number of
4265 hosts and all are timing out (e.g. when the network connections are timing
4266 out). The value of the overall timeout can be changed by specifying an
4267 additional sub-option for "callout", called "maxwait". For example:
4269 verify = sender/callout=5s,maxwait=20s
4271 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4272 to the end before writing, but this should make it even safer).
4274 22. Exim was forgetting that it had advertised PIPELINING for the second and
4275 subsequent messages on an SMTP connection. It was also not resetting its
4276 memory on STARTTLS and an internal HELO.
4278 23. When Exim logs an SMTP synchronization error within a session, it now
4279 records whether PIPELINING has been advertised or not.
4281 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4282 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4283 rather than long int.
4285 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4287 26. Added the never_mail option to autoreply.
4293 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4294 crash if the getsockname() call failed; this can happen if a connection is
4295 closed very soon after it is established. The problem was simply in the
4296 order in which certain operations were done, causing Exim to try to write
4297 to the SMTP stream before it had set up the file descriptor. The bug has
4298 been fixed by making things happen in the correct order.
4304 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4305 before the connection was closed, thus losing the rejection response.
4307 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4308 some early Solaris releases, but causes trouble in current releases where
4309 socklen_t is defined.
4311 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4314 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4317 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4318 that releasing the top of it at the end releases what was used for sub-
4319 expansions (unless the block got too big). However, discard this block if
4320 the first thing is a variable or header, so that we can use its block when
4321 it is dynamic (useful for very large $message_headers, for example).
4323 6. Lookups now cache *every* query, not just the most recent. A new, separate
4324 store pool is used for this. It can be recovered when all lookup caches are
4325 flushed. Lookups now release memory at the end of their result strings.
4326 This has involved some general refactoring of the lookup sources.
4328 7. Some code has been added to the store_xxx() functions to reduce the amount
4329 of flapping under certain conditions.
4331 8. log_incoming_interface used to affect only the <= reception log lines. Now
4332 it causes the local interface and port to be added to several more SMTP log
4333 lines, for example "SMTP connection from", and rejection lines.
4335 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4337 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4339 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4340 could be overwritten at the end of the current message (or the start of a
4341 new message if it was set in a HELO ACL). The value is now preserved for
4342 the duration of the SMTP connection.
4344 12. If a transport had a headers_rewrite setting, and a matching header line
4345 contained an unqualified address, that address was qualified, even if it
4346 did not match any rewriting rules. The underlying bug was that the values
4347 of the flags that permit the existence of unqualified sender and recipient
4348 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4349 non-local messages, and by -bnq for local messages) were not being
4350 preserved with the message after it was received.
4352 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4353 "next input=" as part of the text comprising the host identity instead of
4354 the correct text. The code was using the same buffer for two different
4355 strings. However, depending on which order the printing function evaluated
4356 its arguments, the bug did not always show up. Under Linux, for example, my
4357 test suite worked just fine.
4359 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4360 doesn't work with some older versions of Perl. It has been changed to "my",
4361 which in any case is probably the better facility to use.
4363 15. A really picky compiler found some instances of statements for creating
4364 error messages that either had too many or two few arguments for the format
4367 16. The size of the buffer for calls to the DNS resolver has been increased
4368 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4369 for addresses that have a lot of PTR records. This alleviates a problem; it
4370 does not fully solve it.
4372 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4373 buffer now truncates the list and logs the incident, which is the same
4374 action as happens when Exim is looking up a host name and its aliases.
4375 Previously in this situation something unpredictable would happen;
4376 sometimes it was "internal error: store_reset failed".
4378 18. If a server dropped the connection unexpectedly when an Exim client was
4379 using GnuTLS and trying to read a response, the client delivery process
4380 crashed while trying to generate an error log message.
4382 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4383 string, for example:
4385 warn message = H1: something\nH2: something
4387 the text was added as a single header line from Exim's point of view
4388 though it ended up OK in the delivered message. However, searching for the
4389 second and subsequent header lines using $h_h2: did not work. This has been
4390 fixed. Similarly, if a system filter added multiple headers in this way,
4391 the routers could not see them.
4393 20. Expanded the error message when iplsearch is called with an invalid key to
4394 suggest using net-iplsearch in a host list.
4396 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4397 ACLs are no longer actually imposed (and a message to that effect is
4400 22. If a "gecos" field in a passwd entry contained escaped characters, in
4401 particular, if it contained a \" sequence, Exim got it wrong when building
4402 a From: or a Sender: header from that name. A second bug also caused
4403 incorrect handling when an unquoted " was present following a character
4404 that needed quoting.
4406 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4407 was not being matched caselessly.
4409 24. Arranged for all hyphens in the exim.8 source to be escaped with
4412 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4413 properly. Recipient callout cache records were still being keyed to include
4414 the sender, even when use_sender was set false. This led to far more
4415 callouts that were necessary. The sender is no longer included in the key
4416 when use_sender is false.
4418 26. Added "control = submission" modifier to ACLs.
4420 27. Added the ${base62d: operator to decode base 62 numbers.
4422 28. dnsdb lookups can now access SRV records.
4424 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4425 the configuration file.
4427 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4428 (-v) mode. This makes the output for a verbose queue run more intelligible.
4430 31. Added a use_postmaster feature to recipient callouts.
4432 32. Added the $body_zerocount variable, containing the number of binary zero
4433 bytes in the message body.
4435 33. The time of last modification of the "new" subdirectory is now used as the
4436 "mailbox time last read" when there is a quota error for a maildir
4439 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4441 35. Added +ignore_unknown as a special item in host lists.
4443 36. Code for decoding IPv6 addresses in host lists is now included, even if
4444 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4445 address was recognized as an IP address, but was then not correctly decoded
4446 into binary, causing unexpected and incorrect effects when compared with
4453 1. Very minor rewording of debugging text in manualroute to say "list of
4454 hosts" instead of "hostlist".
4456 2. If verify=header_syntax was set, and a header line with an unqualified
4457 address (no domain) and a large number of spaces between the end of the
4458 name and the colon was received, the reception process suffered a buffer
4459 overflow, and (when I tested it) crashed. This was caused by some obsolete
4460 code that should have been removed. The fix is to remove it!
4462 3. When running in the test harness, delay a bit after writing a bounce
4463 message to get a bit more predictability in the log output.
4465 4. Added a call to search_tidyup() just before forking a reception process. In
4466 theory, someone could use a lookup in the expansion of smtp_accept_max_
4467 per_host which, without the tidyup, could leave open a database connection.
4469 5. Added the variables $recipient_data and $sender_data which get set from a
4470 lookup success in an ACL "recipients" or "senders" condition, or a router
4471 "senders" option, similar to $domain_data and $local_part_data.
4473 6. Moved the writing of debug_print from before to after the "senders" test
4476 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4477 problems for message scanning, either using a data ACL, or using
4478 local_scan() because the Received: header was not generated till after they
4479 were called (in order to set the time as the time of reception completion).
4480 I have revised the way this works. The header is now generated after the
4481 body is received, but before the ACL or local_scan() are called. After they
4482 are run, the timestamp in the header is updated.
4488 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4489 before starting a queue runner without re-exec. This happened only when
4490 deliver_drop_privilege was set or when the Exim user was set to root. The
4491 effect of the bug was that timeouts during subsequent deliveries caused
4492 crashes instead of being properly handled. The handler is now left at its
4493 default (and expected) setting.
4495 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4496 message, again when deliver_drop_privilege is set or Exim is run as root.
4497 The bug described in (1) was not present in this case, but the tidying up
4498 of the other signals was missing. I have made the two cases consistent.
4500 3. The ignore_target_hosts setting on a manualroute router was being ignored
4501 for hosts that were looked up using the /MX notation.
4503 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4506 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4507 operated on the sender address. After changing the $sender_address to <>
4508 for the sender address verify, Exim was re-instated it as the original
4509 (before rewriting) address, but remembering that it had rewritten it, so it
4510 wasn't rewriting it again. This bug also had the effect of breaking the
4511 sender address verification caching when the sender address was rewritten.
4513 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4514 This has been changed so that if the ip literal address matches
4515 ignore_target_hosts, the router declines.
4517 7. Added expansion conditions match_domain, match_address, and match_local_
4518 part (NOT match_host).
4520 8. The placeholder for the Received: header didn't have a length field set.
4522 9. Added code to Exim itself and to exim_lock to test for a specific race
4523 condition that could lead to file corruption when using MBX delivery. The
4524 issue is with the lockfile that is created in /tmp. If this file is removed
4525 after a process has opened it but before that process has acquired a lock,
4526 there is the potential for a second process to recreate the file and also
4527 acquire a lock. This could lead to two Exim processes writing to the file
4528 at the same time. The added code performs the same test as UW imapd; it
4529 checks after acquiring the lock that its file descriptor still refers to
4530 the same named file.
4532 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4533 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4536 11. Added the smtp_active_hostname option. If used, this will typically be made
4537 to depend on the incoming interface address. Because $interface_address is
4538 not set up until the daemon has forked a reception process, error responses
4539 that can happen earlier (such as "too many connections") no longer contain
4542 12. If an expansion in a condition on a "warn" statement fails because a lookup
4543 defers, the "warn" statement is abandoned, and the next ACL statement is
4544 processed. Previously this caused the whole ACL to be aborted.
4546 13. Added the iplsearch lookup type.
4548 14. Added ident_timeout as a log selector.
4550 15. Added tls_certificate_verified as a log selector.
4552 16. Added a global option tls_require_ciphers (compare the smtp transport
4553 option of the same name). This controls incoming TLS connections.
4555 17. I finally figured out how to make tls_require_ciphers do a similar thing
4556 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4557 before starting the TLS session.
4559 18. Tabs are now shown as \t in -bP output.
4561 19. If the log selector return_path_on_delivery was set, Exim crashed when
4562 bouncing a message because it had too many Received: header lines.
4564 20. If two routers both had headers_remove settings, and the first one included
4565 a superfluous trailing colon, the final name in the first list and the
4566 first name in the second list were incorrectly joined into one item (with a
4567 colon in the middle).
4573 1. Added -C and -D options to the exinext utility, mainly to make it easier
4574 to include in the automated testing, but these could be helpful when
4575 multiple configurations are in use.
4577 2. The exinext utility was not formatting the output nicely when there was
4578 an alternate port involved in the retry record key, nor when there was a
4579 message id as well (for retries that were specific to a specific message
4580 and a specific host). It was also confused by IPv6 addresses, because of
4581 the additional colons they contain. I have fixed the IPv4 problem, and
4582 patched it up to do a reasonable job for IPv6.
4584 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4585 delivery, the log line now contains "pipelined" if PIPELINING was used.
4587 4. An SMTP transport process used to panic and die if the bind() call to set
4588 an explicit outgoing interface failed. This has been changed; it is now
4589 treated in the same way as a connect() failure.
4591 5. A reference to $sender_host_name in the part of a conditional expansion
4592 that was being skipped was still causing a DNS lookup. This no longer
4595 6. The def: expansion condition was not recognizing references to header lines
4596 that used bh_ and bheader_.
4598 7. Added the _cache feature to named lists.
4600 8. The code for checking quota_filecount in the appendfile transport was
4601 allowing one more file than it should have been.
4603 9. For compatibility with Sendmail, the command line option
4611 and sets the incoming protocol and host name (for trusted callers). The
4612 host name and its colon can be omitted when only the protocol is to be set.
4613 Note the Exim already has two private options, -pd and -ps, that refer to
4614 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4615 "s", but I don't think that's a major issue.
4617 10. A number of refactoring changes to the code, none of which should affect
4620 (a) The number of logging options was getting close to filling up the
4621 32-bit word that was used as a bit map. I have split them into two classes:
4622 those that are passed in the argument to log_write(), and those that are
4623 only ever tested independently outside of that function. These are now in
4624 separate 32-bit words, so there is plenty of room for expansion again.
4625 There is no change in the user interface or the logging behaviour.
4627 (b) When building, for example, log lines, the code previously used a
4628 macro that called string_cat() twice, in order to add two strings. This is
4629 not really sufficiently general. Furthermore, there was one instance where
4630 it was actually wrong because one of the argument was used twice, and in
4631 one call a function was used. (As it happened, calling the function twice
4632 did not affect the overall behaviour.) The macro has been replaced by a
4633 function that can join an arbitrary number of extra strings onto a growing
4636 (c) The code for expansion conditions now uses a table and a binary chop
4637 instead of a serial search (which was left over from when there were very
4638 few conditions). Also, it now recognizes conditions like "pam" even when
4639 the relevant support is not compiled in: a suitably worded error message is
4640 given if an attempt is made to use such a condition.
4642 11. Added ${time_interval:xxxxx}.
4644 12. A bug was causing one of the ddress fields not to be passed back correctly
4645 from remote delivery subprocesses. The field in question was not being
4646 subsequently used, so this caused to problems in practice.
4648 13. Added new log selectors queue_time and deliver_time.
4650 14. Might have fixed a bug in maildirsizefile handling that threw up
4651 "unexpected character" debug warnings, and recalculated the data
4652 unnecessarily. In any case, I expanded the warning message to give more
4655 15. Added the message "Restricted characters in address" to the statements in
4656 the default ACL that block characters like @ and % in local parts.
4658 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4659 Three changes have been made:
4661 (a) There was a serious bug; a negative response to MAIL caused the whole
4662 recipient domain to be cached as invalid, thereby blocking all messages
4663 to all local parts at the same domain, from all senders. This bug has
4664 been fixed. The domain is no longer cached after a negative response to
4665 MAIL if the sender used is not empty.
4667 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4670 (c) A new callout option, "use_sender" has been added for people who want
4671 the modified behaviour.
4677 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4680 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4681 indeed breaks things for older releases.
4683 3. Added additional logging to the case where there is a problem reading data
4684 from a filter that is running in a subprocess using a pipe, in order to
4685 try to track down a specific problem.
4687 4. Testing facility fudge: when running in the test harness and attempting
4688 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4689 getting "No route to host". Convert this to a timeout.
4691 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4694 6. Some OS don't have socklen_t but use size_t instead. This affects the
4695 fifth argument of getsockopt() amongst other things. This is now
4696 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4697 can be set for individual OS. I have set it for SunOS5, OSF1, and
4698 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4699 some earlier ones do not.
4701 7. Change 4.30/15 was not doing the test caselessly.
4703 8. The standard form for an IPv6 address literal was being rejected by address
4704 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4705 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4706 this, as well as the form without the "IPv6" on the front (but only when
4707 address literals are enabled, of course).
4709 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4711 10. Exim crashed if a message with an empty sender address specified by -f
4712 encountered a router with an errors_to setting. This could be provoked only
4713 by a command such as
4717 where an empty string was supplied; "<>" did not hit this bug.
4719 11. Installed PCRE release 4.5.
4721 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4722 remained set. It is now erased.
4724 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4725 times from message ids (which are base 36 rather than the normal 62).
4727 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4728 were being counted as actual protocol errors, and logged if the log
4729 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4730 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4731 and DATA following a set of rejected RCPTs do not count as protocol errors.
4732 In other words, Exim assumes they were pipelined, though this may not
4733 actually be the case. Of course, in all cases the client gets an
4734 appropriate error code.
4736 15. If a lookup fails in an ACL condition, a message about the failure may
4737 be available; it is used if testing the ACL cannot continue, because most
4738 such messages specify what the cause of the deferral is. However, some
4739 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4740 that caused an old message to be retained and used if a later statement
4741 caused a defer, replacing the real cause of the deferral.
4743 16. If an IP address had so many PTR records that the DNS lookup buffer
4744 was not large enough to hold them, Exim could crash while trying to process
4745 the truncated data. It now detects and logs this case.
4747 17. Further to 4.21/58, another change has been made: if (and only if) the
4748 first line of a message (the first header line) ends with CRLF, a bare LF
4749 in a subsequent header line has a space inserted after it, so as not to
4750 terminate the header.
4752 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4753 unnecessarily, used atoi() instead of strtol(), and didn't check the
4754 termination when getting file sizes from file names by regex.
4756 19. Completely re-implemented the support for maildirsize files, in the light
4757 of a number of problems with the previous contributed implementation
4758 (4.30/29). In particular:
4760 . If the quota is zero, the maildirsize file is maintained, but no quota is
4763 . If the maildir directory does not exist, it is created before any attempt
4764 to write a maildirsize file.
4766 . The quota value in the file is just a cache; if the quota is changed in
4767 the transport, the new value overrides.
4769 . A regular expression is available for excluding directories from the
4772 20. The autoreply transport checks the characters in options that define the
4773 message's headers; it allows continued headers, but it was checking with
4774 isspace() after an embedded newline instead of explicitly looking for a
4777 21. If all the "regular" hosts to which an address was routed had passed their
4778 expiry times, and had not reached their retry times, the address was
4779 bounced, even if fallback hosts were defined. Now Exim should go on to try
4782 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4783 equivalent code in the SMTP transport. Some hosts send humungous responses
4784 to HELO/EHLO, more than 1024 it seems.
4786 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4787 gives compiler warnings in some environments. I've now done it "properly",
4790 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4791 (because of problems with the built-in one) was declared to return uschar *
4792 instead of char *, causing compiler failure.
4794 25. Fixed a file descriptor leak when processing alias/forward files.
4796 26. Fixed a minor format string issue in dbfn.c.
4798 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4800 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4801 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4804 29. When a sender address is verified, it is cached, to save repeating the test
4805 when there is more than one recipient in a message. However, when the
4806 verification involves a callout, it is possible for different callout
4807 options to be set for different recipients. It is too complicated to keep
4808 track of this in the cache, so now Exim always runs a verification when a
4809 callout is required, relying on the callout cache for the optimization.
4810 The overhead is duplication of the address routing, but this should not be
4813 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4814 to be verified with callout=postmaster, and the main callout worked but the
4815 postmaster check failed, the verification correctly failed. However, if a
4816 subsequent RCPT command asked for sender verification *without* the
4817 postmaster check, incorrect caching caused this verification also to fail,
4820 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4821 it was not caching the DNS options (qualify_single, search_parents) that
4822 were used when the lookup failed. A subsequent lookup with different
4823 options therefore always gave the same answer, though there were cases
4824 where it should not have. (Example: a "domains = !$mx_any" option on a
4825 dnslookup router: the "domains" option is always processed without any
4826 widening, but the router might have qualify_single set.) Now Exim uses the
4827 cached value only when the same options are set.
4829 32. Added John Jetmore's "exipick" utility to the distribution.
4831 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4832 than a timeout (e.g. a certificate is required, and is not provided), an
4833 Exim server now closes the connection immediately. Previously it waited for
4834 the client to close - but if the client is SSL, it seems that they each
4835 wait for each other, leading to a delay before one of them times out.
4837 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4838 maintained 0.8.x compatibility because I don't think many are using it, and
4839 it is clearly obsolete.
4841 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4844 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4845 client certificate was expired. A simple patch fixes this, though I don't
4846 understand the full logic of why the verify callback is called multiple
4849 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4850 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4851 which causes problems with some clients (such as the Certicom SSL Plus
4852 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4853 disables the coutermeasure allowing Eudora to connect."
4855 38. Exim was not checking that a write() to a log file succeeded. This could
4856 lead to Bad Things if a log got too big, in particular if it hit a file
4857 size limit. Exim now panics and dies if it cannot write to a log file, just
4858 as it does if it cannot open a log file.
4860 39. Modified OS/Makefile-Linux so that it now contains
4862 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4864 The two -D definitions ensure that Exim is compiled with large file
4865 support, which makes it possible to handle log files that are bigger than
4868 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4869 instance) a domain was checked against a named list that involved a lookup,
4870 causing $domain_data to be set, then another domain was checked against the
4871 same list, then the first domain was re-checked, the value of $domain_data
4872 after the final check could be wrong. In particular, if the second check
4873 failed, it could be set empty. This bug probably also applied to
4876 41. The strip_trailing_dot option was not being applied to the address given
4877 with the -f command-line option.
4879 42. The code for reading a message's header from the spool was incrementing
4880 $received_count, but never initializing it. This meant that the value was
4881 incorrect (doubled) while delivering a message in the same process in which
4882 it was received. In the most common configuration of Exim, this never
4883 happens - a fresh exec is done - but it can happen when
4884 deliver_drop_privilege is set.
4886 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4887 it now includes up to 150 characters of the unexpected data in the log
4890 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4891 and building data strings. The size of both of these buffers was 10 000
4892 bytes - far larger than anybody would *ever* want, thought I. Needless to
4893 say, somebody hit the limit. I have increased the maximum line length to
4894 20 000 and the maximum data length of concatenated lines to 100 000. I have
4895 also fixed two bugs, because there was no checking on these buffers. Tsk,
4896 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4897 buffer is too small.
4899 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4900 lsearch lookups. Now it does.
4902 46. When parsing a route_list item in a manualroute router, a fixed-length
4903 buffer was used for the list of hosts. I made this 1024 bytes long,
4904 thinking that nobody would ever have a list of hosts that long. Wrong.
4905 Somebody had a whole pile of complicated expansion conditions, and the
4906 string was silently truncated, leading to an expansion error. It turns out
4907 that it is easier to change to an unlimited length (owing to other changes
4908 that have happened since this code was originally written) than to build
4909 structure for giving a limitation error. The length of the item that
4910 expands into the list of hosts is now unlimited.
4912 47. The lsearch lookup could not handle data where the length of text line was
4913 more than 4095 characters. Such lines were truncated, leading to shortened
4914 data being returned. It should now handle lines of any length.
4916 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4917 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4920 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4921 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4922 no longer happens in comment lines. A list of replacements is now placed
4923 at the head of all of the source files, except those whose only change is
4924 to replace PERL_COMMAND in the very first #! line.
4926 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4927 messages on the queue, with a bottom-up merge sort, using code contributed
4928 by Michael Haardt. This should make operations like -bp somewhat faster on
4929 large queues. It won't affect queue runners, except when queue_run_in_order
4932 51. Installed eximstats 1.31 in the distribution.
4934 52. Added support for SRV lookups to the dnslookup router.
4936 53. If an ACL referred to $message_body or $message_body_end, the value was not
4937 reset for any messages that followed in the same SMTP session.
4939 54. The store-handling optimization for building very long strings was not
4940 differentiating between the different store pools. I don't think this
4941 actually made any difference in practice, but I've tidied it.
4943 55. While running the routers to verify a sender address, $sender_address
4944 was still set to the sender address. This is wrong, because when routing to
4945 send a bounce to the sender, it would be empty. Therefore, I have changed
4946 it so that, while verifying a sender address, $sender_address is set to <>.
4947 (There is no change to what happens when verifying a recipient address.)
4949 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4950 target A or AAAA records (if not already returned) without resetting the
4951 qualify_single or search_parents options of the DNS resolver. These are
4952 inappropriate in this case because the targets of MX and SRV records must
4953 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4954 target that, when qualified, matched something in the local domain. These
4955 two options are now turned off when doing these lookups.
4957 57. It seems that at least some releases of Reiserfs (which does not have the
4958 concept of a fixed number of inodes) returns zero and not -1 for the
4959 number of available inodes. This interacted badly with check_spool_inodes,
4960 which assumed that -1 was the "no such thing" setting. What I have done is
4961 to check that the total number of inodes is greater than zero before doing
4962 the test of how many are available.
4964 58. When a "warn" ACL statement has a log_message modifier, the message is
4965 remembered, and not repeated. This is to avoid a lot of repetition when a
4966 message has many recipients that cause the same warning to be written.
4967 Howewer, Exim was preserving the list of already written lines for an
4968 entire SMTP session, which doesn't seem right. The memory is now reset if a
4969 new message is started.
4971 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4972 debugging output unless log_rewrite was also set.
4974 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4975 of (int)(handle) when we know that handle contains (void *)(-1).
4977 61. The Exim daemon panic-logs an error return when it closes the incoming
4978 connection. However "connection reset by peer" seems to be common, and
4979 isn't really an error worthy of noting specially, so that particular error
4982 62. When Exim is trying to find all the local interfaces, it used to panic and
4983 die if the ioctl to get the interface flags failed. However, it seems that
4984 on at least one OS (Solaris 9) it is possible to have an interface that is
4985 included in the list of interfaces, but for which you get a failure error
4986 for this call. This happens when the interface is not "plumbed" into a
4987 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4988 failure of the "get flags" call assumes that the interface is down.
4990 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4991 makes life easier for people who are doing arithmetic on fields extracted
4992 from dates, where you often get leading zeros that should not be
4993 interpreted as octal.
4995 64. Added qualify_domain to the redirect router, to override the global
4998 65. If a pathologically long header line contained very many addresses (the
4999 report of this problem mentioned 10 000) and each of them was rewritten,
5000 Exim could use up a very large amount of memory. (It kept on making new
5001 copies of the header line as it rewrote, and never released the old ones.)
5002 At the expense of a bit more processing, the header rewriting function has
5003 been changed so that it no longer eats memory in this way.
5005 66. The generation of the Received: header has been moved from the time that a
5006 message starts to be received, to the time that it finishes. The timestamp
5007 in the Received: header should now be very close to that of the <= log
5008 line. There are two side-effects of this change:
5010 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5011 logged header lines no longer include the local Received: line, because
5012 it has not yet been created. The same applies to a copy of the message
5013 that is returned to a non-SMTP sender when a message is rejected.
5015 (b) When a filter file is tested using -bf, no additional Received: header
5016 is added to the test message. After some thought, I decided that this
5019 This change does not affect the value of $received_for. It is still set
5020 after address rewriting, but before local_scan() is called.
5022 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5024 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5025 gave an unhelpful panic error message, and a defer error. I have managed to
5026 change this behaviour so that it now rejects any supplied certificate,
5027 which seems right, as the list of acceptable certificates is empty.
5029 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5030 gave an unhelpful defer error. I have not managed to make this reject any
5031 supplied certificates, but the error message it gives is "no certificate
5032 supplied", which is not helpful.
5034 70. exigrep's output now also includes lines that are not associated with any
5035 message, but which match the given pattern. Implemented by a patch from
5036 Martin Sluka, which also tidied up the Perl a bit.
5038 71. Recipient callout verification, like sender verification, was using <> in
5039 the MAIL FROM command. This isn't really the right thing, since the actual
5040 sender may affect whether the remote host accepts the recipient or not. I
5041 have changed it to use the actual sender in the callout; this means that
5042 the cache record is now keyed on a recipient/sender pair, not just the
5043 recipient address. There doesn't seem to be a real danger of callout loops,
5044 since a callout by the remote host to check the sender would use <>.
5045 [SEE ABOVE: changed after hitting problems.]
5047 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5048 temporary errors. However, in the case of such a code being given after
5049 the end of a data transmission (i.e. after ".") Exim was failing to write
5050 a retry record for the message. (Yes, there was some broken host that was
5051 actually sending 8xx at this point.)
5053 73. An unknown lookup type in a host list could cause Exim to panic-die when
5054 the list was checked. (An example that provoked this was putting <; in the
5055 middle of a list instead of at the start.) If this happened during a DATA
5056 ACL check, a -D file could be left lying around. This kind of configuration
5057 error no longer causes Exim to die; instead it causes a defer errror. The
5058 incident is still logged to the main and panic logs.
5060 74. Buglet left over from Exim 3 conversion. The message "too many messages
5061 in one connection" was written to the rejectlog but not the mainlog, except
5062 when address rewriting (yes!) was being logged.
5064 75. Added write_rejectlog option.
5066 76. When a system filter was run not as root (that is, when system_filter_user
5067 was set), the values of the $n variables were not being returned to the
5068 main process; thus, they were not subsequently available in the $sn
5071 77. Added +return_path_on_delivery log selector.
5073 78. A connection timeout was being treated differently from recipients deferred
5074 when testing hosts_max_try with a message that was older than the host's
5075 retry timeout. (The host should not be counted, thus allowing all hosts to
5076 be tried at least once before bouncing.) This may have been the cause of an
5077 occasionally reported bug whereby a message would remain on the queue
5078 longer than the retry timeout, but would be bounced if a delivery was
5079 forced. I say "may" because I never totally pinned down the problem;
5080 setting up timeout/retry tests is difficult. See also the next item.
5082 79. The ultimate address timeout was not being applied to errors that involved
5083 a combination of host plus message (for example, a timeout on a MAIL
5084 command). When an address resolved to a number of possible hosts, and they
5085 were not all tried for each delivery (e.g. because of hosts_max_try), a
5086 message could remain on the queue longer than the retry timeout.
5088 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5091 81. Fixed an obscure SMTP outgoing bug which required at least the following
5092 conditions: (a) there was another message waiting for the same server;
5093 (b) the server returned 5xx to all RCPT commands in the first message so
5094 that the message was not completed; (c) the server dropped the connection
5095 or gave a negative response to the RSET that Exim sends to abort the
5096 transaction. The observed case was a dropped connection after DATA that had
5097 been sent in pipelining mode. That is, the server had advertised PIPELINING
5098 but was not implementing it correctly. The effect of the bug was incorrect
5099 behaviour, such as trying another host, and this could lead to a crash.
5105 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5106 and daemon.c were passed as pointers to ints; they should have been
5107 pointers to socklen_t variables (which are typically unsigned ints).
5109 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5112 3. Fixed a really odd bug that affected only the testing scheme; patching a
5113 certain fixed string in the binary changed the value of another string that
5114 happened to be identical to the end of the original first string.
5116 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5117 name", it returns that address as the IP address. On some operating
5118 systems (e.g. Solaris), it also passes back the IP address string as the
5119 "host name". However, on others (e.g. Linux), it passes back an empty
5120 string. Exim wasn't checking for this, and was changing the host name to an
5121 empty string, assuming it had been canonicized.
5123 5. Although rare, it is permitted to have more than one PTR record for a given
5124 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5125 all the names associated with an address, because they do in Solaris.
5126 However, it seems that they do not in Linux for data that comes from the
5127 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5128 I found this out when I moved to a new Linux workstation and tried to run
5129 the Exim test suite.
5131 To get round this problem I have changed the code so that it now does its
5132 own call to the DNS to look up PTR records when searching for a host name.
5133 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5134 addresses that are only in /etc/hosts are still found.
5136 This behaviour is, however, controlled by an option called host_lookup_
5137 order, which defaults to "bydns:byaddr". If people want to use the other
5138 order, or indeed, just use one or the other means of lookup, they can
5139 specify it in this variable.
5141 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5142 some operating systems, this comes back from gethostbyaddr() as an empty
5143 string, and this is what Exim used to test for. However, it seems that in
5144 other systems, "." is yielded. Exim now tests for this case too.
5146 7. The values of check_spool_space and check_log_space are now held internally
5147 as a number of kilobytes instead of an absolute number of bytes. If a
5148 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5149 kilobyte. This means that much larger values can be stored.
5151 8. Exim monitor: an attempt to get the action menu when not actually pointing
5152 at a message produces an empty menu entitled "No message selected". This
5153 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5154 no entries in it ("Shell widget menu has zero width and/or height"). So I
5155 have added a single, blank menu entry in this case.
5157 9. Added ${quote_local_part.
5159 10. MIME decoding is now applied to the contents of Subject: header lines when
5162 11. Now that a reference to $sender_host_address automatically causes a reverse
5163 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5164 host lookup before query-style lookups in lists that might use this
5165 variable. This has therefore been abolished, and the "net-" prefix is no
5166 longer necessary for query-style lookups.
5168 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5169 have been a typo for LFLAGS, so it has been changed.
5171 13. The install script calls Exim with "-C /dev/null" in order to find the
5172 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5173 to be output. Howeve, since Exim outputs its version number before the
5174 error, it didn't break the script. It just looked ugly. I fixed this by
5175 always allowing "-C /dev/null" if the caller is root.
5177 14. Ignore overlarge ACL variable number when reading spool file - insurance
5178 against a later release with more variables having written the file.
5180 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5181 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5182 this, as well as the form without the "IPv6" on the front.
5184 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5185 OS/Makefile-Darwin file.
5187 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5188 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5190 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5191 variable. However, some OS use ints for inodes. I've added cast to long int
5192 to get rid of the compiler warning.
5194 19. I had forgotten to lock out "/../" in configuration file names when
5195 ALT_CONFIG_PREFIX was set.
5197 20. Routers used for verification do not need to specify transports. However,
5198 if such a router generated a host list, and callout was configured, Exim
5199 crashed, because it could not find a port number from the (non-existent)
5200 transport. It now assumes port 25 in this circumstance.
5202 21. Added the -t option to exigrep.
5204 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5205 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5206 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5207 them is set, LOOKUP_LSEARCH is forced.
5209 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5211 24. Added sender and host information to the "rejected by local_scan()" log
5212 line; previously there was no indication of these.
5214 25. Added .include_if_exists.
5216 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5217 Linux. It turns out that not all file systems support this. Apparently some
5218 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5219 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5220 not supported on the file descriptor, is now ignored when Exim is trying to
5221 sync a directory. This applies only to Linux.
5223 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5225 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5226 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5229 29. Added support for maildirsize files from supplied patch (modified a bit).
5231 30. The use of :fail: followed by an empty string could lead Exim to respond to
5232 sender verification failures with (e.g.):
5234 550 Verification failed for <xxx>
5235 550 Sender verify failed
5237 where the first response line was missing the '-' that indicates it is not
5238 the final line of the response.
5240 31. The loop for finding the name of the user that called Exim had a hardwired
5241 limit of 10; it now uses the value of finduser_retries, which is used for
5242 all other user lookups.
5244 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5247 33. Exim was neglecting to zero errno before one call of strtol() when
5248 expanding a string and expecting an integer value. On some systems this
5249 resulted in spurious "integer overflow" errors. Also, it was casting the
5250 result into an int without checking.
5252 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5253 did not work. The code looks as if it has *never* worked, though it appears
5254 to have been documented since at least releast 1.62. I have made it work.
5256 35. The "timeout_DNS" error in retry rules, also documented since at least
5257 1.62, also never worked. As it isn't clear exactly what this means, and
5258 clearly it isn't a major issue, I have abolished the feature by treating it
5259 as "timeout", and writing a warning to the main and panic logs.
5261 36. The display of retry rules for -brt wasn't always showing the error code
5264 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5265 timeout_connect_A, timeout_connect_MX.
5267 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5268 to the empty sender.
5270 39. The daemon was not analysing the content of -oX till after it had closed
5271 stderr and disconnected from the controlling terminal. This meant that any
5272 syntax errors were only noted on the panic log, and the return code from
5273 the command was 0. By re-arranging the code a little, I've made the
5274 decoding happen first, so such errors now appear on stderr, and the return
5275 code is 1. However, the actual setting up of the sockets still happens in
5276 the disconnected process, so errors there are still only recorded on the
5279 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5280 connections (as happens on some IP stacks) was logged at start up time as
5281 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5282 it from "IPv6 and IPv4", which means that two separate sockets are being
5285 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5286 says whether AF_INET or AF_INET6 was passed as an argument.
5288 42. Exiwhat output was messed up when time zones were included in log
5291 43. Exiwhat now gives more information about the daemon's listening ports,
5292 and whether -tls-on-connect was used.
5294 44. The "port" option of the smtp transport is now expanded.
5296 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5297 silently ignored. Now an error message is written to the main and panic
5300 46. There's a new ACL modifier called "logwrite" which writes to a log file
5301 as soon as it is encountered.
5303 47. Added $local_user_uid and $local_user_gid at routing time.
5305 48. Exim crashed when trying to verify a sender address that was being
5308 49. Exim was recognizing only a space character after ".include". It now also
5309 recognizes a tab character.
5311 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5312 extracting the relevant information from the specification. The man page no
5313 longer contains scrambled data for the -d option, and I've added a section
5314 at the front about calling Exim under different names.
5316 51. Added "extra_headers" argument to the "mail" command in filter files.
5318 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5321 53. Installed eximstats 1.29.
5323 54. Added transport_filter_timeout as a generic transport option.
5325 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5326 Cc: header lines. This was required by RFC 822, but it not required by RFC
5329 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5330 incoming messages that did not have them. Now it does so only if the
5331 message originates locally, that is, if there is no associated remote host
5332 address. When Resent- header lines are present, this applies to the Resent-
5333 lines rather than the non-Resent- lines.
5335 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5336 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5338 58. Messages for configuration errors now include the name of the main
5339 configuration file - useful now that there may be more than one file in a
5340 list (.included file names were always shown).
5342 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5343 for those rare installations that do not start the daemon as root or run it
5344 setuid root. I've cut out the call to initgroups() if the daemon is not
5347 60. The Exim user and group can now be bound into the binary as text strings
5348 that are looked up at the start of Exim's processing.
5350 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5352 62. Added $mailstore_basename variable.
5354 63. Installed patch to sieve.c from Michael Haardt.
5356 64. When Exim failed to open the panic log after failing to open the main log,
5357 the original message it was trying to log was written to stderr and debug
5358 output, but if they were not available (the usual case in production), it
5359 was lost. Now it is written to syslog before the two lines that record the
5360 failures to open the logs.
5362 65. Users' Exim filters run in subprocesses under the user's uid. It is
5363 possible for a "deliver" command or an alias in a "personal" command to
5364 provoke an address rewrite. If logging of address rewriting is configured,
5365 this fails because the process is not running as root or exim. There may be
5366 a better way of dealing with this, but for the moment (because 4.30 needs
5367 to be released), I have disabled address rewrite logging when running a
5368 filter in a non-root, non-exim process.
5374 1. The buildconfig auxiliary program wasn't quoting the value set for
5375 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5376 not defined. This bug was masked in 4.22 by the effect that was fixed in
5379 2. Some messages that were rejected after a message id was allocated were
5380 shown as "incomplete" by exigrep. It no longer does this for messages that
5381 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5383 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5384 have allow_domain_literals set, the ID did not get logged in the <= line.
5385 Domain literals are now always recognized in Message-ID: header lines.
5387 4. The first argument for a ${extract expansion item is the key name or field
5388 number. Leading and trailing spaces in this item were not being ignored,
5389 causing some misleading effects.
5391 5. When deliver_drop_privilege was set, single queue runner processes started
5392 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5393 same command in the process it spins off) were not dropping privilege.
5395 6. When the daemon running as "exim" started a queue runner, it always
5396 re-executed Exim in the spun-off process. This is a waste of effort when
5397 deliver_drop_privilege is set. The new process now just calls the
5398 queue-runner function directly.
5404 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5407 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5408 ignored. Though the use of -oP was forcing the writing of a pid file, it
5409 was always written to the default place.
5411 3. If the message "no IP address found for host xxxx" is generated during
5412 incoming verification, it is now followed by identification of the incoming
5413 connection (so you can more easily find what provoked it).
5415 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5417 5. Added some features to "harden" Exim a bit more against certain attacks:
5419 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5420 be put in Local/Makefile. This is like the never_users runtime option,
5421 but it cannot be overridden. The default setting is "root".
5423 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5424 prefix string with which any file named in a -C command line option
5427 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5428 is retained for -C and -D only if the caller of Exim is root. Without
5429 it, the exim user may also use -C and -D and retain privilege.
5431 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5432 command line option is disabled.
5434 6. Macro names set by the -D option must start with an upper case letter, just
5435 like macro names defined in the configuration file.
5437 7. Added "dereference=" facility to LDAP.
5439 8. Two instances of the typo "uknown" in the source files are fixed.
5441 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5442 the Configure-Makefile script screwed up while processing it.
5444 10. Incorporated PCRE 4.4.
5446 11. The SMTP synchronization check was not operating right at the start of an
5447 SMTP session. For example, it could not catch a HELO sent before the client
5448 waited for the greeting. There is now a check for outstanding input at the
5449 point when the greeting is written. Because of the duplex, asynchronous
5450 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5451 way, but not yet received, when the check is performed.
5453 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5454 on TCP/IP sockets, because this apparently causes some broken clients to
5457 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5458 unchanged) from the Cygwin maintainer.
5460 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5461 was supported instead of testing for mbx. Effectively a typo.
5463 15. The spa authenticator server code was not checking that the input it
5464 received was valid base64.
5466 16. The debug output line for the "set" modifier in ACLs was not showing the
5467 name of the variable that was being set.
5469 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5471 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5472 lookup on its own. Something else was needed to trigger the lookup. For
5473 example, a match in host_lookup or the need for a host name in a host list.
5474 Now, if $sender_host_name is referenced and the host name has not yet been
5475 looked up, a lookup is performed. If the lookup fails, the variable remains
5476 empty, and $host_lookup_failed is set to "1".
5478 19. Added "eqi" as a case-independent comparison operator.
5480 20. The saslauthd authentication condition could segfault if neither service
5481 nor realm was specified.
5483 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5484 error message that was logged was misleading, and incoming SMTP
5485 connections were dropped. The message is now more accurate, and temporary
5486 errors are given to SMTP connections.
5488 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5489 (except RSET) with a 421 error, until QUIT is received. However, it was
5490 failing to send a response to QUIT.
5492 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5493 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5494 ACL is now run after the helo_try_verify_hosts code.
5496 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5497 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5498 case-independent, so other case variants are also recognized). Apparently
5499 some systems use these upper case variants.
5501 25. If more than two messages were waiting for the same host, and a transport
5502 filter was specified for the transport, Exim sent two messages over the
5503 same TCP/IP connection, and then failed with "socket operation on non-
5504 socket" when it tried to send the third.
5506 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5508 27. The extern definition of crypt16() in expand.c was not being excluded when
5509 the OS had its own crypt16() function.
5511 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5512 as a preferred synonym for return_size_limit, both as an option and as an
5515 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5517 30. Changed the default configuration ACL to relax the local part checking rule
5518 for addresses that are not in any local domains. For these addresses,
5519 slashes and pipe symbols are allowed within local parts, but the sequence
5520 /../ is explicitly forbidden.
5522 31. SPA server authentication was not clearing the challenge buffer before
5525 32. log_message in a "warn" ACL statement was writing to the reject log as
5526 well as to the main log, which contradicts the documentation and doesn't
5527 seem right (because no rejection is happening). So I have stopped it.
5529 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5530 However, I am unable to do any testing of this.
5532 34. Fixed an infelicity in the appendfile transport. When checking directories
5533 for a mailbox, to see if any needed to be created, it was accidentally
5534 using path names with one or more superfluous leading slashes; tracing
5535 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5537 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5538 contents are added to the log line that is written for every discarded
5539 recipient. (Previously a log_message setting was ignored.)
5541 36. The ${quote: operator now quotes the string if it is empty.
5543 37. The install script runs exim in order to find its version number. If for
5544 some reason other than non-existence or emptiness, which it checks, it
5545 could not run './exim', it was installing it with an empty version number,
5546 i.e. as "exim-". This error state is now caught, and the installation is
5549 38. An argument was missing from the function that creates an error message
5550 when Exim fails to connect to the socket for saslauthd authentication.
5551 This could cause Exim to crash, or give a corrupted message.
5553 39. Added isip, isip4, and isip6 to ${if conditions.
5555 40. The ACL variables $acl_xx are now saved with the message, and can be
5556 accessed later in routers, transports, and filters.
5558 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5559 strings in the file are not string-expanded.
5561 42. If a MAIL command specified a SIZE value that was too large to fit into an
5562 int variable, the check against message_size_limit failed. Such values are
5563 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5564 day this will have to be increased, but I don't think I want to be around
5565 when emails are that large.
5572 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5573 iconv() is not standard in FreeBSD.
5575 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5576 IPv6 enabled. The observed symptom was a segmentation fault on return from
5577 the function os_common_find_running_interfaces() in src/os.c.
5579 3. In the check_special_case() function in daemon.c I had used "errno" as an
5580 argument name, which causes warnings on some systems. This was basically a
5581 typo, since it was named "eno" in the comments!
5583 4. The code that waits for the clock to tick (at a resolution of some fraction
5584 of a second) so as to ensure message-id uniqueness was always waiting for
5585 at least one whole tick, when it could have waited for less. [This is
5586 almost certainly not relevant at current processor speeds, where it is
5587 unlikely to ever wait at all. But we try to future-proof.]
5589 5. The function that sleeps for a time interval that includes fractions of a
5590 second contained a race. It did not block SIGALRM between setting the
5591 timer, and suspending (a couple of lines later). If the interval was short
5592 and the sigsuspend() was delayed until after it had expired, the suspension
5593 never ended. On busy systems this could lead to processes getting stuck for
5596 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5597 process, before it forks any delivery processes. The open lookup caching
5598 mechanism meant that the open file or database connection was passed into
5599 the delivery process. The problem was that delivery processes always tidy
5600 up cached lookup data. This could cause a problem for the next delivery
5601 process started by the queue runner, because the external queue runner
5602 process does not know about the closure. So the next delivery process
5603 still has data in the lookup cache. In the case of a file lookup, there was
5604 no problem because closing a file descriptor in a subprocess doesn't affect
5605 the parent. However, if the lookup was caching a connection to a database,
5606 the connection was closed, and the second delivery process was likely to
5607 see errors such as "PGSQL: query failed: server closed the connection
5608 unexpectedly". The problem has been fixed by closing all cached lookups
5609 in a queue runner before running a delivery process.
5611 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5612 seem to have the "const" qualifier which it has on other OS. I've
5615 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5616 *of the same type* (client or server) with the same public name that an
5617 error should be diagnosed.
5619 9. When Exim looked up a host name for an IP address, but failed to find the
5620 original IP address when looking up the host name (a safety check), it
5621 output the message "<ip address> does not match any IP for NULL", which was
5622 confusing, to say the least. The bug was that the host name should have
5623 appeared instead of "NULL".
5625 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5626 user that is built into the binary, and the -C or -D options is used, root
5627 privilege is dropped before the configuration file is read. In addition,
5628 logging is switched to stderr instead of the normal log files. If the
5629 configuration then re-defines the Exim user, the unprivileged environment
5630 is probably not what is expected, so Exim logs a panic warning message (but
5633 However, if deliver_drop_privilege is set, the unprivileged state may well
5634 be exactly what is intended, so the warning has been cut out in that case,
5635 and Exim is allowed to try to write to its normal log files.
5641 1. smtp_return_error_details was not giving details for temporary sender
5642 or receiver verification errors.
5644 2. Diagnose a configuration error if two authenticators have the same public
5647 3. Exim used not to create the message log file for a message until the first
5648 delivery attempt. This could be confusing when incoming messages were held
5649 for policy or load reasons. The message log file is now created at the time
5650 the message is received, and an initial "Received" line is written to it.
5652 4. The automatically generated man page for command line options had a minor
5653 bug that caused no ill effects; however, a more serious problem was that
5654 the procedure for building the man page automatically didn't always
5655 operate. Consequently, release 4.20 contains an out-of-date version. This
5656 shouldn't happen again.
5658 5. When building Exim with embedded Perl support, the script that builds the
5659 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5660 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5662 6. The freeze_tell option was not being used for messages that were frozen on
5663 arrival, either by an ACL or by local_scan().
5665 7. Added the smtp_incomplete_transaction log selector.
5667 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5668 was accepting AUTH without a new EHLO.
5670 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5671 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5674 10. Logging of TCP/IP connections (when configured) now happens in the main
5675 daemon process instead of the child process, so that the TCP/IP connection
5676 count is more accurate (but it can never be perfect).
5678 11. The use of "drop" in a nested ACL was not being handled correctly in the
5679 outer ACL. Now, if condition failure induced by the nested "drop" causes
5680 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5681 or "require"), the connection is dropped.
5683 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5684 that yield "discard" can now be used with an "accept" or a "discard" verb,
5685 but an error is generated for any others (because I can't see a useful way
5686 to define what should happen).
5688 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5689 are now processed in the same way as lines in the Exim configuration file.
5690 In particular, continuation lines are supported.
5692 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5694 15. Added -ti meaning -t -i.
5696 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5697 domains, and warn by logging if others are found.
5699 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5700 returned from GIFCONF when Exim is trying to find the list of interfaces on
5701 a host. The code in os.c has been modified to copy each ifreq to an aligned
5702 structure in all cases.
5704 Also, in some cases, the returned ifreq's were being copied to a 'struct
5705 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5706 means the last couple of bytes of an IPv6 address could be chopped if the
5707 ifreq contained only a normal sockaddr (14 bytes storage).
5709 18. Named domain lists were not supported in the hosts_treat_as_local option.
5710 An entry such as +xxxx was not recognized, and was treated as a literal
5713 19. Ensure that header lines added by a DATA ACL are included in the reject log
5714 if the ACL subsequently rejects the message.
5716 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5717 MD5 (which is deprecated).
5719 21. When testing a filter file using -bf, Exim was writing a message when it
5720 took the sender from a "From " line in the message, but it was not doing so
5721 when it took $return_path from a Return-Path: header line. It now does.
5723 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5724 with a valid header line field name (a series of printing characters
5725 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5727 23. Changed "disc" in the source to "disk" to conform to the documentation and
5728 the book and for uniformity.
5730 24. Ignore Sendmail's -Ooption=value command line item.
5732 25. When execve() failed while trying to run a command in a pipe transport,
5733 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5734 could be confused with a return value of 69 from the command itself. This
5735 has been changed to 127, the value the shell returns if it is asked to run
5736 a non-existent command. The wording for the related log line suggests a
5737 non-existent command as the problem.
5739 26. If received_header_text expands to an empty string, do not add a Received:
5740 header line to the message. (Well, it adds a token one on the spool, but
5741 marks it "old" so that it doesn't get used or transmitted.)
5743 27. Installed eximstats 1.28 (addition of -nt option).
5745 28. There was no check for failure on the call to getsockname() in the daemon
5746 code. This can fail if there is a shortage of resources on the system, with
5747 ENOMEM, for example. A temporary error is now given on failure.
5749 29. Contrary to the C standard, it seems that in some environments, the
5750 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5751 program. Exim now does this explicitly; it affects the formatting of
5752 timestamps using strftime().
5754 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5755 complaints. I've now initialized all the variables, to avoid this.
5757 32. Header lines added by a system filter were not being "seen" during
5758 transport-time rewrites.
5760 33. The info_callback() function passed to OpenSSL is set up with type void
5761 (*)(SSL *, int, int), as described somewhere. However, when calling the
5762 function (actually a macro) that sets it up, the type void(*)() is
5763 expected. I've put in a cast to prevent warnings from picky compilers.
5765 34. If a DNS black list lookup found a CNAME record, but there were no A
5766 records associated with the domain it pointed at, Exim crashed.
5768 35. If a DNS black list lookup returned more than one A record, Exim ignored
5769 all but the first. It now scans all returned addresses if a particular IP
5770 value is being sought. In this situation, the contents of the
5771 $dnslist_value variable are a list of all the addresses, separated by a
5774 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5775 to accept a host name and all its aliases if the forward lookup for any of
5776 them yielded the IP address of the incoming connection. Now it accepts only
5777 those names whose forward lookup yields the correct IP address. Any other
5778 names are discarded. This closes a loophole whereby a rogue DNS
5779 administrator could create reverse DNS records to break through a
5780 wildcarded host restriction in an ACL.
5782 37. If a user filter or a system filter that ran in a subprocess used any of
5783 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5784 the wrong values were passed to the pipe command ($thisaddress had the
5785 value of $0, $0 had the value of $1, etc). This bug was introduced by
5786 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5788 38. Improved the line breaking for long SMTP error messages from ACLs.
5789 Previously, if there was no break point between 40 and 75 characters, Exim
5790 left the rest of the message alone. Two changes have been made: (a) I've
5791 reduced the minimum length to 35 characters; (b) if it can't find a break
5792 point between 35 and 75 characters, it looks ahead and uses the first one
5793 that it finds. This may give the occasional overlong line, but at least the
5794 remaining text gets split now.
5796 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5797 file descriptors might be low, and that setting 1000 would always raise it.
5798 It turns out that in some environments, the limit is already over 1000 and
5799 that lowering it causes trouble. So now Exim takes care not to decrease it.
5801 40. When delivering a message, the value of $return_path is set to $sender_
5802 address at the start of routing (routers may change the value). By an
5803 oversight, this default was not being set up when an address was tested by
5804 -bt or -bv, which affected the outcome if any router or filter referred to
5807 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5808 log only when "message" or "log_message" are set. However, if one of the
5809 conditions was an address verification, or a call to a nested ACL, the
5810 messages generated by the underlying test were being passed through. This
5811 no longer happens. The underlying message is available in $acl_verify_
5812 message for both "message" and "log_message" expansions, so it can be
5813 passed through if needed.
5815 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5816 new expansion $bh_ to give the encoded byte string without charset
5817 translation. Translation happens only if iconv() is available; HAVE_ICONV
5818 indicates this at build time. HEADERS_CHARSET gives the charset to
5819 translate to; headers_charset can change it in the configuration, and
5820 "headers charset" can change it in an individual filter file.
5822 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5823 that creates RFC 2047 encoded "words" labels them as that charset instead
5824 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5825 expansion operator; (ii) when Exim creates a From: line for a local
5826 message; (iii) when a header line is rewritten to include a "phrase" part.
5828 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5829 buggy, causing it to skip the first lines of messages whose message ID
5830 ended in 'D'. This would not have bitten before Exim release 4.14, because
5831 message IDs were unlikely to end in 'D' before then. The effect was to have
5832 incorrect size information for certain domains.
5834 45. #include "config.h" was missing at the start of the crypt16.c module. This
5835 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5838 46. If there was a timeout during a "random" callout check, Exim treated it as
5839 a failure of the random address, and carried on sending RSET and the real
5840 address. If the delay was just some slowness somewhere, the response to the
5841 original RCPT would be taken as a response to RSET and so on, causing
5842 mayhem of various kinds.
5844 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5845 when I implemented it. It didn't allow for the fact that some option values
5846 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5847 the right test for positive values.
5849 48. Domain names in DNS records are case-independent. Exim always looks them up
5850 in lower case. Some resolvers return domain names in exactly the case they
5851 appear in the zone file, that is, they may contain uppercase letters. Not
5852 all resolvers do this - some return always lower case. Exim was treating a
5853 change of case by a resolver as a change of domain, similar to a widening
5854 of a domain abbreviation. This triggered its re-routing code and so it was
5855 trying to route what was effectively the same domain again. This normally
5856 caused routing to fail (because the router wouldn't handle the domain
5857 twice). Now Exim checks for this case specially, and just changes the
5858 casing of the domain that it ultimately uses when it transmits the message
5861 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5864 50. If a filter generated a file delivery with a non-absolute name (possible if
5865 no home directory exists for the router), the forbid_file option was not
5868 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5869 the existing equality matching.
5871 52. Exim was using ints instead of ino_t variables in some places where it was
5872 dealing with inode numbers.
5874 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5875 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5876 called TMPDIR, and if it finds it is different, it changes its value.
5878 54. The smtp_printf() function is now made available to local_scan() so
5879 additional output lines can be written before returning. There is also an
5880 smtp_fflush() function to enable the detection of a dropped connection.
5881 The variables smtp_input and smtp_batched_input are exported to
5884 55. Changed the default runtime configuration: the message "Unknown user"
5885 has been removed from the ACL, and instead placed on the localuser router,
5886 using the cannot_route_message feature. This means that any verification
5887 failures that generate their own messages won't get overridden. Similarly,
5888 the "Unrouteable address" message that was in the ACL for unverifiable
5889 relay addresses has also been removed.
5891 56. Added hosts_avoid_esmtp to the smtp transport.
5893 57. The exicyclog script was not checking for the esoteric option
5894 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5895 will work only if exicyclog is run under the appropriate euid.
5897 58. Following a discussion on the list, the rules by which Exim recognises line
5898 endings on incoming messages have been changed. The -dropcr and drop_cr
5899 options are now no-ops, retained only for backwards compatibility. The
5900 following line terminators are recognized: LF CRLF CR. However, special
5901 processing applies to CR:
5903 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5904 nor a local message in the state where . is a terminator.
5906 (ii) If a bare CR is encountered in a header line, an extra space is added
5907 after the line terminator so as not to end the header. The reasoning
5908 behind this is that bare CRs in header lines are most likely either
5909 to be mistakes, or people trying to play silly games.
5911 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5912 was being incorrectly given as 18 bytes larger than it should have been.
5913 This is a VOB (very old bug).
5915 60. This may never have affected anything current, but just in case it has:
5916 When the local host is found other than at the start of a list of hosts,
5917 the local host, those with the same MX, and any that follow, are discarded.
5918 When the list in question was part of a longer list of hosts, the following
5919 hosts (not currently being processed) were also being discarded. This no
5920 longer happens. I'm not sure if this situation could ever has previously
5923 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5926 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5927 header. This is something that is recommended in a new Internet Draft, and
5928 is something that is documented as being done by Sendmail. There are two
5929 possible values. For messages generated by the autoreply transport, Exim
5932 Auto-Submitted: auto-replied
5934 whereas for all other generated messages (e.g. bounces) it adds
5936 Auto-Submitted: auto-generated
5938 63. The "personal" condition in filters now includes a test for the
5939 Auto-Submitted: header. If it contains the string "auto-" the message it
5940 not considered personal.
5942 64. Added rcpt_include_affixes as a generic transport option.
5944 65. Added queue_only_override (default true).
5946 66. Added the syslog_duplication option.
5948 67. If what should have been the first header line of a message consisted of
5949 a space followed by a colon, Exim was mis-interpreting it as a header line.
5950 It isn't of course - it is syntactically invalid and should therefore be
5951 treated as the start of the message body. The misbehaviour could have
5952 caused a number of strange effects, including loss of data in subsequent
5953 header lines, and spool format errors.
5955 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5956 client host had authenticated. This control can now be exercised by an ACL
5957 for more flexibility.
5959 69. By default, callouts do not happen when testing with -bh. There is now a
5960 variant, -bhc, which does actually run the callout code, including
5961 consulting and updating the callout cache.
5963 70. Added support for saslauthd authentication, courtesy of Alexander
5966 71. If statvfs() failed on the spool or log directories while checking their
5967 size for availability, Exim confusingly gave the error "space shortage".
5968 Furthermore, in debugging mode it crashed with a floating point exception.
5969 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5970 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5971 serious problem, Exim now writes to the main and panic logs when this
5972 happens, with details of the failure. It then refuses to accept the
5973 incoming message, giving the message "spool directory problem" or "log
5974 directory problem" with a 421 code for SMTP messages.
5976 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5977 0, 1, and 2 exist, because some OS complain for execs without them (see
5978 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5979 descriptors. However, the code omitted to check that the open succeeded,
5980 causing mysterious errors if for some reason the permissions on /dev/null
5981 got screwed. Now Exim writes a message to the main and panic logs, and
5982 bombs out if it can't open /dev/null.
5984 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5985 interact so that it is all more flexible. It is supposed to remain
5986 backwards compatible. Also added extra_local_interfaces.
5988 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5989 to bomb out with an assertion failure - to the client this appears as a
5990 connection drop. This problem occurs in the part of the code that was taken
5991 from the Samba project. Fortunately, the assertion is in a very simple
5992 function, so I have fixed this by reproducing the function inline in the
5993 one place where it is called, and arranging for authentication to fail
5994 instead of killing the process with assert().
5996 75. The SPA client code was not working when the server requested OEM rather
5997 than Unicode encoding.
5999 76. Added code to make require_files with a specific uid setting more usable in
6000 the case where statting the file as root fails - usually a non-root-mounted
6001 NFS file system. When this happens and the failure is EACCES, Exim now
6002 forks a subprocess and does the per-uid checking as the relevant uid.
6004 77. Added process_log_path.
6006 78. If log_file_path was not explicitly set, a setting of check_log_space or
6007 check_log_inodes was ignored.
6009 79. If a space check for the spool or log partitions fails, the incident is now
6010 logged. Of course, in the latter case the data may get lost...
6012 80. Added the %p formatting code to string_format() so that it can be used to
6013 print addresses in debug_print(). Adjusted all the address printing in the
6014 debugging in store.c to use %p rather than %d.
6016 81. There was a concern that a line of code in smtp_in.c could overflow a
6017 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6018 initially expressed, the concern was not well-founded, because trailing
6019 spaces are removed early. However, if the trailing spaces were followed by
6020 a NULL, they did not get removed, so the overflow was possible. Two fixes
6023 (a) I re-wrote the offending code in a cleaner fashion.
6024 (b) If an incoming SMTP command contains a NULL character, it is rejected
6027 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6028 runs initgroups(), so that if the Exim user is in any additional groups,
6029 they will be used during message reception.
6035 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob