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. Is a backward incompatibility if
46 the ACL tests for either of these two results. Patch contributed by
47 user bes-internal on the mailing list.
52 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
54 PP/02 Make -n do something, by making it not do something.
55 When combined with -bP, the name of an option is not output.
57 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
60 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
61 $sender_host_name and config options to manage this, and basic check
64 PP/05 DSCP support for outbound connections and control modifier for inbound.
66 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
67 (Only plugin which currently uses this is kerberos4, which nobody should
68 be using, but we should make it available and other future plugins might
69 conceivably use it, even though it would break NAT; stuff *should* be
70 using channel bindings instead).
72 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
73 name; added for Sendmail compatibility; requires admin caller.
74 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
75 just ignore it); requires trusted caller.
76 Also parse but ignore: -Ac -Am -X<logfile>
79 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
81 TL/02 Add +smtp_confirmation as a default logging option.
83 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
84 Patch by Magnus Holmgren from 2007-02-20.
86 TL/04 Bugzilla 1281 - Spec typo.
87 Bugzilla 1283 - Spec typo.
88 Bugzilla 1290 - Spec grammar fixes.
90 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
92 TL/06 Add Experimental DMARC support using libopendmarc libraries.
94 TL/07 Fix an out of order global option causing a segfault. Reported to dev
95 mailing list by by Dmitry Isaikin.
97 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
99 JH/02 Support "G" suffix to numbers in ${if comparisons.
101 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
103 NM/01 Bugzilla 1197 - Spec typo
104 Bugzilla 1196 - Spec examples corrections
106 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
108 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
109 gnutls_enable_pkcs11, but renamed to more accurately indicate its
112 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
113 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
115 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
116 "acl {{name}{arg}...}", and optional args on acl condition
119 JH/05 Permit multiple router/transport headers_add/remove lines.
121 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
123 JH/07 Avoid using a waiting database for a single-message-only transport.
124 Performance patch from Paul Fisher. Bugzilla 1262.
126 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
129 JH/09 Add $headers_added variable, with content from use of ACL modifier
130 add_header (but not yet added to the message). Bugzilla 199.
132 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
133 Pulled from Bugzilla 817 by Wolfgang Breyha.
135 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
137 (nb: this is the same fix as in Exim 4.80.1)
139 JH/11 Add A= logging on delivery lines, and a client_set_id option on
142 JH/12 Add optional authenticated_sender logging to A= and a log_selector
145 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
147 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
148 advertise SMTP AUTH mechanism to us, instead of a generic
149 protocol violation error. Also, make Exim more robust to bad
150 data from the Dovecot auth socket.
152 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
154 When a queue runner is handling a message, Exim first routes the
155 recipient addresses, during which it prunes them based on the retry
156 hints database. After that it attempts to deliver the message to
157 any remaining recipients. It then updates the hints database using
160 So if a recipient address works intermittently, it can get repeatedly
161 deferred at routing time. The retry hints record remains fresh so the
162 address never reaches the final cutoff time.
164 This is a fairly common occurrence when a user is bumping up against
165 their storage quota. Exim had some logic in its local delivery code
166 to deal with this. However it did not apply to per-recipient defers
167 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
169 This change adds a proper retry rule check during routing so that the
170 final cutoff time is checked against the message's age. We only do
171 this check if there is an address retry record and there is not a
172 domain retry record; this implies that previous attempts to handle
173 the address had the retry_use_local_parts option turned on. We use
174 this as an approximation for the destination being like a local
175 delivery, as in LMTP.
177 I suspect this new check makes the old local delivery cutoff check
178 redundant, but I have not verified this so I left the code in place.
180 TF/02 Correct gecos expansion when From: is a prefix of the username.
182 Test 0254 submits a message to Exim with the header
186 When I ran the test suite under the user fanf2, Exim expanded
187 the header to contain my full name, whereas it should have added
188 a Resent-Sender: header. It erroneously treats any prefix of the
189 username as equal to the username.
191 This change corrects that bug.
193 GF/01 DCC debug and logging tidyup
194 Error conditions log to paniclog rather than rejectlog.
195 Debug lines prefixed by "DCC: " to remove any ambiguity.
197 TF/03 Avoid unnecessary rebuilds of lookup-related code.
199 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
200 Bug spotted by Jeremy Harris; was flawed since initial commit.
201 Would have resulted in OCSP responses post-SNI triggering an Exim
202 NULL dereference and crash.
204 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
206 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
207 Bug detection, analysis and fix by Samuel Thibault.
208 Bugzilla 1331, Debian bug #698092.
210 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
212 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
213 Server implementation by Todd Lyons, client by JH.
214 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
215 config variable "prdr_enable" controls whether the server
216 advertises the facility. If the client requests PRDR a new
217 acl_data_smtp_prdr ACL is called once for each recipient, after
218 the body content is received and before the acl_smtp_data ACL.
219 The client is controlled by bolth of: a hosts_try_prdr option
220 on the smtp transport, and the server advertisement.
221 Default client logging of deliveries and rejections involving
222 PRDR are flagged with the string "PRDR".
224 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
225 fclose(). Diagnosis by Todd Lyons.
227 PP/17 Update configure.default to handle IPv6 localhost better.
228 Patch by Alain Williams (plus minor tweaks).
231 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
232 This is now consistent with GnuTLS, and is now documented: the
233 previous undocumented portable approach to treating the option as
234 unset was to force an expansion failure. That still works, and
235 an empty string is now equivalent.
237 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
238 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
239 not performing validation itself.
241 PP/20 Added force_command boolean option to pipe transport.
242 Patch from Nick Koston, of cPanel Inc.
244 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
247 TF/04 Added udpsend ACL modifer and hexquote expansion operator
249 PP/21 Fix eximon continuous updating with timestamped log-files.
250 Broken in a format-string cleanup in 4.80, missed when I repaired the
251 other false fix of the same issue.
252 Report and fix from Heiko Schlichting.
255 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
256 Report from Prashanth Katuri.
258 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
259 It's SecureTransport, so affects any MacOS clients which use the
260 system-integrated TLS libraries, including email clients.
262 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
263 using a MIME ACL for non-SMTP local injection.
264 Report and assistance in diagnosis by Warren Baker.
266 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
268 JH/16 Fix comparisons for 64b. Bugzilla 1385.
270 TL/09 Add expansion variable $authenticated_fail_id to keep track of
271 last id that failed so it may be referenced in subsequent ACL's.
273 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
276 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
277 ldap library initialization, allowing self-signed CA's to be
278 used. Also properly sets require_cert option later in code by
279 using NULL (global ldap config) instead of ldap handle (per
280 session). Bug diagnosis and testing by alxgomz.
282 TL/12 Enhanced documentation in the ratelimit.pl script provided in
283 the src/util/ subdirectory.
285 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
286 renamed to Transport Post Delivery Action by Jeremy Harris, as
289 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
290 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
291 redis_servers = needs to be configured which will be used by the redis
292 lookup. Patch from Warren Baker, of The Packet Hub.
294 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
296 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
297 hostname or reverse DNS when processing a host list. Used suggestions
298 from multiple comments on this bug.
300 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
302 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
303 Missed a few lines, added it to make the runtest require no keyboard
306 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
307 contains upper case chars. Make router use caseful_local_part.
309 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
310 support when GnuTLS has been built with p11-kit.
316 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
318 This, or similar/improved, will also be change PP/11 of 4.82.
324 PP/01 Handle short writes when writing local log-files.
325 In practice, only affects FreeBSD (8 onwards).
326 Bugzilla 1053, with thanks to Dmitry Isaikin.
328 NM/01 Bugzilla 949 - Documentation tweak
330 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
333 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
335 PP/02 Implemented gsasl authenticator.
337 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
339 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
340 `pkg-config foo` for cflags/libs.
342 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
343 with rest of GSASL and with heimdal_gssapi.
345 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
346 `pkg-config foo` for cflags/libs for the TLS implementation.
348 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
349 properties get this fed in as external SSF. A number of robustness
350 and debugging improvements to the cyrus_sasl authenticator.
352 PP/08 cyrus_sasl server now expands the server_realm option.
354 PP/09 Bugzilla 1214 - Log authentication information in reject log.
355 Patch by Jeremy Harris.
357 PP/10 Added dbmjz lookup type.
359 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
361 PP/12 MAIL args handles TAB as well as SP, for better interop with
362 non-compliant senders.
363 Analysis and variant patch by Todd Lyons.
365 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
366 Bug report from Lars Müller <lars@samba.org> (via SUSE),
367 Patch from Dirk Mueller <dmueller@suse.com>
369 PP/13 tls_peerdn now print-escaped for spool files.
370 Observed some $tls_peerdn in wild which contained \n, which resulted
371 in spool file corruption.
373 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
374 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
375 or write after TLS renegotiation, which otherwise led to messages
378 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
379 as a tracking header (ie: a signed header comes before the signature).
380 Patch from Wolfgang Breyha.
382 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
383 comma-sep list; embedded commas doubled.
385 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
387 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
389 Report and patch from Dmitry Banschikov.
391 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
392 Removed SSL_clear() after SSL_new() which led to protocol negotiation
393 failures. We appear to now support TLS1.1+ with Exim.
395 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
396 lets Exim select keys and certificates based upon TLS SNI from client.
397 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
398 before an outbound SMTP session. New log_selector, +tls_sni.
400 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
401 NULL dereference. Report and patch from Alun Jones.
403 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
404 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
405 Not seeing resolver debug output on NetBSD, but suspect this is a
406 resolver implementation change.
408 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
409 Left warnings. Added "eximon gdb" invocation mode.
411 PP/21 Defaulting "accept_8bitmime" to true, not false.
413 PP/22 Added -bw for inetd wait mode support.
415 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
416 locate the relevant includes and libraries. Made this the default.
418 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
419 Bugzilla 1246, report and most of solution from Tomasz Kusy.
421 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
422 This may cause build issues on older platforms.
424 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
425 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
426 gnutls_require_mac & gnutls_require_protocols (no longer supported).
427 Added SNI support via GnuTLS too.
428 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
430 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
432 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
433 Applied second patch from Janne, implementing suggestion to default
434 multiple-strings-in-record handling to match SPF spec.
436 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
438 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
439 read-only, out of scope).
440 Patch from Wolfgang Breyha, report from Stuart Northfield.
442 PP/29 Fix three issues highlighted by clang analyser static analysis.
443 Only crash-plausible issue would require the Cambridge-specific
444 iplookup router and a misconfiguration.
445 Report from Marcin Mirosław.
447 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
449 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
450 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
451 As part of this, removing so much warning spew let me fix some minor
452 real issues in debug logging.
454 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
455 assignment on my part. Fixed.
457 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
458 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
459 Janne Snabb (who went above and beyond: thank you).
461 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
462 string otherwise requires a connection and a bunch more work and it's
463 relatively easy to get wrong. Should also expose TLS library linkage
466 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
467 64-bit ${eval} (JH/03).
469 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
470 GNU libc to support some of the 64-bit stuff, should not lead to
471 conflicts. Defined before os.h is pulled in, so if a given platform
472 needs to override this, it can.
474 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
475 protection layer was required, which is not implemented.
476 Bugzilla 1254, patch from Wolfgang Breyha.
478 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
479 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
480 tls_dhparam take prime identifiers. Also unbreak combination of
481 OpenSSL+DH_params+TLSSNI.
483 PP/39 Disable SSLv2 by default in OpenSSL support.
489 PP/01 Solaris build fix for Oracle's LDAP libraries.
490 Bugzilla 1109, patch from Stephen Usher.
492 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
494 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
497 TF/02 Fix a couple more cases where we did not log the error message
498 when unlink() failed. See also change 4.74-TF/03.
500 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
501 lock up or crash if it happened to be inside a call to libc when it
502 got a SIGUSR1 from exiwhat.
504 The SIGUSR1 handler appends the current process status to the process
505 log which is later printed by exiwhat. It used to use the general
506 purpose logging code to do this, but several functions it calls are
507 not safe for signals.
509 The new output code in the SIGUSR1 handler is specific to the process
510 log, and simple enough that it's easy to inspect for signal safety.
511 Removing some special cases also simplifies the general logging code.
512 Removing the spurious timestamps from the process log simplifies
515 TF/04 Improved ratelimit ACL condition.
517 The /noupdate option has been deprecated in favour of /readonly which
518 has clearer semantics. The /leaky, /strict, and /readonly update modes
519 are mutually exclusive. The update mode is no longer included in the
520 database key; it just determines when the database is updated. (This
521 means that when you upgrde Exim will forget old rate measurements.)
523 Exim now checks that the per_* options are used with an update mode that
524 makes sense for the current ACL. For example, when Exim is processing a
525 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
526 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
527 must specify per_mail/readonly. If you omit the update mode it defaults to
528 /leaky where that makes sense (as before) or /readonly where required.
530 The /noupdate option is now undocumented but still supported for
531 backwards compatibility. It is equivalent to /readonly except that in
532 ACLs where /readonly is required you may specify /leaky/noupdate or
533 /strict/noupdate which are treated the same as /readonly.
535 A useful new feature is the /count= option. This is a generalization
536 of the per_byte option, so that you can measure the throughput of other
537 aggregate values. For example, the per_byte option is now equivalent
538 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
540 The per_rcpt option has been generalized using the /count= mechanism
541 (though it's more complicated than the per_byte equivalence). When it is
542 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
543 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
544 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
545 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
546 Note that using per_rcpt with a non-readonly update mode in more than
547 one ACL will cause the recipients to be double-counted. (The per_mail
548 and per_byte options don't have this problem.)
550 The handling of very low rates has changed slightly. If the computed rate
551 is less than the event's count (usually one) then this event is the first
552 after a long gap. In this case the rate is set to the same as this event's
553 count, so that the first message of a spam run is counted properly.
555 The major new feature is a mechanism for counting the rate of unique
556 events. The new per_addr option counts the number of different
557 recipients that someone has sent messages to in the last time period. It
558 behaves like per_rcpt if all the recipient addresses are different, but
559 duplicate recipient addresses do not increase the measured rate. Like
560 the /count= option this is a general mechanism, so the per_addr option
561 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
562 example, measure the rate that a client uses different sender addresses
563 with the options per_mail/unique=$sender_address. There are further
564 details in the main documentation.
566 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
568 TF/06 Removed a few PCRE remnants.
570 TF/07 Automatically extract Exim's version number from tags in the git
571 repository when doing development or release builds.
573 PP/02 Raise smtp_cmd_buffer_size to 16kB.
574 Bugzilla 879. Patch from Paul Fisher.
576 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
577 Heavily based on revision 40f9a89a from Simon Arlott's tree.
580 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
582 PP/05 Variable $av_failed, true if the AV scanner deferred.
583 Bugzilla 1078. Patch from John Horne.
585 PP/06 Stop make process more reliably on build failure.
586 Bugzilla 1087. Patch from Heiko Schlittermann.
588 PP/07 Make maildir_use_size_file an _expandable_ boolean.
589 Bugzilla 1089. Patch from Heiko Schlittermann.
591 PP/08 Handle ${run} returning more data than OS pipe buffer size.
592 Bugzilla 1131. Patch from Holger Weiß.
594 PP/09 Handle IPv6 addresses with SPF.
595 Bugzilla 860. Patch from Wolfgang Breyha.
597 PP/10 GnuTLS: support TLS 1.2 & 1.1.
599 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
602 PP/11 match_* no longer expand right-hand-side by default.
603 New compile-time build option, EXPAND_LISTMATCH_RHS.
604 New expansion conditions, "inlist", "inlisti".
606 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
608 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
610 PP/14 fix log_write() format string regression from TF/03.
611 Bugzilla 1152. Patch from Dmitry Isaikin.
617 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
619 PP/02 Harmonised TLS library version reporting; only show if debugging.
620 Layout now matches that introduced for other libraries in 4.74 PP/03.
622 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
624 PP/04 New "dns_use_edns0" global option.
626 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
629 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
630 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
632 TK/01 Updated PolarSSL code to 0.14.2.
633 Bugzilla 1097. Patch from Andreas Metzler.
635 PP/07 Catch divide-by-zero in ${eval:...}.
638 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
641 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
642 format-string attack -- SECURITY: remote arbitrary code execution.
644 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
645 time unintentionally subject to list matching rules, letting the header
646 cause arbitrary Exim lookups (of items which can occur in lists, *not*
647 arbitrary string expansion). This allowed for information disclosure.
649 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
650 INT_MIN/-1 -- value coerced to INT_MAX.
656 NM/01 Workround for PCRE version dependency in version reporting
659 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
660 This fixes portability to compilers other than gcc, notably
661 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
663 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
664 makefiles for portability to HP-UX and POSIX correctness.
666 PP/01 Permit LOOKUP_foo enabling on the make command-line.
667 Also via indented variable definition in the Makefile.
668 (Debugging by Oliver Heesakkers).
670 PP/02 Restore caching of spamd results with expanded spamd_address.
671 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
673 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
674 Improves build reliability. Fix from: Frank Elsner
676 NM/02 Fix wide character breakage in the rfc2047 coding
677 Fixes bug 1064. Patch from Andrey N. Oktyabrski
679 NM/03 Allow underscore in dnslist lookups
680 Fixes bug 1026. Patch from Graeme Fowler
682 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
683 Code patches from Adam Ciarcinski of NetBSD.
685 NM/04 Fixed exiqgrep to cope with mailq missing size issue
688 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
689 is logged, to avoid truncation. Patch from John Horne.
691 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
692 Patch from Jakob Hirsch.
694 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
695 SQL string expansion failure details.
696 Patch from Andrey Oktyabrski.
698 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
699 Patch from Simon Arlott.
701 PP/09 New lookups functionality failed to compile on old gcc which rejects
702 extern declarations in function scope.
703 Patch from Oliver Fleischmann
705 PP/10 Use sig_atomic_t for flags set from signal handlers.
706 Check getgroups() return and improve debugging.
707 Fixed developed for diagnosis in bug 927 (which turned out to be
710 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
711 Patch from Mark Zealey.
713 PP/12 Bugzilla 1056: Improved spamd server selection.
714 Patch from Mark Zealey.
716 PP/13 Bugzilla 1086: Deal with maildir quota file races.
717 Based on patch from Heiko Schlittermann.
719 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
720 Patch from Uwe Doering, sign-off by Michael Haardt.
722 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
723 variable declaration deep within a block. Bug and patch from
726 PP/15 lookups-Makefile IRIX compatibilty coercion.
728 PP/16 Make DISABLE_DKIM build knob functional.
730 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
731 Patch by Simon Arlott
733 TF/03 Fix valgrind.h portability to C89 compilers that do not support
734 variable argument macros. Our copy now differs from upstream.
740 TF/01 Failure to get a lock on a hints database can have serious
741 consequences so log it to the panic log.
743 TF/02 Log LMTP confirmation messages in the same way as SMTP,
744 controlled using the smtp_confirmation log selector.
746 TF/03 Include the error message when we fail to unlink a spool file.
748 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
749 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
750 for maintaining out-of-tree patches for some time.
752 PP/01 Bugzilla 139: Documentation and portability issues.
753 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
754 Handle per-OS dynamic-module compilation flags.
756 PP/02 Let /dev/null have normal permissions.
757 The 4.73 fixes were a little too stringent and complained about the
758 permissions on /dev/null. Exempt it from some checks.
759 Reported by Andreas M. Kirchwitz.
761 PP/03 Report version information for many libraries, including
762 Exim version information for dynamically loaded libraries. Created
763 version.h, now support a version extension string for distributors
764 who patch heavily. Dynamic module ABI change.
766 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
767 privilege escalation vulnerability whereby the Exim run-time user
768 can cause root to append content of the attacker's choosing to
771 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
774 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
775 If dropping privileges for untrusted macros, we disabled normal logging
776 on the basis that it would fail; for the Exim run-time user, this is not
777 the case, and it resulted in successful deliveries going unlogged.
778 Fixed. Reported by Andreas Metzler.
784 PP/01 Date: & Message-Id: revert to normally being appended to a message,
785 only prepend for the Resent-* case. Fixes regression introduced in
786 Exim 4.70 by NM/22 for Bugzilla 607.
788 PP/02 Include check_rfc2047_length in configure.default because we're seeing
789 increasing numbers of administrators be bitten by this.
791 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
793 PP/03 Bugzilla 994: added openssl_options main configuration option.
795 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
797 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
799 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
801 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
802 without a peer certificate, leading to a segfault because of an
803 assumption that peers always have certificates. Be a little more
804 paranoid. Problem reported by Martin Tscholak.
806 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
807 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
808 NB: ClamAV planning to remove STREAM in "middle of 2010".
809 CL also introduces -bmalware, various -d+acl logging additions and
810 more caution in buffer sizes.
812 PP/09 Implemented reverse_ip expansion operator.
814 PP/10 Bugzilla 937: provide a "debug" ACL control.
816 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
818 PP/12 Bugzilla 973: Implement --version.
820 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
822 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
824 PP/15 Bugzilla 816: support multiple condition rules on Routers.
826 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
827 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
828 ignore trailing whitespace.
830 JJ/02 prevent non-panic DKIM error from being sent to paniclog
832 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
835 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
836 Notification from Dr Andrew Aitchison.
838 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
839 ExtendedDetectionInfo response format.
840 Notification from John Horne.
842 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
845 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
846 XSL and documented dependency on system catalogs, with examples of how
849 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
852 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
853 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
854 configuration file which is writeable by the Exim user or group.
856 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
857 of configuration files to cover files specified with the -C option if
858 they are going to be used with root privileges, not just the default
861 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
862 option (effectively making it always true).
864 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
865 files to be used while preserving root privileges.
867 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
868 that rogue child processes cannot use them.
870 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
871 run-time user, instead of root.
873 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
874 Exim run-time user without dropping privileges.
876 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
877 result string, instead of calling string_vformat() twice with the same
880 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
881 for other users. Others should always drop root privileges if they use
882 -C on the command line, even for a whitelisted configure file.
884 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
886 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
892 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
893 $data_path, and $header_path variables; fixed documentation bugs and
896 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
897 exipick to access non-standard spools, including the "frozen" queue
900 NM/01 Bugzilla 965: Support mysql stored procedures.
901 Patch from Alain Williams
903 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
905 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
906 Patch from Andreas Metzler
908 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
909 Patch from Kirill Miazine
911 NM/05 Bugzilla 671: Added umask to procmail example.
913 JJ/03 installed exipick 20100323.0, fixing doc bug
915 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
916 directory. Notification and patch from Dan Rosenberg.
918 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
920 TK/02 Improve log output when DKIM signing operation fails.
922 MH/01 Treat the transport option dkim_domain as a colon separated
923 list, not as a single string, and sign the message with each element,
924 omitting multiple occurences of the same signer.
926 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
927 Bugzilla 985, 986. Patch by Simon Arlott
929 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
930 Patch by Simon Arlott
932 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
933 MBX locking. Notification from Dan Rosenberg.
939 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
941 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
943 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
945 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
947 NM/04 Bugzilla 915: Flag broken perl installation during build.
953 TK/01 Added patch by Johannes Berg that expands the main option
954 "spamd_address" if it starts with a dollar sign.
956 TK/02 Write list of recipients to X-Envelope-Sender header when building
957 the mbox-format spool file for content scanning (suggested by Jakob
960 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
961 (http://www.dcc-servers.net/) support via dccifd. Activated by
962 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
964 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
965 by Mark Daniel Reidel <mr@df.eu>.
967 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
968 When building exim an external PCRE library is now needed -
969 PCRE is a system library on the majority of modern systems.
970 See entry on PCRE_LIBS in EDITME file.
972 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
973 conversation. Added nologin parameter to request.
974 Patch contributed by Kirill Miazine.
976 TF/01 Do not log submission mode rewrites if they do not change the address.
978 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
980 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
981 log files in place. Contributed by Roberto Lima.
983 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
985 TF/03 Bugzilla 615: When checking the local_parts router precondition
986 after a local_part_suffix or local_part_prefix option, Exim now
987 does not use the address's named list lookup cache, since this
988 contains cached lookups for the whole local part.
990 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
991 Robert Millan. Documentation is in experimental-spec.txt.
993 TF/04 Bugzilla 668: Fix parallel build (make -j).
995 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
997 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
998 Patch provided by Jan Srzednicki.
1000 TF/05 Leading white space used to be stripped from $spam_report which
1001 wrecked the formatting. Now it is preserved.
1003 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1004 that they are available at delivery time.
1006 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1008 TF/08 TLS error reporting now respects the incoming_interface and
1009 incoming_port log selectors.
1011 TF/09 Produce a more useful error message if an SMTP transport's hosts
1012 setting expands to an empty string.
1014 NM/06 Bugzilla 744: EXPN did not work under TLS.
1015 Patch provided by Phil Pennock.
1017 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1018 Patch provided by Richard Godbee.
1020 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1021 acl_smtp_notquit, added index entry.
1023 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1024 Patch provided by Eugene Bujak.
1026 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1027 accept(). Patch provided by Maxim Dounin.
1029 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1030 Patch provided by Phil Pennock.
1032 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1034 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1035 Patch provided by Brad "anomie" Jorsch.
1037 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1038 Patch provided by Dean Brooks.
1040 TK/05 Add native DKIM support (does not depend on external libraries).
1042 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1043 Patch provided by Graeme Fowler.
1045 NM/16 Bugzilla 851: Documentation example syntax fix.
1047 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1049 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1052 NM/19 Bugzilla 745: TLS version reporting.
1053 Patch provided by Phil Pennock.
1055 NM/20 Bugzilla 167: bool: condition support.
1056 Patch provided by Phil Pennock.
1058 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1059 clients. Patch provided by Phil Pennock.
1061 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1062 Patch provided by Brad "anomie" Jorsch.
1064 NM/23 Bugzilla 687: Fix misparses in eximstats.
1065 Patch provided by Heiko Schlittermann.
1067 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1068 Patch provided by Heiko Schlittermann.
1070 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1071 plus update to original patch.
1073 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1075 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1076 Patch provided by David Brownlee.
1078 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1080 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1082 NM/30 Bugzilla 888: TLS documentation bugfixes.
1084 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1086 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1087 Unlike the original bugzilla I have changed all shell scripts in src tree.
1089 NM/33 Bugzilla 898: Transport filter timeout fix.
1090 Patch by Todd Rinaldo.
1092 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1093 Patch by Serge Demonchaux.
1095 NM/35 Bugzilla 39: Base64 decode bug fixes.
1096 Patch by Jakob Hirsch.
1098 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1100 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1102 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1104 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1110 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1111 ALT-N's libdkim that I have put here:
1112 http://duncanthrax.net/exim-experimental/
1114 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1115 were called 'true' and it seems that C99 defines that as a reserved
1116 keyword to be used with 'bool' variable types. That means you could
1117 not include C99-style headers which use bools without triggering
1118 build errors in sieve.c.
1120 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1121 as mailq or other aliases. Changed the --help handling significantly
1122 to do whats expected. exim_usage() emits usage/help information.
1124 SC/01 Added the -bylocaldomain option to eximstats.
1126 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1128 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1130 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1136 PH/01 Another patch from the Sieve maintainer.
1138 PH/02 When an IPv6 address is converted to a string for single-key lookup
1139 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1140 dots are used instead of colons so that keys in lsearch files need not
1141 contain colons. This was done some time before quoting was made available
1142 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1143 (notated using the quote facility) so as to distinguish them from IPv4
1144 keys. This meant that lookups for IP addresses in host lists did not work
1145 for iplsearch lookups.
1147 This has been fixed by arranging for IPv6 addresses to be expressed with
1148 colons if the lookup type is iplsearch. This is not incompatible, because
1149 previously such lookups could never work.
1151 The situation is now rather anomolous, since one *can* have colons in
1152 ordinary lsearch keys. However, making the change in all cases is
1153 incompatible and would probably break a number of configurations.
1155 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1158 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1159 conversion specification without a maximum field width, thereby enabling
1160 a rogue spamd server to cause a buffer overflow. While nobody in their
1161 right mind would setup Exim to query an untrusted spamd server, an
1162 attacker that gains access to a server running spamd could potentially
1163 exploit this vulnerability to run arbitrary code as the Exim user.
1165 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1166 $primary_hostname instead of what libspf2 thinks the hosts name is.
1168 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1169 a directory entry by the name of the lookup key. Previously, if a
1170 symlink pointed to a non-existing file or a file in a directory that
1171 Exim lacked permissions to read, a lookup for a key matching that
1172 symlink would fail. Now it is enough that a matching directory entry
1173 exists, symlink or not. (Bugzilla 503.)
1175 PH/03 The body_linecount and body_zerocount variables are now exported in the
1178 PH/04 Added the $dnslist_matched variable.
1180 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1181 This means they are set thereafter only if the connection becomes
1184 PH/06 Added the client_condition to authenticators so that some can be skipped
1185 by clients under certain conditions.
1187 PH/07 The error message for a badly-placed control=no_multiline_responses left
1188 "_responses" off the end of the name.
1190 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1192 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1193 (without spaces) instead of just copying the configuration text.
1195 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1197 PH/11 Added $max_received_linelength.
1199 PH/12 Added +ignore_defer and +include_defer to host lists.
1201 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1202 way in which PCRE > 7.0 is built.
1204 PH/14 Implemented queue_only_load_latch.
1206 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1207 MAIL command. The effect was to mangle the value on 64-bit systems.
1209 PH/16 Another patch from the Sieve maintainer.
1211 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1213 PH/18 If a system quota error occurred while trying to create the file for
1214 a maildir delivery, the message "Mailbox is full" was not appended to the
1215 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1216 only to a quota excession during the actual writing of the file.
1218 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1219 characters?) which causes problems in log lines. The DN values are now
1220 passed through string_printing() before being added to log lines.
1222 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1223 and InterBase are left for another time.)
1225 PH/21 Added message_body_newlines option.
1227 PH/22 Guard against possible overflow in moan_check_errorcopy().
1229 PH/23 POSIX allows open() to be a macro; guard against that.
1231 PH/24 If the recipient of an error message contained an @ in the local part
1232 (suitably quoted, of course), incorrect values were put in $domain and
1233 $local_part during the evaluation of errors_copy.
1239 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1240 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1243 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1244 issue a MAIL command.
1246 PH/02 In an ACL statement such as
1248 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1250 if a client was not listed at all, or was listed with a value other than
1251 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1252 the condition was not true (as it should be), so access was not denied.
1253 The bug was that the ! inversion was incorrectly passed on to the second
1254 item. This has been fixed.
1256 PH/03 Added additional dnslists conditions == and =& which are different from
1257 = and & when the dns lookup returns more than one IP address.
1259 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1260 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1262 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1263 FSYNC, which compiles an option called disable_fsync that allows for
1264 bypassing fsync(). The documentation is heavily laced with warnings.
1266 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1268 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1269 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1270 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1271 including adding "make clean"; (3) Added -fPIC when compiling the test
1272 dynamically loaded module, to get rid of a warning.
1274 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1275 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1276 = 0s. The bug is otherwise harmless.
1278 PH/07 There was a bug in the dovecot authenticator such that the value of
1279 $auth1 could be overwritten, and so not correctly preserved, after a
1280 successful authentication. This usually meant that the value preserved by
1281 the server_setid option was incorrect.
1283 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1285 PH/09 Installed PCRE release 7.0.
1287 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1288 run for batched SMTP input. It is now run at the start of every message
1289 in the batch. While fixing this I discovered that the process information
1290 (output by running exiwhat) was not always getting set for -bs and -bS
1291 input. This is fixed, and it now also says "batched" for BSMTP.
1293 PH/11 Added control=no_pipelining.
1295 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1296 patch, slightly modified), and move the expansion of helo_data till after
1297 the connection is made in the smtp transport (so it can use these
1300 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1302 PH/14 Added log_selector = +pid.
1304 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1306 PH/16 Add ${if forany and ${if forall.
1308 PH/17 Added dsn_from option to vary the From: line in DSNs.
1310 PH/18 Flush SMTP output before performing a callout, unless control =
1311 no_callout_flush is set.
1313 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1314 was true (the default) a successful delivery failed to delete the retry
1315 item, thus causing premature timeout of the address. The bug is now
1318 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1320 PH/21 Long custom messages for fakedefer and fakereject are now split up
1321 into multiline reponses in the same way that messages for "deny" and
1322 other ACL rejections are.
1324 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1325 with slight modification.
1327 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1328 draft, changing the syntax and factoring some duplicate code".
1330 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1331 for deliveries of the second and subsequent messages over the same SMTP
1334 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1335 ${reduce, with only minor "tidies".
1337 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1339 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1340 expansion side effects.
1342 PH/27 When a message times out after an over-quota error from an Exim-imposed
1343 quota, the bounce message says "mailbox is full". This message was not
1344 being given when it was a system quota that was exceeded. It now should
1347 MH/03 Made $recipients available in local_scan(). local_scan() already has
1348 better access to the recipient list through recipients_list[], but
1349 $recipients can be useful in postmaster-provided expansion strings.
1351 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1352 in the case of a MAIL command with additional options following the
1353 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1354 were accidentally chopped off.
1356 PH/29 SMTP synchronization checks are implemented when a command is read -
1357 there is a check that no more input is waiting when there shouldn't be
1358 any. However, for some commands, a delay in an ACL can mean that it is
1359 some time before the response is written. In this time, more input might
1360 arrive, invalidly. So now there are extra checks after an ACL has run for
1361 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1362 pipelining has not been advertised.
1364 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1366 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1367 returned in the SMTP response when smtp_return_error_details was false.
1368 This has been fixed.
1370 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1371 instead of the C I/O that was originally supplied, because problems were
1372 reported on Solaris.
1374 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1375 Exim which did not show up earlier: it was assuming that a call to
1376 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1377 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1378 was a macro that became an assignment, so it seemed to work. This has
1379 changed to a proper function call with a void return, hence the compile
1380 error. Exim's code has been fixed.
1382 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1385 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1387 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1389 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1390 because it turns out to be too big or there is a timeout) there is no
1391 "Completed" line in the log. When some messages of this type were
1392 selected by exigrep, they were listed as "not completed". Others were
1393 picked up by some special patterns. I have improved the selection
1394 criteria to be more general.
1396 PH/38 The host_find_failed option in the manualroute router can now be set
1397 to "ignore", to completely ignore a host whose IP address cannot be
1398 found. If all hosts are ignored, the behaviour is controlled by the new
1399 host_all_ignored option.
1401 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1402 homing or because of multiple MX records with /mx) generated more than
1403 one IP address, and the following item turned out to be the local host,
1404 all the secondary addresses of the first item were incorrectly removed
1405 from the list, along with the local host and any following hosts (which
1406 is what is supposed to happen).
1408 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1409 whoever called Exim into the -H file. In the case of the daemon it was
1410 behaving confusingly. When first started, it used values for whoever
1411 started the daemon, but after a SIGHUP it used the Exim user (because it
1412 calls itself on a restart). I have changed the code so that it now always
1415 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1416 message are rejected with the same error (e.g. no authentication or bad
1417 sender address), and a DATA command is nevertheless sent (as can happen
1418 with PIPELINING or a stupid MUA), the error message that was given to the
1419 RCPT commands is included in the rejection of the DATA command. This is
1420 intended to be helpful for MUAs that show only the final error to their
1423 PH/42 Another patch from the Sieve maintainer.
1425 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1426 Eximstats - Fixed some broken HTML links and added missing column headers
1428 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1429 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1431 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1433 PH/43 Yet another patch from the Sieve maintainer.
1435 PH/44 I found a way to check for a TCP/IP connection going away before sending
1436 the response to the final '.' that terminates a message, but only in the
1437 case where the client has not sent further data following the '.'
1438 (unfortunately, this is allowed). However, in many cases there won't be
1439 any further data because there won't be any more messages to send. A call
1440 to select() can be used: if it shows that the input is "ready", there is
1441 either input waiting, or the socket has been closed. An attempt to read
1442 the next input character can distinguish the two cases. Previously, Exim
1443 would have sent an OK response which the client would never have see.
1444 This could lead to message repetition. This fix should cure that, at
1445 least in a lot of common cases.
1447 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1448 advertised in response to EHLO.
1454 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1455 fixed by 4.65/MH/01 (is this a record?) are fixed:
1457 (i) An empty string was always treated as zero by the numeric comparison
1458 operators. This behaviour has been restored.
1460 (ii) It is documented that the numeric comparison operators always treat
1461 their arguments as decimal numbers. This was broken in that numbers
1462 starting with 0 were being interpreted as octal.
1464 While fixing these problems I realized that there was another issue that
1465 hadn't been noticed. Values of message_size_limit (both the global option
1466 and the transport option) were treated as octal if they started with 0.
1467 The documentation was vague. These values are now always treated as
1468 decimal, and I will make that clear in the documentation.
1474 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1475 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1478 MH/01 Don't check that the operands of numeric comparison operators are
1479 integers when their expansion is in "skipping" mode (fixes bug
1480 introduced by 4.64-PH/07).
1482 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1483 child addresses, Exim now panics and dies. Previously, because the count
1484 is held in a short int, deliveries were likely to be lost. As such a
1485 large number of recipients for a single message is ridiculous
1486 (performance will be very, very poor), I have chosen to impose a limit
1487 rather than extend the field.
1493 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1494 leftover -K file (the existence of which was triggered by #402).
1495 While we were at it, introduced process PID as part of the -K
1496 filename. This should rule out race conditions when creating
1499 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1500 processing considerably. Previous code took too long for large mails,
1501 triggering a timeout which in turn triggers #401.
1503 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1504 in the DK code in transports.c. sendfile() is not really portable,
1505 hence the _LINUX specificness.
1507 TF/01 In the add_headers option to the mail command in an Exim filter,
1508 there was a bug that Exim would claim a syntax error in any
1509 header after the first one which had an odd number of characters
1512 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1513 callout verification, Exim cached a "reject" for the entire domain. This
1514 is correct for most verifications, but it is not correct for a recipient
1515 verification with use_sender or use_postmaster set, because in that case
1516 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1517 case of MAIL FROM:<> rejection from other early rejections (e.g.
1518 rejection of HELO). When verifying a recipient using a non-null MAIL
1519 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1520 Whatever the result of the callout, the value of the domain cache is
1521 left unchanged (for any other kind of callout, getting as far as trying
1522 RCPT means that the domain itself is ok).
1524 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1527 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1528 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1529 session, the daemon ignores ECONNECT errors and logs others; it now
1530 ignores EPIPE as well.
1532 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1533 (quoted-printable decoding).
1535 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1536 later the small subsequent patch to fix an introduced bug.
1538 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1540 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1542 PH/08 An error is now given if message_size_limit is specified negative.
1544 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1545 to be given (somewhat) arbitrary names.
1547 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1550 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1551 miscellaneous code fixes
1553 PH/10 Added the log_reject_target ACL modifier to specify where to log
1556 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1557 hostname. This is wrong, because it relates to the incoming message (and
1558 probably the interface on which it is arriving) and not to the outgoing
1559 callout (which could be using a different interface). This has been
1560 changed to use the value of the helo_data option from the smtp transport
1561 instead - this is what is used when a message is actually being sent. If
1562 there is no remote transport (possible with a router that sets up host
1563 addresses), $smtp_active_hostname is used.
1565 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1566 tweaks were necessary in order to get it to work (see also 21 below):
1567 (a) The code assumed that strncpy() returns a negative number on buffer
1568 overflow, which isn't the case. Replaced with Exim's string_format()
1570 (b) There were several signed/unsigned issues. I just did the minimum
1571 hacking in of casts. There is scope for a larger refactoring.
1572 (c) The code used strcasecmp() which is not a standard C function.
1573 Replaced with Exim's strcmpic() function.
1574 (d) The code set only $1; it now sets $auth1 as well.
1575 (e) A simple test gave the error "authentication client didn't specify
1576 service in request". It would seem that Dovecot has changed its
1577 interface. Fortunately there's a specification; I followed it and
1578 changed what the client sends and it appears to be working now.
1580 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1583 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1584 address A is aliased to B and C, where B exists and C does not. Without
1585 -v the output is "A verified" because verification stops after a
1586 successful redirection if more than one address is generated. However,
1587 with -v the child addresses are also verified. Exim was outputting "A
1588 failed to verify" and then showing the successful verification for C,
1589 with its parentage. It now outputs "B failed to verify", showing B's
1590 parentage before showing the successful verification of C.
1592 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1593 look up a TXT record in a specific list after matching in a combined
1596 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1597 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1598 they consult the DNS. I had assumed they would set it the way they
1599 wanted; and indeed my experiments on Linux seem to show that in some
1600 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1601 To be on the safe side, however, I have now made the interface to
1602 host_find_byname() similar to host_find_bydns(), with an argument
1603 containing the DNS resolver options. The host_find_byname() function now
1604 sets these options at its start, just as host_find_bydns() does. The smtp
1605 transport options dns_qualify_single and dns_search_parents are passed to
1606 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1607 of host_find_byname() use the default settings of RES_DEFNAMES
1608 (qualify_single) but not RES_DNSRCH (search_parents).
1610 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1611 spool_read_header() do less string testing, by means of a preliminary
1612 switch on the second character of optional "-foo" lines. (This is
1613 overdue, caused by the large number of possibilities that now exist.
1614 Originally there were few.) While I was there, I also converted the
1615 str(n)cmp tests so they don't re-test the leading "-" and the first
1616 character, in the hope this might squeeze out yet more improvement.
1618 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1619 flag allowing group syntax was set by the header_syntax check but not
1620 turned off, possible causing trouble later; (2) The flag was not being
1621 set at all for the header_verify test, causing "group"-style headers to
1622 be rejected. I have now set it in this case, and also caused header_
1623 verify to ignore an empty address taken from a group. While doing this, I
1624 came across some other cases where the code for allowing group syntax
1625 while scanning a header line wasn't quite right (mostly, not resetting
1626 the flag correctly in the right place). These bugs could have caused
1627 trouble for malformed header lines. I hope it is now all correct.
1629 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1630 with the "reply" argument non-NULL. The code, however (which originally
1631 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1632 but it didn't always do it. This confused somebody who was copying the
1633 code for some other use. I have removed all the tests.
1635 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1636 feature that was used to support insecure browsers during the U.S. crypto
1637 embargo. It requires special client support, and Exim is probably the
1638 only MTA that supported it -- and would never use it because real RSA is
1639 always available. This code has been removed, because it had the bad
1640 effect of slowing Exim down by computing (never used) parameters for the
1641 RSA_EXPORT functionality.
1643 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1644 authenticator to fail if there's a tab character in the incoming data
1645 (there should never be unless someone is messing about, as it's supposed
1646 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1647 if the connection is using TLS or if the remote IP is the same as the
1648 local IP, and the "valid-client-cert option" if a client certificate has
1651 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1652 authenticators. This can be used for authorization after authentication
1653 succeeds. (In the case of plaintext, it servers for both authentication
1656 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1657 if any retry times were supplied.
1659 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1660 connection, where there is no client IP address to check. In this
1661 situation, the verify now always succeeds.
1663 PH/25 Applied John Jetmore's -Mset patch.
1665 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1667 PH/27 In a string expansion for a processed (not raw) header when multiple
1668 headers of the same name were present, leading whitespace was being
1669 removed from all of them, but trailing whitespace was being removed only
1670 from the last one. Now trailing whitespace is removed from each header
1671 before concatenation. Completely empty headers in a concatenation (as
1672 before) are ignored.
1674 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1675 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1677 PH/29 [Removed. This was a change that I later backed out, and forgot to
1678 correct the ChangeLog entry (that I had efficiently created) before
1679 committing the later change.]
1681 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1682 address errors (4xx response to RCPT) over the same connection as other
1683 messages routed to the same hosts. Such deliveries are always "forced",
1684 so retry times are not inspected. This resulted in far too many retries
1685 for the affected addresses. The effect occurred only when there were more
1686 hosts than the hosts_max_try setting in the smtp transport when it had
1687 the 4xx errors. Those hosts that it had tried were not added to the list
1688 of hosts for which the message was waiting, so if all were tried, there
1689 was no problem. Two fixes have been applied:
1691 (i) If there are any address or message errors in an SMTP delivery, none
1692 of the hosts (tried or untried) are now added to the list of hosts
1693 for which the message is waiting, so the message should not be a
1694 candidate for sending over the same connection that was used for a
1695 successful delivery of some other message. This seems entirely
1696 reasonable: after all the message is NOT "waiting for some host".
1697 This is so "obvious" that I'm not sure why it wasn't done
1698 previously. Hope I haven't missed anything, but it can't do any
1699 harm, as the worst effect is to miss an optimization.
1701 (ii) If, despite (i), such a delivery is accidentally attempted, the
1702 routing retry time is respected, so at least it doesn't keep
1703 hammering the server.
1705 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1706 in ${readsocket because some servers need this prod.
1708 PH/32 Added some extra debug output when updating a wait-xxx database.
1710 PH/33 The hint "could be header name not terminated by colon", which has been
1711 given for certain expansion errors for a long time, was not being given
1712 for the ${if def:h_colon_omitted{... case.
1714 PH/34 The spec says: "With one important exception, whenever a domain list is
1715 being scanned, $domain contains the subject domain." There was at least
1716 one case where this was not true.
1718 PH/35 The error "getsockname() failed: connection reset by peer" was being
1719 written to the panic log as well as the main log, but it isn't really
1720 panic-worthy as it just means the connection died rather early on. I have
1721 removed the panic log writing for the ECONNRESET error when getsockname()
1724 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1725 runs only) independently of the message's sender address. This meant
1726 that, if the 4xx error was in fact related to the sender, a different
1727 message to the same recipient with a different sender could confuse
1728 things. In particualar, this can happen when sending to a greylisting
1729 server, but other circumstances could also provoke similar problems.
1730 I have changed the default so that the retry time for these errors is now
1731 based a combination of the sender and recipient addresses. This change
1732 can be overridden by setting address_retry_include_sender=false in the
1735 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1736 remote server are returned as part of bounce messages. This was not
1737 happening for LMTP over a pipe (the lmtp transport), but now it is the
1738 same for both kinds of LMTP.
1740 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1741 in header lines that were in fact CNAMEs. This is no longer the case.
1743 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1744 and queue runs started by the daemon processed all messages. This has
1745 been fixed so that -R and -S can now usefully be given with -q<time>.
1747 PH/40 Import PCRE release 6.7 (fixes some bugs).
1749 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1751 PH/42 Give an error if -q is specified more than once.
1753 PH/43 Renamed the variables $interface_address and $interface_port as
1754 $received_ip_address and $received_port, to make it clear that these
1755 values apply to message reception, and not to the outgoing interface when
1756 a message is delivered. (The old names remain recognized, of course.)
1758 PH/44 There was no timeout on the connect() call when using a Unix domain
1759 socket in the ${readsocket expansion. There now is.
1761 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1762 be meaningful with "accept".
1764 SC/01 Eximstats V1.43
1765 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1767 SC/02 Eximstats V1.44
1768 Use a glob alias rather than an array ref in the generated
1769 parser. This improves both readability and performance.
1771 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1772 Collect SpamAssassin and rejection statistics.
1773 Don't display local sender or destination tables unless
1774 there is data to show.
1775 Added average volumes into the top table text output.
1777 SC/04 Eximstats V1.46
1778 Collect data on the number of addresses (recipients)
1779 as well as the number of messages.
1781 SC/05 Eximstats V1.47
1782 Added 'Message too big' to the list of mail rejection
1783 reasons (thanks to Marco Gaiarin).
1785 SC/06 Eximstats V1.48
1786 Mainlog lines which have GMT offsets and are too short to
1787 have a flag are now skipped.
1789 SC/07 Eximstats V1.49 (Alain Williams)
1790 Added the -emptyok flag.
1792 SC/08 Eximstats V1.50
1793 Fixes for obtaining the IP address from reject messages.
1795 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1796 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1797 whitesspace changes from 4.64-PH/27
1799 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1802 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1803 are found, allow negative numbers in numeric criteria)
1805 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1807 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1808 to match changes made in 4.64-PH/43
1810 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1812 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1813 "domain" is actually an IP address, and force a failure. This locks out
1814 those revolvers/nameservers that support "A-for-A" lookups, in
1815 contravention of the specifications.
1817 PH/48 When a host name was looked up from an IP address, and the subsequent
1818 forward lookup of the name timed out, the host name was left in
1819 $sender_host_name, contrary to the specification.
1821 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1822 restricted to single-key lookups, Exim was not diagnosing an error if
1823 * or *@ was used with a query-style lookup.
1825 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1827 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1828 long ago, but noone interested enough thought of it. Let's just say that
1829 the "1.1" means that there are some new functions that weren't there at
1830 some point in the past.
1832 PH/51 Error processing for expansion failure of helo_data from an smtp
1833 transport during callout processing was broken.
1835 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1836 tested/used via the -bh/-bhc/-bs options.
1838 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1839 bug, fixed in subsequent PCRE releases).
1841 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1842 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1844 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1850 SC/01 Use a glob alias rather than an array ref in eximstats generated
1851 parser. This improves both readability and performance.
1853 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1854 Don't display local sender or destination tables in eximstats unless
1855 there is data to show.
1856 Added average volumes into the eximstats top table text output.
1858 SC/03 Collect data on the number of addresses (recipients) as well
1859 as the number of messages in eximstats.
1861 TF/01 Correct an error in the documentation for the redirect router. Exim
1862 does (usually) call initgroups() when daemonizing.
1864 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1865 with consistent privilege compared to when running as a daemon.
1867 TF/03 Note in the spec that $authenticated_id is not set for local
1868 submissions from trusted users.
1870 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1871 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1873 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1874 by adding some example configuration directives to the default
1875 configuration file. A little bit of work is required to uncomment the
1876 directives and define how usernames and passwords are checked, but
1877 there is now a framework to start from.
1879 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1880 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1881 without this. I don't know how relevant this is to other LDAP libraries.
1883 PH/02 Add the verb name to the "unknown ACL verb" error.
1885 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1887 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1889 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1890 directory not expanded when it should be if an expanded home directory
1891 was set for the address (which is overridden by the transport).
1893 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1896 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1897 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1898 because it is too late at that time, and has no effect.
1900 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1901 security issue with \' (bugzilla #107). I could not use the
1902 PQescapeStringConn() function, because it needs a PGconn value as one of
1905 PH/08 When testing addresses using -bt, indicate those final addresses that
1906 are duplicates that would not cause an additional delivery. At least one
1907 person was confused, thinking that -bt output corresponded to deliveries.
1908 (Suppressing duplicates isn't a good idea as you lose the information
1909 about possibly different redirections that led to the duplicates.)
1911 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1912 systems where poll() doesn't work, in particular OS X.
1914 PH/10 Added more information to debugging output for retry time not reached.
1916 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1917 operations in malware.c.
1919 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1922 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1923 syslog_duplication set false, log lines that would normally be written
1924 both the the main log and to the reject log were not written to syslog at
1927 PH/14 In the default configuration, change the use of "message" in ACL warn
1928 statements to "add_header".
1930 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1931 not followed by a command (e.g. "seen endif").
1933 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1934 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1937 PH/17 Added extra conditions to the default value of delay_warning_condition
1941 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1942 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1943 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1946 The Auto-Submitted: and various List- headers are standardised, whereas I
1947 don't think Precedence: ever was.
1949 PH/18 Refactored debugging code in route_finduser() to show more information,
1950 in particular, the error code if getpwnam() issues one.
1952 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1953 This is apparently needed in addition to the PH/07 change above to avoid
1954 any possible encoding problems.
1956 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1957 but not after initializing Perl.
1959 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1960 output them only if debugging. By default they are written stderr,
1961 apparently, which is not desirable.
1963 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1966 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1969 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1971 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1972 authenticated or an ident call has been made. Suppress the default
1973 values for $authenticated_id and $authenticated_sender (but permit -oMai
1974 and -oMas) when testing with -bh.
1976 PH/24 Re-jigged the order of the tests in the default configuration so that the
1977 tests for valid domains and recipients precede the DNS black list and CSA
1978 tests, on the grounds that those ones are more expensive.
1980 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1981 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1982 This bug exists in every version of Exim that I still have, right back to
1985 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1986 However, an attempt to turn on case-sensitivity in a regex key by
1987 including (?-i) didn't work because the subject string was already
1988 lowercased, and the effects were non-intuitive. It turns out that a
1989 one-line patch can be used to allow (?-i) to work as expected.
1995 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1996 other effects) broke the use of negated acl sub-conditions.
1998 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2001 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2002 "Deny" causes Exim to reject the incoming connection with a 554 error.
2003 Unfortunately, if there is a major crisis, such as a disk failure,
2004 tcp-wrappers gives "deny", whereas what one would like would be some
2005 kind of temporary error. A kludge has been added to help with this.
2006 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2007 554 error is used if errno is still zero or contains ENOENT (which occurs
2008 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2011 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2013 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2014 errors. Otherwise a message that provokes a temporary error (when other
2015 messages do not) can cause a whole host to time out.
2017 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2018 addresses were routed directly to files or pipes from a redirect router.
2019 File deliveries just didn't batch; pipe deliveries might have suffered
2022 PH/06 A failure to get a lock for a hints database would erroneously always say
2023 "Failed to get write lock", even when it was really a read lock.
2025 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2026 of 0600. This has been changed to use the value of the lockfile_mode
2027 option (which defaults to 0600).
2029 PH/08 Applied small patch from the Sieve maintainer.
2031 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2032 folder from quota calculations, a direct delivery into this folder messed
2033 up the contents of the maildirsize file. This was because the regex was
2034 used only to exclude .Trash (or whatever) when the size of the mailbox
2035 was calculated. There was no check that a delivery was happening into an
2036 excluded directory. This bug has been fixed by ignoring all quota
2037 processing for deliveries into excluded directories.
2039 PH/10 Added the maildirfolder_create_regex option to appendfile.
2045 PH/01 The code for finding all the local interface addresses on a FreeBSD
2046 system running IPv6 was broken. This may well have applied to all BSD
2047 systems, as well as to others that have similar system calls. The broken
2048 code found IPv4 interfaces correctly, but gave incorrect values for the
2049 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2050 that it would not match correctly against @[] and not recognize the IPv6
2053 PH/02 The ipliteral router was not recognizing addresses of the form user@
2054 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2056 PH/03 Added disable_ipv6.
2058 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2059 decoded form, because it is most often used to construct To: headers
2060 lines in autoreplies, and the decoded form may well be syntactically
2061 invalid. However, $reply_address has leading white space removed, and all
2062 newlines turned into spaces so that the autoreply transport does not
2065 PH/05 If group was specified without a user on a router, and no group or user
2066 was specified on a transport, the group from the router was ignored.
2068 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2069 for visible compile-time settings that can be used to change these
2070 numbers, for those that want even more. Backwards compatibility with old
2071 spool files has been maintained. However, going back to a previous Exim
2072 release will lost any variables that are in spool files.
2074 PH/07 Two small changes when running in the test harness: increase delay when
2075 passing a TCP/IP connection to a new process, in case the original
2076 process has to generate a bounce, and remove special handling of
2077 127.0.0.2 (sic), which is no longer necessary.
2079 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2080 be the same on different OS.
2082 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2085 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2086 whether --show-vars was specified or not
2088 JJ/02 exipick: Added support for new ACL variable spool format introduced
2091 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2092 syntactically invalid From: or Reply-to: line, and a filter used this to
2093 generate an autoreply, and therefore failed to obtain an address for the
2094 autoreply, Exim could try to deliver to a non-existent relative file
2095 name, causing unrelated and misleading errors. What now happens is that
2096 it logs this as a hard delivery error, but does not attempt to create a
2099 PH/11 The exinext utility has a -C option for testing purposes, but although
2100 the given file was scanned by exinext itself; it wasn't being passed on
2101 when Exim was called.
2103 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2104 an end-of-file indication when reading a command response.
2106 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2107 compiled. In many other places in Exim, IPv6 addresses are always
2108 recognized, so I have changed this. It also means that IPv4 domain
2109 literals of the form [IPV4:n.n.n.n] are now always recognized.
2111 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2112 used if the router is not running as root, for example, when verifying at
2113 ACL time, or when using -bh. The debugging output from this situation was
2114 non-existent - all you got was a failure to exec. I have made two
2117 (a) Failures to set uid/gid, the current directory, or a process leader
2118 in a subprocess such as that created by queryprogram now generate
2119 suitable debugging ouput when -d is set.
2121 (b) The queryprogram router detects when it is not running as root,
2122 outputs suitable debugging information if -d is set, and then runs
2123 the subprocess without attempting to change uid/gid.
2125 PH/15 Minor change to Makefile for building test_host (undocumented testing
2128 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2129 additional section of a DNS packet that returns MX or SRV records.
2130 Instead, it always explicitly searches for A/AAAA records. This avoids
2131 major problems that occur when a DNS server includes only records of one
2132 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2133 fixed another bug: if SRV records were looked up and the corresponding
2134 address records were *not* found in the additional section, the port
2135 values from the SRV records were lost.
2137 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2138 using the correct key (the original address) when searching the retry
2139 rules in order to find which one to use for generating the retry hint.
2141 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2142 adding the default one. Similarly, if it contains a Reply-To: header, the
2143 errors_reply_to option, if set, is not used.
2145 PH/19 When calculating a retry time, Exim used to measure the "time since
2146 failure" by looking at the "first failed" field in the retry record. Now
2147 it does not use this if it is later than than the arrival time of the
2148 message. Instead it uses the arrival time. This makes for better
2149 behaviour in cases where some deliveries succeed, thus re-setting the
2150 "first failed" field. An example is a quota failure for a huge message
2151 when small messages continue to be delivered. Without this change, the
2152 "time since failure" will always be short, possible causing more frequent
2153 delivery attempts for the huge message than are intended.
2154 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2156 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2157 $1, $2, $3) because the numerical variables can be reset during some
2158 expansion items (e.g. "match"), thereby losing the authentication data.
2160 PH/21 Make -bV show the size of off_t variables so that the test suite can
2161 decide whether to run tests for quotas > 2G.
2163 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2164 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2165 filecount value is greater than 2G or if a quota value is greater than 2G
2166 on a system where the size of off_t is not greater than 4, a panic error
2169 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2170 never match. The debug and -bh output now contains an explicit error
2171 message indicating a malformed IPv4 address or mask.
2173 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2174 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2175 PH/23 above applies.
2177 PH/25 Do not write to syslog when running in the test harness. The only
2178 occasion when this arises is a failure to open the main or panic logs
2179 (for which there is an explicit test).
2181 PH/26 Added the /no_tell option to "control=freeze".
2183 PH/27 If a host name lookup failed very early in a connection, for example, if
2184 the IP address matched host_lookup and the reverse lookup yielded a name
2185 that did not have a forward lookup, an error message of the form "no IP
2186 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2187 could be logged. Now it outputs the IP address instead of "NULL".
2189 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2190 allows the sender and the authenticated sender to be set when
2191 submitting a message from within Exim. Since child_open_exim() is
2192 documented for local_scan(), the new function should be too.
2194 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2195 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2196 results in an empty string is now treated as unset.
2198 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2200 PH/31 Added sender_verify_fail logging option.
2202 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2203 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2204 tidied the source and removed it altogether.
2206 PH/33 When a queue run was abandoned because the load average was too high, a
2207 log line was always written; now it is written only if the queue_run log
2208 selector is set. In addition, the log line for abandonment now contains
2209 information about the queue run such as the pid. This is always present
2210 in "start" and "stop" lines but was omitted from the "abandon" line.
2212 PH/34 Omit spaces between a header name and the colon in the error message that
2213 is given when verify = headers_syntax fails (if there are lots of them,
2214 the message gets confusing).
2216 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2217 names, as there are now some PTR records that contain slashes. This check
2218 is only to protect against broken name servers that fall over on strange
2219 characters, so the fact that it applies to all lookups doesn't matter.
2221 PH/36 Now that the new test suite is complete, we can remove some of the
2222 special code in Exim that was needed for the old test suite. For example,
2223 sorting DNS records because real resolvers return them in an arbitrary
2224 order. The new test suite's fake resolver always returns records in the
2227 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2228 bounces) except when queue_only is set, to avoid logging races between
2229 the different processes.
2231 PH/38 Panic-die if .include specifies a non-absolute path.
2233 PH/39 A tweak to the "H" retry rule from its user.
2235 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2236 a label. They prevented compilation on older perls.
2238 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2239 a warning to be raised on newish perls.
2241 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2242 on queue. Changes to match documented behaviour of showing count of
2243 messages matching specified criteria.
2245 PH/40 Changed the default ident timeout from 30s to 5s.
2247 PH/41 Added support for the use of login_cap features, on those BSD systems
2248 that have them, for controlling the resources used by pipe deliveries.
2250 PH/42 The content-scanning code uses fopen() to create files in which to put
2251 message data. Previously it was not paying any attention to the mode of
2252 the files. Exim runs with umask(0) because the rest of the code creates
2253 files with open(), and sets the required mode explicitly. Thus, these
2254 files were ending up world-writeable. This was not a big issue, because,
2255 being within the spool directory, they were not world-accessible. I have
2256 created a function called modefopen, which takes an additional mode
2257 argument. It sets umask(777), creates the file, chmods it to the required
2258 mode, then resets the umask. All the relevant calls to fopen() in the
2259 content scanning code have been changed to use this function.
2261 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2262 to 24 hours. This avoids potential overflow problems when processing G
2263 and H retry rules. I suspect nobody ever tinkers with this value.
2265 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2267 PH/45 When the plaintext authenticator is running as a client, the server's
2268 challenges are checked to ensure they are valid base64 strings. By
2269 default, the authentication attempt is cancelled if an invalid string is
2270 received. Setting client_ignore_invalid_base64 true ignores these errors.
2271 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2272 they are received. Thus, the responses can be made to depend on the
2273 challenges. If an invalid string is ignored, an empty string is placed in
2276 PH/46 Messages that are created by the autoreply transport now contains a
2277 References: header, in accordance with RFCs 2822 and 3834.
2279 PH/47 Added authenticated_sender_force to the smtp transport.
2281 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2283 PH/49 Installed latest patch from the Sieve maintainer.
2285 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2286 was also set, the appendfile transport was unnecessarily scanning a
2287 directory of message files (e.g. for maildir delivery) to find the count
2288 of files (along with the size), even though it did not need this
2289 information. It now does the scan only if it needs to find either the
2290 size of the count of files.
2292 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2294 PH/52 Two bugs concerned with error handling when the smtp transport is
2297 (i) Exim was not creating retry information for temporary errors given
2298 for individual recipients after the DATA command when the smtp transport
2299 was used in LMTP mode. This meant that they could be retried too
2300 frequently, and not timed out correctly.
2302 (ii) Exim was setting the flag that allows error details to be returned
2303 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2304 recipients that were returned after the DATA command.
2306 PH/53 This is related to PH/52, but is more general: for any failing address,
2307 when detailed error information was permitted to be returned to the
2308 sender, but the error was temporary, then after the final timeout, only
2309 "retry timeout exceeded" was returned. Now it returns the full error as
2310 well as "retry timeout exceeded".
2312 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2313 do this, and (what is worse) MTAs that accept it.
2315 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2316 will now be deprecated.
2318 PH/56 New os.c-cygwin from the Cygwin maintainer.
2320 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2321 formats (previously only available in exim formats via -bpr, -bpru,
2322 and -bpra. Now also available in native and exiqgrep formats)
2324 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2325 with very large, slow to parse queues
2327 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2329 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2331 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2332 responses to authentication challenges, though it was showing the
2333 challenges; (ii) I've removed the CR characters from the debug output for
2336 PH/58 Allow for the insertion of a newline as well as a space when a string
2337 is turned into more than one encoded-word during RFC 2047 encoding. The
2338 Sieve code now uses this.
2340 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2341 data_4xx, lost_connection, tls_required.
2343 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2344 message was being sent as an SMTP response.
2346 PH/61 Add -l and -k options to exicyclog.
2348 PH/62 When verifying, if an address was redirected to one new address, so that
2349 verification continued, and the new address failed or deferred after
2350 having set something in $address_data, the value of $address_data was not
2351 passed back to the ACL. This was different to the case when no
2352 redirection occurred. The value is now passed back in both cases.
2354 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2355 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2356 use login_cap.h, so on its own it isn't the distinguishing feature. The
2357 new name refers directly to the setclassresources() function.
2359 PH/65 Added configuration files for NetBSD3.
2361 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2363 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2364 is preferred over IPv4.
2366 PH/68 The bounce_return_message and bounce_return_body options were not being
2367 honoured for bounces generated during the reception of non-SMTP messages.
2368 In particular, this applied to messages rejected by the ACL. This bug has
2369 been fixed. However, if bounce_return_message is true and bounce_return_
2370 body is false, the headers that are returned for a non-SMTP message
2371 include only those that have been read before the error was detected.
2372 (In the case of an ACL rejection, they have all been read.)
2374 PH/69 The HTML version of the specification is now built in a directory called
2375 spec_html instead of spec.html, because the latter looks like a path with
2376 a MIME-type, and this confuses some software.
2378 PH/70 Catch two compiler warnings in sieve.c.
2380 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2381 function verify_get_ident() calls ip_connect() to connect a socket, but
2382 if the "connect()" function timed out, ip_connect() used to close the
2383 socket. However, verify_get_ident() also closes the socket later, and in
2384 between Exim writes to the log, which may get opened at this point. When
2385 the socket was closed in ip_connect(), the log could get the same file
2386 descriptor number as the socket. This naturally causes chaos. The fix is
2387 not to close the socket in ip_connect(); the socket should be closed by
2388 the function that creates it. There was only one place in the code where
2389 this was missing, in the iplookup router, which I don't think anybody now
2390 uses, but I've fixed it anyway.
2392 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2393 well as to direct DNS lookups. Otherwise the handling of names in host
2394 lists is inconsistent and therefore confusing.
2400 PH/01 Two changes to the default runtime configuration:
2402 (1) Move the checks for relay_from_hosts and authenticated clients from
2403 after to before the (commented out) DNS black list checks.
2405 (2) Add control=submission to the relay_from_hosts and authenticated
2406 clients checks, on the grounds that messages accepted by these
2407 statements are most likely to be submissions.
2409 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2411 (1) Generate an error if the third argument for the ${prvs expansion is
2414 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2417 (3) Reset the variables that are obtained from the first argument of
2418 ${prvscheck and used in the second argument before leaving the code,
2419 because their memory is reclaimed, so using them afterwards may do
2422 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2423 one (it's much easier than Tom thought :-).
2425 (5) Because of (4), we can now allow for the use of $prvscheck_result
2426 inside the third argument.
2428 PH/03 For some reason, the default setting of PATH when running a command from
2429 a pipe transport was just "/usr/bin". I have changed it to
2432 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2433 anything to be listed in the output from -bV.
2435 PH/05 When a filter generated an autoreply, the entire To: header line was
2436 quoted in the delivery log line, like this:
2438 => >A.N.Other <ano@some.domain> <original@ddress> ...
2440 This has been changed so that it extracts the operative address. There
2441 may be more than one such address. If so, they are comma-separated, like
2444 => >ano@some.domain,ona@other.domain <original@ddress> ...
2446 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2447 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2448 not being looked up in the rDNS to get a host name, Exim was showing the
2449 IP address twice in Received: lines, even though the IP addresses were
2450 identical. For example:
2452 Received: from [1.2.3.4] (helo=[1.2.3.4])
2454 However, if the real host name was known, it was omitting the HELO data
2455 if it matched the actual IP address. This has been tidied up so that it
2456 doesn't show the same IP address twice.
2458 PH/07 When both +timestamp and +memory debugging was on, the value given by
2459 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2460 called by the debug printing, thereby overwriting the timestamp buffer.
2461 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2463 PH/08 When the original message was included in an autoreply transport, it
2464 always said "this is a copy of the message, including all the headers",
2465 even if body_only or headers_only was set. It now gives an appropriate
2468 PH/09 Applied a patch from the Sieve maintainer which:
2470 o fixes some comments
2471 o adds the (disabled) notify extension core
2472 o adds some debug output for the result of if/elsif tests
2473 o points to the current vacation draft in the documentation
2474 and documents the missing references header update
2478 o fixes a bug in processing the envelope test (when testing
2479 multiple envelope elements, the last element determinted the
2482 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2483 Electronic Mail") by including:
2485 Auto-submitted: auto-generated
2487 in the messages that it generates (bounce messages and others, such as
2488 warnings). In the case of bounce messages for non-SMTP mesages, there was
2489 also a typo: it was using "Auto_submitted" (underscore instead of
2490 hyphen). Since every message generated by Exim is necessarily in response
2491 to another message, thes have all been changed to:
2493 Auto-Submitted: auto-replied
2495 in accordance with these statements in the RFC:
2497 The auto-replied keyword:
2499 - SHOULD be used on messages sent in direct response to another
2500 message by an automatic process,
2502 - MUST NOT be used on manually-generated messages,
2504 - MAY be used on Delivery Status Notifications (DSNs) and Message
2505 Disposition Notifications (MDNs),
2507 - MUST NOT be used on messages generated by automatic or periodic
2508 processes, except for messages which are automatic responses to
2511 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2512 to the default Received: header definition.
2514 PH/12 Added log selector acl_warn_skipped (default on).
2516 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2517 variables because (a) they are in the wrong storage pool and (b) even if
2518 they were copied, it wouldn't work properly because of the caching.
2520 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2521 checking when decoding. Apparently there are clients that generate
2522 overlong encoded strings. Why am I not surprised?
2524 PH/15 If the first argument of "${if match_address" was not empty, but did not
2525 contain an "@" character, Exim crashed. Now it writes a panic log message
2526 and treats the condition as false.
2528 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2530 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2531 extension "envelope-auth". The code is finished and in agreement with
2532 other implementations, but there is no documentation so far and in fact,
2533 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2534 not changing the active code.
2536 Print executed "if" and "elsif" statements when debugging is used. This
2537 helps a great deal to understand what a filter does.
2539 Document more things not specified clearly in RFC3028. I had all this
2540 sorted out, when out of a sudden new issues came to my mind. Oops."
2542 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2545 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2546 canonical form (as documented). However, after a host name lookup from
2547 the IP address, check_host() was doing a simple string comparison with
2548 addresses acquired from the DNS when checking that the found name did
2549 have the original IP as one of its addresses. Since any found IPv6
2550 addresses are likely to be in abbreviated form, the comparison could
2551 fail. Luckily, there already exists a function for doing the comparison
2552 by converting both addresses to binary, so now that is used instead of
2553 the text comparison.
2555 PH/20 There was another similar case to PH/19, when a complete host name was
2556 given in a host list; looking up its IP address could give an abbreviated
2557 form, whereas the current host's name might or might not be abbreviated.
2558 The same fix has been applied.
2564 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2565 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2568 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2569 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2571 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2573 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2574 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2575 Also, for "Auto-Submitted", treat anything other than "no" as
2576 non-personal, in accordance with RFC 3834. (Previously it treated
2577 anything starting "auto-" as non-personal.)
2579 TF/01 The control=submission/name=... option had a problem with syntax
2580 errors if the name included a slash character. The /name= option
2581 now slurps the rest of the string, so it can include any characters
2582 but it must come last in the list of options (after /sender_retain
2585 PH/05 Some modifications to the interface to the fake nameserver for the new
2593 TK/01 Added the "success_on_redirect" address verification option. See
2594 NewStuff for rationale and an example.
2596 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2598 PH/02 Patch to exigrep to allow it to work on syslog lines.
2600 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2601 fread() to skip over the body file's header line, because in Cygwin the
2602 header line is locked and is inaccessible.
2604 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2605 co-exist for some time) to make it clear that it is the Exim ID that is
2606 referenced, not the Message-ID: header line.
2608 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2609 string_format() function, because snprintf() does not exist on all
2612 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2613 ${stat: expansion item.
2615 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2616 protocol synchronization error", to keep the pedants happy.
2618 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2619 well as for IRIX systems, when gcc is being used. See the host.c source
2622 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2624 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2627 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2628 transport and to the smtp transport in LMTP mode.
2630 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2632 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2633 run a filter in a subprocess. This could lead to confusion in subsequent
2634 lookups in the parent process. There should also be a search_tidyup() at
2635 the end of the subprocess.
2637 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2638 only if the host matched helo_try_verify_hosts, which caused the
2639 verification to occur when the EHLO/HELO command was issued. The ACL just
2640 tested the remembered result. Now, if a previous verification attempt has
2641 not happened, "verify = helo" does it there and then.
2643 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2645 TK/03 Fix log output including CR from clamd.
2647 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2648 exist provoked a memory error which could cause a segfault.
2650 PH/15 Installed PCRE 6.2
2652 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2654 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2655 of the problem. Specifically, suggested +O2 rather than +O1 for the
2658 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2660 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2661 record, the port was not correctly logged when the outgoing_port log
2662 selector was set (it logged the transort's default port).
2664 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2665 fallback_hosts, and "hosts" in the smtp transport.
2667 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2668 host errors such as "Connection refused".
2670 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2671 authentication with radiusclient 0.4.9:
2673 - Error returned from rc_read_config was caught wrongly
2674 - Username/password not passed on to radius server due to wrong length.
2676 The presumption is that some radiusclient API changes for 4.51/PH/17
2677 were not taken care of correctly. The code is still untested by me (my
2678 Linux distribution still has 0.3.2 of radiusclient), but it was
2679 contributed by a Radius user.
2681 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2682 expanding the "port" option of the smtp transport.
2684 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2685 while reading a MIME header. Thanks to Tom Hughes for a patch.
2687 PH/24 Include config.h inside local_scan.h so that configuration settings are
2690 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2691 that in an ACL for RCPT (for example), you can examine exactly what was
2694 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2695 commands, but it was not correctly comparing the address with the actual
2696 client host address. Thus, it would show the EHLO address in Received:
2697 header lines when this was not necessary.
2699 PH/27 Added the % operator to ${eval:}.
2701 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2702 it should be ignoring failures (because with -C, for example, it has lost
2703 privilege). It wasn't ignoring creation failures other than "already
2706 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2709 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2710 because an input line was too long, either on its own, or by virtue of
2711 too many continuations, the temporary file was not being removed, and the
2712 return code was incorrect.
2714 PH/31 Missing "BOOL" in function definition in filtertest.c.
2716 PH/32 Applied Sieve patches from the maintainer.
2718 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2720 PH/33 Added "verify = not_blind".
2722 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2723 Local/Makefile (with some defaults set). These are used in built scripts
2724 such as exicyclog, but they have never been used in the exim_install
2725 script (though there are many overriding facilities there). I have
2726 arranged that the exim_install script now takes note of these two
2729 PH/35 Installed configuration files for Dragonfly.
2731 PH/36 When a locally submitted message by a trusted user did not contain a
2732 From: header, and the sender address was obtained from -f or from an SMTP
2733 MAIL command, and the trusted user did not use -F to supply a sender
2734 name, $originator_name was incorrectly used when constructing a From:
2735 header. Furthermore, $originator_name was used for submission mode
2736 messages from external hosts without From: headers in a similar way,
2737 which is clearly wrong.
2739 PH/37 Added control=suppress_local_fixups.
2741 PH/38 When log_selector = +received_sender was set, and the addition of the
2742 sender made the log line's construction buffer exactly full, or one byte
2743 less than full, an overflow happened when the terminating "\n" was
2746 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2747 when the result of a list match is failure because a DNS lookup failed.
2749 PH/40 RM_COMMAND is now used in the building process.
2751 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2752 the "build-* directories that it finds.
2754 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2755 address in a domain literal was a prefix of an interface address.
2757 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2758 when verifying a sender address, unless rewrite_headers is false.
2760 PH/44 Wrote a long comment about why errors_to addresses are verified as
2761 recipients, not senders.
2763 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2764 the ratelimit ACL was added.
2766 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2768 PH/46 Added extra information about PostgreSQL errors to the error string.
2770 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2771 avoiding the need to install special zones in a real server. This is
2772 backwards compatible; if it can't find the fake resolver, it drops back.
2773 Thus, both old and new test suites can be run.
2775 TF/02 Added util/ratelimit.pl
2777 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2778 clock is set back in time.
2780 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2781 Candler <B.Candler@pobox.com>.
2783 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2784 OK for addresses that are the result of redirections.
2786 PH/48 A number of further additions for the benefit of the new test suite,
2787 including a fake gethostbyname() that interfaces to the fake DNS resolver
2790 TF/06 The fix for widen_domains has also been applied to qualify_single and
2791 search_parents which are the other dnslookup options that can cause
2794 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2797 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2799 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2800 service" if it encountered an MX record with an empty target hostname.
2801 The message is now "an MX or SRV record indicated no SMTP service".
2803 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2804 if the DNS of the sending site is misconfigured. This is quite a
2805 common situation. This change restores the behaviour of treating a
2806 helo verification defer as a failure.
2808 PH/51 If self=fail was set on a router, the bounce message did not include the
2809 actual error message.
2815 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2817 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2818 command itself ended in error, the underlying message about the transport
2819 filter timeout was being overwritten with the pipe command error. Now the
2820 underlying error message should be appended to the second error message.
2822 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2824 PH/02 Reduce the amount of output that "make" produces by default. Full output
2825 can still be requested.
2827 PH/03 The warning log line about a condition test deferring for a "warn" verb
2828 was being output only once per connection, rather than after each
2829 occurrence (because it was using the same function as for successful
2830 "warn" verbs). This seems wrong, so I have changed it.
2832 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2833 it should not have, which might have caused a crash in the right
2834 circumstances, but probably never did.
2836 PH/04 Installed a modified version of Tony Finch's patch to make submission
2837 mode fix the return path as well as the Sender: header line, and to
2838 add a /name= option so that you can make the user's friendly name appear
2841 TF/03 Added the control = fakedefer ACL modifier.
2843 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2844 Mark Lowes for thorough testing.
2846 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2848 TK/03 Merged latest SRS patch from Miles Wilton.
2850 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2851 with the definition in sysexits.h (which is #included earlier).
2852 Fortunately, Exim does not actually use EX_OK. The code used to try to
2853 preserve the sysexits.h value, by assumimg that macro definitions were
2854 scanned for macro replacements. I have been disabused of this notion,
2855 so now the code just undefines EX_OK before #including unistd.h.
2857 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2858 in the smtp transport. When a block could not be written in a single
2859 write() function, the timeout was being re-applied to each part-write.
2860 This seems wrong - if the receiver was accepting one byte at a time it
2861 would take for ever. The timeout is now adjusted when this happens. It
2862 doesn't have to be particularly precise.
2864 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2865 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2867 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2868 without a domain if the check to <postmaster@domain> fails.
2870 SC/01 Eximstats: added -xls and the ability to specify output files
2871 (patch written by Frank Heydlauf).
2873 SC/02 Eximstats: use FileHandles for outputing results.
2875 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2877 SC/04 Eximstats: fixed display of large numbers with -nvr option
2879 SC/05 Eximstats: fixed merging of reports with empty tables.
2881 SC/06 Eximstats: added the -include_original_destination flag
2883 SC/07 Eximstats: removed tabs and trailing whitespace.
2885 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2887 TK/06 MBOX spool code: Add real "From " MBOX separator line
2888 so the .eml file is really in mbox format (even though
2889 most programs do not really care). Patch from Alex Miller.
2891 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2892 The latter is generated from $received_to and is only set if the
2893 message has one envelope recipient. SA can use these headers,
2894 obviously out-of-the-box. Patch from Alex Miller.
2896 PH/08 The ${def test on a variable was returning false if the variable's
2897 value was "0", contrary to what the specification has always said!
2898 The result should be true unless the variable is empty.
2900 PH/09 The syntax error of a character other than { following "${if
2901 def:variable_name" (after optional whitespace) was not being diagnosed.
2902 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2903 accidental colon was present, for example, could give incorrect results.
2905 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2906 result is used (not including appendfile, where other changes are about
2909 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2910 This involved changing a lot of size variables from int to off_t. It
2911 should work with maildirs and everything.
2913 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2914 spamd dying while we are connected to it.
2916 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2919 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2920 function for BDB 4.3.
2922 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2924 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2925 the sources for PCRE have been re-arranged and more files are now
2928 PH/15 The code I had for printing potentially long long variables in PH/11
2929 above was not the best (it lost precision). The length of off_t variables
2930 is now inspected at build time, and an appropriate printing format (%ld
2931 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2932 to be "long long int" or "long int". This is needed for the internal
2933 formatting function string_vformat().
2935 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2936 the configuration file to be ":syslog", then the script "guesses" where
2937 the logs files are, rather than using the compiled in default. In our
2938 case the guess is not the same as the compiled default, so the script
2939 suddenly stopped working when I started to use syslog. The patch checks
2940 to see if log_file_path is "". If so, it attempts to read it from exim
2941 with no configuration file to get the compiled in version, before it
2942 falls back to the previous guessing code."
2944 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2945 implementing BATV in an Exim configuration. See NewStuff for the gory
2948 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2949 Makefile that are specific to HP-UX.
2951 PH/18 If the "use_postmaster" option was set for a recipient callout together
2952 with the "random" option, the postmaster address was used as the MAIL
2953 FROM address for the random test, but not for the subsequent recipient
2954 test. It is now used for both.
2956 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2957 patch removes a few documentation additions to RFC 3028, because the
2958 latest draft now contains them. It adds the new en;ascii-case comparator
2959 and a new error check for 8bit text in MIME parts. Comparator and
2960 require names are now matched exactly. I enabled the subaddress
2961 extension, but it is not well tested yet (read: it works for me)."
2963 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2964 rework some of the code of TK/09 above to avoid the hardwired use of
2965 "%lld" and "long long". Replaced the call to snprintf() with a call to
2968 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2969 records point to non-existent hosts", "retry timeout exceeded", and
2970 "retry time not reached for any host after a long failure period".
2972 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2973 experimental DomainKeys support:
2975 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2976 (2) On an error such as an illegally used "control", the wrong name for
2977 the control was given.
2979 These problems did NOT occur unless DomainKeys support was compiled.
2981 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2983 PH/24 Added ${if match_ip condition.
2985 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2986 databases so that it will be absolutely obvious if a crash occurs in the
2987 DB library. This is a regular occurrence (often caused by mis-matched
2990 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2991 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2992 picked up on a user's system that detects such things. There doesn't seem
2993 to be a gcc warning option for this - only an attribute that has to be
2994 put on the function's prototype. It seems that in Fedora Core 4 they have
2995 set this on a number of new functions. No doubt there will be more in due
2998 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2999 specify a transport. However, if an address that was verified by such a
3000 router was the subject of a callout, Exim crashed because it tried to
3001 read the rcpt_include_affixes from the non-existent transport. Now it
3002 just assumes that the setting of that option is false. This bug was
3003 introduced by 4.51/PH/31.
3005 PH/28 Changed -d+all to exclude +memory, because that information is very
3006 rarely of interest, but it makes the output a lot bigger. People tend to
3007 do -d+all out of habit.
3009 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3010 code in os-type was giving problems when libc.so lives in lib64, like on
3013 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3014 aren't the modern standard, and it seems that some systems' include files
3015 don't always have them. Exim was already checking for some of the newer
3016 ones like T_AAAA, and defining it itself. I've added checks for all the
3017 record types that Exim uses.
3019 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3020 not automatically generating a new one, as it is supposed to. This
3021 prevented TLS from working. If the file did exist, but contained invalid
3022 data, a new version was generated, as expected. It was only the case of a
3023 non-existent file that was broken.
3025 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3026 with a change in libdomainkeys > 0.64.
3028 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3029 from DNS. If the selector record carries the flag, it now has
3030 precedence over the domain-wide flag.
3032 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3034 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3035 the use of a port name that isn't defined in /etc/services) occurred, the
3036 message was deferred as in a normal delivery, and thus remained on the
3037 spool, instead of being failed because of the mua_wrapper setting. This
3038 is now fixed, and I tidied up some of the mua_wrapper messages at the
3041 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3042 the messages in a hash of arrays rather than using individual hashes.
3043 This is a bit cleaner and results in dramatic memory savings, albeit
3044 at a slight CPU cost.
3046 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3047 as requested by Marc Sherman.
3049 SC/10 Eximstats: added histograms for user specified patterns as requested
3052 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3054 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3055 fopen() in the content-scanning modules that did not already have it.
3061 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3062 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3064 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3066 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3068 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3069 to test Sieve filters that use "vacation".
3071 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3072 that changes the way the GnuTLS parameters are stored in the cache file.
3073 The new format can be generated externally. For backward compatibility,
3074 if the data in the cache doesn't make sense, Exim assumes it has read an
3075 old-format file, and it generates new data and writes a new file. This
3076 means that you can't go back to an older release without removing the
3079 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3080 work if there are any delivery delays because "one_time" forces the
3081 parent to be marked "delivered", so its unseen clone is never tried
3082 again. For this reason, Exim now forbids the simultaneous setting of
3085 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3086 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3087 entry if you want to know the details. The fix, however, neglected to
3088 consider the case when local delivery batching is involved. The test for
3089 "previously delivered" was not happening when checking to see if an
3090 address could be batched with a previous (undelivered) one; under
3091 certain circumstances this could lead to multiple deliveries to the same
3094 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3095 in its include files, and this causes problems building Exim.
3097 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3098 header_syntax) but Exim was just ignoring anything given after a slash.
3099 In particular, this caused confusion with an attempt to use "verify =
3100 reverse_host_lookup/defer_ok". An error is now given when options are
3101 supplied for verify items that do not have them. (Maybe reverse_host_
3102 lookup should have a defer_ok option, but that's a different point.)
3104 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3105 defined by RFC 821) to 2048, because there were problems with some AUTH
3106 commands, and RFC 1869 says the size should be increased for extended
3107 SMTP commands that take arguments.
3109 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3112 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3113 "unknown" error; now it says that the functionality isn't in the binary.
3115 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3116 an address' error message when a string expansion fails (syntax or
3117 whatever). Otherwise the password may appear in the log. Following change
3118 PH/42 below, there is no longer a chance of it appearing in a bounce
3121 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3123 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3124 of its addresses was ever tried. (Bugzilla bug #2.)
3126 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3127 the result incorrectly in the debug output. (It correctly added a newline
3128 to what was transported.)
3130 TF/01 Added $received_time.
3132 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3133 commented out examples of how to interface to a virus scanner and to
3134 SpamAssassin. Also added commented examples of av_scanner and
3135 spamd_address settings.
3137 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3138 and controls are allowed in which ACLs. There were a couple of minor
3139 errors. Some of the entries in the conditions table (which is a table of
3140 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3141 negation of where the condition IS allowed.
3143 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3145 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3146 header file does not have a version number, so I've had to invent a new
3147 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3148 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3149 radiusclient), but it was contributed by a Radius user.
3151 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3152 files or directories, for OpenSSL.
3154 PH/19 When an Exim process that is running as root has to create an Exim log
3155 file, it does so in a subprocess that runs as exim:exim so as to get the
3156 ownership right at creation (otherwise, other Exim processes might see
3157 the file with the wrong ownership). There was no test for failure of this
3158 fork() call, which would lead to the process getting stuck as it waited
3159 for a non-existent subprocess. Forks do occasionally fail when resources
3160 run out. I reviewed all the other calls to fork(); they all seem to check
3163 PH/20 When checking for unexpected SMTP input at connect time (before writing
3164 the banner), Exim was not dealing correctly with a non-positive return
3165 from the read() function. If the client had disconnected by this time,
3166 the result was a log entry for a synchronization error with an empty
3167 string after "input=" when read() returned zero. If read() returned -1
3168 (an event I could not check), uninitialized data bytes were printed.
3169 There were reports of junk text (parts of files, etc) appearing after
3172 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3174 PH/22 Added support for macro redefinition, and (re)definition in between
3175 driver and ACL definitions.
3177 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3178 forgetting to use the resulting value; it was using the unexpanded value.
3180 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3181 hadn't been configured. The fix is from Juergen Kreileder, who
3182 understands it better than I do:
3184 "Here's what I see happening with three configured cyrus_sasl
3185 authenticators configured (plain, login, cram-md5):
3187 On startup auth_cyrus_sasl_init() gets called for each of these.
3188 This means three calls to sasl_listmech() without a specified mech_list.
3189 => SASL tests which mechs of all available mechs actually work
3190 => three warnings about OTP not working
3191 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3193 With the patch, sasl_listmech() also gets called three times. But now
3194 SASL's mech_list option is set to the server_mech specified in the the
3195 authenticator. Or in other words, the answer from sasl_listmech()
3196 gets limited to just the mech you're testing for (which is different
3198 => the return list contains just 'plain' or 'login', 'cram-md5' or
3199 nothing depending on the value of ob->server_mech.
3201 I've just tested the patch: Authentication still works fine,
3202 unavailable mechs specified in the exim configuration are still
3203 caught, and the auth.log warnings about OTP are gone."
3205 PH/25 When debugging is enabled, the contents of the command line are added
3206 to the debugging output, even when log_selector=+arguments is not
3209 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3210 answer is "GNU", and only if the return is "GNU/something" is the answer
3213 PH/27 $acl_verify_message is now set immediately after the failure of a
3214 verification in an ACL, and so is available in subsequent modifiers. In
3215 particular, the message can be preserved by coding like this:
3217 warn !verify = sender
3218 set acl_m0 = $acl_verify_message
3220 Previously, $acl_verify_message was set only while expanding "message"
3221 and "log_message" when a very denied access.
3223 PH/28 Modified OS/os.c-Linux with
3225 -#ifndef OS_LOAD_AVERAGE
3226 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3228 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3229 nomenclature these days.)
3231 PH/29 Installed patch from the Sieve maintainer that adds the options
3232 sieve_useraddress and sieve_subaddress to the redirect router.
3234 PH/30 In these circumstances:
3235 . Two addresses routed to the same list of hosts;
3236 . First host does not offer TLS;
3237 . First host accepts first address;
3238 . First host gives temporary error to second address;
3239 . Second host offers TLS and a TLS session is established;
3240 . Second host accepts second address.
3241 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3242 and peerdn, if requested) that were in fact used only for the second
3245 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3246 attention to any local part prefix or suffix that was matched by the
3247 router that accepted the address. It now behaves in the same way as it
3248 does for delivery: the affixes are removed from the local part unless
3249 rcpt_include_affixes is set on the transport.
3251 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3252 timeout during the DATA phase of an incoming message.
3254 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3255 applied a patch sanctioned by the Sieve maintainer.
3257 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3258 the uid or gid is negative. A case of a negative gid caused this to be
3259 noticed. The fix allows for either to be negative.
3261 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3262 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3265 PH/36 The change PH/12 above was broken. Fixed it.
3267 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3268 the grounds that routing the same address twice would always produce the
3269 same answer. This might have been true once, but it is certainly no
3270 longer true now. Routing a child address may depend on the previous
3271 routing that produced that child. Some complicated redirection strategies
3272 went wrong when messages had multiple recipients, and made Exim's
3273 behaviour dependent on the order in which the addresses were given.
3275 I have moved the duplicate checking until after the routing is complete.
3276 Exim scans the addresses that are assigned to local and remote
3277 transports, and removes any duplicates. This means that more work will be
3278 done, as duplicates will always all be routed, but duplicates are
3279 presumably rare, so I don't expect this is of any significance.
3281 For deliveries to pipes, files, and autoreplies, the duplicate checking
3282 still happens during the routing process, since they are not going to be
3285 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3286 It corrects a timeout issue with spamd. This is Ian's comment: "The
3287 background is that sometimes spamd either never reads data from a
3288 connection it has accepted, or it never writes response data. The exiscan
3289 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3290 blindly assumes that writes won't block so it may never time out."
3292 PH/39 Allow G after quota size as well as K and M.
3294 PH/40 The value set for $authenticated_id in an authenticator may not contain
3295 binary zeroes or newlines because the value is written to log lines and
3296 to spool files. There was no check on this. Now the value is run through
3297 the string_printing() function so that such characters are converted to
3298 printable escape sequences.
3300 PH/41 $message_linecount is a new variable that contains the total number of
3301 lines in the message. Compare $body_linecount, which is the count for the
3304 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3305 bounce and delay warning messages, except in certain special cases, which
3308 (a) An SMTP error message from a remote host;
3309 (b) A message specified in a :fail: redirection;
3310 (c) A message specified in a "fail" command in a system filter;
3311 (d) A message specified in a FAIL return from the queryprogram router;
3312 (e) A message specified by the cannot_route_message router option.
3314 In these cases only, Exim does include the error details in bounce and
3315 warning messages. There are also a few cases where bland messages such
3316 as "unrouteable address" or "local delivery error" are given.
3318 PH/43 $value is now also set for the "else" part of a ${run expansion.
3320 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3321 being worked on, but at least Exim now implements the latest version to
3324 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3325 process to complete was treated as a delivery failure, a timeout while
3326 writing the message to the pipe was logged, but erroneously treated as a
3327 successful delivery. Such timeouts include transport filter timeouts. For
3328 consistency with the overall process timeout, these timeouts are now
3329 treated as errors, giving rise to delivery failures by default. However,
3330 there is now a new Boolean option for the pipe transport called
3331 timeout_defer, which, if set TRUE, converts the failures into defers for
3332 both kinds of timeout. A transport filter timeout is now identified in
3335 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3336 systems where "make" and "gmake" are different, calling "gmake" at top
3337 level broke things. I've arranged for the value of $(MAKE) to be passed
3338 from the Makefile to this script so that it can call the same version of
3342 A note about Exim versions 4.44 and 4.50
3343 ----------------------------------------
3345 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3346 changes of various kinds. As a consequence, a big documentation update was
3347 needed. This delayed the release for rather longer than seemed good, especially
3348 in the light of a couple of (minor) security issues. Therefore, the changes
3349 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3350 release. So 4.44 and 4.50 are in effect two different branches that both start
3353 I have left the 4.50 change log unchanged; it contains all the changes since
3354 4.43. The change log for 4.44 is below; many of its items are identical to
3355 those for 4.50. This seems to be the most sensible way to preserve the
3356 historical information.
3362 1. Minor wording change to the doc/README.SIEVE file.
3364 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3365 computation of the current number of files was incorrect.
3367 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3368 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3371 4. Give more explanation in the error message when the command for a transport
3372 filter fails to execute.
3374 5. There are several places where Exim runs a non-Exim command in a
3375 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3376 was being done only for the command run by the queryprogram router. It is
3377 now done for all such subprocesses. The other cases are: ${run, transport
3378 filters, and the commands run by the lmtp and pipe transports.
3380 6. Added CONFIGURE_GROUP build-time option.
3382 7. Some older OS have a limit of 256 on the maximum number of file
3383 descriptors. Exim was using setrlimit() to set 1000 as a large value
3384 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3385 systems. I've change it so that if it can't get 1000, it tries for 256.
3387 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3388 was an oversight, and furthermore, ever since the addition of extra
3389 controls (e.g. 4.43/32), the checks on when to allow different forms of
3390 "control" were broken. There should now be diagnostics for all cases when a
3391 control that does not make sense is encountered.
3393 9. Added the /retain_sender option to "control=submission".
3395 10. $recipients is now available in the predata ACL (oversight).
3397 11. Tidy the search cache before the fork to do a delivery from a message
3398 received from the command line. Otherwise the child will trigger a lookup
3399 failure and thereby defer the delivery if it tries to use (for example) a
3400 cached ldap connection that the parent has called unbind on.
3402 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3403 of $address_data from the recipient verification was clobbered by the
3404 sender verification.
3406 13. The value of address_data from a sender verification is now available in
3407 $sender_address_data in subsequent conditions in the ACL statement.
3409 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3411 15. Added a new option "connect=<time>" to callout options, to set a different
3414 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3415 was its contents. (It was OK if the option was not defined at all.)
3417 17. A "Completed" log line is now written for messages that are removed from
3418 the spool by the -Mrm option.
3420 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3421 information about exactly what failed.
3423 19. Added -dd to debug only the daemon process.
3425 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3426 handles timeouts, both on the server side and network timeouts. Renamed the
3427 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3429 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3430 It is now set to "smtps".
3432 22. $host_address is now set to the target address during the checking of
3433 ignore_target_hosts.
3435 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3436 being passed; this would have caused $sender_host_name to have been used if
3437 matching the list had actually called for a host name (not very likely,
3438 since this list is usually IP addresses). A host name is now passed as
3441 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3442 code that specifies a non-restarting handler (typically sigaction() in
3443 modern systems) in an attempt to fix a rare and obscure crash bug.
3445 25. Narrowed the window for a race in the daemon that could cause it to ignore
3446 SIGCHLD signals. This is not a major problem, because they are used only to
3447 wake it up if nothing else does.
3449 26. A malformed maildirsize file could cause Exim to calculate negative values
3450 for the mailbox size or file count. Odd effects could occur as a result.
3451 The maildirsize information is now recalculated if the size or filecount
3454 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3455 support for a long time. Removed HAVE_SYS_VFS_H.
3457 28. Installed the latest version of exipick from John Jetmore.
3459 29. In an address list, if the pattern was not a regular expression, an empty
3460 subject address (from a bounce message) matched only if the pattern was an
3461 empty string. Non-empty patterns were not even tested. This was the wrong
3462 because it is perfectly reasonable to use an empty address as part of a
3463 database query. An empty address is now tested by patterns that are
3464 lookups. However, all the other forms of pattern expect the subject to
3465 contain a local part and a domain, and therefore, for them, an empty
3466 address still always fails if the pattern is not itself empty.
3468 30. Exim went into a mad DNS loop when attempting to do a callout where the
3469 host was specified on an smtp transport, and looking it up yielded more
3470 than one IP address.
3472 31. Re-factored the code for checking spool and log partition space into a
3473 function that finds that data and another that does the check. The former
3474 is then used to implement four new variables: $spool_space, $log_space,
3475 $spool_inodes, and $log_inodes.
3477 32. The RFC2047 encoding function was originally intended for short strings
3478 such as real names; it was not keeping to the 75-character limit for
3479 encoded words that the RFC imposes. It now respects the limit, and
3480 generates multiple encoded words if necessary. To be on the safe side, I
3481 have increased the buffer size for the ${rfc2047: expansion operator from
3484 33. It is now permitted to omit both strings after an "if" condition; if the
3485 condition is true, the result is "true". As before, when the second string
3486 is omitted, a false condition yields an empty string. This makes it less
3487 cumbersome to write custom ACL and router conditions.
3489 34. Failure to deliver a bounce message always caused it to be frozen, even if
3490 there was an errors_to setting on the router. The errors_to setting is now
3493 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3494 canonical form (fully expanded) before being placed in
3495 $sender_host_address.
3497 36. The table in the code that translates DNS record types into text (T_A to
3498 "A" for instance) was missing entries for NS and CNAME. It is just possible
3499 that this could have caused confusion if both these types were looked up
3500 for the same domain, because the text type is used as part of Exim's
3501 per-process caching. But the chance of anyone hitting this buglet seems
3504 37. The dnsdb lookup has been extended in a number of ways.
3506 (1) There is a new type, "zns", which walks up the domain tree until it
3507 finds some nameserver records. It should be used with care.
3509 (2) There is a new type, "mxh", which is like "mx" except that it returns
3510 just the host names, not the priorities.
3512 (3) It is now possible to give a list of domains (or IP addresses) to be
3513 looked up. The behaviour when one of the lookups defers can be
3514 controlled by a keyword.
3516 (4) It is now possible to specify the separator character for use when
3517 multiple records are returned.
3519 38. The dnslists ACL condition has been extended: it is now possible to supply
3520 a list of IP addresses and/or domains to be looked up in a particular DNS
3523 39. Added log_selector=+queue_time_overall.
3525 40. When running the queue in the test harness, wait just a tad after forking a
3526 delivery process, to get repeatability of debugging output.
3528 41. Include certificate and key file names in error message when GnuTLS fails
3529 to set them up, because the GnuTLS error message doesn't include the name
3530 of the failing file when there is a problem reading it.
3532 42. Allow both -bf and -bF in the same test run.
3534 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3536 44. The "Exiscan patch" is now merged into the mainline Exim source.
3538 45. Sometimes the final signoff response after QUIT could fail to get
3539 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3540 < 0 before doing a fflush(). This bug looks as though it goes back to the
3541 introduction of TLS in release 3.20, but "sometimes" must have been rare
3542 because the tests only now provoked it.
3544 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3545 (this can affect the format of dates).
3547 47. exim_tidydb, when checking for the continued existence of a message for
3548 which it has found a message-specific retry record, was not finding
3549 messages that were in split spool directories. Consequently, it was
3550 deleting retry records that should have stayed in existence.
3552 48. Steve fixed some bugs in eximstats.
3554 49. The SPA authentication driver was not abandoning authentication and moving
3555 on to the next authenticator when an expansion was forced to fail,
3556 contradicting the general specification for all authenticators. Instead it
3557 was generating a temporary error. It now behaves as specified.
3559 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3560 (the order specifies the preference for clients). The order is now AES256,
3561 AES128, 3DES, ARCFOUR128.
3563 51. Small patch to Sieve code - explicitly set From: when generating an
3566 52. Exim crashed if a remote delivery caused a very long error message to be
3567 recorded - for instance if somebody sent an entire SpamAssassin report back
3568 as a large number of 550 error lines. This bug was coincidentally fixed by
3569 increasing the size of one of Exim's internal buffers (big_buffer) that
3570 happened as part of the Exiscan merge. However, to be on the safe side, I
3571 have made the code more robust (and fixed the comments that describe what
3574 53. Now that there can be additional text after "Completed" in log lines (if
3575 the queue_time_overall log selector is set), a one-byte patch to exigrep
3576 was needed to allow it to recognize "Completed" as not the last thing in
3579 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3580 patch that reportedly fixes this has been added. I am not expert enough to
3581 create a test for it. This is what the patch creator wrote:
3583 "I found a little strange behaviour of ldap code when working with
3584 Windows 2003 AD Domain, where users was placed in more than one
3585 Organization Units. When I tried to give exim partial DN, the exit code
3586 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3587 But simultaneously result of request was absolutely normal ldap result,
3588 so I produce this patch..."
3590 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3591 REFERENCE, so I have modified the code to exclude the patch when that macro
3594 55. Some experimental protocols are using DNS PTR records for new purposes. The
3595 keys for these records are domain names, not reversed IP addresses. The
3596 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3597 leaves it alone. Component reversal etc. now happens only for IP addresses.
3600 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3602 57. Double the size of the debug message buffer (to 2048) so that more of very
3603 long debug lines gets shown.
3605 58. The exicyclog utility now does better if the number of log files to keep
3606 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3608 59. Two changes related to the smtp_active_hostname option:
3610 (1) $smtp_active_hostname is now available as a variable.
3611 (2) The default for smtp_banner uses $smtp_active_hostname instead
3612 of $primary_hostname.
3614 60. The host_aton() function is supposed to be passed a string that is known
3615 to be a valid IP address. However, in the case of IPv6 addresses, it was
3616 not checking this. This is a hostage to fortune. Exim now panics and dies
3617 if the condition is not met. A case was found where this could be provoked
3618 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3619 components; fortuitously, this particular loophole had already been fixed
3620 by change 4.50/55 above.
3622 If there are any other similar loopholes, the new check in host_aton()
3623 itself should stop them being exploited. The report I received stated that
3624 data on the command line could provoke the exploit when Exim was running as
3625 exim, but did not say which command line option was involved. All I could
3626 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3627 running as the user.
3630 61. There was a buffer overflow vulnerability in the SPA authentication code
3631 (which came originally from the Samba project). I have added a test to the
3632 spa_base64_to_bits() function which I hope fixes it.
3635 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3636 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3638 63. The daemon start-up calls getloadavg() while still root for those OS that
3639 need the first call to be done as root, but it missed one case: when
3640 deliver_queue_load_max is set with deliver_drop_privilege. This is
3641 necessary for the benefit of the queue runner, because there is no re-exec
3642 when deliver_drop_privilege is set.
3644 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3645 This has been fixed.
3647 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3648 was in use, was not putting the data itself into the right store pool;
3649 consequently, it could be overwritten for a subsequent message in the same
3650 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3653 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3655 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3656 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3657 the function were treating the return as a boolean value, which happened to
3658 work because 0=false and not-0=true, but is not correct code.
3660 68. The host_aton() function was not handling scoped IPv6 addresses (those
3661 with, for example, "%eth0" on the end) correctly.
3663 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3664 negated items (that is, ~something) in unsigned ints. Some compilers
3665 apparently mutter when there is no cast.
3667 70. If an address verification called from an ACL failed, and did not produce a
3668 user-specific message (i.e. there was only a "system" message), nothing was
3669 put in $acl_verify_message. In this situation, it now puts the system
3672 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3673 session; change 4.31/43 added the unwanted input to the log line - except
3674 that it did not do this in the start of session case. It now does.
3676 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3677 This is wrong and can cause the other end to generate a synchronization
3678 error if it is another Exim or anything else that does the synchronization
3679 check. A QUIT command is no longer sent after a timeout.
3681 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3682 during host lookups.
3684 74. The defer_ok option of callout verification was not working if it was used
3685 when verifying addresses in header lines, that is, for this case:
3687 verify = header_sender/callout=defer_ok
3689 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3690 those file descriptors could be used for SMTP connections. If anything
3691 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3692 could be sent to the SMTP client, causing chaos. The daemon now opens
3693 stdin, stdout, and stderr to /dev/null when it puts itself into the
3696 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3697 log by default. The user can override this with suitable Perl magic.
3699 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3700 the log message when discard triggers, was not working for the DATA ACL or
3701 for the non-SMTP ACL.
3703 78. Error message wording change in sieve.c.
3705 79. If smtp_accept_max_per_host was set, the number of connections could be
3706 restricted to fewer than expected, because the daemon was trying to set up
3707 a new connection before checking whether the processes handling previous
3708 connections had finished. The check for completed processes is now done
3709 earlier. On busy systems, this bug wouldn't be noticed because something
3710 else would have woken the daemon, and it would have reaped the completed
3713 80. If a message was submitted locally by a user whose login name contained one
3714 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3715 It caused a spool format error. I have fixed the spool reading code. A
3716 related problem was that the "from" clause in the Received: line became
3717 illegal because of the space(s). It is now covered by ${quote_local_part.
3719 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3722 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3724 83. Patch from Sieve maintainer to fix unterminated string problem in
3725 "vacation" handling.
3727 84. Some minor changes to the Linux configuration files to help with other
3728 OS variants using glibc.
3730 85. One more patch for Sieve to update vacation handling to latest spec.
3733 ----------------------------------------------------
3734 See the note above about the 4.44 and 4.50 releases.
3735 ----------------------------------------------------
3741 1. Change 4.43/35 introduced a bug that caused file counts to be
3742 incorrectly computed when quota_filecount was set in an appendfile
3745 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3746 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3749 3. Give more explanation in the error message when the command for a transport
3750 filter fails to execute.
3752 4. There are several places where Exim runs a non-Exim command in a
3753 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3754 was being done only for the command run by the queryprogram router. It is
3755 now done for all such subprocesses. The other cases are: ${run, transport
3756 filters, and the commands run by the lmtp and pipe transports.
3758 5. Some older OS have a limit of 256 on the maximum number of file
3759 descriptors. Exim was using setrlimit() to set 1000 as a large value
3760 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3761 systems. I've change it so that if it can't get 1000, it tries for 256.
3763 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3764 was an oversight, and furthermore, ever since the addition of extra
3765 controls (e.g. 4.43/32), the checks on when to allow different forms of
3766 "control" were broken. There should now be diagnostics for all cases when a
3767 control that does not make sense is encountered.
3769 7. $recipients is now available in the predata ACL (oversight).
3771 8. Tidy the search cache before the fork to do a delivery from a message
3772 received from the command line. Otherwise the child will trigger a lookup
3773 failure and thereby defer the delivery if it tries to use (for example) a
3774 cached ldap connection that the parent has called unbind on.
3776 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3777 of $address_data from the recipient verification was clobbered by the
3778 sender verification.
3780 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3781 was its contents. (It was OK if the option was not defined at all.)
3783 11. A "Completed" log line is now written for messages that are removed from
3784 the spool by the -Mrm option.
3786 12. $host_address is now set to the target address during the checking of
3787 ignore_target_hosts.
3789 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3790 being passed; this would have caused $sender_host_name to have been used if
3791 matching the list had actually called for a host name (not very likely,
3792 since this list is usually IP addresses). A host name is now passed as
3795 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3796 code that specifies a non-restarting handler (typically sigaction() in
3797 modern systems) in an attempt to fix a rare and obscure crash bug.
3799 15. Narrowed the window for a race in the daemon that could cause it to ignore
3800 SIGCHLD signals. This is not a major problem, because they are used only to
3801 wake it up if nothing else does.
3803 16. A malformed maildirsize file could cause Exim to calculate negative values
3804 for the mailbox size or file count. Odd effects could occur as a result.
3805 The maildirsize information is now recalculated if the size or filecount
3808 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3809 support for a long time. Removed HAVE_SYS_VFS_H.
3811 18. Updated exipick to current release from John Jetmore.
3813 19. Allow an empty sender to be matched against a lookup in an address list.
3814 Previously the only cases considered were a regular expression, or an
3817 20. Exim went into a mad DNS lookup loop when doing a callout where the
3818 host was specified on the transport, if the DNS lookup yielded more than
3821 21. The RFC2047 encoding function was originally intended for short strings
3822 such as real names; it was not keeping to the 75-character limit for
3823 encoded words that the RFC imposes. It now respects the limit, and
3824 generates multiple encoded words if necessary. To be on the safe side, I
3825 have increased the buffer size for the ${rfc2047: expansion operator from
3828 22. Failure to deliver a bounce message always caused it to be frozen, even if
3829 there was an errors_to setting on the router. The errors_to setting is now
3832 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3833 canonical form (fully expanded) before being placed in
3834 $sender_host_address.
3836 24. Updated eximstats to version 1.33
3838 25. Include certificate and key file names in error message when GnuTLS fails
3839 to set them up, because the GnuTLS error message doesn't include the name
3840 of the failing file when there is a problem reading it.
3842 26. Expand error message when OpenSSL has problems setting up cert/key files.
3845 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3846 (this can affect the format of dates).
3848 28. exim_tidydb, when checking for the continued existence of a message for
3849 which it has found a message-specific retry record, was not finding
3850 messages that were in split spool directories. Consequently, it was
3851 deleting retry records that should have stayed in existence.
3853 29. eximstats updated to version 1.35
3854 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3855 1.35 - bugfix such that pie charts by volume are generated correctly
3857 30. The SPA authentication driver was not abandoning authentication and moving
3858 on to the next authenticator when an expansion was forced to fail,
3859 contradicting the general specification for all authenticators. Instead it
3860 was generating a temporary error. It now behaves as specified.
3862 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3863 (the order specifies the preference for clients). The order is now AES256,
3864 AES128, 3DES, ARCFOUR128.
3866 31. Small patch to Sieve code - explicitly set From: when generating an
3869 32. Exim crashed if a remote delivery caused a very long error message to be
3870 recorded - for instance if somebody sent an entire SpamAssassin report back
3871 as a large number of 550 error lines. This bug was coincidentally fixed by
3872 increasing the size of one of Exim's internal buffers (big_buffer) that
3873 happened as part of the Exiscan merge. However, to be on the safe side, I
3874 have made the code more robust (and fixed the comments that describe what
3877 33. Some experimental protocols are using DNS PTR records for new purposes. The
3878 keys for these records are domain names, not reversed IP addresses. The
3879 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3880 leaves it alone. Component reversal etc. now happens only for IP addresses.
3883 34. The host_aton() function is supposed to be passed a string that is known
3884 to be a valid IP address. However, in the case of IPv6 addresses, it was
3885 not checking this. This is a hostage to fortune. Exim now panics and dies
3886 if the condition is not met. A case was found where this could be provoked
3887 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3888 components; fortuitously, this particular loophole had already been fixed
3889 by change 4.50/55 or 4.44/33 above.
3891 If there are any other similar loopholes, the new check in host_aton()
3892 itself should stop them being exploited. The report I received stated that
3893 data on the command line could provoke the exploit when Exim was running as
3894 exim, but did not say which command line option was involved. All I could
3895 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3896 running as the user.
3899 35. There was a buffer overflow vulnerability in the SPA authentication code
3900 (which came originally from the Samba project). I have added a test to the
3901 spa_base64_to_bits() function which I hope fixes it.
3904 36. The daemon start-up calls getloadavg() while still root for those OS that
3905 need the first call to be done as root, but it missed one case: when
3906 deliver_queue_load_max is set with deliver_drop_privilege. This is
3907 necessary for the benefit of the queue runner, because there is no re-exec
3908 when deliver_drop_privilege is set.
3910 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3911 was in use, was not putting the data itself into the right store pool;
3912 consequently, it could be overwritten for a subsequent message in the same
3913 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3916 38. Sometimes the final signoff response after QUIT could fail to get
3917 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3918 < 0 before doing a fflush(). This bug looks as though it goes back to the
3919 introduction of TLS in release 3.20, but "sometimes" must have been rare
3920 because the tests only now provoked it.
3926 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3927 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3928 Now it returns an integer. A number of calls were still expecting a T/F
3929 return. Fortuitously, in all cases, the tests worked in OK situations,
3930 which is the norm. However, things would have gone wrong on any write
3931 failures on the smtp file descriptor. This function is used when sending
3932 messages over SMTP and also when doing verify callouts.
3934 2. When Exim is called to do synchronous delivery of a locally submitted
3935 message (the -odf or -odi options), it no longer closes stderr before doing
3938 3. Implemented the mua_wrapper option.
3940 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3942 5. Implemented the functions header_remove(), header_testname(),
3943 header_add_at_position(), and receive_remove_recipient(), and exported them
3946 6. If an ACL "warn" statement specified the addition of headers, Exim already
3947 inserted X-ACL-Warn: at the start if there was no header name. However, it
3948 was not making this test for the second and subsequent header lines if
3949 there were newlines in the string. This meant that an invalid header could
3950 be inserted if Exim was badly configured.
3952 7. Allow an ACL "warn" statement to add header lines at the start or after all
3953 the Received: headers, as well as at the end.
3955 8. Added the rcpt_4xx retry error code.
3957 9. Added postmaster_mailfrom=xxx to callout verification option.
3959 10. Added mailfrom=xxxx to the callout verification option, for verify=
3962 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3963 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3965 12. Inserted some casts to stop certain compilers warning when using pointer
3966 differences as field lengths or precisions in printf-type calls (mostly
3967 affecting debugging statements).
3969 13. Added optional readline() support for -be (dynamically loaded).
3971 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3972 same clock tick as a message's arrival, so that its received time was the
3973 same as the "first fail" time on the retry record, and that message
3974 remained on the queue past the ultimate address timeout, every queue runner
3975 would try a delivery (because it was past the ultimate address timeout) but
3976 after another failure, the ultimate address timeout, which should have then
3977 bounced the address, did not kick in. This was a "< instead of <=" error;
3978 in most cases the first failure would have been in the next clock tick
3979 after the received time, and all would be well.
3981 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3982 being recognized when the domain list was tested by the match_domain
3983 condition in an expansion string.
3985 16. Added the ${str2b64: operator.
3987 17. Exim was always calling setrlimit() to set a large limit for the number of
3988 processes, without checking whether the existing limit was already
3989 adequate. (It did check for the limit on file descriptors.) Furthermore,
3990 errors from getrlimit() and setrlimit() were being ignored. Now they are
3991 logged to the main and panic logs, but Exim does carry on, to try to do its
3992 job under whatever limits there are.
3994 18. Imported PCRE 5.0.
3996 19. Trivial typo in log message " temporarily refused connection" (the leading
3999 20. If the log selector return_path_on_delivery was set and an address was
4000 redirected to /dev/null, the delivery process crashed because it assumed
4001 that a return path would always be set for a "successful" delivery. In this
4002 case, the whole delivery is bypassed as an optimization, and therefore no
4005 21. Internal re-arrangement: the function for sending a challenge and reading
4006 a response while authentication was assuming a zero-terminated challenge
4007 string. It's now changed to take a pointer and a length, to allow for
4008 binary data in such strings.
4010 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4012 23. Exim was not respecting finduser_retries when seeking the login of the
4013 uid under which it was called; it was always trying 10 times. (The default
4014 setting of finduser_retries is zero.) Also, it was sleeping after the final
4015 failure, which is pointless.
4017 24. Implemented tls_on_connect_ports.
4019 25. Implemented acl_smtp_predata.
4021 26. If the domain in control=submission is set empty, Exim assumes that the
4022 authenticated id is a complete email address when it generates From: or
4023 Sender: header lines.
4025 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4026 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4027 chown and chgrp in /bin and hostname in /usr/bin.
4029 28. Exim was keeping the "process log" file open after each use, just as it
4030 does for the main log. This opens the possibility of it remaining open for
4031 long periods when the USR1 signal hits a daemon. Occasional processlog
4032 errors were reported, that could have been caused by this. Anyway, it seems
4033 much more sensible not to leave this file open at all, so that is what now
4036 29. The long-running daemon process does not normally write to the log once it
4037 has entered its main loop, and it closes the log before doing so. This is
4038 so that log files can straightforwardly be renamed and moved. However,
4039 there are a couple of unusual error situations where the daemon does write
4040 log entries, and I had neglected to close the log afterwards.
4042 30. The text of an SMTP error response that was received during a remote
4043 delivery was being truncated at 512 bytes. This is too short for some of
4044 the long messages that one sometimes sees. I've increased the limit to
4047 31. It is now possible to make retry rules that apply only when a message has a
4048 specific sender, in particular, an empty sender.
4050 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4051 it possible to be selective about when SMTP synchronization is enforced.
4053 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4055 32. Implemented hosts_connection_nolog.
4057 33. Added an ACL for QUIT.
4059 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4062 35. Added mailbox_size and mailbox_filecount to appendfile.
4064 36. Added control = no_multiline_responses to ACLs.
4066 37. There was a bug in the logic of the code that waits for the clock to tick
4067 in the case where the clock went backwards by a substantial amount such
4068 that the microsecond fraction of "now" was more than the microsecond
4069 fraction of "then" (but the whole seconds number was less).
4071 38. Added support for the libradius Radius client library this is found on
4072 FreeBSD (previously only the radiusclient library was supported).
4078 1. When certain lookups returned multiple values in the form name=value, the
4079 quoting of the values was not always being done properly. Specifically:
4080 (a) If the value started with a double quote, but contained no whitespace,
4082 (b) If the value contained whitespace other than a space character (i.e.
4083 tabs or newlines or carriage returns) it was not quoted.
4084 This fix has been applied to the mysql and pgsql lookups by writing a
4085 separate quoting function and calling it from the lookup code. The fix
4086 should probably also be applied to nisplus, ibase and oracle lookups, but
4087 since I cannot test any of those, I have not disturbed their existing code.
4089 2. A hit in the callout cache for a specific address caused a log line with no
4090 reason for rejecting RCPT. Now it says "Previous (cached) callout
4091 verification failure".
4093 3. There was an off-by-one bug in the queryprogram router. An over-long
4094 return line was truncated at 256 instead of 255 characters, thereby
4095 overflowing its buffer with the terminating zero. As well as fixing this, I
4096 have increased the buffer size to 1024 (and made a note to document this).
4098 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4099 when Exim is waiting for an SMTP response from a remote server, Exim
4100 restarts its select() call on the socket, thereby resetting its timeout.
4101 This is not a problem when such interrupts are rare. Somebody set up a cron
4102 job to run exiwhat every 2 minutes, which is less than the normal select()
4103 timeout (5 or 10 minutes). This meant that the select() timeout never
4104 kicked in because it was always reset. I have fixed this by comparing the
4105 time when an interrupt arrives with the time at the start of the first call
4106 to select(). If more time than the timeout has elapsed, the interrupt is
4107 treated as a timeout.
4109 5. Some internal re-factoring in preparation for the addition of Sieve
4110 extensions (by MH). In particular, the "personal" test is moved to a
4111 separate function, and given an option for scanning Cc: and Bcc: (which is
4112 not set for Exim filters).
4114 6. When Exim created an email address using the login of the caller as the
4115 local part (e.g. when creating a From: or Sender: header line), it was not
4116 quoting the local part when it contained special characters such as @.
4118 7. Installed new OpenBSD configuration files.
4120 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4121 try to make them clearer.
4123 9. Callout options, other than the timeout value, were being ignored when
4124 verifying sender addresses in header lines. For example, when using
4126 verify = header_sender/callout=no_cache
4128 the cache was (incorrectly) being used.
4130 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4131 only the Cygwin environment.
4133 11. When return_path_on_delivery was set as a log selector, if different remote
4134 addresses in the same message used different return paths and parallel
4135 remote delivery occurred, the wrong values would sometimes be logged.
4136 (Whenever a remote delivery process finished, the return path value from
4137 the most recently started remote delivery process was logged.)
4139 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4140 lines when AUTH and/or TLS are in use. This is the "received protocol"
4141 field. Exim used to use "asmtp" for authenticated SMTP, without any
4142 indication (in the protocol name) for TLS use. Now it follows the RFC and
4143 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4144 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4145 names appear in log lines as well as in Received: header lines.
4147 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4148 extensions, and comparison tests, and to fix some bugs.
4150 14. Changes to the "personal" filter test:
4152 (1) The test was buggy in that it was just doing the equivalent of
4153 "contains" tests on header lines. For example, if a user's address was
4154 anne@some.where, the "personal" test would incorrectly be true for
4156 To: susanne@some.where
4158 This test is now done by extracting each address from the header in turn,
4159 and checking the entire address. Other tests that are part of "personal"
4160 are now done using regular expressions (for example, to check local parts
4161 of addresses in From: header lines).
4163 (2) The list of non-personal local parts in From: addresses has been
4164 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4165 taken from the Sieve specification recommendations.
4167 (3) If the message contains any header line starting with "List-" it is
4168 treated as non-personal.
4170 (4) The test for "circular" in the Subject: header line has been removed
4171 because it now seems ill-conceived.
4173 15. Minor typos in src/EDITME comments corrected.
4175 16. Installed latest exipick from John Jetmore.
4177 17. If headers_add on a router specified a text string that was too long for
4178 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4179 of string_sprintf() is now avoided.
4181 18. $message_body_size was not set (it was always zero) when running the DATA
4182 ACL and the local_scan() function.
4184 19. For the "mail" command in an Exim filter, no default was being set for
4185 the once_repeat time, causing a random time value to be used if "once" was
4186 specified. (If the value happened to be <= 0, no repeat happened.) The
4187 default is now 0s, meaning "never repeat". The "vacation" command was OK
4188 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4189 (I found it when inspecting the code).
4191 20. There is now an overall timeout for performing a callout verification. It
4192 defaults to 4 times the callout timeout, which applies to individual SMTP
4193 commands during the callout. The overall timeout applies when there is more
4194 than one host that can be tried. The timeout is checked before trying the
4195 next host. This prevents very long delays if there are a large number of
4196 hosts and all are timing out (e.g. when the network connections are timing
4197 out). The value of the overall timeout can be changed by specifying an
4198 additional sub-option for "callout", called "maxwait". For example:
4200 verify = sender/callout=5s,maxwait=20s
4202 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4203 to the end before writing, but this should make it even safer).
4205 22. Exim was forgetting that it had advertised PIPELINING for the second and
4206 subsequent messages on an SMTP connection. It was also not resetting its
4207 memory on STARTTLS and an internal HELO.
4209 23. When Exim logs an SMTP synchronization error within a session, it now
4210 records whether PIPELINING has been advertised or not.
4212 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4213 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4214 rather than long int.
4216 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4218 26. Added the never_mail option to autoreply.
4224 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4225 crash if the getsockname() call failed; this can happen if a connection is
4226 closed very soon after it is established. The problem was simply in the
4227 order in which certain operations were done, causing Exim to try to write
4228 to the SMTP stream before it had set up the file descriptor. The bug has
4229 been fixed by making things happen in the correct order.
4235 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4236 before the connection was closed, thus losing the rejection response.
4238 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4239 some early Solaris releases, but causes trouble in current releases where
4240 socklen_t is defined.
4242 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4245 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4248 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4249 that releasing the top of it at the end releases what was used for sub-
4250 expansions (unless the block got too big). However, discard this block if
4251 the first thing is a variable or header, so that we can use its block when
4252 it is dynamic (useful for very large $message_headers, for example).
4254 6. Lookups now cache *every* query, not just the most recent. A new, separate
4255 store pool is used for this. It can be recovered when all lookup caches are
4256 flushed. Lookups now release memory at the end of their result strings.
4257 This has involved some general refactoring of the lookup sources.
4259 7. Some code has been added to the store_xxx() functions to reduce the amount
4260 of flapping under certain conditions.
4262 8. log_incoming_interface used to affect only the <= reception log lines. Now
4263 it causes the local interface and port to be added to several more SMTP log
4264 lines, for example "SMTP connection from", and rejection lines.
4266 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4268 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4270 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4271 could be overwritten at the end of the current message (or the start of a
4272 new message if it was set in a HELO ACL). The value is now preserved for
4273 the duration of the SMTP connection.
4275 12. If a transport had a headers_rewrite setting, and a matching header line
4276 contained an unqualified address, that address was qualified, even if it
4277 did not match any rewriting rules. The underlying bug was that the values
4278 of the flags that permit the existence of unqualified sender and recipient
4279 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4280 non-local messages, and by -bnq for local messages) were not being
4281 preserved with the message after it was received.
4283 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4284 "next input=" as part of the text comprising the host identity instead of
4285 the correct text. The code was using the same buffer for two different
4286 strings. However, depending on which order the printing function evaluated
4287 its arguments, the bug did not always show up. Under Linux, for example, my
4288 test suite worked just fine.
4290 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4291 doesn't work with some older versions of Perl. It has been changed to "my",
4292 which in any case is probably the better facility to use.
4294 15. A really picky compiler found some instances of statements for creating
4295 error messages that either had too many or two few arguments for the format
4298 16. The size of the buffer for calls to the DNS resolver has been increased
4299 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4300 for addresses that have a lot of PTR records. This alleviates a problem; it
4301 does not fully solve it.
4303 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4304 buffer now truncates the list and logs the incident, which is the same
4305 action as happens when Exim is looking up a host name and its aliases.
4306 Previously in this situation something unpredictable would happen;
4307 sometimes it was "internal error: store_reset failed".
4309 18. If a server dropped the connection unexpectedly when an Exim client was
4310 using GnuTLS and trying to read a response, the client delivery process
4311 crashed while trying to generate an error log message.
4313 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4314 string, for example:
4316 warn message = H1: something\nH2: something
4318 the text was added as a single header line from Exim's point of view
4319 though it ended up OK in the delivered message. However, searching for the
4320 second and subsequent header lines using $h_h2: did not work. This has been
4321 fixed. Similarly, if a system filter added multiple headers in this way,
4322 the routers could not see them.
4324 20. Expanded the error message when iplsearch is called with an invalid key to
4325 suggest using net-iplsearch in a host list.
4327 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4328 ACLs are no longer actually imposed (and a message to that effect is
4331 22. If a "gecos" field in a passwd entry contained escaped characters, in
4332 particular, if it contained a \" sequence, Exim got it wrong when building
4333 a From: or a Sender: header from that name. A second bug also caused
4334 incorrect handling when an unquoted " was present following a character
4335 that needed quoting.
4337 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4338 was not being matched caselessly.
4340 24. Arranged for all hyphens in the exim.8 source to be escaped with
4343 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4344 properly. Recipient callout cache records were still being keyed to include
4345 the sender, even when use_sender was set false. This led to far more
4346 callouts that were necessary. The sender is no longer included in the key
4347 when use_sender is false.
4349 26. Added "control = submission" modifier to ACLs.
4351 27. Added the ${base62d: operator to decode base 62 numbers.
4353 28. dnsdb lookups can now access SRV records.
4355 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4356 the configuration file.
4358 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4359 (-v) mode. This makes the output for a verbose queue run more intelligible.
4361 31. Added a use_postmaster feature to recipient callouts.
4363 32. Added the $body_zerocount variable, containing the number of binary zero
4364 bytes in the message body.
4366 33. The time of last modification of the "new" subdirectory is now used as the
4367 "mailbox time last read" when there is a quota error for a maildir
4370 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4372 35. Added +ignore_unknown as a special item in host lists.
4374 36. Code for decoding IPv6 addresses in host lists is now included, even if
4375 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4376 address was recognized as an IP address, but was then not correctly decoded
4377 into binary, causing unexpected and incorrect effects when compared with
4384 1. Very minor rewording of debugging text in manualroute to say "list of
4385 hosts" instead of "hostlist".
4387 2. If verify=header_syntax was set, and a header line with an unqualified
4388 address (no domain) and a large number of spaces between the end of the
4389 name and the colon was received, the reception process suffered a buffer
4390 overflow, and (when I tested it) crashed. This was caused by some obsolete
4391 code that should have been removed. The fix is to remove it!
4393 3. When running in the test harness, delay a bit after writing a bounce
4394 message to get a bit more predictability in the log output.
4396 4. Added a call to search_tidyup() just before forking a reception process. In
4397 theory, someone could use a lookup in the expansion of smtp_accept_max_
4398 per_host which, without the tidyup, could leave open a database connection.
4400 5. Added the variables $recipient_data and $sender_data which get set from a
4401 lookup success in an ACL "recipients" or "senders" condition, or a router
4402 "senders" option, similar to $domain_data and $local_part_data.
4404 6. Moved the writing of debug_print from before to after the "senders" test
4407 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4408 problems for message scanning, either using a data ACL, or using
4409 local_scan() because the Received: header was not generated till after they
4410 were called (in order to set the time as the time of reception completion).
4411 I have revised the way this works. The header is now generated after the
4412 body is received, but before the ACL or local_scan() are called. After they
4413 are run, the timestamp in the header is updated.
4419 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4420 before starting a queue runner without re-exec. This happened only when
4421 deliver_drop_privilege was set or when the Exim user was set to root. The
4422 effect of the bug was that timeouts during subsequent deliveries caused
4423 crashes instead of being properly handled. The handler is now left at its
4424 default (and expected) setting.
4426 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4427 message, again when deliver_drop_privilege is set or Exim is run as root.
4428 The bug described in (1) was not present in this case, but the tidying up
4429 of the other signals was missing. I have made the two cases consistent.
4431 3. The ignore_target_hosts setting on a manualroute router was being ignored
4432 for hosts that were looked up using the /MX notation.
4434 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4437 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4438 operated on the sender address. After changing the $sender_address to <>
4439 for the sender address verify, Exim was re-instated it as the original
4440 (before rewriting) address, but remembering that it had rewritten it, so it
4441 wasn't rewriting it again. This bug also had the effect of breaking the
4442 sender address verification caching when the sender address was rewritten.
4444 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4445 This has been changed so that if the ip literal address matches
4446 ignore_target_hosts, the router declines.
4448 7. Added expansion conditions match_domain, match_address, and match_local_
4449 part (NOT match_host).
4451 8. The placeholder for the Received: header didn't have a length field set.
4453 9. Added code to Exim itself and to exim_lock to test for a specific race
4454 condition that could lead to file corruption when using MBX delivery. The
4455 issue is with the lockfile that is created in /tmp. If this file is removed
4456 after a process has opened it but before that process has acquired a lock,
4457 there is the potential for a second process to recreate the file and also
4458 acquire a lock. This could lead to two Exim processes writing to the file
4459 at the same time. The added code performs the same test as UW imapd; it
4460 checks after acquiring the lock that its file descriptor still refers to
4461 the same named file.
4463 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4464 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4467 11. Added the smtp_active_hostname option. If used, this will typically be made
4468 to depend on the incoming interface address. Because $interface_address is
4469 not set up until the daemon has forked a reception process, error responses
4470 that can happen earlier (such as "too many connections") no longer contain
4473 12. If an expansion in a condition on a "warn" statement fails because a lookup
4474 defers, the "warn" statement is abandoned, and the next ACL statement is
4475 processed. Previously this caused the whole ACL to be aborted.
4477 13. Added the iplsearch lookup type.
4479 14. Added ident_timeout as a log selector.
4481 15. Added tls_certificate_verified as a log selector.
4483 16. Added a global option tls_require_ciphers (compare the smtp transport
4484 option of the same name). This controls incoming TLS connections.
4486 17. I finally figured out how to make tls_require_ciphers do a similar thing
4487 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4488 before starting the TLS session.
4490 18. Tabs are now shown as \t in -bP output.
4492 19. If the log selector return_path_on_delivery was set, Exim crashed when
4493 bouncing a message because it had too many Received: header lines.
4495 20. If two routers both had headers_remove settings, and the first one included
4496 a superfluous trailing colon, the final name in the first list and the
4497 first name in the second list were incorrectly joined into one item (with a
4498 colon in the middle).
4504 1. Added -C and -D options to the exinext utility, mainly to make it easier
4505 to include in the automated testing, but these could be helpful when
4506 multiple configurations are in use.
4508 2. The exinext utility was not formatting the output nicely when there was
4509 an alternate port involved in the retry record key, nor when there was a
4510 message id as well (for retries that were specific to a specific message
4511 and a specific host). It was also confused by IPv6 addresses, because of
4512 the additional colons they contain. I have fixed the IPv4 problem, and
4513 patched it up to do a reasonable job for IPv6.
4515 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4516 delivery, the log line now contains "pipelined" if PIPELINING was used.
4518 4. An SMTP transport process used to panic and die if the bind() call to set
4519 an explicit outgoing interface failed. This has been changed; it is now
4520 treated in the same way as a connect() failure.
4522 5. A reference to $sender_host_name in the part of a conditional expansion
4523 that was being skipped was still causing a DNS lookup. This no longer
4526 6. The def: expansion condition was not recognizing references to header lines
4527 that used bh_ and bheader_.
4529 7. Added the _cache feature to named lists.
4531 8. The code for checking quota_filecount in the appendfile transport was
4532 allowing one more file than it should have been.
4534 9. For compatibility with Sendmail, the command line option
4542 and sets the incoming protocol and host name (for trusted callers). The
4543 host name and its colon can be omitted when only the protocol is to be set.
4544 Note the Exim already has two private options, -pd and -ps, that refer to
4545 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4546 "s", but I don't think that's a major issue.
4548 10. A number of refactoring changes to the code, none of which should affect
4551 (a) The number of logging options was getting close to filling up the
4552 32-bit word that was used as a bit map. I have split them into two classes:
4553 those that are passed in the argument to log_write(), and those that are
4554 only ever tested independently outside of that function. These are now in
4555 separate 32-bit words, so there is plenty of room for expansion again.
4556 There is no change in the user interface or the logging behaviour.
4558 (b) When building, for example, log lines, the code previously used a
4559 macro that called string_cat() twice, in order to add two strings. This is
4560 not really sufficiently general. Furthermore, there was one instance where
4561 it was actually wrong because one of the argument was used twice, and in
4562 one call a function was used. (As it happened, calling the function twice
4563 did not affect the overall behaviour.) The macro has been replaced by a
4564 function that can join an arbitrary number of extra strings onto a growing
4567 (c) The code for expansion conditions now uses a table and a binary chop
4568 instead of a serial search (which was left over from when there were very
4569 few conditions). Also, it now recognizes conditions like "pam" even when
4570 the relevant support is not compiled in: a suitably worded error message is
4571 given if an attempt is made to use such a condition.
4573 11. Added ${time_interval:xxxxx}.
4575 12. A bug was causing one of the ddress fields not to be passed back correctly
4576 from remote delivery subprocesses. The field in question was not being
4577 subsequently used, so this caused to problems in practice.
4579 13. Added new log selectors queue_time and deliver_time.
4581 14. Might have fixed a bug in maildirsizefile handling that threw up
4582 "unexpected character" debug warnings, and recalculated the data
4583 unnecessarily. In any case, I expanded the warning message to give more
4586 15. Added the message "Restricted characters in address" to the statements in
4587 the default ACL that block characters like @ and % in local parts.
4589 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4590 Three changes have been made:
4592 (a) There was a serious bug; a negative response to MAIL caused the whole
4593 recipient domain to be cached as invalid, thereby blocking all messages
4594 to all local parts at the same domain, from all senders. This bug has
4595 been fixed. The domain is no longer cached after a negative response to
4596 MAIL if the sender used is not empty.
4598 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4601 (c) A new callout option, "use_sender" has been added for people who want
4602 the modified behaviour.
4608 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4611 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4612 indeed breaks things for older releases.
4614 3. Added additional logging to the case where there is a problem reading data
4615 from a filter that is running in a subprocess using a pipe, in order to
4616 try to track down a specific problem.
4618 4. Testing facility fudge: when running in the test harness and attempting
4619 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4620 getting "No route to host". Convert this to a timeout.
4622 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4625 6. Some OS don't have socklen_t but use size_t instead. This affects the
4626 fifth argument of getsockopt() amongst other things. This is now
4627 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4628 can be set for individual OS. I have set it for SunOS5, OSF1, and
4629 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4630 some earlier ones do not.
4632 7. Change 4.30/15 was not doing the test caselessly.
4634 8. The standard form for an IPv6 address literal was being rejected by address
4635 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4636 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4637 this, as well as the form without the "IPv6" on the front (but only when
4638 address literals are enabled, of course).
4640 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4642 10. Exim crashed if a message with an empty sender address specified by -f
4643 encountered a router with an errors_to setting. This could be provoked only
4644 by a command such as
4648 where an empty string was supplied; "<>" did not hit this bug.
4650 11. Installed PCRE release 4.5.
4652 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4653 remained set. It is now erased.
4655 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4656 times from message ids (which are base 36 rather than the normal 62).
4658 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4659 were being counted as actual protocol errors, and logged if the log
4660 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4661 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4662 and DATA following a set of rejected RCPTs do not count as protocol errors.
4663 In other words, Exim assumes they were pipelined, though this may not
4664 actually be the case. Of course, in all cases the client gets an
4665 appropriate error code.
4667 15. If a lookup fails in an ACL condition, a message about the failure may
4668 be available; it is used if testing the ACL cannot continue, because most
4669 such messages specify what the cause of the deferral is. However, some
4670 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4671 that caused an old message to be retained and used if a later statement
4672 caused a defer, replacing the real cause of the deferral.
4674 16. If an IP address had so many PTR records that the DNS lookup buffer
4675 was not large enough to hold them, Exim could crash while trying to process
4676 the truncated data. It now detects and logs this case.
4678 17. Further to 4.21/58, another change has been made: if (and only if) the
4679 first line of a message (the first header line) ends with CRLF, a bare LF
4680 in a subsequent header line has a space inserted after it, so as not to
4681 terminate the header.
4683 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4684 unnecessarily, used atoi() instead of strtol(), and didn't check the
4685 termination when getting file sizes from file names by regex.
4687 19. Completely re-implemented the support for maildirsize files, in the light
4688 of a number of problems with the previous contributed implementation
4689 (4.30/29). In particular:
4691 . If the quota is zero, the maildirsize file is maintained, but no quota is
4694 . If the maildir directory does not exist, it is created before any attempt
4695 to write a maildirsize file.
4697 . The quota value in the file is just a cache; if the quota is changed in
4698 the transport, the new value overrides.
4700 . A regular expression is available for excluding directories from the
4703 20. The autoreply transport checks the characters in options that define the
4704 message's headers; it allows continued headers, but it was checking with
4705 isspace() after an embedded newline instead of explicitly looking for a
4708 21. If all the "regular" hosts to which an address was routed had passed their
4709 expiry times, and had not reached their retry times, the address was
4710 bounced, even if fallback hosts were defined. Now Exim should go on to try
4713 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4714 equivalent code in the SMTP transport. Some hosts send humungous responses
4715 to HELO/EHLO, more than 1024 it seems.
4717 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4718 gives compiler warnings in some environments. I've now done it "properly",
4721 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4722 (because of problems with the built-in one) was declared to return uschar *
4723 instead of char *, causing compiler failure.
4725 25. Fixed a file descriptor leak when processing alias/forward files.
4727 26. Fixed a minor format string issue in dbfn.c.
4729 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4731 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4732 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4735 29. When a sender address is verified, it is cached, to save repeating the test
4736 when there is more than one recipient in a message. However, when the
4737 verification involves a callout, it is possible for different callout
4738 options to be set for different recipients. It is too complicated to keep
4739 track of this in the cache, so now Exim always runs a verification when a
4740 callout is required, relying on the callout cache for the optimization.
4741 The overhead is duplication of the address routing, but this should not be
4744 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4745 to be verified with callout=postmaster, and the main callout worked but the
4746 postmaster check failed, the verification correctly failed. However, if a
4747 subsequent RCPT command asked for sender verification *without* the
4748 postmaster check, incorrect caching caused this verification also to fail,
4751 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4752 it was not caching the DNS options (qualify_single, search_parents) that
4753 were used when the lookup failed. A subsequent lookup with different
4754 options therefore always gave the same answer, though there were cases
4755 where it should not have. (Example: a "domains = !$mx_any" option on a
4756 dnslookup router: the "domains" option is always processed without any
4757 widening, but the router might have qualify_single set.) Now Exim uses the
4758 cached value only when the same options are set.
4760 32. Added John Jetmore's "exipick" utility to the distribution.
4762 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4763 than a timeout (e.g. a certificate is required, and is not provided), an
4764 Exim server now closes the connection immediately. Previously it waited for
4765 the client to close - but if the client is SSL, it seems that they each
4766 wait for each other, leading to a delay before one of them times out.
4768 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4769 maintained 0.8.x compatibility because I don't think many are using it, and
4770 it is clearly obsolete.
4772 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4775 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4776 client certificate was expired. A simple patch fixes this, though I don't
4777 understand the full logic of why the verify callback is called multiple
4780 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4781 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4782 which causes problems with some clients (such as the Certicom SSL Plus
4783 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4784 disables the coutermeasure allowing Eudora to connect."
4786 38. Exim was not checking that a write() to a log file succeeded. This could
4787 lead to Bad Things if a log got too big, in particular if it hit a file
4788 size limit. Exim now panics and dies if it cannot write to a log file, just
4789 as it does if it cannot open a log file.
4791 39. Modified OS/Makefile-Linux so that it now contains
4793 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4795 The two -D definitions ensure that Exim is compiled with large file
4796 support, which makes it possible to handle log files that are bigger than
4799 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4800 instance) a domain was checked against a named list that involved a lookup,
4801 causing $domain_data to be set, then another domain was checked against the
4802 same list, then the first domain was re-checked, the value of $domain_data
4803 after the final check could be wrong. In particular, if the second check
4804 failed, it could be set empty. This bug probably also applied to
4807 41. The strip_trailing_dot option was not being applied to the address given
4808 with the -f command-line option.
4810 42. The code for reading a message's header from the spool was incrementing
4811 $received_count, but never initializing it. This meant that the value was
4812 incorrect (doubled) while delivering a message in the same process in which
4813 it was received. In the most common configuration of Exim, this never
4814 happens - a fresh exec is done - but it can happen when
4815 deliver_drop_privilege is set.
4817 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4818 it now includes up to 150 characters of the unexpected data in the log
4821 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4822 and building data strings. The size of both of these buffers was 10 000
4823 bytes - far larger than anybody would *ever* want, thought I. Needless to
4824 say, somebody hit the limit. I have increased the maximum line length to
4825 20 000 and the maximum data length of concatenated lines to 100 000. I have
4826 also fixed two bugs, because there was no checking on these buffers. Tsk,
4827 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4828 buffer is too small.
4830 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4831 lsearch lookups. Now it does.
4833 46. When parsing a route_list item in a manualroute router, a fixed-length
4834 buffer was used for the list of hosts. I made this 1024 bytes long,
4835 thinking that nobody would ever have a list of hosts that long. Wrong.
4836 Somebody had a whole pile of complicated expansion conditions, and the
4837 string was silently truncated, leading to an expansion error. It turns out
4838 that it is easier to change to an unlimited length (owing to other changes
4839 that have happened since this code was originally written) than to build
4840 structure for giving a limitation error. The length of the item that
4841 expands into the list of hosts is now unlimited.
4843 47. The lsearch lookup could not handle data where the length of text line was
4844 more than 4095 characters. Such lines were truncated, leading to shortened
4845 data being returned. It should now handle lines of any length.
4847 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4848 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4851 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4852 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4853 no longer happens in comment lines. A list of replacements is now placed
4854 at the head of all of the source files, except those whose only change is
4855 to replace PERL_COMMAND in the very first #! line.
4857 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4858 messages on the queue, with a bottom-up merge sort, using code contributed
4859 by Michael Haardt. This should make operations like -bp somewhat faster on
4860 large queues. It won't affect queue runners, except when queue_run_in_order
4863 51. Installed eximstats 1.31 in the distribution.
4865 52. Added support for SRV lookups to the dnslookup router.
4867 53. If an ACL referred to $message_body or $message_body_end, the value was not
4868 reset for any messages that followed in the same SMTP session.
4870 54. The store-handling optimization for building very long strings was not
4871 differentiating between the different store pools. I don't think this
4872 actually made any difference in practice, but I've tidied it.
4874 55. While running the routers to verify a sender address, $sender_address
4875 was still set to the sender address. This is wrong, because when routing to
4876 send a bounce to the sender, it would be empty. Therefore, I have changed
4877 it so that, while verifying a sender address, $sender_address is set to <>.
4878 (There is no change to what happens when verifying a recipient address.)
4880 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4881 target A or AAAA records (if not already returned) without resetting the
4882 qualify_single or search_parents options of the DNS resolver. These are
4883 inappropriate in this case because the targets of MX and SRV records must
4884 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4885 target that, when qualified, matched something in the local domain. These
4886 two options are now turned off when doing these lookups.
4888 57. It seems that at least some releases of Reiserfs (which does not have the
4889 concept of a fixed number of inodes) returns zero and not -1 for the
4890 number of available inodes. This interacted badly with check_spool_inodes,
4891 which assumed that -1 was the "no such thing" setting. What I have done is
4892 to check that the total number of inodes is greater than zero before doing
4893 the test of how many are available.
4895 58. When a "warn" ACL statement has a log_message modifier, the message is
4896 remembered, and not repeated. This is to avoid a lot of repetition when a
4897 message has many recipients that cause the same warning to be written.
4898 Howewer, Exim was preserving the list of already written lines for an
4899 entire SMTP session, which doesn't seem right. The memory is now reset if a
4900 new message is started.
4902 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4903 debugging output unless log_rewrite was also set.
4905 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4906 of (int)(handle) when we know that handle contains (void *)(-1).
4908 61. The Exim daemon panic-logs an error return when it closes the incoming
4909 connection. However "connection reset by peer" seems to be common, and
4910 isn't really an error worthy of noting specially, so that particular error
4913 62. When Exim is trying to find all the local interfaces, it used to panic and
4914 die if the ioctl to get the interface flags failed. However, it seems that
4915 on at least one OS (Solaris 9) it is possible to have an interface that is
4916 included in the list of interfaces, but for which you get a failure error
4917 for this call. This happens when the interface is not "plumbed" into a
4918 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4919 failure of the "get flags" call assumes that the interface is down.
4921 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4922 makes life easier for people who are doing arithmetic on fields extracted
4923 from dates, where you often get leading zeros that should not be
4924 interpreted as octal.
4926 64. Added qualify_domain to the redirect router, to override the global
4929 65. If a pathologically long header line contained very many addresses (the
4930 report of this problem mentioned 10 000) and each of them was rewritten,
4931 Exim could use up a very large amount of memory. (It kept on making new
4932 copies of the header line as it rewrote, and never released the old ones.)
4933 At the expense of a bit more processing, the header rewriting function has
4934 been changed so that it no longer eats memory in this way.
4936 66. The generation of the Received: header has been moved from the time that a
4937 message starts to be received, to the time that it finishes. The timestamp
4938 in the Received: header should now be very close to that of the <= log
4939 line. There are two side-effects of this change:
4941 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4942 logged header lines no longer include the local Received: line, because
4943 it has not yet been created. The same applies to a copy of the message
4944 that is returned to a non-SMTP sender when a message is rejected.
4946 (b) When a filter file is tested using -bf, no additional Received: header
4947 is added to the test message. After some thought, I decided that this
4950 This change does not affect the value of $received_for. It is still set
4951 after address rewriting, but before local_scan() is called.
4953 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4955 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4956 gave an unhelpful panic error message, and a defer error. I have managed to
4957 change this behaviour so that it now rejects any supplied certificate,
4958 which seems right, as the list of acceptable certificates is empty.
4960 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4961 gave an unhelpful defer error. I have not managed to make this reject any
4962 supplied certificates, but the error message it gives is "no certificate
4963 supplied", which is not helpful.
4965 70. exigrep's output now also includes lines that are not associated with any
4966 message, but which match the given pattern. Implemented by a patch from
4967 Martin Sluka, which also tidied up the Perl a bit.
4969 71. Recipient callout verification, like sender verification, was using <> in
4970 the MAIL FROM command. This isn't really the right thing, since the actual
4971 sender may affect whether the remote host accepts the recipient or not. I
4972 have changed it to use the actual sender in the callout; this means that
4973 the cache record is now keyed on a recipient/sender pair, not just the
4974 recipient address. There doesn't seem to be a real danger of callout loops,
4975 since a callout by the remote host to check the sender would use <>.
4976 [SEE ABOVE: changed after hitting problems.]
4978 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4979 temporary errors. However, in the case of such a code being given after
4980 the end of a data transmission (i.e. after ".") Exim was failing to write
4981 a retry record for the message. (Yes, there was some broken host that was
4982 actually sending 8xx at this point.)
4984 73. An unknown lookup type in a host list could cause Exim to panic-die when
4985 the list was checked. (An example that provoked this was putting <; in the
4986 middle of a list instead of at the start.) If this happened during a DATA
4987 ACL check, a -D file could be left lying around. This kind of configuration
4988 error no longer causes Exim to die; instead it causes a defer errror. The
4989 incident is still logged to the main and panic logs.
4991 74. Buglet left over from Exim 3 conversion. The message "too many messages
4992 in one connection" was written to the rejectlog but not the mainlog, except
4993 when address rewriting (yes!) was being logged.
4995 75. Added write_rejectlog option.
4997 76. When a system filter was run not as root (that is, when system_filter_user
4998 was set), the values of the $n variables were not being returned to the
4999 main process; thus, they were not subsequently available in the $sn
5002 77. Added +return_path_on_delivery log selector.
5004 78. A connection timeout was being treated differently from recipients deferred
5005 when testing hosts_max_try with a message that was older than the host's
5006 retry timeout. (The host should not be counted, thus allowing all hosts to
5007 be tried at least once before bouncing.) This may have been the cause of an
5008 occasionally reported bug whereby a message would remain on the queue
5009 longer than the retry timeout, but would be bounced if a delivery was
5010 forced. I say "may" because I never totally pinned down the problem;
5011 setting up timeout/retry tests is difficult. See also the next item.
5013 79. The ultimate address timeout was not being applied to errors that involved
5014 a combination of host plus message (for example, a timeout on a MAIL
5015 command). When an address resolved to a number of possible hosts, and they
5016 were not all tried for each delivery (e.g. because of hosts_max_try), a
5017 message could remain on the queue longer than the retry timeout.
5019 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5022 81. Fixed an obscure SMTP outgoing bug which required at least the following
5023 conditions: (a) there was another message waiting for the same server;
5024 (b) the server returned 5xx to all RCPT commands in the first message so
5025 that the message was not completed; (c) the server dropped the connection
5026 or gave a negative response to the RSET that Exim sends to abort the
5027 transaction. The observed case was a dropped connection after DATA that had
5028 been sent in pipelining mode. That is, the server had advertised PIPELINING
5029 but was not implementing it correctly. The effect of the bug was incorrect
5030 behaviour, such as trying another host, and this could lead to a crash.
5036 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5037 and daemon.c were passed as pointers to ints; they should have been
5038 pointers to socklen_t variables (which are typically unsigned ints).
5040 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5043 3. Fixed a really odd bug that affected only the testing scheme; patching a
5044 certain fixed string in the binary changed the value of another string that
5045 happened to be identical to the end of the original first string.
5047 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5048 name", it returns that address as the IP address. On some operating
5049 systems (e.g. Solaris), it also passes back the IP address string as the
5050 "host name". However, on others (e.g. Linux), it passes back an empty
5051 string. Exim wasn't checking for this, and was changing the host name to an
5052 empty string, assuming it had been canonicized.
5054 5. Although rare, it is permitted to have more than one PTR record for a given
5055 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5056 all the names associated with an address, because they do in Solaris.
5057 However, it seems that they do not in Linux for data that comes from the
5058 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5059 I found this out when I moved to a new Linux workstation and tried to run
5060 the Exim test suite.
5062 To get round this problem I have changed the code so that it now does its
5063 own call to the DNS to look up PTR records when searching for a host name.
5064 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5065 addresses that are only in /etc/hosts are still found.
5067 This behaviour is, however, controlled by an option called host_lookup_
5068 order, which defaults to "bydns:byaddr". If people want to use the other
5069 order, or indeed, just use one or the other means of lookup, they can
5070 specify it in this variable.
5072 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5073 some operating systems, this comes back from gethostbyaddr() as an empty
5074 string, and this is what Exim used to test for. However, it seems that in
5075 other systems, "." is yielded. Exim now tests for this case too.
5077 7. The values of check_spool_space and check_log_space are now held internally
5078 as a number of kilobytes instead of an absolute number of bytes. If a
5079 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5080 kilobyte. This means that much larger values can be stored.
5082 8. Exim monitor: an attempt to get the action menu when not actually pointing
5083 at a message produces an empty menu entitled "No message selected". This
5084 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5085 no entries in it ("Shell widget menu has zero width and/or height"). So I
5086 have added a single, blank menu entry in this case.
5088 9. Added ${quote_local_part.
5090 10. MIME decoding is now applied to the contents of Subject: header lines when
5093 11. Now that a reference to $sender_host_address automatically causes a reverse
5094 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5095 host lookup before query-style lookups in lists that might use this
5096 variable. This has therefore been abolished, and the "net-" prefix is no
5097 longer necessary for query-style lookups.
5099 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5100 have been a typo for LFLAGS, so it has been changed.
5102 13. The install script calls Exim with "-C /dev/null" in order to find the
5103 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5104 to be output. Howeve, since Exim outputs its version number before the
5105 error, it didn't break the script. It just looked ugly. I fixed this by
5106 always allowing "-C /dev/null" if the caller is root.
5108 14. Ignore overlarge ACL variable number when reading spool file - insurance
5109 against a later release with more variables having written the file.
5111 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5112 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5113 this, as well as the form without the "IPv6" on the front.
5115 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5116 OS/Makefile-Darwin file.
5118 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5119 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5121 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5122 variable. However, some OS use ints for inodes. I've added cast to long int
5123 to get rid of the compiler warning.
5125 19. I had forgotten to lock out "/../" in configuration file names when
5126 ALT_CONFIG_PREFIX was set.
5128 20. Routers used for verification do not need to specify transports. However,
5129 if such a router generated a host list, and callout was configured, Exim
5130 crashed, because it could not find a port number from the (non-existent)
5131 transport. It now assumes port 25 in this circumstance.
5133 21. Added the -t option to exigrep.
5135 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5136 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5137 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5138 them is set, LOOKUP_LSEARCH is forced.
5140 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5142 24. Added sender and host information to the "rejected by local_scan()" log
5143 line; previously there was no indication of these.
5145 25. Added .include_if_exists.
5147 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5148 Linux. It turns out that not all file systems support this. Apparently some
5149 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5150 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5151 not supported on the file descriptor, is now ignored when Exim is trying to
5152 sync a directory. This applies only to Linux.
5154 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5156 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5157 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5160 29. Added support for maildirsize files from supplied patch (modified a bit).
5162 30. The use of :fail: followed by an empty string could lead Exim to respond to
5163 sender verification failures with (e.g.):
5165 550 Verification failed for <xxx>
5166 550 Sender verify failed
5168 where the first response line was missing the '-' that indicates it is not
5169 the final line of the response.
5171 31. The loop for finding the name of the user that called Exim had a hardwired
5172 limit of 10; it now uses the value of finduser_retries, which is used for
5173 all other user lookups.
5175 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5178 33. Exim was neglecting to zero errno before one call of strtol() when
5179 expanding a string and expecting an integer value. On some systems this
5180 resulted in spurious "integer overflow" errors. Also, it was casting the
5181 result into an int without checking.
5183 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5184 did not work. The code looks as if it has *never* worked, though it appears
5185 to have been documented since at least releast 1.62. I have made it work.
5187 35. The "timeout_DNS" error in retry rules, also documented since at least
5188 1.62, also never worked. As it isn't clear exactly what this means, and
5189 clearly it isn't a major issue, I have abolished the feature by treating it
5190 as "timeout", and writing a warning to the main and panic logs.
5192 36. The display of retry rules for -brt wasn't always showing the error code
5195 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5196 timeout_connect_A, timeout_connect_MX.
5198 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5199 to the empty sender.
5201 39. The daemon was not analysing the content of -oX till after it had closed
5202 stderr and disconnected from the controlling terminal. This meant that any
5203 syntax errors were only noted on the panic log, and the return code from
5204 the command was 0. By re-arranging the code a little, I've made the
5205 decoding happen first, so such errors now appear on stderr, and the return
5206 code is 1. However, the actual setting up of the sockets still happens in
5207 the disconnected process, so errors there are still only recorded on the
5210 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5211 connections (as happens on some IP stacks) was logged at start up time as
5212 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5213 it from "IPv6 and IPv4", which means that two separate sockets are being
5216 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5217 says whether AF_INET or AF_INET6 was passed as an argument.
5219 42. Exiwhat output was messed up when time zones were included in log
5222 43. Exiwhat now gives more information about the daemon's listening ports,
5223 and whether -tls-on-connect was used.
5225 44. The "port" option of the smtp transport is now expanded.
5227 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5228 silently ignored. Now an error message is written to the main and panic
5231 46. There's a new ACL modifier called "logwrite" which writes to a log file
5232 as soon as it is encountered.
5234 47. Added $local_user_uid and $local_user_gid at routing time.
5236 48. Exim crashed when trying to verify a sender address that was being
5239 49. Exim was recognizing only a space character after ".include". It now also
5240 recognizes a tab character.
5242 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5243 extracting the relevant information from the specification. The man page no
5244 longer contains scrambled data for the -d option, and I've added a section
5245 at the front about calling Exim under different names.
5247 51. Added "extra_headers" argument to the "mail" command in filter files.
5249 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5252 53. Installed eximstats 1.29.
5254 54. Added transport_filter_timeout as a generic transport option.
5256 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5257 Cc: header lines. This was required by RFC 822, but it not required by RFC
5260 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5261 incoming messages that did not have them. Now it does so only if the
5262 message originates locally, that is, if there is no associated remote host
5263 address. When Resent- header lines are present, this applies to the Resent-
5264 lines rather than the non-Resent- lines.
5266 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5267 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5269 58. Messages for configuration errors now include the name of the main
5270 configuration file - useful now that there may be more than one file in a
5271 list (.included file names were always shown).
5273 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5274 for those rare installations that do not start the daemon as root or run it
5275 setuid root. I've cut out the call to initgroups() if the daemon is not
5278 60. The Exim user and group can now be bound into the binary as text strings
5279 that are looked up at the start of Exim's processing.
5281 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5283 62. Added $mailstore_basename variable.
5285 63. Installed patch to sieve.c from Michael Haardt.
5287 64. When Exim failed to open the panic log after failing to open the main log,
5288 the original message it was trying to log was written to stderr and debug
5289 output, but if they were not available (the usual case in production), it
5290 was lost. Now it is written to syslog before the two lines that record the
5291 failures to open the logs.
5293 65. Users' Exim filters run in subprocesses under the user's uid. It is
5294 possible for a "deliver" command or an alias in a "personal" command to
5295 provoke an address rewrite. If logging of address rewriting is configured,
5296 this fails because the process is not running as root or exim. There may be
5297 a better way of dealing with this, but for the moment (because 4.30 needs
5298 to be released), I have disabled address rewrite logging when running a
5299 filter in a non-root, non-exim process.
5305 1. The buildconfig auxiliary program wasn't quoting the value set for
5306 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5307 not defined. This bug was masked in 4.22 by the effect that was fixed in
5310 2. Some messages that were rejected after a message id was allocated were
5311 shown as "incomplete" by exigrep. It no longer does this for messages that
5312 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5314 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5315 have allow_domain_literals set, the ID did not get logged in the <= line.
5316 Domain literals are now always recognized in Message-ID: header lines.
5318 4. The first argument for a ${extract expansion item is the key name or field
5319 number. Leading and trailing spaces in this item were not being ignored,
5320 causing some misleading effects.
5322 5. When deliver_drop_privilege was set, single queue runner processes started
5323 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5324 same command in the process it spins off) were not dropping privilege.
5326 6. When the daemon running as "exim" started a queue runner, it always
5327 re-executed Exim in the spun-off process. This is a waste of effort when
5328 deliver_drop_privilege is set. The new process now just calls the
5329 queue-runner function directly.
5335 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5338 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5339 ignored. Though the use of -oP was forcing the writing of a pid file, it
5340 was always written to the default place.
5342 3. If the message "no IP address found for host xxxx" is generated during
5343 incoming verification, it is now followed by identification of the incoming
5344 connection (so you can more easily find what provoked it).
5346 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5348 5. Added some features to "harden" Exim a bit more against certain attacks:
5350 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5351 be put in Local/Makefile. This is like the never_users runtime option,
5352 but it cannot be overridden. The default setting is "root".
5354 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5355 prefix string with which any file named in a -C command line option
5358 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5359 is retained for -C and -D only if the caller of Exim is root. Without
5360 it, the exim user may also use -C and -D and retain privilege.
5362 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5363 command line option is disabled.
5365 6. Macro names set by the -D option must start with an upper case letter, just
5366 like macro names defined in the configuration file.
5368 7. Added "dereference=" facility to LDAP.
5370 8. Two instances of the typo "uknown" in the source files are fixed.
5372 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5373 the Configure-Makefile script screwed up while processing it.
5375 10. Incorporated PCRE 4.4.
5377 11. The SMTP synchronization check was not operating right at the start of an
5378 SMTP session. For example, it could not catch a HELO sent before the client
5379 waited for the greeting. There is now a check for outstanding input at the
5380 point when the greeting is written. Because of the duplex, asynchronous
5381 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5382 way, but not yet received, when the check is performed.
5384 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5385 on TCP/IP sockets, because this apparently causes some broken clients to
5388 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5389 unchanged) from the Cygwin maintainer.
5391 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5392 was supported instead of testing for mbx. Effectively a typo.
5394 15. The spa authenticator server code was not checking that the input it
5395 received was valid base64.
5397 16. The debug output line for the "set" modifier in ACLs was not showing the
5398 name of the variable that was being set.
5400 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5402 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5403 lookup on its own. Something else was needed to trigger the lookup. For
5404 example, a match in host_lookup or the need for a host name in a host list.
5405 Now, if $sender_host_name is referenced and the host name has not yet been
5406 looked up, a lookup is performed. If the lookup fails, the variable remains
5407 empty, and $host_lookup_failed is set to "1".
5409 19. Added "eqi" as a case-independent comparison operator.
5411 20. The saslauthd authentication condition could segfault if neither service
5412 nor realm was specified.
5414 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5415 error message that was logged was misleading, and incoming SMTP
5416 connections were dropped. The message is now more accurate, and temporary
5417 errors are given to SMTP connections.
5419 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5420 (except RSET) with a 421 error, until QUIT is received. However, it was
5421 failing to send a response to QUIT.
5423 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5424 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5425 ACL is now run after the helo_try_verify_hosts code.
5427 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5428 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5429 case-independent, so other case variants are also recognized). Apparently
5430 some systems use these upper case variants.
5432 25. If more than two messages were waiting for the same host, and a transport
5433 filter was specified for the transport, Exim sent two messages over the
5434 same TCP/IP connection, and then failed with "socket operation on non-
5435 socket" when it tried to send the third.
5437 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5439 27. The extern definition of crypt16() in expand.c was not being excluded when
5440 the OS had its own crypt16() function.
5442 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5443 as a preferred synonym for return_size_limit, both as an option and as an
5446 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5448 30. Changed the default configuration ACL to relax the local part checking rule
5449 for addresses that are not in any local domains. For these addresses,
5450 slashes and pipe symbols are allowed within local parts, but the sequence
5451 /../ is explicitly forbidden.
5453 31. SPA server authentication was not clearing the challenge buffer before
5456 32. log_message in a "warn" ACL statement was writing to the reject log as
5457 well as to the main log, which contradicts the documentation and doesn't
5458 seem right (because no rejection is happening). So I have stopped it.
5460 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5461 However, I am unable to do any testing of this.
5463 34. Fixed an infelicity in the appendfile transport. When checking directories
5464 for a mailbox, to see if any needed to be created, it was accidentally
5465 using path names with one or more superfluous leading slashes; tracing
5466 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5468 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5469 contents are added to the log line that is written for every discarded
5470 recipient. (Previously a log_message setting was ignored.)
5472 36. The ${quote: operator now quotes the string if it is empty.
5474 37. The install script runs exim in order to find its version number. If for
5475 some reason other than non-existence or emptiness, which it checks, it
5476 could not run './exim', it was installing it with an empty version number,
5477 i.e. as "exim-". This error state is now caught, and the installation is
5480 38. An argument was missing from the function that creates an error message
5481 when Exim fails to connect to the socket for saslauthd authentication.
5482 This could cause Exim to crash, or give a corrupted message.
5484 39. Added isip, isip4, and isip6 to ${if conditions.
5486 40. The ACL variables $acl_xx are now saved with the message, and can be
5487 accessed later in routers, transports, and filters.
5489 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5490 strings in the file are not string-expanded.
5492 42. If a MAIL command specified a SIZE value that was too large to fit into an
5493 int variable, the check against message_size_limit failed. Such values are
5494 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5495 day this will have to be increased, but I don't think I want to be around
5496 when emails are that large.
5503 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5504 iconv() is not standard in FreeBSD.
5506 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5507 IPv6 enabled. The observed symptom was a segmentation fault on return from
5508 the function os_common_find_running_interfaces() in src/os.c.
5510 3. In the check_special_case() function in daemon.c I had used "errno" as an
5511 argument name, which causes warnings on some systems. This was basically a
5512 typo, since it was named "eno" in the comments!
5514 4. The code that waits for the clock to tick (at a resolution of some fraction
5515 of a second) so as to ensure message-id uniqueness was always waiting for
5516 at least one whole tick, when it could have waited for less. [This is
5517 almost certainly not relevant at current processor speeds, where it is
5518 unlikely to ever wait at all. But we try to future-proof.]
5520 5. The function that sleeps for a time interval that includes fractions of a
5521 second contained a race. It did not block SIGALRM between setting the
5522 timer, and suspending (a couple of lines later). If the interval was short
5523 and the sigsuspend() was delayed until after it had expired, the suspension
5524 never ended. On busy systems this could lead to processes getting stuck for
5527 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5528 process, before it forks any delivery processes. The open lookup caching
5529 mechanism meant that the open file or database connection was passed into
5530 the delivery process. The problem was that delivery processes always tidy
5531 up cached lookup data. This could cause a problem for the next delivery
5532 process started by the queue runner, because the external queue runner
5533 process does not know about the closure. So the next delivery process
5534 still has data in the lookup cache. In the case of a file lookup, there was
5535 no problem because closing a file descriptor in a subprocess doesn't affect
5536 the parent. However, if the lookup was caching a connection to a database,
5537 the connection was closed, and the second delivery process was likely to
5538 see errors such as "PGSQL: query failed: server closed the connection
5539 unexpectedly". The problem has been fixed by closing all cached lookups
5540 in a queue runner before running a delivery process.
5542 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5543 seem to have the "const" qualifier which it has on other OS. I've
5546 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5547 *of the same type* (client or server) with the same public name that an
5548 error should be diagnosed.
5550 9. When Exim looked up a host name for an IP address, but failed to find the
5551 original IP address when looking up the host name (a safety check), it
5552 output the message "<ip address> does not match any IP for NULL", which was
5553 confusing, to say the least. The bug was that the host name should have
5554 appeared instead of "NULL".
5556 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5557 user that is built into the binary, and the -C or -D options is used, root
5558 privilege is dropped before the configuration file is read. In addition,
5559 logging is switched to stderr instead of the normal log files. If the
5560 configuration then re-defines the Exim user, the unprivileged environment
5561 is probably not what is expected, so Exim logs a panic warning message (but
5564 However, if deliver_drop_privilege is set, the unprivileged state may well
5565 be exactly what is intended, so the warning has been cut out in that case,
5566 and Exim is allowed to try to write to its normal log files.
5572 1. smtp_return_error_details was not giving details for temporary sender
5573 or receiver verification errors.
5575 2. Diagnose a configuration error if two authenticators have the same public
5578 3. Exim used not to create the message log file for a message until the first
5579 delivery attempt. This could be confusing when incoming messages were held
5580 for policy or load reasons. The message log file is now created at the time
5581 the message is received, and an initial "Received" line is written to it.
5583 4. The automatically generated man page for command line options had a minor
5584 bug that caused no ill effects; however, a more serious problem was that
5585 the procedure for building the man page automatically didn't always
5586 operate. Consequently, release 4.20 contains an out-of-date version. This
5587 shouldn't happen again.
5589 5. When building Exim with embedded Perl support, the script that builds the
5590 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5591 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5593 6. The freeze_tell option was not being used for messages that were frozen on
5594 arrival, either by an ACL or by local_scan().
5596 7. Added the smtp_incomplete_transaction log selector.
5598 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5599 was accepting AUTH without a new EHLO.
5601 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5602 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5605 10. Logging of TCP/IP connections (when configured) now happens in the main
5606 daemon process instead of the child process, so that the TCP/IP connection
5607 count is more accurate (but it can never be perfect).
5609 11. The use of "drop" in a nested ACL was not being handled correctly in the
5610 outer ACL. Now, if condition failure induced by the nested "drop" causes
5611 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5612 or "require"), the connection is dropped.
5614 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5615 that yield "discard" can now be used with an "accept" or a "discard" verb,
5616 but an error is generated for any others (because I can't see a useful way
5617 to define what should happen).
5619 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5620 are now processed in the same way as lines in the Exim configuration file.
5621 In particular, continuation lines are supported.
5623 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5625 15. Added -ti meaning -t -i.
5627 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5628 domains, and warn by logging if others are found.
5630 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5631 returned from GIFCONF when Exim is trying to find the list of interfaces on
5632 a host. The code in os.c has been modified to copy each ifreq to an aligned
5633 structure in all cases.
5635 Also, in some cases, the returned ifreq's were being copied to a 'struct
5636 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5637 means the last couple of bytes of an IPv6 address could be chopped if the
5638 ifreq contained only a normal sockaddr (14 bytes storage).
5640 18. Named domain lists were not supported in the hosts_treat_as_local option.
5641 An entry such as +xxxx was not recognized, and was treated as a literal
5644 19. Ensure that header lines added by a DATA ACL are included in the reject log
5645 if the ACL subsequently rejects the message.
5647 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5648 MD5 (which is deprecated).
5650 21. When testing a filter file using -bf, Exim was writing a message when it
5651 took the sender from a "From " line in the message, but it was not doing so
5652 when it took $return_path from a Return-Path: header line. It now does.
5654 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5655 with a valid header line field name (a series of printing characters
5656 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5658 23. Changed "disc" in the source to "disk" to conform to the documentation and
5659 the book and for uniformity.
5661 24. Ignore Sendmail's -Ooption=value command line item.
5663 25. When execve() failed while trying to run a command in a pipe transport,
5664 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5665 could be confused with a return value of 69 from the command itself. This
5666 has been changed to 127, the value the shell returns if it is asked to run
5667 a non-existent command. The wording for the related log line suggests a
5668 non-existent command as the problem.
5670 26. If received_header_text expands to an empty string, do not add a Received:
5671 header line to the message. (Well, it adds a token one on the spool, but
5672 marks it "old" so that it doesn't get used or transmitted.)
5674 27. Installed eximstats 1.28 (addition of -nt option).
5676 28. There was no check for failure on the call to getsockname() in the daemon
5677 code. This can fail if there is a shortage of resources on the system, with
5678 ENOMEM, for example. A temporary error is now given on failure.
5680 29. Contrary to the C standard, it seems that in some environments, the
5681 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5682 program. Exim now does this explicitly; it affects the formatting of
5683 timestamps using strftime().
5685 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5686 complaints. I've now initialized all the variables, to avoid this.
5688 32. Header lines added by a system filter were not being "seen" during
5689 transport-time rewrites.
5691 33. The info_callback() function passed to OpenSSL is set up with type void
5692 (*)(SSL *, int, int), as described somewhere. However, when calling the
5693 function (actually a macro) that sets it up, the type void(*)() is
5694 expected. I've put in a cast to prevent warnings from picky compilers.
5696 34. If a DNS black list lookup found a CNAME record, but there were no A
5697 records associated with the domain it pointed at, Exim crashed.
5699 35. If a DNS black list lookup returned more than one A record, Exim ignored
5700 all but the first. It now scans all returned addresses if a particular IP
5701 value is being sought. In this situation, the contents of the
5702 $dnslist_value variable are a list of all the addresses, separated by a
5705 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5706 to accept a host name and all its aliases if the forward lookup for any of
5707 them yielded the IP address of the incoming connection. Now it accepts only
5708 those names whose forward lookup yields the correct IP address. Any other
5709 names are discarded. This closes a loophole whereby a rogue DNS
5710 administrator could create reverse DNS records to break through a
5711 wildcarded host restriction in an ACL.
5713 37. If a user filter or a system filter that ran in a subprocess used any of
5714 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5715 the wrong values were passed to the pipe command ($thisaddress had the
5716 value of $0, $0 had the value of $1, etc). This bug was introduced by
5717 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5719 38. Improved the line breaking for long SMTP error messages from ACLs.
5720 Previously, if there was no break point between 40 and 75 characters, Exim
5721 left the rest of the message alone. Two changes have been made: (a) I've
5722 reduced the minimum length to 35 characters; (b) if it can't find a break
5723 point between 35 and 75 characters, it looks ahead and uses the first one
5724 that it finds. This may give the occasional overlong line, but at least the
5725 remaining text gets split now.
5727 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5728 file descriptors might be low, and that setting 1000 would always raise it.
5729 It turns out that in some environments, the limit is already over 1000 and
5730 that lowering it causes trouble. So now Exim takes care not to decrease it.
5732 40. When delivering a message, the value of $return_path is set to $sender_
5733 address at the start of routing (routers may change the value). By an
5734 oversight, this default was not being set up when an address was tested by
5735 -bt or -bv, which affected the outcome if any router or filter referred to
5738 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5739 log only when "message" or "log_message" are set. However, if one of the
5740 conditions was an address verification, or a call to a nested ACL, the
5741 messages generated by the underlying test were being passed through. This
5742 no longer happens. The underlying message is available in $acl_verify_
5743 message for both "message" and "log_message" expansions, so it can be
5744 passed through if needed.
5746 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5747 new expansion $bh_ to give the encoded byte string without charset
5748 translation. Translation happens only if iconv() is available; HAVE_ICONV
5749 indicates this at build time. HEADERS_CHARSET gives the charset to
5750 translate to; headers_charset can change it in the configuration, and
5751 "headers charset" can change it in an individual filter file.
5753 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5754 that creates RFC 2047 encoded "words" labels them as that charset instead
5755 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5756 expansion operator; (ii) when Exim creates a From: line for a local
5757 message; (iii) when a header line is rewritten to include a "phrase" part.
5759 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5760 buggy, causing it to skip the first lines of messages whose message ID
5761 ended in 'D'. This would not have bitten before Exim release 4.14, because
5762 message IDs were unlikely to end in 'D' before then. The effect was to have
5763 incorrect size information for certain domains.
5765 45. #include "config.h" was missing at the start of the crypt16.c module. This
5766 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5769 46. If there was a timeout during a "random" callout check, Exim treated it as
5770 a failure of the random address, and carried on sending RSET and the real
5771 address. If the delay was just some slowness somewhere, the response to the
5772 original RCPT would be taken as a response to RSET and so on, causing
5773 mayhem of various kinds.
5775 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5776 when I implemented it. It didn't allow for the fact that some option values
5777 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5778 the right test for positive values.
5780 48. Domain names in DNS records are case-independent. Exim always looks them up
5781 in lower case. Some resolvers return domain names in exactly the case they
5782 appear in the zone file, that is, they may contain uppercase letters. Not
5783 all resolvers do this - some return always lower case. Exim was treating a
5784 change of case by a resolver as a change of domain, similar to a widening
5785 of a domain abbreviation. This triggered its re-routing code and so it was
5786 trying to route what was effectively the same domain again. This normally
5787 caused routing to fail (because the router wouldn't handle the domain
5788 twice). Now Exim checks for this case specially, and just changes the
5789 casing of the domain that it ultimately uses when it transmits the message
5792 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5795 50. If a filter generated a file delivery with a non-absolute name (possible if
5796 no home directory exists for the router), the forbid_file option was not
5799 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5800 the existing equality matching.
5802 52. Exim was using ints instead of ino_t variables in some places where it was
5803 dealing with inode numbers.
5805 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5806 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5807 called TMPDIR, and if it finds it is different, it changes its value.
5809 54. The smtp_printf() function is now made available to local_scan() so
5810 additional output lines can be written before returning. There is also an
5811 smtp_fflush() function to enable the detection of a dropped connection.
5812 The variables smtp_input and smtp_batched_input are exported to
5815 55. Changed the default runtime configuration: the message "Unknown user"
5816 has been removed from the ACL, and instead placed on the localuser router,
5817 using the cannot_route_message feature. This means that any verification
5818 failures that generate their own messages won't get overridden. Similarly,
5819 the "Unrouteable address" message that was in the ACL for unverifiable
5820 relay addresses has also been removed.
5822 56. Added hosts_avoid_esmtp to the smtp transport.
5824 57. The exicyclog script was not checking for the esoteric option
5825 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5826 will work only if exicyclog is run under the appropriate euid.
5828 58. Following a discussion on the list, the rules by which Exim recognises line
5829 endings on incoming messages have been changed. The -dropcr and drop_cr
5830 options are now no-ops, retained only for backwards compatibility. The
5831 following line terminators are recognized: LF CRLF CR. However, special
5832 processing applies to CR:
5834 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5835 nor a local message in the state where . is a terminator.
5837 (ii) If a bare CR is encountered in a header line, an extra space is added
5838 after the line terminator so as not to end the header. The reasoning
5839 behind this is that bare CRs in header lines are most likely either
5840 to be mistakes, or people trying to play silly games.
5842 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5843 was being incorrectly given as 18 bytes larger than it should have been.
5844 This is a VOB (very old bug).
5846 60. This may never have affected anything current, but just in case it has:
5847 When the local host is found other than at the start of a list of hosts,
5848 the local host, those with the same MX, and any that follow, are discarded.
5849 When the list in question was part of a longer list of hosts, the following
5850 hosts (not currently being processed) were also being discarded. This no
5851 longer happens. I'm not sure if this situation could ever has previously
5854 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5857 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5858 header. This is something that is recommended in a new Internet Draft, and
5859 is something that is documented as being done by Sendmail. There are two
5860 possible values. For messages generated by the autoreply transport, Exim
5863 Auto-Submitted: auto-replied
5865 whereas for all other generated messages (e.g. bounces) it adds
5867 Auto-Submitted: auto-generated
5869 63. The "personal" condition in filters now includes a test for the
5870 Auto-Submitted: header. If it contains the string "auto-" the message it
5871 not considered personal.
5873 64. Added rcpt_include_affixes as a generic transport option.
5875 65. Added queue_only_override (default true).
5877 66. Added the syslog_duplication option.
5879 67. If what should have been the first header line of a message consisted of
5880 a space followed by a colon, Exim was mis-interpreting it as a header line.
5881 It isn't of course - it is syntactically invalid and should therefore be
5882 treated as the start of the message body. The misbehaviour could have
5883 caused a number of strange effects, including loss of data in subsequent
5884 header lines, and spool format errors.
5886 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5887 client host had authenticated. This control can now be exercised by an ACL
5888 for more flexibility.
5890 69. By default, callouts do not happen when testing with -bh. There is now a
5891 variant, -bhc, which does actually run the callout code, including
5892 consulting and updating the callout cache.
5894 70. Added support for saslauthd authentication, courtesy of Alexander
5897 71. If statvfs() failed on the spool or log directories while checking their
5898 size for availability, Exim confusingly gave the error "space shortage".
5899 Furthermore, in debugging mode it crashed with a floating point exception.
5900 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5901 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5902 serious problem, Exim now writes to the main and panic logs when this
5903 happens, with details of the failure. It then refuses to accept the
5904 incoming message, giving the message "spool directory problem" or "log
5905 directory problem" with a 421 code for SMTP messages.
5907 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5908 0, 1, and 2 exist, because some OS complain for execs without them (see
5909 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5910 descriptors. However, the code omitted to check that the open succeeded,
5911 causing mysterious errors if for some reason the permissions on /dev/null
5912 got screwed. Now Exim writes a message to the main and panic logs, and
5913 bombs out if it can't open /dev/null.
5915 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5916 interact so that it is all more flexible. It is supposed to remain
5917 backwards compatible. Also added extra_local_interfaces.
5919 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5920 to bomb out with an assertion failure - to the client this appears as a
5921 connection drop. This problem occurs in the part of the code that was taken
5922 from the Samba project. Fortunately, the assertion is in a very simple
5923 function, so I have fixed this by reproducing the function inline in the
5924 one place where it is called, and arranging for authentication to fail
5925 instead of killing the process with assert().
5927 75. The SPA client code was not working when the server requested OEM rather
5928 than Unicode encoding.
5930 76. Added code to make require_files with a specific uid setting more usable in
5931 the case where statting the file as root fails - usually a non-root-mounted
5932 NFS file system. When this happens and the failure is EACCES, Exim now
5933 forks a subprocess and does the per-uid checking as the relevant uid.
5935 77. Added process_log_path.
5937 78. If log_file_path was not explicitly set, a setting of check_log_space or
5938 check_log_inodes was ignored.
5940 79. If a space check for the spool or log partitions fails, the incident is now
5941 logged. Of course, in the latter case the data may get lost...
5943 80. Added the %p formatting code to string_format() so that it can be used to
5944 print addresses in debug_print(). Adjusted all the address printing in the
5945 debugging in store.c to use %p rather than %d.
5947 81. There was a concern that a line of code in smtp_in.c could overflow a
5948 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5949 initially expressed, the concern was not well-founded, because trailing
5950 spaces are removed early. However, if the trailing spaces were followed by
5951 a NULL, they did not get removed, so the overflow was possible. Two fixes
5954 (a) I re-wrote the offending code in a cleaner fashion.
5955 (b) If an incoming SMTP command contains a NULL character, it is rejected
5958 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5959 runs initgroups(), so that if the Exim user is in any additional groups,
5960 they will be used during message reception.
5966 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob