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
42 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
44 PP/02 Make -n do something, by making it not do something.
45 When combined with -bP, the name of an option is not output.
47 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
50 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
51 $sender_host_name and config options to manage this, and basic check
54 PP/05 DSCP support for outbound connections and control modifier for inbound.
56 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
57 (Only plugin which currently uses this is kerberos4, which nobody should
58 be using, but we should make it available and other future plugins might
59 conceivably use it, even though it would break NAT; stuff *should* be
60 using channel bindings instead).
62 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
63 name; added for Sendmail compatibility; requires admin caller.
64 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
65 just ignore it); requires trusted caller.
66 Also parse but ignore: -Ac -Am -X<logfile>
69 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
71 TL/02 Add +smtp_confirmation as a default logging option.
73 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
74 Patch by Magnus Holmgren from 2007-02-20.
76 TL/04 Bugzilla 1281 - Spec typo.
77 Bugzilla 1283 - Spec typo.
78 Bugzilla 1290 - Spec grammar fixes.
80 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
82 TL/06 Add Experimental DMARC support using libopendmarc libraries.
84 TL/07 Fix an out of order global option causing a segfault. Reported to dev
85 mailing list by by Dmitry Isaikin.
87 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
89 JH/02 Support "G" suffix to numbers in ${if comparisons.
91 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
93 NM/01 Bugzilla 1197 - Spec typo
94 Bugzilla 1196 - Spec examples corrections
96 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
98 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
99 gnutls_enable_pkcs11, but renamed to more accurately indicate its
102 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
103 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
105 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
106 "acl {{name}{arg}...}", and optional args on acl condition
109 JH/05 Permit multiple router/transport headers_add/remove lines.
111 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
113 JH/07 Avoid using a waiting database for a single-message-only transport.
114 Performance patch from Paul Fisher. Bugzilla 1262.
116 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
119 JH/09 Add $headers_added variable, with content from use of ACL modifier
120 add_header (but not yet added to the message). Bugzilla 199.
122 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
123 Pulled from Bugzilla 817 by Wolfgang Breyha.
125 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
127 (nb: this is the same fix as in Exim 4.80.1)
129 JH/11 Add A= logging on delivery lines, and a client_set_id option on
132 JH/12 Add optional authenticated_sender logging to A= and a log_selector
135 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
137 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
138 advertise SMTP AUTH mechanism to us, instead of a generic
139 protocol violation error. Also, make Exim more robust to bad
140 data from the Dovecot auth socket.
142 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
144 When a queue runner is handling a message, Exim first routes the
145 recipient addresses, during which it prunes them based on the retry
146 hints database. After that it attempts to deliver the message to
147 any remaining recipients. It then updates the hints database using
150 So if a recipient address works intermittently, it can get repeatedly
151 deferred at routing time. The retry hints record remains fresh so the
152 address never reaches the final cutoff time.
154 This is a fairly common occurrence when a user is bumping up against
155 their storage quota. Exim had some logic in its local delivery code
156 to deal with this. However it did not apply to per-recipient defers
157 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
159 This change adds a proper retry rule check during routing so that the
160 final cutoff time is checked against the message's age. We only do
161 this check if there is an address retry record and there is not a
162 domain retry record; this implies that previous attempts to handle
163 the address had the retry_use_local_parts option turned on. We use
164 this as an approximation for the destination being like a local
165 delivery, as in LMTP.
167 I suspect this new check makes the old local delivery cutoff check
168 redundant, but I have not verified this so I left the code in place.
170 TF/02 Correct gecos expansion when From: is a prefix of the username.
172 Test 0254 submits a message to Exim with the header
176 When I ran the test suite under the user fanf2, Exim expanded
177 the header to contain my full name, whereas it should have added
178 a Resent-Sender: header. It erroneously treats any prefix of the
179 username as equal to the username.
181 This change corrects that bug.
183 GF/01 DCC debug and logging tidyup
184 Error conditions log to paniclog rather than rejectlog.
185 Debug lines prefixed by "DCC: " to remove any ambiguity.
187 TF/03 Avoid unnecessary rebuilds of lookup-related code.
189 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
190 Bug spotted by Jeremy Harris; was flawed since initial commit.
191 Would have resulted in OCSP responses post-SNI triggering an Exim
192 NULL dereference and crash.
194 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
196 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
197 Bug detection, analysis and fix by Samuel Thibault.
198 Bugzilla 1331, Debian bug #698092.
200 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
202 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
203 Server implementation by Todd Lyons, client by JH.
204 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
205 config variable "prdr_enable" controls whether the server
206 advertises the facility. If the client requests PRDR a new
207 acl_data_smtp_prdr ACL is called once for each recipient, after
208 the body content is received and before the acl_smtp_data ACL.
209 The client is controlled by bolth of: a hosts_try_prdr option
210 on the smtp transport, and the server advertisement.
211 Default client logging of deliveries and rejections involving
212 PRDR are flagged with the string "PRDR".
214 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
215 fclose(). Diagnosis by Todd Lyons.
217 PP/17 Update configure.default to handle IPv6 localhost better.
218 Patch by Alain Williams (plus minor tweaks).
221 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
222 This is now consistent with GnuTLS, and is now documented: the
223 previous undocumented portable approach to treating the option as
224 unset was to force an expansion failure. That still works, and
225 an empty string is now equivalent.
227 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
228 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
229 not performing validation itself.
231 PP/20 Added force_command boolean option to pipe transport.
232 Patch from Nick Koston, of cPanel Inc.
234 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
237 TF/04 Added udpsend ACL modifer and hexquote expansion operator
239 PP/21 Fix eximon continuous updating with timestamped log-files.
240 Broken in a format-string cleanup in 4.80, missed when I repaired the
241 other false fix of the same issue.
242 Report and fix from Heiko Schlichting.
245 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
246 Report from Prashanth Katuri.
248 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
249 It's SecureTransport, so affects any MacOS clients which use the
250 system-integrated TLS libraries, including email clients.
252 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
253 using a MIME ACL for non-SMTP local injection.
254 Report and assistance in diagnosis by Warren Baker.
256 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
258 JH/16 Fix comparisons for 64b. Bugzilla 1385.
260 TL/09 Add expansion variable $authenticated_fail_id to keep track of
261 last id that failed so it may be referenced in subsequent ACL's.
263 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
266 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
267 ldap library initialization, allowing self-signed CA's to be
268 used. Also properly sets require_cert option later in code by
269 using NULL (global ldap config) instead of ldap handle (per
270 session). Bug diagnosis and testing by alxgomz.
272 TL/12 Enhanced documentation in the ratelimit.pl script provided in
273 the src/util/ subdirectory.
275 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
276 renamed to Transport Post Delivery Action by Jeremy Harris, as
279 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
280 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
281 redis_servers = needs to be configured which will be used by the redis
282 lookup. Patch from Warren Baker, of The Packet Hub.
284 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
286 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
287 hostname or reverse DNS when processing a host list. Used suggestions
288 from multiple comments on this bug.
290 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
292 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
293 Missed a few lines, added it to make the runtest require no keyboard
296 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
297 contains upper case chars. Make router use caseful_local_part.
299 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
300 support when GnuTLS has been built with p11-kit.
306 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
308 This, or similar/improved, will also be change PP/11 of 4.82.
314 PP/01 Handle short writes when writing local log-files.
315 In practice, only affects FreeBSD (8 onwards).
316 Bugzilla 1053, with thanks to Dmitry Isaikin.
318 NM/01 Bugzilla 949 - Documentation tweak
320 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
323 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
325 PP/02 Implemented gsasl authenticator.
327 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
329 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
330 `pkg-config foo` for cflags/libs.
332 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
333 with rest of GSASL and with heimdal_gssapi.
335 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
336 `pkg-config foo` for cflags/libs for the TLS implementation.
338 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
339 properties get this fed in as external SSF. A number of robustness
340 and debugging improvements to the cyrus_sasl authenticator.
342 PP/08 cyrus_sasl server now expands the server_realm option.
344 PP/09 Bugzilla 1214 - Log authentication information in reject log.
345 Patch by Jeremy Harris.
347 PP/10 Added dbmjz lookup type.
349 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
351 PP/12 MAIL args handles TAB as well as SP, for better interop with
352 non-compliant senders.
353 Analysis and variant patch by Todd Lyons.
355 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
356 Bug report from Lars Müller <lars@samba.org> (via SUSE),
357 Patch from Dirk Mueller <dmueller@suse.com>
359 PP/13 tls_peerdn now print-escaped for spool files.
360 Observed some $tls_peerdn in wild which contained \n, which resulted
361 in spool file corruption.
363 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
364 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
365 or write after TLS renegotiation, which otherwise led to messages
368 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
369 as a tracking header (ie: a signed header comes before the signature).
370 Patch from Wolfgang Breyha.
372 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
373 comma-sep list; embedded commas doubled.
375 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
377 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
379 Report and patch from Dmitry Banschikov.
381 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
382 Removed SSL_clear() after SSL_new() which led to protocol negotiation
383 failures. We appear to now support TLS1.1+ with Exim.
385 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
386 lets Exim select keys and certificates based upon TLS SNI from client.
387 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
388 before an outbound SMTP session. New log_selector, +tls_sni.
390 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
391 NULL dereference. Report and patch from Alun Jones.
393 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
394 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
395 Not seeing resolver debug output on NetBSD, but suspect this is a
396 resolver implementation change.
398 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
399 Left warnings. Added "eximon gdb" invocation mode.
401 PP/21 Defaulting "accept_8bitmime" to true, not false.
403 PP/22 Added -bw for inetd wait mode support.
405 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
406 locate the relevant includes and libraries. Made this the default.
408 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
409 Bugzilla 1246, report and most of solution from Tomasz Kusy.
411 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
412 This may cause build issues on older platforms.
414 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
415 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
416 gnutls_require_mac & gnutls_require_protocols (no longer supported).
417 Added SNI support via GnuTLS too.
418 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
420 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
422 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
423 Applied second patch from Janne, implementing suggestion to default
424 multiple-strings-in-record handling to match SPF spec.
426 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
428 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
429 read-only, out of scope).
430 Patch from Wolfgang Breyha, report from Stuart Northfield.
432 PP/29 Fix three issues highlighted by clang analyser static analysis.
433 Only crash-plausible issue would require the Cambridge-specific
434 iplookup router and a misconfiguration.
435 Report from Marcin Mirosław.
437 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
439 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
440 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
441 As part of this, removing so much warning spew let me fix some minor
442 real issues in debug logging.
444 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
445 assignment on my part. Fixed.
447 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
448 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
449 Janne Snabb (who went above and beyond: thank you).
451 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
452 string otherwise requires a connection and a bunch more work and it's
453 relatively easy to get wrong. Should also expose TLS library linkage
456 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
457 64-bit ${eval} (JH/03).
459 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
460 GNU libc to support some of the 64-bit stuff, should not lead to
461 conflicts. Defined before os.h is pulled in, so if a given platform
462 needs to override this, it can.
464 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
465 protection layer was required, which is not implemented.
466 Bugzilla 1254, patch from Wolfgang Breyha.
468 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
469 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
470 tls_dhparam take prime identifiers. Also unbreak combination of
471 OpenSSL+DH_params+TLSSNI.
473 PP/39 Disable SSLv2 by default in OpenSSL support.
479 PP/01 Solaris build fix for Oracle's LDAP libraries.
480 Bugzilla 1109, patch from Stephen Usher.
482 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
484 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
487 TF/02 Fix a couple more cases where we did not log the error message
488 when unlink() failed. See also change 4.74-TF/03.
490 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
491 lock up or crash if it happened to be inside a call to libc when it
492 got a SIGUSR1 from exiwhat.
494 The SIGUSR1 handler appends the current process status to the process
495 log which is later printed by exiwhat. It used to use the general
496 purpose logging code to do this, but several functions it calls are
497 not safe for signals.
499 The new output code in the SIGUSR1 handler is specific to the process
500 log, and simple enough that it's easy to inspect for signal safety.
501 Removing some special cases also simplifies the general logging code.
502 Removing the spurious timestamps from the process log simplifies
505 TF/04 Improved ratelimit ACL condition.
507 The /noupdate option has been deprecated in favour of /readonly which
508 has clearer semantics. The /leaky, /strict, and /readonly update modes
509 are mutually exclusive. The update mode is no longer included in the
510 database key; it just determines when the database is updated. (This
511 means that when you upgrde Exim will forget old rate measurements.)
513 Exim now checks that the per_* options are used with an update mode that
514 makes sense for the current ACL. For example, when Exim is processing a
515 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
516 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
517 must specify per_mail/readonly. If you omit the update mode it defaults to
518 /leaky where that makes sense (as before) or /readonly where required.
520 The /noupdate option is now undocumented but still supported for
521 backwards compatibility. It is equivalent to /readonly except that in
522 ACLs where /readonly is required you may specify /leaky/noupdate or
523 /strict/noupdate which are treated the same as /readonly.
525 A useful new feature is the /count= option. This is a generalization
526 of the per_byte option, so that you can measure the throughput of other
527 aggregate values. For example, the per_byte option is now equivalent
528 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
530 The per_rcpt option has been generalized using the /count= mechanism
531 (though it's more complicated than the per_byte equivalence). When it is
532 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
533 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
534 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
535 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
536 Note that using per_rcpt with a non-readonly update mode in more than
537 one ACL will cause the recipients to be double-counted. (The per_mail
538 and per_byte options don't have this problem.)
540 The handling of very low rates has changed slightly. If the computed rate
541 is less than the event's count (usually one) then this event is the first
542 after a long gap. In this case the rate is set to the same as this event's
543 count, so that the first message of a spam run is counted properly.
545 The major new feature is a mechanism for counting the rate of unique
546 events. The new per_addr option counts the number of different
547 recipients that someone has sent messages to in the last time period. It
548 behaves like per_rcpt if all the recipient addresses are different, but
549 duplicate recipient addresses do not increase the measured rate. Like
550 the /count= option this is a general mechanism, so the per_addr option
551 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
552 example, measure the rate that a client uses different sender addresses
553 with the options per_mail/unique=$sender_address. There are further
554 details in the main documentation.
556 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
558 TF/06 Removed a few PCRE remnants.
560 TF/07 Automatically extract Exim's version number from tags in the git
561 repository when doing development or release builds.
563 PP/02 Raise smtp_cmd_buffer_size to 16kB.
564 Bugzilla 879. Patch from Paul Fisher.
566 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
567 Heavily based on revision 40f9a89a from Simon Arlott's tree.
570 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
572 PP/05 Variable $av_failed, true if the AV scanner deferred.
573 Bugzilla 1078. Patch from John Horne.
575 PP/06 Stop make process more reliably on build failure.
576 Bugzilla 1087. Patch from Heiko Schlittermann.
578 PP/07 Make maildir_use_size_file an _expandable_ boolean.
579 Bugzilla 1089. Patch from Heiko Schlittermann.
581 PP/08 Handle ${run} returning more data than OS pipe buffer size.
582 Bugzilla 1131. Patch from Holger Weiß.
584 PP/09 Handle IPv6 addresses with SPF.
585 Bugzilla 860. Patch from Wolfgang Breyha.
587 PP/10 GnuTLS: support TLS 1.2 & 1.1.
589 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
592 PP/11 match_* no longer expand right-hand-side by default.
593 New compile-time build option, EXPAND_LISTMATCH_RHS.
594 New expansion conditions, "inlist", "inlisti".
596 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
598 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
600 PP/14 fix log_write() format string regression from TF/03.
601 Bugzilla 1152. Patch from Dmitry Isaikin.
607 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
609 PP/02 Harmonised TLS library version reporting; only show if debugging.
610 Layout now matches that introduced for other libraries in 4.74 PP/03.
612 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
614 PP/04 New "dns_use_edns0" global option.
616 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
619 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
620 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
622 TK/01 Updated PolarSSL code to 0.14.2.
623 Bugzilla 1097. Patch from Andreas Metzler.
625 PP/07 Catch divide-by-zero in ${eval:...}.
628 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
631 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
632 format-string attack -- SECURITY: remote arbitrary code execution.
634 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
635 time unintentionally subject to list matching rules, letting the header
636 cause arbitrary Exim lookups (of items which can occur in lists, *not*
637 arbitrary string expansion). This allowed for information disclosure.
639 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
640 INT_MIN/-1 -- value coerced to INT_MAX.
646 NM/01 Workround for PCRE version dependency in version reporting
649 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
650 This fixes portability to compilers other than gcc, notably
651 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
653 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
654 makefiles for portability to HP-UX and POSIX correctness.
656 PP/01 Permit LOOKUP_foo enabling on the make command-line.
657 Also via indented variable definition in the Makefile.
658 (Debugging by Oliver Heesakkers).
660 PP/02 Restore caching of spamd results with expanded spamd_address.
661 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
663 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
664 Improves build reliability. Fix from: Frank Elsner
666 NM/02 Fix wide character breakage in the rfc2047 coding
667 Fixes bug 1064. Patch from Andrey N. Oktyabrski
669 NM/03 Allow underscore in dnslist lookups
670 Fixes bug 1026. Patch from Graeme Fowler
672 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
673 Code patches from Adam Ciarcinski of NetBSD.
675 NM/04 Fixed exiqgrep to cope with mailq missing size issue
678 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
679 is logged, to avoid truncation. Patch from John Horne.
681 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
682 Patch from Jakob Hirsch.
684 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
685 SQL string expansion failure details.
686 Patch from Andrey Oktyabrski.
688 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
689 Patch from Simon Arlott.
691 PP/09 New lookups functionality failed to compile on old gcc which rejects
692 extern declarations in function scope.
693 Patch from Oliver Fleischmann
695 PP/10 Use sig_atomic_t for flags set from signal handlers.
696 Check getgroups() return and improve debugging.
697 Fixed developed for diagnosis in bug 927 (which turned out to be
700 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
701 Patch from Mark Zealey.
703 PP/12 Bugzilla 1056: Improved spamd server selection.
704 Patch from Mark Zealey.
706 PP/13 Bugzilla 1086: Deal with maildir quota file races.
707 Based on patch from Heiko Schlittermann.
709 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
710 Patch from Uwe Doering, sign-off by Michael Haardt.
712 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
713 variable declaration deep within a block. Bug and patch from
716 PP/15 lookups-Makefile IRIX compatibilty coercion.
718 PP/16 Make DISABLE_DKIM build knob functional.
720 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
721 Patch by Simon Arlott
723 TF/03 Fix valgrind.h portability to C89 compilers that do not support
724 variable argument macros. Our copy now differs from upstream.
730 TF/01 Failure to get a lock on a hints database can have serious
731 consequences so log it to the panic log.
733 TF/02 Log LMTP confirmation messages in the same way as SMTP,
734 controlled using the smtp_confirmation log selector.
736 TF/03 Include the error message when we fail to unlink a spool file.
738 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
739 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
740 for maintaining out-of-tree patches for some time.
742 PP/01 Bugzilla 139: Documentation and portability issues.
743 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
744 Handle per-OS dynamic-module compilation flags.
746 PP/02 Let /dev/null have normal permissions.
747 The 4.73 fixes were a little too stringent and complained about the
748 permissions on /dev/null. Exempt it from some checks.
749 Reported by Andreas M. Kirchwitz.
751 PP/03 Report version information for many libraries, including
752 Exim version information for dynamically loaded libraries. Created
753 version.h, now support a version extension string for distributors
754 who patch heavily. Dynamic module ABI change.
756 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
757 privilege escalation vulnerability whereby the Exim run-time user
758 can cause root to append content of the attacker's choosing to
761 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
764 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
765 If dropping privileges for untrusted macros, we disabled normal logging
766 on the basis that it would fail; for the Exim run-time user, this is not
767 the case, and it resulted in successful deliveries going unlogged.
768 Fixed. Reported by Andreas Metzler.
774 PP/01 Date: & Message-Id: revert to normally being appended to a message,
775 only prepend for the Resent-* case. Fixes regression introduced in
776 Exim 4.70 by NM/22 for Bugzilla 607.
778 PP/02 Include check_rfc2047_length in configure.default because we're seeing
779 increasing numbers of administrators be bitten by this.
781 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
783 PP/03 Bugzilla 994: added openssl_options main configuration option.
785 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
787 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
789 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
791 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
792 without a peer certificate, leading to a segfault because of an
793 assumption that peers always have certificates. Be a little more
794 paranoid. Problem reported by Martin Tscholak.
796 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
797 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
798 NB: ClamAV planning to remove STREAM in "middle of 2010".
799 CL also introduces -bmalware, various -d+acl logging additions and
800 more caution in buffer sizes.
802 PP/09 Implemented reverse_ip expansion operator.
804 PP/10 Bugzilla 937: provide a "debug" ACL control.
806 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
808 PP/12 Bugzilla 973: Implement --version.
810 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
812 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
814 PP/15 Bugzilla 816: support multiple condition rules on Routers.
816 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
817 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
818 ignore trailing whitespace.
820 JJ/02 prevent non-panic DKIM error from being sent to paniclog
822 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
825 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
826 Notification from Dr Andrew Aitchison.
828 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
829 ExtendedDetectionInfo response format.
830 Notification from John Horne.
832 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
835 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
836 XSL and documented dependency on system catalogs, with examples of how
839 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
842 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
843 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
844 configuration file which is writeable by the Exim user or group.
846 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
847 of configuration files to cover files specified with the -C option if
848 they are going to be used with root privileges, not just the default
851 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
852 option (effectively making it always true).
854 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
855 files to be used while preserving root privileges.
857 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
858 that rogue child processes cannot use them.
860 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
861 run-time user, instead of root.
863 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
864 Exim run-time user without dropping privileges.
866 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
867 result string, instead of calling string_vformat() twice with the same
870 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
871 for other users. Others should always drop root privileges if they use
872 -C on the command line, even for a whitelisted configure file.
874 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
876 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
882 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
883 $data_path, and $header_path variables; fixed documentation bugs and
886 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
887 exipick to access non-standard spools, including the "frozen" queue
890 NM/01 Bugzilla 965: Support mysql stored procedures.
891 Patch from Alain Williams
893 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
895 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
896 Patch from Andreas Metzler
898 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
899 Patch from Kirill Miazine
901 NM/05 Bugzilla 671: Added umask to procmail example.
903 JJ/03 installed exipick 20100323.0, fixing doc bug
905 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
906 directory. Notification and patch from Dan Rosenberg.
908 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
910 TK/02 Improve log output when DKIM signing operation fails.
912 MH/01 Treat the transport option dkim_domain as a colon separated
913 list, not as a single string, and sign the message with each element,
914 omitting multiple occurences of the same signer.
916 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
917 Bugzilla 985, 986. Patch by Simon Arlott
919 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
920 Patch by Simon Arlott
922 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
923 MBX locking. Notification from Dan Rosenberg.
929 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
931 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
933 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
935 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
937 NM/04 Bugzilla 915: Flag broken perl installation during build.
943 TK/01 Added patch by Johannes Berg that expands the main option
944 "spamd_address" if it starts with a dollar sign.
946 TK/02 Write list of recipients to X-Envelope-Sender header when building
947 the mbox-format spool file for content scanning (suggested by Jakob
950 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
951 (http://www.dcc-servers.net/) support via dccifd. Activated by
952 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
954 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
955 by Mark Daniel Reidel <mr@df.eu>.
957 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
958 When building exim an external PCRE library is now needed -
959 PCRE is a system library on the majority of modern systems.
960 See entry on PCRE_LIBS in EDITME file.
962 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
963 conversation. Added nologin parameter to request.
964 Patch contributed by Kirill Miazine.
966 TF/01 Do not log submission mode rewrites if they do not change the address.
968 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
970 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
971 log files in place. Contributed by Roberto Lima.
973 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
975 TF/03 Bugzilla 615: When checking the local_parts router precondition
976 after a local_part_suffix or local_part_prefix option, Exim now
977 does not use the address's named list lookup cache, since this
978 contains cached lookups for the whole local part.
980 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
981 Robert Millan. Documentation is in experimental-spec.txt.
983 TF/04 Bugzilla 668: Fix parallel build (make -j).
985 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
987 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
988 Patch provided by Jan Srzednicki.
990 TF/05 Leading white space used to be stripped from $spam_report which
991 wrecked the formatting. Now it is preserved.
993 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
994 that they are available at delivery time.
996 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
998 TF/08 TLS error reporting now respects the incoming_interface and
999 incoming_port log selectors.
1001 TF/09 Produce a more useful error message if an SMTP transport's hosts
1002 setting expands to an empty string.
1004 NM/06 Bugzilla 744: EXPN did not work under TLS.
1005 Patch provided by Phil Pennock.
1007 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1008 Patch provided by Richard Godbee.
1010 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1011 acl_smtp_notquit, added index entry.
1013 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1014 Patch provided by Eugene Bujak.
1016 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1017 accept(). Patch provided by Maxim Dounin.
1019 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1020 Patch provided by Phil Pennock.
1022 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1024 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1025 Patch provided by Brad "anomie" Jorsch.
1027 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1028 Patch provided by Dean Brooks.
1030 TK/05 Add native DKIM support (does not depend on external libraries).
1032 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1033 Patch provided by Graeme Fowler.
1035 NM/16 Bugzilla 851: Documentation example syntax fix.
1037 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1039 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1042 NM/19 Bugzilla 745: TLS version reporting.
1043 Patch provided by Phil Pennock.
1045 NM/20 Bugzilla 167: bool: condition support.
1046 Patch provided by Phil Pennock.
1048 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1049 clients. Patch provided by Phil Pennock.
1051 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1052 Patch provided by Brad "anomie" Jorsch.
1054 NM/23 Bugzilla 687: Fix misparses in eximstats.
1055 Patch provided by Heiko Schlittermann.
1057 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1058 Patch provided by Heiko Schlittermann.
1060 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1061 plus update to original patch.
1063 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1065 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1066 Patch provided by David Brownlee.
1068 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1070 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1072 NM/30 Bugzilla 888: TLS documentation bugfixes.
1074 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1076 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1077 Unlike the original bugzilla I have changed all shell scripts in src tree.
1079 NM/33 Bugzilla 898: Transport filter timeout fix.
1080 Patch by Todd Rinaldo.
1082 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1083 Patch by Serge Demonchaux.
1085 NM/35 Bugzilla 39: Base64 decode bug fixes.
1086 Patch by Jakob Hirsch.
1088 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1090 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1092 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1094 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1100 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1101 ALT-N's libdkim that I have put here:
1102 http://duncanthrax.net/exim-experimental/
1104 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1105 were called 'true' and it seems that C99 defines that as a reserved
1106 keyword to be used with 'bool' variable types. That means you could
1107 not include C99-style headers which use bools without triggering
1108 build errors in sieve.c.
1110 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1111 as mailq or other aliases. Changed the --help handling significantly
1112 to do whats expected. exim_usage() emits usage/help information.
1114 SC/01 Added the -bylocaldomain option to eximstats.
1116 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1118 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1120 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1126 PH/01 Another patch from the Sieve maintainer.
1128 PH/02 When an IPv6 address is converted to a string for single-key lookup
1129 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1130 dots are used instead of colons so that keys in lsearch files need not
1131 contain colons. This was done some time before quoting was made available
1132 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1133 (notated using the quote facility) so as to distinguish them from IPv4
1134 keys. This meant that lookups for IP addresses in host lists did not work
1135 for iplsearch lookups.
1137 This has been fixed by arranging for IPv6 addresses to be expressed with
1138 colons if the lookup type is iplsearch. This is not incompatible, because
1139 previously such lookups could never work.
1141 The situation is now rather anomolous, since one *can* have colons in
1142 ordinary lsearch keys. However, making the change in all cases is
1143 incompatible and would probably break a number of configurations.
1145 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1148 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1149 conversion specification without a maximum field width, thereby enabling
1150 a rogue spamd server to cause a buffer overflow. While nobody in their
1151 right mind would setup Exim to query an untrusted spamd server, an
1152 attacker that gains access to a server running spamd could potentially
1153 exploit this vulnerability to run arbitrary code as the Exim user.
1155 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1156 $primary_hostname instead of what libspf2 thinks the hosts name is.
1158 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1159 a directory entry by the name of the lookup key. Previously, if a
1160 symlink pointed to a non-existing file or a file in a directory that
1161 Exim lacked permissions to read, a lookup for a key matching that
1162 symlink would fail. Now it is enough that a matching directory entry
1163 exists, symlink or not. (Bugzilla 503.)
1165 PH/03 The body_linecount and body_zerocount variables are now exported in the
1168 PH/04 Added the $dnslist_matched variable.
1170 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1171 This means they are set thereafter only if the connection becomes
1174 PH/06 Added the client_condition to authenticators so that some can be skipped
1175 by clients under certain conditions.
1177 PH/07 The error message for a badly-placed control=no_multiline_responses left
1178 "_responses" off the end of the name.
1180 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1182 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1183 (without spaces) instead of just copying the configuration text.
1185 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1187 PH/11 Added $max_received_linelength.
1189 PH/12 Added +ignore_defer and +include_defer to host lists.
1191 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1192 way in which PCRE > 7.0 is built.
1194 PH/14 Implemented queue_only_load_latch.
1196 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1197 MAIL command. The effect was to mangle the value on 64-bit systems.
1199 PH/16 Another patch from the Sieve maintainer.
1201 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1203 PH/18 If a system quota error occurred while trying to create the file for
1204 a maildir delivery, the message "Mailbox is full" was not appended to the
1205 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1206 only to a quota excession during the actual writing of the file.
1208 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1209 characters?) which causes problems in log lines. The DN values are now
1210 passed through string_printing() before being added to log lines.
1212 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1213 and InterBase are left for another time.)
1215 PH/21 Added message_body_newlines option.
1217 PH/22 Guard against possible overflow in moan_check_errorcopy().
1219 PH/23 POSIX allows open() to be a macro; guard against that.
1221 PH/24 If the recipient of an error message contained an @ in the local part
1222 (suitably quoted, of course), incorrect values were put in $domain and
1223 $local_part during the evaluation of errors_copy.
1229 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1230 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1233 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1234 issue a MAIL command.
1236 PH/02 In an ACL statement such as
1238 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1240 if a client was not listed at all, or was listed with a value other than
1241 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1242 the condition was not true (as it should be), so access was not denied.
1243 The bug was that the ! inversion was incorrectly passed on to the second
1244 item. This has been fixed.
1246 PH/03 Added additional dnslists conditions == and =& which are different from
1247 = and & when the dns lookup returns more than one IP address.
1249 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1250 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1252 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1253 FSYNC, which compiles an option called disable_fsync that allows for
1254 bypassing fsync(). The documentation is heavily laced with warnings.
1256 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1258 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1259 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1260 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1261 including adding "make clean"; (3) Added -fPIC when compiling the test
1262 dynamically loaded module, to get rid of a warning.
1264 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1265 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1266 = 0s. The bug is otherwise harmless.
1268 PH/07 There was a bug in the dovecot authenticator such that the value of
1269 $auth1 could be overwritten, and so not correctly preserved, after a
1270 successful authentication. This usually meant that the value preserved by
1271 the server_setid option was incorrect.
1273 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1275 PH/09 Installed PCRE release 7.0.
1277 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1278 run for batched SMTP input. It is now run at the start of every message
1279 in the batch. While fixing this I discovered that the process information
1280 (output by running exiwhat) was not always getting set for -bs and -bS
1281 input. This is fixed, and it now also says "batched" for BSMTP.
1283 PH/11 Added control=no_pipelining.
1285 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1286 patch, slightly modified), and move the expansion of helo_data till after
1287 the connection is made in the smtp transport (so it can use these
1290 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1292 PH/14 Added log_selector = +pid.
1294 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1296 PH/16 Add ${if forany and ${if forall.
1298 PH/17 Added dsn_from option to vary the From: line in DSNs.
1300 PH/18 Flush SMTP output before performing a callout, unless control =
1301 no_callout_flush is set.
1303 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1304 was true (the default) a successful delivery failed to delete the retry
1305 item, thus causing premature timeout of the address. The bug is now
1308 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1310 PH/21 Long custom messages for fakedefer and fakereject are now split up
1311 into multiline reponses in the same way that messages for "deny" and
1312 other ACL rejections are.
1314 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1315 with slight modification.
1317 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1318 draft, changing the syntax and factoring some duplicate code".
1320 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1321 for deliveries of the second and subsequent messages over the same SMTP
1324 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1325 ${reduce, with only minor "tidies".
1327 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1329 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1330 expansion side effects.
1332 PH/27 When a message times out after an over-quota error from an Exim-imposed
1333 quota, the bounce message says "mailbox is full". This message was not
1334 being given when it was a system quota that was exceeded. It now should
1337 MH/03 Made $recipients available in local_scan(). local_scan() already has
1338 better access to the recipient list through recipients_list[], but
1339 $recipients can be useful in postmaster-provided expansion strings.
1341 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1342 in the case of a MAIL command with additional options following the
1343 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1344 were accidentally chopped off.
1346 PH/29 SMTP synchronization checks are implemented when a command is read -
1347 there is a check that no more input is waiting when there shouldn't be
1348 any. However, for some commands, a delay in an ACL can mean that it is
1349 some time before the response is written. In this time, more input might
1350 arrive, invalidly. So now there are extra checks after an ACL has run for
1351 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1352 pipelining has not been advertised.
1354 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1356 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1357 returned in the SMTP response when smtp_return_error_details was false.
1358 This has been fixed.
1360 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1361 instead of the C I/O that was originally supplied, because problems were
1362 reported on Solaris.
1364 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1365 Exim which did not show up earlier: it was assuming that a call to
1366 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1367 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1368 was a macro that became an assignment, so it seemed to work. This has
1369 changed to a proper function call with a void return, hence the compile
1370 error. Exim's code has been fixed.
1372 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1375 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1377 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1379 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1380 because it turns out to be too big or there is a timeout) there is no
1381 "Completed" line in the log. When some messages of this type were
1382 selected by exigrep, they were listed as "not completed". Others were
1383 picked up by some special patterns. I have improved the selection
1384 criteria to be more general.
1386 PH/38 The host_find_failed option in the manualroute router can now be set
1387 to "ignore", to completely ignore a host whose IP address cannot be
1388 found. If all hosts are ignored, the behaviour is controlled by the new
1389 host_all_ignored option.
1391 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1392 homing or because of multiple MX records with /mx) generated more than
1393 one IP address, and the following item turned out to be the local host,
1394 all the secondary addresses of the first item were incorrectly removed
1395 from the list, along with the local host and any following hosts (which
1396 is what is supposed to happen).
1398 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1399 whoever called Exim into the -H file. In the case of the daemon it was
1400 behaving confusingly. When first started, it used values for whoever
1401 started the daemon, but after a SIGHUP it used the Exim user (because it
1402 calls itself on a restart). I have changed the code so that it now always
1405 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1406 message are rejected with the same error (e.g. no authentication or bad
1407 sender address), and a DATA command is nevertheless sent (as can happen
1408 with PIPELINING or a stupid MUA), the error message that was given to the
1409 RCPT commands is included in the rejection of the DATA command. This is
1410 intended to be helpful for MUAs that show only the final error to their
1413 PH/42 Another patch from the Sieve maintainer.
1415 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1416 Eximstats - Fixed some broken HTML links and added missing column headers
1418 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1419 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1421 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1423 PH/43 Yet another patch from the Sieve maintainer.
1425 PH/44 I found a way to check for a TCP/IP connection going away before sending
1426 the response to the final '.' that terminates a message, but only in the
1427 case where the client has not sent further data following the '.'
1428 (unfortunately, this is allowed). However, in many cases there won't be
1429 any further data because there won't be any more messages to send. A call
1430 to select() can be used: if it shows that the input is "ready", there is
1431 either input waiting, or the socket has been closed. An attempt to read
1432 the next input character can distinguish the two cases. Previously, Exim
1433 would have sent an OK response which the client would never have see.
1434 This could lead to message repetition. This fix should cure that, at
1435 least in a lot of common cases.
1437 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1438 advertised in response to EHLO.
1444 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1445 fixed by 4.65/MH/01 (is this a record?) are fixed:
1447 (i) An empty string was always treated as zero by the numeric comparison
1448 operators. This behaviour has been restored.
1450 (ii) It is documented that the numeric comparison operators always treat
1451 their arguments as decimal numbers. This was broken in that numbers
1452 starting with 0 were being interpreted as octal.
1454 While fixing these problems I realized that there was another issue that
1455 hadn't been noticed. Values of message_size_limit (both the global option
1456 and the transport option) were treated as octal if they started with 0.
1457 The documentation was vague. These values are now always treated as
1458 decimal, and I will make that clear in the documentation.
1464 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1465 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1468 MH/01 Don't check that the operands of numeric comparison operators are
1469 integers when their expansion is in "skipping" mode (fixes bug
1470 introduced by 4.64-PH/07).
1472 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1473 child addresses, Exim now panics and dies. Previously, because the count
1474 is held in a short int, deliveries were likely to be lost. As such a
1475 large number of recipients for a single message is ridiculous
1476 (performance will be very, very poor), I have chosen to impose a limit
1477 rather than extend the field.
1483 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1484 leftover -K file (the existence of which was triggered by #402).
1485 While we were at it, introduced process PID as part of the -K
1486 filename. This should rule out race conditions when creating
1489 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1490 processing considerably. Previous code took too long for large mails,
1491 triggering a timeout which in turn triggers #401.
1493 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1494 in the DK code in transports.c. sendfile() is not really portable,
1495 hence the _LINUX specificness.
1497 TF/01 In the add_headers option to the mail command in an Exim filter,
1498 there was a bug that Exim would claim a syntax error in any
1499 header after the first one which had an odd number of characters
1502 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1503 callout verification, Exim cached a "reject" for the entire domain. This
1504 is correct for most verifications, but it is not correct for a recipient
1505 verification with use_sender or use_postmaster set, because in that case
1506 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1507 case of MAIL FROM:<> rejection from other early rejections (e.g.
1508 rejection of HELO). When verifying a recipient using a non-null MAIL
1509 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1510 Whatever the result of the callout, the value of the domain cache is
1511 left unchanged (for any other kind of callout, getting as far as trying
1512 RCPT means that the domain itself is ok).
1514 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1517 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1518 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1519 session, the daemon ignores ECONNECT errors and logs others; it now
1520 ignores EPIPE as well.
1522 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1523 (quoted-printable decoding).
1525 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1526 later the small subsequent patch to fix an introduced bug.
1528 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1530 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1532 PH/08 An error is now given if message_size_limit is specified negative.
1534 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1535 to be given (somewhat) arbitrary names.
1537 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1540 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1541 miscellaneous code fixes
1543 PH/10 Added the log_reject_target ACL modifier to specify where to log
1546 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1547 hostname. This is wrong, because it relates to the incoming message (and
1548 probably the interface on which it is arriving) and not to the outgoing
1549 callout (which could be using a different interface). This has been
1550 changed to use the value of the helo_data option from the smtp transport
1551 instead - this is what is used when a message is actually being sent. If
1552 there is no remote transport (possible with a router that sets up host
1553 addresses), $smtp_active_hostname is used.
1555 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1556 tweaks were necessary in order to get it to work (see also 21 below):
1557 (a) The code assumed that strncpy() returns a negative number on buffer
1558 overflow, which isn't the case. Replaced with Exim's string_format()
1560 (b) There were several signed/unsigned issues. I just did the minimum
1561 hacking in of casts. There is scope for a larger refactoring.
1562 (c) The code used strcasecmp() which is not a standard C function.
1563 Replaced with Exim's strcmpic() function.
1564 (d) The code set only $1; it now sets $auth1 as well.
1565 (e) A simple test gave the error "authentication client didn't specify
1566 service in request". It would seem that Dovecot has changed its
1567 interface. Fortunately there's a specification; I followed it and
1568 changed what the client sends and it appears to be working now.
1570 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1573 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1574 address A is aliased to B and C, where B exists and C does not. Without
1575 -v the output is "A verified" because verification stops after a
1576 successful redirection if more than one address is generated. However,
1577 with -v the child addresses are also verified. Exim was outputting "A
1578 failed to verify" and then showing the successful verification for C,
1579 with its parentage. It now outputs "B failed to verify", showing B's
1580 parentage before showing the successful verification of C.
1582 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1583 look up a TXT record in a specific list after matching in a combined
1586 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1587 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1588 they consult the DNS. I had assumed they would set it the way they
1589 wanted; and indeed my experiments on Linux seem to show that in some
1590 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1591 To be on the safe side, however, I have now made the interface to
1592 host_find_byname() similar to host_find_bydns(), with an argument
1593 containing the DNS resolver options. The host_find_byname() function now
1594 sets these options at its start, just as host_find_bydns() does. The smtp
1595 transport options dns_qualify_single and dns_search_parents are passed to
1596 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1597 of host_find_byname() use the default settings of RES_DEFNAMES
1598 (qualify_single) but not RES_DNSRCH (search_parents).
1600 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1601 spool_read_header() do less string testing, by means of a preliminary
1602 switch on the second character of optional "-foo" lines. (This is
1603 overdue, caused by the large number of possibilities that now exist.
1604 Originally there were few.) While I was there, I also converted the
1605 str(n)cmp tests so they don't re-test the leading "-" and the first
1606 character, in the hope this might squeeze out yet more improvement.
1608 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1609 flag allowing group syntax was set by the header_syntax check but not
1610 turned off, possible causing trouble later; (2) The flag was not being
1611 set at all for the header_verify test, causing "group"-style headers to
1612 be rejected. I have now set it in this case, and also caused header_
1613 verify to ignore an empty address taken from a group. While doing this, I
1614 came across some other cases where the code for allowing group syntax
1615 while scanning a header line wasn't quite right (mostly, not resetting
1616 the flag correctly in the right place). These bugs could have caused
1617 trouble for malformed header lines. I hope it is now all correct.
1619 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1620 with the "reply" argument non-NULL. The code, however (which originally
1621 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1622 but it didn't always do it. This confused somebody who was copying the
1623 code for some other use. I have removed all the tests.
1625 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1626 feature that was used to support insecure browsers during the U.S. crypto
1627 embargo. It requires special client support, and Exim is probably the
1628 only MTA that supported it -- and would never use it because real RSA is
1629 always available. This code has been removed, because it had the bad
1630 effect of slowing Exim down by computing (never used) parameters for the
1631 RSA_EXPORT functionality.
1633 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1634 authenticator to fail if there's a tab character in the incoming data
1635 (there should never be unless someone is messing about, as it's supposed
1636 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1637 if the connection is using TLS or if the remote IP is the same as the
1638 local IP, and the "valid-client-cert option" if a client certificate has
1641 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1642 authenticators. This can be used for authorization after authentication
1643 succeeds. (In the case of plaintext, it servers for both authentication
1646 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1647 if any retry times were supplied.
1649 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1650 connection, where there is no client IP address to check. In this
1651 situation, the verify now always succeeds.
1653 PH/25 Applied John Jetmore's -Mset patch.
1655 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1657 PH/27 In a string expansion for a processed (not raw) header when multiple
1658 headers of the same name were present, leading whitespace was being
1659 removed from all of them, but trailing whitespace was being removed only
1660 from the last one. Now trailing whitespace is removed from each header
1661 before concatenation. Completely empty headers in a concatenation (as
1662 before) are ignored.
1664 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1665 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1667 PH/29 [Removed. This was a change that I later backed out, and forgot to
1668 correct the ChangeLog entry (that I had efficiently created) before
1669 committing the later change.]
1671 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1672 address errors (4xx response to RCPT) over the same connection as other
1673 messages routed to the same hosts. Such deliveries are always "forced",
1674 so retry times are not inspected. This resulted in far too many retries
1675 for the affected addresses. The effect occurred only when there were more
1676 hosts than the hosts_max_try setting in the smtp transport when it had
1677 the 4xx errors. Those hosts that it had tried were not added to the list
1678 of hosts for which the message was waiting, so if all were tried, there
1679 was no problem. Two fixes have been applied:
1681 (i) If there are any address or message errors in an SMTP delivery, none
1682 of the hosts (tried or untried) are now added to the list of hosts
1683 for which the message is waiting, so the message should not be a
1684 candidate for sending over the same connection that was used for a
1685 successful delivery of some other message. This seems entirely
1686 reasonable: after all the message is NOT "waiting for some host".
1687 This is so "obvious" that I'm not sure why it wasn't done
1688 previously. Hope I haven't missed anything, but it can't do any
1689 harm, as the worst effect is to miss an optimization.
1691 (ii) If, despite (i), such a delivery is accidentally attempted, the
1692 routing retry time is respected, so at least it doesn't keep
1693 hammering the server.
1695 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1696 in ${readsocket because some servers need this prod.
1698 PH/32 Added some extra debug output when updating a wait-xxx database.
1700 PH/33 The hint "could be header name not terminated by colon", which has been
1701 given for certain expansion errors for a long time, was not being given
1702 for the ${if def:h_colon_omitted{... case.
1704 PH/34 The spec says: "With one important exception, whenever a domain list is
1705 being scanned, $domain contains the subject domain." There was at least
1706 one case where this was not true.
1708 PH/35 The error "getsockname() failed: connection reset by peer" was being
1709 written to the panic log as well as the main log, but it isn't really
1710 panic-worthy as it just means the connection died rather early on. I have
1711 removed the panic log writing for the ECONNRESET error when getsockname()
1714 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1715 runs only) independently of the message's sender address. This meant
1716 that, if the 4xx error was in fact related to the sender, a different
1717 message to the same recipient with a different sender could confuse
1718 things. In particualar, this can happen when sending to a greylisting
1719 server, but other circumstances could also provoke similar problems.
1720 I have changed the default so that the retry time for these errors is now
1721 based a combination of the sender and recipient addresses. This change
1722 can be overridden by setting address_retry_include_sender=false in the
1725 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1726 remote server are returned as part of bounce messages. This was not
1727 happening for LMTP over a pipe (the lmtp transport), but now it is the
1728 same for both kinds of LMTP.
1730 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1731 in header lines that were in fact CNAMEs. This is no longer the case.
1733 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1734 and queue runs started by the daemon processed all messages. This has
1735 been fixed so that -R and -S can now usefully be given with -q<time>.
1737 PH/40 Import PCRE release 6.7 (fixes some bugs).
1739 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1741 PH/42 Give an error if -q is specified more than once.
1743 PH/43 Renamed the variables $interface_address and $interface_port as
1744 $received_ip_address and $received_port, to make it clear that these
1745 values apply to message reception, and not to the outgoing interface when
1746 a message is delivered. (The old names remain recognized, of course.)
1748 PH/44 There was no timeout on the connect() call when using a Unix domain
1749 socket in the ${readsocket expansion. There now is.
1751 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1752 be meaningful with "accept".
1754 SC/01 Eximstats V1.43
1755 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1757 SC/02 Eximstats V1.44
1758 Use a glob alias rather than an array ref in the generated
1759 parser. This improves both readability and performance.
1761 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1762 Collect SpamAssassin and rejection statistics.
1763 Don't display local sender or destination tables unless
1764 there is data to show.
1765 Added average volumes into the top table text output.
1767 SC/04 Eximstats V1.46
1768 Collect data on the number of addresses (recipients)
1769 as well as the number of messages.
1771 SC/05 Eximstats V1.47
1772 Added 'Message too big' to the list of mail rejection
1773 reasons (thanks to Marco Gaiarin).
1775 SC/06 Eximstats V1.48
1776 Mainlog lines which have GMT offsets and are too short to
1777 have a flag are now skipped.
1779 SC/07 Eximstats V1.49 (Alain Williams)
1780 Added the -emptyok flag.
1782 SC/08 Eximstats V1.50
1783 Fixes for obtaining the IP address from reject messages.
1785 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1786 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1787 whitesspace changes from 4.64-PH/27
1789 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1792 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1793 are found, allow negative numbers in numeric criteria)
1795 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1797 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1798 to match changes made in 4.64-PH/43
1800 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1802 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1803 "domain" is actually an IP address, and force a failure. This locks out
1804 those revolvers/nameservers that support "A-for-A" lookups, in
1805 contravention of the specifications.
1807 PH/48 When a host name was looked up from an IP address, and the subsequent
1808 forward lookup of the name timed out, the host name was left in
1809 $sender_host_name, contrary to the specification.
1811 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1812 restricted to single-key lookups, Exim was not diagnosing an error if
1813 * or *@ was used with a query-style lookup.
1815 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1817 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1818 long ago, but noone interested enough thought of it. Let's just say that
1819 the "1.1" means that there are some new functions that weren't there at
1820 some point in the past.
1822 PH/51 Error processing for expansion failure of helo_data from an smtp
1823 transport during callout processing was broken.
1825 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1826 tested/used via the -bh/-bhc/-bs options.
1828 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1829 bug, fixed in subsequent PCRE releases).
1831 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1832 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1834 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1840 SC/01 Use a glob alias rather than an array ref in eximstats generated
1841 parser. This improves both readability and performance.
1843 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1844 Don't display local sender or destination tables in eximstats unless
1845 there is data to show.
1846 Added average volumes into the eximstats top table text output.
1848 SC/03 Collect data on the number of addresses (recipients) as well
1849 as the number of messages in eximstats.
1851 TF/01 Correct an error in the documentation for the redirect router. Exim
1852 does (usually) call initgroups() when daemonizing.
1854 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1855 with consistent privilege compared to when running as a daemon.
1857 TF/03 Note in the spec that $authenticated_id is not set for local
1858 submissions from trusted users.
1860 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1861 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1863 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1864 by adding some example configuration directives to the default
1865 configuration file. A little bit of work is required to uncomment the
1866 directives and define how usernames and passwords are checked, but
1867 there is now a framework to start from.
1869 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1870 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1871 without this. I don't know how relevant this is to other LDAP libraries.
1873 PH/02 Add the verb name to the "unknown ACL verb" error.
1875 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1877 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1879 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1880 directory not expanded when it should be if an expanded home directory
1881 was set for the address (which is overridden by the transport).
1883 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1886 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1887 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1888 because it is too late at that time, and has no effect.
1890 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1891 security issue with \' (bugzilla #107). I could not use the
1892 PQescapeStringConn() function, because it needs a PGconn value as one of
1895 PH/08 When testing addresses using -bt, indicate those final addresses that
1896 are duplicates that would not cause an additional delivery. At least one
1897 person was confused, thinking that -bt output corresponded to deliveries.
1898 (Suppressing duplicates isn't a good idea as you lose the information
1899 about possibly different redirections that led to the duplicates.)
1901 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1902 systems where poll() doesn't work, in particular OS X.
1904 PH/10 Added more information to debugging output for retry time not reached.
1906 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1907 operations in malware.c.
1909 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1912 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1913 syslog_duplication set false, log lines that would normally be written
1914 both the the main log and to the reject log were not written to syslog at
1917 PH/14 In the default configuration, change the use of "message" in ACL warn
1918 statements to "add_header".
1920 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1921 not followed by a command (e.g. "seen endif").
1923 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1924 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1927 PH/17 Added extra conditions to the default value of delay_warning_condition
1931 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1932 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1933 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1936 The Auto-Submitted: and various List- headers are standardised, whereas I
1937 don't think Precedence: ever was.
1939 PH/18 Refactored debugging code in route_finduser() to show more information,
1940 in particular, the error code if getpwnam() issues one.
1942 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1943 This is apparently needed in addition to the PH/07 change above to avoid
1944 any possible encoding problems.
1946 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1947 but not after initializing Perl.
1949 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1950 output them only if debugging. By default they are written stderr,
1951 apparently, which is not desirable.
1953 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1956 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1959 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1961 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1962 authenticated or an ident call has been made. Suppress the default
1963 values for $authenticated_id and $authenticated_sender (but permit -oMai
1964 and -oMas) when testing with -bh.
1966 PH/24 Re-jigged the order of the tests in the default configuration so that the
1967 tests for valid domains and recipients precede the DNS black list and CSA
1968 tests, on the grounds that those ones are more expensive.
1970 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1971 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1972 This bug exists in every version of Exim that I still have, right back to
1975 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1976 However, an attempt to turn on case-sensitivity in a regex key by
1977 including (?-i) didn't work because the subject string was already
1978 lowercased, and the effects were non-intuitive. It turns out that a
1979 one-line patch can be used to allow (?-i) to work as expected.
1985 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1986 other effects) broke the use of negated acl sub-conditions.
1988 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1991 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1992 "Deny" causes Exim to reject the incoming connection with a 554 error.
1993 Unfortunately, if there is a major crisis, such as a disk failure,
1994 tcp-wrappers gives "deny", whereas what one would like would be some
1995 kind of temporary error. A kludge has been added to help with this.
1996 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1997 554 error is used if errno is still zero or contains ENOENT (which occurs
1998 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2001 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2003 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2004 errors. Otherwise a message that provokes a temporary error (when other
2005 messages do not) can cause a whole host to time out.
2007 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2008 addresses were routed directly to files or pipes from a redirect router.
2009 File deliveries just didn't batch; pipe deliveries might have suffered
2012 PH/06 A failure to get a lock for a hints database would erroneously always say
2013 "Failed to get write lock", even when it was really a read lock.
2015 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2016 of 0600. This has been changed to use the value of the lockfile_mode
2017 option (which defaults to 0600).
2019 PH/08 Applied small patch from the Sieve maintainer.
2021 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2022 folder from quota calculations, a direct delivery into this folder messed
2023 up the contents of the maildirsize file. This was because the regex was
2024 used only to exclude .Trash (or whatever) when the size of the mailbox
2025 was calculated. There was no check that a delivery was happening into an
2026 excluded directory. This bug has been fixed by ignoring all quota
2027 processing for deliveries into excluded directories.
2029 PH/10 Added the maildirfolder_create_regex option to appendfile.
2035 PH/01 The code for finding all the local interface addresses on a FreeBSD
2036 system running IPv6 was broken. This may well have applied to all BSD
2037 systems, as well as to others that have similar system calls. The broken
2038 code found IPv4 interfaces correctly, but gave incorrect values for the
2039 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2040 that it would not match correctly against @[] and not recognize the IPv6
2043 PH/02 The ipliteral router was not recognizing addresses of the form user@
2044 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2046 PH/03 Added disable_ipv6.
2048 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2049 decoded form, because it is most often used to construct To: headers
2050 lines in autoreplies, and the decoded form may well be syntactically
2051 invalid. However, $reply_address has leading white space removed, and all
2052 newlines turned into spaces so that the autoreply transport does not
2055 PH/05 If group was specified without a user on a router, and no group or user
2056 was specified on a transport, the group from the router was ignored.
2058 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2059 for visible compile-time settings that can be used to change these
2060 numbers, for those that want even more. Backwards compatibility with old
2061 spool files has been maintained. However, going back to a previous Exim
2062 release will lost any variables that are in spool files.
2064 PH/07 Two small changes when running in the test harness: increase delay when
2065 passing a TCP/IP connection to a new process, in case the original
2066 process has to generate a bounce, and remove special handling of
2067 127.0.0.2 (sic), which is no longer necessary.
2069 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2070 be the same on different OS.
2072 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2075 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2076 whether --show-vars was specified or not
2078 JJ/02 exipick: Added support for new ACL variable spool format introduced
2081 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2082 syntactically invalid From: or Reply-to: line, and a filter used this to
2083 generate an autoreply, and therefore failed to obtain an address for the
2084 autoreply, Exim could try to deliver to a non-existent relative file
2085 name, causing unrelated and misleading errors. What now happens is that
2086 it logs this as a hard delivery error, but does not attempt to create a
2089 PH/11 The exinext utility has a -C option for testing purposes, but although
2090 the given file was scanned by exinext itself; it wasn't being passed on
2091 when Exim was called.
2093 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2094 an end-of-file indication when reading a command response.
2096 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2097 compiled. In many other places in Exim, IPv6 addresses are always
2098 recognized, so I have changed this. It also means that IPv4 domain
2099 literals of the form [IPV4:n.n.n.n] are now always recognized.
2101 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2102 used if the router is not running as root, for example, when verifying at
2103 ACL time, or when using -bh. The debugging output from this situation was
2104 non-existent - all you got was a failure to exec. I have made two
2107 (a) Failures to set uid/gid, the current directory, or a process leader
2108 in a subprocess such as that created by queryprogram now generate
2109 suitable debugging ouput when -d is set.
2111 (b) The queryprogram router detects when it is not running as root,
2112 outputs suitable debugging information if -d is set, and then runs
2113 the subprocess without attempting to change uid/gid.
2115 PH/15 Minor change to Makefile for building test_host (undocumented testing
2118 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2119 additional section of a DNS packet that returns MX or SRV records.
2120 Instead, it always explicitly searches for A/AAAA records. This avoids
2121 major problems that occur when a DNS server includes only records of one
2122 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2123 fixed another bug: if SRV records were looked up and the corresponding
2124 address records were *not* found in the additional section, the port
2125 values from the SRV records were lost.
2127 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2128 using the correct key (the original address) when searching the retry
2129 rules in order to find which one to use for generating the retry hint.
2131 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2132 adding the default one. Similarly, if it contains a Reply-To: header, the
2133 errors_reply_to option, if set, is not used.
2135 PH/19 When calculating a retry time, Exim used to measure the "time since
2136 failure" by looking at the "first failed" field in the retry record. Now
2137 it does not use this if it is later than than the arrival time of the
2138 message. Instead it uses the arrival time. This makes for better
2139 behaviour in cases where some deliveries succeed, thus re-setting the
2140 "first failed" field. An example is a quota failure for a huge message
2141 when small messages continue to be delivered. Without this change, the
2142 "time since failure" will always be short, possible causing more frequent
2143 delivery attempts for the huge message than are intended.
2144 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2146 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2147 $1, $2, $3) because the numerical variables can be reset during some
2148 expansion items (e.g. "match"), thereby losing the authentication data.
2150 PH/21 Make -bV show the size of off_t variables so that the test suite can
2151 decide whether to run tests for quotas > 2G.
2153 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2154 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2155 filecount value is greater than 2G or if a quota value is greater than 2G
2156 on a system where the size of off_t is not greater than 4, a panic error
2159 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2160 never match. The debug and -bh output now contains an explicit error
2161 message indicating a malformed IPv4 address or mask.
2163 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2164 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2165 PH/23 above applies.
2167 PH/25 Do not write to syslog when running in the test harness. The only
2168 occasion when this arises is a failure to open the main or panic logs
2169 (for which there is an explicit test).
2171 PH/26 Added the /no_tell option to "control=freeze".
2173 PH/27 If a host name lookup failed very early in a connection, for example, if
2174 the IP address matched host_lookup and the reverse lookup yielded a name
2175 that did not have a forward lookup, an error message of the form "no IP
2176 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2177 could be logged. Now it outputs the IP address instead of "NULL".
2179 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2180 allows the sender and the authenticated sender to be set when
2181 submitting a message from within Exim. Since child_open_exim() is
2182 documented for local_scan(), the new function should be too.
2184 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2185 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2186 results in an empty string is now treated as unset.
2188 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2190 PH/31 Added sender_verify_fail logging option.
2192 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2193 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2194 tidied the source and removed it altogether.
2196 PH/33 When a queue run was abandoned because the load average was too high, a
2197 log line was always written; now it is written only if the queue_run log
2198 selector is set. In addition, the log line for abandonment now contains
2199 information about the queue run such as the pid. This is always present
2200 in "start" and "stop" lines but was omitted from the "abandon" line.
2202 PH/34 Omit spaces between a header name and the colon in the error message that
2203 is given when verify = headers_syntax fails (if there are lots of them,
2204 the message gets confusing).
2206 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2207 names, as there are now some PTR records that contain slashes. This check
2208 is only to protect against broken name servers that fall over on strange
2209 characters, so the fact that it applies to all lookups doesn't matter.
2211 PH/36 Now that the new test suite is complete, we can remove some of the
2212 special code in Exim that was needed for the old test suite. For example,
2213 sorting DNS records because real resolvers return them in an arbitrary
2214 order. The new test suite's fake resolver always returns records in the
2217 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2218 bounces) except when queue_only is set, to avoid logging races between
2219 the different processes.
2221 PH/38 Panic-die if .include specifies a non-absolute path.
2223 PH/39 A tweak to the "H" retry rule from its user.
2225 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2226 a label. They prevented compilation on older perls.
2228 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2229 a warning to be raised on newish perls.
2231 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2232 on queue. Changes to match documented behaviour of showing count of
2233 messages matching specified criteria.
2235 PH/40 Changed the default ident timeout from 30s to 5s.
2237 PH/41 Added support for the use of login_cap features, on those BSD systems
2238 that have them, for controlling the resources used by pipe deliveries.
2240 PH/42 The content-scanning code uses fopen() to create files in which to put
2241 message data. Previously it was not paying any attention to the mode of
2242 the files. Exim runs with umask(0) because the rest of the code creates
2243 files with open(), and sets the required mode explicitly. Thus, these
2244 files were ending up world-writeable. This was not a big issue, because,
2245 being within the spool directory, they were not world-accessible. I have
2246 created a function called modefopen, which takes an additional mode
2247 argument. It sets umask(777), creates the file, chmods it to the required
2248 mode, then resets the umask. All the relevant calls to fopen() in the
2249 content scanning code have been changed to use this function.
2251 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2252 to 24 hours. This avoids potential overflow problems when processing G
2253 and H retry rules. I suspect nobody ever tinkers with this value.
2255 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2257 PH/45 When the plaintext authenticator is running as a client, the server's
2258 challenges are checked to ensure they are valid base64 strings. By
2259 default, the authentication attempt is cancelled if an invalid string is
2260 received. Setting client_ignore_invalid_base64 true ignores these errors.
2261 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2262 they are received. Thus, the responses can be made to depend on the
2263 challenges. If an invalid string is ignored, an empty string is placed in
2266 PH/46 Messages that are created by the autoreply transport now contains a
2267 References: header, in accordance with RFCs 2822 and 3834.
2269 PH/47 Added authenticated_sender_force to the smtp transport.
2271 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2273 PH/49 Installed latest patch from the Sieve maintainer.
2275 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2276 was also set, the appendfile transport was unnecessarily scanning a
2277 directory of message files (e.g. for maildir delivery) to find the count
2278 of files (along with the size), even though it did not need this
2279 information. It now does the scan only if it needs to find either the
2280 size of the count of files.
2282 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2284 PH/52 Two bugs concerned with error handling when the smtp transport is
2287 (i) Exim was not creating retry information for temporary errors given
2288 for individual recipients after the DATA command when the smtp transport
2289 was used in LMTP mode. This meant that they could be retried too
2290 frequently, and not timed out correctly.
2292 (ii) Exim was setting the flag that allows error details to be returned
2293 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2294 recipients that were returned after the DATA command.
2296 PH/53 This is related to PH/52, but is more general: for any failing address,
2297 when detailed error information was permitted to be returned to the
2298 sender, but the error was temporary, then after the final timeout, only
2299 "retry timeout exceeded" was returned. Now it returns the full error as
2300 well as "retry timeout exceeded".
2302 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2303 do this, and (what is worse) MTAs that accept it.
2305 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2306 will now be deprecated.
2308 PH/56 New os.c-cygwin from the Cygwin maintainer.
2310 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2311 formats (previously only available in exim formats via -bpr, -bpru,
2312 and -bpra. Now also available in native and exiqgrep formats)
2314 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2315 with very large, slow to parse queues
2317 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2319 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2321 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2322 responses to authentication challenges, though it was showing the
2323 challenges; (ii) I've removed the CR characters from the debug output for
2326 PH/58 Allow for the insertion of a newline as well as a space when a string
2327 is turned into more than one encoded-word during RFC 2047 encoding. The
2328 Sieve code now uses this.
2330 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2331 data_4xx, lost_connection, tls_required.
2333 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2334 message was being sent as an SMTP response.
2336 PH/61 Add -l and -k options to exicyclog.
2338 PH/62 When verifying, if an address was redirected to one new address, so that
2339 verification continued, and the new address failed or deferred after
2340 having set something in $address_data, the value of $address_data was not
2341 passed back to the ACL. This was different to the case when no
2342 redirection occurred. The value is now passed back in both cases.
2344 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2345 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2346 use login_cap.h, so on its own it isn't the distinguishing feature. The
2347 new name refers directly to the setclassresources() function.
2349 PH/65 Added configuration files for NetBSD3.
2351 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2353 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2354 is preferred over IPv4.
2356 PH/68 The bounce_return_message and bounce_return_body options were not being
2357 honoured for bounces generated during the reception of non-SMTP messages.
2358 In particular, this applied to messages rejected by the ACL. This bug has
2359 been fixed. However, if bounce_return_message is true and bounce_return_
2360 body is false, the headers that are returned for a non-SMTP message
2361 include only those that have been read before the error was detected.
2362 (In the case of an ACL rejection, they have all been read.)
2364 PH/69 The HTML version of the specification is now built in a directory called
2365 spec_html instead of spec.html, because the latter looks like a path with
2366 a MIME-type, and this confuses some software.
2368 PH/70 Catch two compiler warnings in sieve.c.
2370 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2371 function verify_get_ident() calls ip_connect() to connect a socket, but
2372 if the "connect()" function timed out, ip_connect() used to close the
2373 socket. However, verify_get_ident() also closes the socket later, and in
2374 between Exim writes to the log, which may get opened at this point. When
2375 the socket was closed in ip_connect(), the log could get the same file
2376 descriptor number as the socket. This naturally causes chaos. The fix is
2377 not to close the socket in ip_connect(); the socket should be closed by
2378 the function that creates it. There was only one place in the code where
2379 this was missing, in the iplookup router, which I don't think anybody now
2380 uses, but I've fixed it anyway.
2382 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2383 well as to direct DNS lookups. Otherwise the handling of names in host
2384 lists is inconsistent and therefore confusing.
2390 PH/01 Two changes to the default runtime configuration:
2392 (1) Move the checks for relay_from_hosts and authenticated clients from
2393 after to before the (commented out) DNS black list checks.
2395 (2) Add control=submission to the relay_from_hosts and authenticated
2396 clients checks, on the grounds that messages accepted by these
2397 statements are most likely to be submissions.
2399 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2401 (1) Generate an error if the third argument for the ${prvs expansion is
2404 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2407 (3) Reset the variables that are obtained from the first argument of
2408 ${prvscheck and used in the second argument before leaving the code,
2409 because their memory is reclaimed, so using them afterwards may do
2412 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2413 one (it's much easier than Tom thought :-).
2415 (5) Because of (4), we can now allow for the use of $prvscheck_result
2416 inside the third argument.
2418 PH/03 For some reason, the default setting of PATH when running a command from
2419 a pipe transport was just "/usr/bin". I have changed it to
2422 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2423 anything to be listed in the output from -bV.
2425 PH/05 When a filter generated an autoreply, the entire To: header line was
2426 quoted in the delivery log line, like this:
2428 => >A.N.Other <ano@some.domain> <original@ddress> ...
2430 This has been changed so that it extracts the operative address. There
2431 may be more than one such address. If so, they are comma-separated, like
2434 => >ano@some.domain,ona@other.domain <original@ddress> ...
2436 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2437 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2438 not being looked up in the rDNS to get a host name, Exim was showing the
2439 IP address twice in Received: lines, even though the IP addresses were
2440 identical. For example:
2442 Received: from [1.2.3.4] (helo=[1.2.3.4])
2444 However, if the real host name was known, it was omitting the HELO data
2445 if it matched the actual IP address. This has been tidied up so that it
2446 doesn't show the same IP address twice.
2448 PH/07 When both +timestamp and +memory debugging was on, the value given by
2449 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2450 called by the debug printing, thereby overwriting the timestamp buffer.
2451 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2453 PH/08 When the original message was included in an autoreply transport, it
2454 always said "this is a copy of the message, including all the headers",
2455 even if body_only or headers_only was set. It now gives an appropriate
2458 PH/09 Applied a patch from the Sieve maintainer which:
2460 o fixes some comments
2461 o adds the (disabled) notify extension core
2462 o adds some debug output for the result of if/elsif tests
2463 o points to the current vacation draft in the documentation
2464 and documents the missing references header update
2468 o fixes a bug in processing the envelope test (when testing
2469 multiple envelope elements, the last element determinted the
2472 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2473 Electronic Mail") by including:
2475 Auto-submitted: auto-generated
2477 in the messages that it generates (bounce messages and others, such as
2478 warnings). In the case of bounce messages for non-SMTP mesages, there was
2479 also a typo: it was using "Auto_submitted" (underscore instead of
2480 hyphen). Since every message generated by Exim is necessarily in response
2481 to another message, thes have all been changed to:
2483 Auto-Submitted: auto-replied
2485 in accordance with these statements in the RFC:
2487 The auto-replied keyword:
2489 - SHOULD be used on messages sent in direct response to another
2490 message by an automatic process,
2492 - MUST NOT be used on manually-generated messages,
2494 - MAY be used on Delivery Status Notifications (DSNs) and Message
2495 Disposition Notifications (MDNs),
2497 - MUST NOT be used on messages generated by automatic or periodic
2498 processes, except for messages which are automatic responses to
2501 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2502 to the default Received: header definition.
2504 PH/12 Added log selector acl_warn_skipped (default on).
2506 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2507 variables because (a) they are in the wrong storage pool and (b) even if
2508 they were copied, it wouldn't work properly because of the caching.
2510 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2511 checking when decoding. Apparently there are clients that generate
2512 overlong encoded strings. Why am I not surprised?
2514 PH/15 If the first argument of "${if match_address" was not empty, but did not
2515 contain an "@" character, Exim crashed. Now it writes a panic log message
2516 and treats the condition as false.
2518 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2520 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2521 extension "envelope-auth". The code is finished and in agreement with
2522 other implementations, but there is no documentation so far and in fact,
2523 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2524 not changing the active code.
2526 Print executed "if" and "elsif" statements when debugging is used. This
2527 helps a great deal to understand what a filter does.
2529 Document more things not specified clearly in RFC3028. I had all this
2530 sorted out, when out of a sudden new issues came to my mind. Oops."
2532 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2535 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2536 canonical form (as documented). However, after a host name lookup from
2537 the IP address, check_host() was doing a simple string comparison with
2538 addresses acquired from the DNS when checking that the found name did
2539 have the original IP as one of its addresses. Since any found IPv6
2540 addresses are likely to be in abbreviated form, the comparison could
2541 fail. Luckily, there already exists a function for doing the comparison
2542 by converting both addresses to binary, so now that is used instead of
2543 the text comparison.
2545 PH/20 There was another similar case to PH/19, when a complete host name was
2546 given in a host list; looking up its IP address could give an abbreviated
2547 form, whereas the current host's name might or might not be abbreviated.
2548 The same fix has been applied.
2554 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2555 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2558 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2559 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2561 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2563 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2564 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2565 Also, for "Auto-Submitted", treat anything other than "no" as
2566 non-personal, in accordance with RFC 3834. (Previously it treated
2567 anything starting "auto-" as non-personal.)
2569 TF/01 The control=submission/name=... option had a problem with syntax
2570 errors if the name included a slash character. The /name= option
2571 now slurps the rest of the string, so it can include any characters
2572 but it must come last in the list of options (after /sender_retain
2575 PH/05 Some modifications to the interface to the fake nameserver for the new
2583 TK/01 Added the "success_on_redirect" address verification option. See
2584 NewStuff for rationale and an example.
2586 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2588 PH/02 Patch to exigrep to allow it to work on syslog lines.
2590 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2591 fread() to skip over the body file's header line, because in Cygwin the
2592 header line is locked and is inaccessible.
2594 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2595 co-exist for some time) to make it clear that it is the Exim ID that is
2596 referenced, not the Message-ID: header line.
2598 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2599 string_format() function, because snprintf() does not exist on all
2602 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2603 ${stat: expansion item.
2605 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2606 protocol synchronization error", to keep the pedants happy.
2608 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2609 well as for IRIX systems, when gcc is being used. See the host.c source
2612 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2614 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2617 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2618 transport and to the smtp transport in LMTP mode.
2620 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2622 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2623 run a filter in a subprocess. This could lead to confusion in subsequent
2624 lookups in the parent process. There should also be a search_tidyup() at
2625 the end of the subprocess.
2627 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2628 only if the host matched helo_try_verify_hosts, which caused the
2629 verification to occur when the EHLO/HELO command was issued. The ACL just
2630 tested the remembered result. Now, if a previous verification attempt has
2631 not happened, "verify = helo" does it there and then.
2633 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2635 TK/03 Fix log output including CR from clamd.
2637 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2638 exist provoked a memory error which could cause a segfault.
2640 PH/15 Installed PCRE 6.2
2642 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2644 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2645 of the problem. Specifically, suggested +O2 rather than +O1 for the
2648 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2650 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2651 record, the port was not correctly logged when the outgoing_port log
2652 selector was set (it logged the transort's default port).
2654 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2655 fallback_hosts, and "hosts" in the smtp transport.
2657 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2658 host errors such as "Connection refused".
2660 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2661 authentication with radiusclient 0.4.9:
2663 - Error returned from rc_read_config was caught wrongly
2664 - Username/password not passed on to radius server due to wrong length.
2666 The presumption is that some radiusclient API changes for 4.51/PH/17
2667 were not taken care of correctly. The code is still untested by me (my
2668 Linux distribution still has 0.3.2 of radiusclient), but it was
2669 contributed by a Radius user.
2671 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2672 expanding the "port" option of the smtp transport.
2674 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2675 while reading a MIME header. Thanks to Tom Hughes for a patch.
2677 PH/24 Include config.h inside local_scan.h so that configuration settings are
2680 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2681 that in an ACL for RCPT (for example), you can examine exactly what was
2684 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2685 commands, but it was not correctly comparing the address with the actual
2686 client host address. Thus, it would show the EHLO address in Received:
2687 header lines when this was not necessary.
2689 PH/27 Added the % operator to ${eval:}.
2691 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2692 it should be ignoring failures (because with -C, for example, it has lost
2693 privilege). It wasn't ignoring creation failures other than "already
2696 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2699 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2700 because an input line was too long, either on its own, or by virtue of
2701 too many continuations, the temporary file was not being removed, and the
2702 return code was incorrect.
2704 PH/31 Missing "BOOL" in function definition in filtertest.c.
2706 PH/32 Applied Sieve patches from the maintainer.
2708 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2710 PH/33 Added "verify = not_blind".
2712 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2713 Local/Makefile (with some defaults set). These are used in built scripts
2714 such as exicyclog, but they have never been used in the exim_install
2715 script (though there are many overriding facilities there). I have
2716 arranged that the exim_install script now takes note of these two
2719 PH/35 Installed configuration files for Dragonfly.
2721 PH/36 When a locally submitted message by a trusted user did not contain a
2722 From: header, and the sender address was obtained from -f or from an SMTP
2723 MAIL command, and the trusted user did not use -F to supply a sender
2724 name, $originator_name was incorrectly used when constructing a From:
2725 header. Furthermore, $originator_name was used for submission mode
2726 messages from external hosts without From: headers in a similar way,
2727 which is clearly wrong.
2729 PH/37 Added control=suppress_local_fixups.
2731 PH/38 When log_selector = +received_sender was set, and the addition of the
2732 sender made the log line's construction buffer exactly full, or one byte
2733 less than full, an overflow happened when the terminating "\n" was
2736 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2737 when the result of a list match is failure because a DNS lookup failed.
2739 PH/40 RM_COMMAND is now used in the building process.
2741 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2742 the "build-* directories that it finds.
2744 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2745 address in a domain literal was a prefix of an interface address.
2747 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2748 when verifying a sender address, unless rewrite_headers is false.
2750 PH/44 Wrote a long comment about why errors_to addresses are verified as
2751 recipients, not senders.
2753 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2754 the ratelimit ACL was added.
2756 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2758 PH/46 Added extra information about PostgreSQL errors to the error string.
2760 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2761 avoiding the need to install special zones in a real server. This is
2762 backwards compatible; if it can't find the fake resolver, it drops back.
2763 Thus, both old and new test suites can be run.
2765 TF/02 Added util/ratelimit.pl
2767 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2768 clock is set back in time.
2770 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2771 Candler <B.Candler@pobox.com>.
2773 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2774 OK for addresses that are the result of redirections.
2776 PH/48 A number of further additions for the benefit of the new test suite,
2777 including a fake gethostbyname() that interfaces to the fake DNS resolver
2780 TF/06 The fix for widen_domains has also been applied to qualify_single and
2781 search_parents which are the other dnslookup options that can cause
2784 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2787 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2789 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2790 service" if it encountered an MX record with an empty target hostname.
2791 The message is now "an MX or SRV record indicated no SMTP service".
2793 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2794 if the DNS of the sending site is misconfigured. This is quite a
2795 common situation. This change restores the behaviour of treating a
2796 helo verification defer as a failure.
2798 PH/51 If self=fail was set on a router, the bounce message did not include the
2799 actual error message.
2805 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2807 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2808 command itself ended in error, the underlying message about the transport
2809 filter timeout was being overwritten with the pipe command error. Now the
2810 underlying error message should be appended to the second error message.
2812 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2814 PH/02 Reduce the amount of output that "make" produces by default. Full output
2815 can still be requested.
2817 PH/03 The warning log line about a condition test deferring for a "warn" verb
2818 was being output only once per connection, rather than after each
2819 occurrence (because it was using the same function as for successful
2820 "warn" verbs). This seems wrong, so I have changed it.
2822 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2823 it should not have, which might have caused a crash in the right
2824 circumstances, but probably never did.
2826 PH/04 Installed a modified version of Tony Finch's patch to make submission
2827 mode fix the return path as well as the Sender: header line, and to
2828 add a /name= option so that you can make the user's friendly name appear
2831 TF/03 Added the control = fakedefer ACL modifier.
2833 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2834 Mark Lowes for thorough testing.
2836 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2838 TK/03 Merged latest SRS patch from Miles Wilton.
2840 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2841 with the definition in sysexits.h (which is #included earlier).
2842 Fortunately, Exim does not actually use EX_OK. The code used to try to
2843 preserve the sysexits.h value, by assumimg that macro definitions were
2844 scanned for macro replacements. I have been disabused of this notion,
2845 so now the code just undefines EX_OK before #including unistd.h.
2847 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2848 in the smtp transport. When a block could not be written in a single
2849 write() function, the timeout was being re-applied to each part-write.
2850 This seems wrong - if the receiver was accepting one byte at a time it
2851 would take for ever. The timeout is now adjusted when this happens. It
2852 doesn't have to be particularly precise.
2854 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2855 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2857 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2858 without a domain if the check to <postmaster@domain> fails.
2860 SC/01 Eximstats: added -xls and the ability to specify output files
2861 (patch written by Frank Heydlauf).
2863 SC/02 Eximstats: use FileHandles for outputing results.
2865 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2867 SC/04 Eximstats: fixed display of large numbers with -nvr option
2869 SC/05 Eximstats: fixed merging of reports with empty tables.
2871 SC/06 Eximstats: added the -include_original_destination flag
2873 SC/07 Eximstats: removed tabs and trailing whitespace.
2875 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2877 TK/06 MBOX spool code: Add real "From " MBOX separator line
2878 so the .eml file is really in mbox format (even though
2879 most programs do not really care). Patch from Alex Miller.
2881 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2882 The latter is generated from $received_to and is only set if the
2883 message has one envelope recipient. SA can use these headers,
2884 obviously out-of-the-box. Patch from Alex Miller.
2886 PH/08 The ${def test on a variable was returning false if the variable's
2887 value was "0", contrary to what the specification has always said!
2888 The result should be true unless the variable is empty.
2890 PH/09 The syntax error of a character other than { following "${if
2891 def:variable_name" (after optional whitespace) was not being diagnosed.
2892 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2893 accidental colon was present, for example, could give incorrect results.
2895 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2896 result is used (not including appendfile, where other changes are about
2899 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2900 This involved changing a lot of size variables from int to off_t. It
2901 should work with maildirs and everything.
2903 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2904 spamd dying while we are connected to it.
2906 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2909 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2910 function for BDB 4.3.
2912 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2914 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2915 the sources for PCRE have been re-arranged and more files are now
2918 PH/15 The code I had for printing potentially long long variables in PH/11
2919 above was not the best (it lost precision). The length of off_t variables
2920 is now inspected at build time, and an appropriate printing format (%ld
2921 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2922 to be "long long int" or "long int". This is needed for the internal
2923 formatting function string_vformat().
2925 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2926 the configuration file to be ":syslog", then the script "guesses" where
2927 the logs files are, rather than using the compiled in default. In our
2928 case the guess is not the same as the compiled default, so the script
2929 suddenly stopped working when I started to use syslog. The patch checks
2930 to see if log_file_path is "". If so, it attempts to read it from exim
2931 with no configuration file to get the compiled in version, before it
2932 falls back to the previous guessing code."
2934 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2935 implementing BATV in an Exim configuration. See NewStuff for the gory
2938 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2939 Makefile that are specific to HP-UX.
2941 PH/18 If the "use_postmaster" option was set for a recipient callout together
2942 with the "random" option, the postmaster address was used as the MAIL
2943 FROM address for the random test, but not for the subsequent recipient
2944 test. It is now used for both.
2946 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2947 patch removes a few documentation additions to RFC 3028, because the
2948 latest draft now contains them. It adds the new en;ascii-case comparator
2949 and a new error check for 8bit text in MIME parts. Comparator and
2950 require names are now matched exactly. I enabled the subaddress
2951 extension, but it is not well tested yet (read: it works for me)."
2953 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2954 rework some of the code of TK/09 above to avoid the hardwired use of
2955 "%lld" and "long long". Replaced the call to snprintf() with a call to
2958 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2959 records point to non-existent hosts", "retry timeout exceeded", and
2960 "retry time not reached for any host after a long failure period".
2962 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2963 experimental DomainKeys support:
2965 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2966 (2) On an error such as an illegally used "control", the wrong name for
2967 the control was given.
2969 These problems did NOT occur unless DomainKeys support was compiled.
2971 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2973 PH/24 Added ${if match_ip condition.
2975 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2976 databases so that it will be absolutely obvious if a crash occurs in the
2977 DB library. This is a regular occurrence (often caused by mis-matched
2980 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2981 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2982 picked up on a user's system that detects such things. There doesn't seem
2983 to be a gcc warning option for this - only an attribute that has to be
2984 put on the function's prototype. It seems that in Fedora Core 4 they have
2985 set this on a number of new functions. No doubt there will be more in due
2988 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2989 specify a transport. However, if an address that was verified by such a
2990 router was the subject of a callout, Exim crashed because it tried to
2991 read the rcpt_include_affixes from the non-existent transport. Now it
2992 just assumes that the setting of that option is false. This bug was
2993 introduced by 4.51/PH/31.
2995 PH/28 Changed -d+all to exclude +memory, because that information is very
2996 rarely of interest, but it makes the output a lot bigger. People tend to
2997 do -d+all out of habit.
2999 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3000 code in os-type was giving problems when libc.so lives in lib64, like on
3003 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3004 aren't the modern standard, and it seems that some systems' include files
3005 don't always have them. Exim was already checking for some of the newer
3006 ones like T_AAAA, and defining it itself. I've added checks for all the
3007 record types that Exim uses.
3009 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3010 not automatically generating a new one, as it is supposed to. This
3011 prevented TLS from working. If the file did exist, but contained invalid
3012 data, a new version was generated, as expected. It was only the case of a
3013 non-existent file that was broken.
3015 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3016 with a change in libdomainkeys > 0.64.
3018 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3019 from DNS. If the selector record carries the flag, it now has
3020 precedence over the domain-wide flag.
3022 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3024 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3025 the use of a port name that isn't defined in /etc/services) occurred, the
3026 message was deferred as in a normal delivery, and thus remained on the
3027 spool, instead of being failed because of the mua_wrapper setting. This
3028 is now fixed, and I tidied up some of the mua_wrapper messages at the
3031 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3032 the messages in a hash of arrays rather than using individual hashes.
3033 This is a bit cleaner and results in dramatic memory savings, albeit
3034 at a slight CPU cost.
3036 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3037 as requested by Marc Sherman.
3039 SC/10 Eximstats: added histograms for user specified patterns as requested
3042 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3044 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3045 fopen() in the content-scanning modules that did not already have it.
3051 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3052 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3054 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3056 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3058 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3059 to test Sieve filters that use "vacation".
3061 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3062 that changes the way the GnuTLS parameters are stored in the cache file.
3063 The new format can be generated externally. For backward compatibility,
3064 if the data in the cache doesn't make sense, Exim assumes it has read an
3065 old-format file, and it generates new data and writes a new file. This
3066 means that you can't go back to an older release without removing the
3069 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3070 work if there are any delivery delays because "one_time" forces the
3071 parent to be marked "delivered", so its unseen clone is never tried
3072 again. For this reason, Exim now forbids the simultaneous setting of
3075 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3076 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3077 entry if you want to know the details. The fix, however, neglected to
3078 consider the case when local delivery batching is involved. The test for
3079 "previously delivered" was not happening when checking to see if an
3080 address could be batched with a previous (undelivered) one; under
3081 certain circumstances this could lead to multiple deliveries to the same
3084 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3085 in its include files, and this causes problems building Exim.
3087 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3088 header_syntax) but Exim was just ignoring anything given after a slash.
3089 In particular, this caused confusion with an attempt to use "verify =
3090 reverse_host_lookup/defer_ok". An error is now given when options are
3091 supplied for verify items that do not have them. (Maybe reverse_host_
3092 lookup should have a defer_ok option, but that's a different point.)
3094 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3095 defined by RFC 821) to 2048, because there were problems with some AUTH
3096 commands, and RFC 1869 says the size should be increased for extended
3097 SMTP commands that take arguments.
3099 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3102 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3103 "unknown" error; now it says that the functionality isn't in the binary.
3105 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3106 an address' error message when a string expansion fails (syntax or
3107 whatever). Otherwise the password may appear in the log. Following change
3108 PH/42 below, there is no longer a chance of it appearing in a bounce
3111 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3113 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3114 of its addresses was ever tried. (Bugzilla bug #2.)
3116 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3117 the result incorrectly in the debug output. (It correctly added a newline
3118 to what was transported.)
3120 TF/01 Added $received_time.
3122 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3123 commented out examples of how to interface to a virus scanner and to
3124 SpamAssassin. Also added commented examples of av_scanner and
3125 spamd_address settings.
3127 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3128 and controls are allowed in which ACLs. There were a couple of minor
3129 errors. Some of the entries in the conditions table (which is a table of
3130 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3131 negation of where the condition IS allowed.
3133 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3135 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3136 header file does not have a version number, so I've had to invent a new
3137 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3138 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3139 radiusclient), but it was contributed by a Radius user.
3141 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3142 files or directories, for OpenSSL.
3144 PH/19 When an Exim process that is running as root has to create an Exim log
3145 file, it does so in a subprocess that runs as exim:exim so as to get the
3146 ownership right at creation (otherwise, other Exim processes might see
3147 the file with the wrong ownership). There was no test for failure of this
3148 fork() call, which would lead to the process getting stuck as it waited
3149 for a non-existent subprocess. Forks do occasionally fail when resources
3150 run out. I reviewed all the other calls to fork(); they all seem to check
3153 PH/20 When checking for unexpected SMTP input at connect time (before writing
3154 the banner), Exim was not dealing correctly with a non-positive return
3155 from the read() function. If the client had disconnected by this time,
3156 the result was a log entry for a synchronization error with an empty
3157 string after "input=" when read() returned zero. If read() returned -1
3158 (an event I could not check), uninitialized data bytes were printed.
3159 There were reports of junk text (parts of files, etc) appearing after
3162 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3164 PH/22 Added support for macro redefinition, and (re)definition in between
3165 driver and ACL definitions.
3167 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3168 forgetting to use the resulting value; it was using the unexpanded value.
3170 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3171 hadn't been configured. The fix is from Juergen Kreileder, who
3172 understands it better than I do:
3174 "Here's what I see happening with three configured cyrus_sasl
3175 authenticators configured (plain, login, cram-md5):
3177 On startup auth_cyrus_sasl_init() gets called for each of these.
3178 This means three calls to sasl_listmech() without a specified mech_list.
3179 => SASL tests which mechs of all available mechs actually work
3180 => three warnings about OTP not working
3181 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3183 With the patch, sasl_listmech() also gets called three times. But now
3184 SASL's mech_list option is set to the server_mech specified in the the
3185 authenticator. Or in other words, the answer from sasl_listmech()
3186 gets limited to just the mech you're testing for (which is different
3188 => the return list contains just 'plain' or 'login', 'cram-md5' or
3189 nothing depending on the value of ob->server_mech.
3191 I've just tested the patch: Authentication still works fine,
3192 unavailable mechs specified in the exim configuration are still
3193 caught, and the auth.log warnings about OTP are gone."
3195 PH/25 When debugging is enabled, the contents of the command line are added
3196 to the debugging output, even when log_selector=+arguments is not
3199 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3200 answer is "GNU", and only if the return is "GNU/something" is the answer
3203 PH/27 $acl_verify_message is now set immediately after the failure of a
3204 verification in an ACL, and so is available in subsequent modifiers. In
3205 particular, the message can be preserved by coding like this:
3207 warn !verify = sender
3208 set acl_m0 = $acl_verify_message
3210 Previously, $acl_verify_message was set only while expanding "message"
3211 and "log_message" when a very denied access.
3213 PH/28 Modified OS/os.c-Linux with
3215 -#ifndef OS_LOAD_AVERAGE
3216 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3218 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3219 nomenclature these days.)
3221 PH/29 Installed patch from the Sieve maintainer that adds the options
3222 sieve_useraddress and sieve_subaddress to the redirect router.
3224 PH/30 In these circumstances:
3225 . Two addresses routed to the same list of hosts;
3226 . First host does not offer TLS;
3227 . First host accepts first address;
3228 . First host gives temporary error to second address;
3229 . Second host offers TLS and a TLS session is established;
3230 . Second host accepts second address.
3231 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3232 and peerdn, if requested) that were in fact used only for the second
3235 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3236 attention to any local part prefix or suffix that was matched by the
3237 router that accepted the address. It now behaves in the same way as it
3238 does for delivery: the affixes are removed from the local part unless
3239 rcpt_include_affixes is set on the transport.
3241 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3242 timeout during the DATA phase of an incoming message.
3244 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3245 applied a patch sanctioned by the Sieve maintainer.
3247 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3248 the uid or gid is negative. A case of a negative gid caused this to be
3249 noticed. The fix allows for either to be negative.
3251 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3252 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3255 PH/36 The change PH/12 above was broken. Fixed it.
3257 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3258 the grounds that routing the same address twice would always produce the
3259 same answer. This might have been true once, but it is certainly no
3260 longer true now. Routing a child address may depend on the previous
3261 routing that produced that child. Some complicated redirection strategies
3262 went wrong when messages had multiple recipients, and made Exim's
3263 behaviour dependent on the order in which the addresses were given.
3265 I have moved the duplicate checking until after the routing is complete.
3266 Exim scans the addresses that are assigned to local and remote
3267 transports, and removes any duplicates. This means that more work will be
3268 done, as duplicates will always all be routed, but duplicates are
3269 presumably rare, so I don't expect this is of any significance.
3271 For deliveries to pipes, files, and autoreplies, the duplicate checking
3272 still happens during the routing process, since they are not going to be
3275 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3276 It corrects a timeout issue with spamd. This is Ian's comment: "The
3277 background is that sometimes spamd either never reads data from a
3278 connection it has accepted, or it never writes response data. The exiscan
3279 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3280 blindly assumes that writes won't block so it may never time out."
3282 PH/39 Allow G after quota size as well as K and M.
3284 PH/40 The value set for $authenticated_id in an authenticator may not contain
3285 binary zeroes or newlines because the value is written to log lines and
3286 to spool files. There was no check on this. Now the value is run through
3287 the string_printing() function so that such characters are converted to
3288 printable escape sequences.
3290 PH/41 $message_linecount is a new variable that contains the total number of
3291 lines in the message. Compare $body_linecount, which is the count for the
3294 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3295 bounce and delay warning messages, except in certain special cases, which
3298 (a) An SMTP error message from a remote host;
3299 (b) A message specified in a :fail: redirection;
3300 (c) A message specified in a "fail" command in a system filter;
3301 (d) A message specified in a FAIL return from the queryprogram router;
3302 (e) A message specified by the cannot_route_message router option.
3304 In these cases only, Exim does include the error details in bounce and
3305 warning messages. There are also a few cases where bland messages such
3306 as "unrouteable address" or "local delivery error" are given.
3308 PH/43 $value is now also set for the "else" part of a ${run expansion.
3310 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3311 being worked on, but at least Exim now implements the latest version to
3314 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3315 process to complete was treated as a delivery failure, a timeout while
3316 writing the message to the pipe was logged, but erroneously treated as a
3317 successful delivery. Such timeouts include transport filter timeouts. For
3318 consistency with the overall process timeout, these timeouts are now
3319 treated as errors, giving rise to delivery failures by default. However,
3320 there is now a new Boolean option for the pipe transport called
3321 timeout_defer, which, if set TRUE, converts the failures into defers for
3322 both kinds of timeout. A transport filter timeout is now identified in
3325 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3326 systems where "make" and "gmake" are different, calling "gmake" at top
3327 level broke things. I've arranged for the value of $(MAKE) to be passed
3328 from the Makefile to this script so that it can call the same version of
3332 A note about Exim versions 4.44 and 4.50
3333 ----------------------------------------
3335 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3336 changes of various kinds. As a consequence, a big documentation update was
3337 needed. This delayed the release for rather longer than seemed good, especially
3338 in the light of a couple of (minor) security issues. Therefore, the changes
3339 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3340 release. So 4.44 and 4.50 are in effect two different branches that both start
3343 I have left the 4.50 change log unchanged; it contains all the changes since
3344 4.43. The change log for 4.44 is below; many of its items are identical to
3345 those for 4.50. This seems to be the most sensible way to preserve the
3346 historical information.
3352 1. Minor wording change to the doc/README.SIEVE file.
3354 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3355 computation of the current number of files was incorrect.
3357 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3358 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3361 4. Give more explanation in the error message when the command for a transport
3362 filter fails to execute.
3364 5. There are several places where Exim runs a non-Exim command in a
3365 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3366 was being done only for the command run by the queryprogram router. It is
3367 now done for all such subprocesses. The other cases are: ${run, transport
3368 filters, and the commands run by the lmtp and pipe transports.
3370 6. Added CONFIGURE_GROUP build-time option.
3372 7. Some older OS have a limit of 256 on the maximum number of file
3373 descriptors. Exim was using setrlimit() to set 1000 as a large value
3374 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3375 systems. I've change it so that if it can't get 1000, it tries for 256.
3377 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3378 was an oversight, and furthermore, ever since the addition of extra
3379 controls (e.g. 4.43/32), the checks on when to allow different forms of
3380 "control" were broken. There should now be diagnostics for all cases when a
3381 control that does not make sense is encountered.
3383 9. Added the /retain_sender option to "control=submission".
3385 10. $recipients is now available in the predata ACL (oversight).
3387 11. Tidy the search cache before the fork to do a delivery from a message
3388 received from the command line. Otherwise the child will trigger a lookup
3389 failure and thereby defer the delivery if it tries to use (for example) a
3390 cached ldap connection that the parent has called unbind on.
3392 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3393 of $address_data from the recipient verification was clobbered by the
3394 sender verification.
3396 13. The value of address_data from a sender verification is now available in
3397 $sender_address_data in subsequent conditions in the ACL statement.
3399 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3401 15. Added a new option "connect=<time>" to callout options, to set a different
3404 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3405 was its contents. (It was OK if the option was not defined at all.)
3407 17. A "Completed" log line is now written for messages that are removed from
3408 the spool by the -Mrm option.
3410 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3411 information about exactly what failed.
3413 19. Added -dd to debug only the daemon process.
3415 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3416 handles timeouts, both on the server side and network timeouts. Renamed the
3417 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3419 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3420 It is now set to "smtps".
3422 22. $host_address is now set to the target address during the checking of
3423 ignore_target_hosts.
3425 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3426 being passed; this would have caused $sender_host_name to have been used if
3427 matching the list had actually called for a host name (not very likely,
3428 since this list is usually IP addresses). A host name is now passed as
3431 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3432 code that specifies a non-restarting handler (typically sigaction() in
3433 modern systems) in an attempt to fix a rare and obscure crash bug.
3435 25. Narrowed the window for a race in the daemon that could cause it to ignore
3436 SIGCHLD signals. This is not a major problem, because they are used only to
3437 wake it up if nothing else does.
3439 26. A malformed maildirsize file could cause Exim to calculate negative values
3440 for the mailbox size or file count. Odd effects could occur as a result.
3441 The maildirsize information is now recalculated if the size or filecount
3444 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3445 support for a long time. Removed HAVE_SYS_VFS_H.
3447 28. Installed the latest version of exipick from John Jetmore.
3449 29. In an address list, if the pattern was not a regular expression, an empty
3450 subject address (from a bounce message) matched only if the pattern was an
3451 empty string. Non-empty patterns were not even tested. This was the wrong
3452 because it is perfectly reasonable to use an empty address as part of a
3453 database query. An empty address is now tested by patterns that are
3454 lookups. However, all the other forms of pattern expect the subject to
3455 contain a local part and a domain, and therefore, for them, an empty
3456 address still always fails if the pattern is not itself empty.
3458 30. Exim went into a mad DNS loop when attempting to do a callout where the
3459 host was specified on an smtp transport, and looking it up yielded more
3460 than one IP address.
3462 31. Re-factored the code for checking spool and log partition space into a
3463 function that finds that data and another that does the check. The former
3464 is then used to implement four new variables: $spool_space, $log_space,
3465 $spool_inodes, and $log_inodes.
3467 32. The RFC2047 encoding function was originally intended for short strings
3468 such as real names; it was not keeping to the 75-character limit for
3469 encoded words that the RFC imposes. It now respects the limit, and
3470 generates multiple encoded words if necessary. To be on the safe side, I
3471 have increased the buffer size for the ${rfc2047: expansion operator from
3474 33. It is now permitted to omit both strings after an "if" condition; if the
3475 condition is true, the result is "true". As before, when the second string
3476 is omitted, a false condition yields an empty string. This makes it less
3477 cumbersome to write custom ACL and router conditions.
3479 34. Failure to deliver a bounce message always caused it to be frozen, even if
3480 there was an errors_to setting on the router. The errors_to setting is now
3483 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3484 canonical form (fully expanded) before being placed in
3485 $sender_host_address.
3487 36. The table in the code that translates DNS record types into text (T_A to
3488 "A" for instance) was missing entries for NS and CNAME. It is just possible
3489 that this could have caused confusion if both these types were looked up
3490 for the same domain, because the text type is used as part of Exim's
3491 per-process caching. But the chance of anyone hitting this buglet seems
3494 37. The dnsdb lookup has been extended in a number of ways.
3496 (1) There is a new type, "zns", which walks up the domain tree until it
3497 finds some nameserver records. It should be used with care.
3499 (2) There is a new type, "mxh", which is like "mx" except that it returns
3500 just the host names, not the priorities.
3502 (3) It is now possible to give a list of domains (or IP addresses) to be
3503 looked up. The behaviour when one of the lookups defers can be
3504 controlled by a keyword.
3506 (4) It is now possible to specify the separator character for use when
3507 multiple records are returned.
3509 38. The dnslists ACL condition has been extended: it is now possible to supply
3510 a list of IP addresses and/or domains to be looked up in a particular DNS
3513 39. Added log_selector=+queue_time_overall.
3515 40. When running the queue in the test harness, wait just a tad after forking a
3516 delivery process, to get repeatability of debugging output.
3518 41. Include certificate and key file names in error message when GnuTLS fails
3519 to set them up, because the GnuTLS error message doesn't include the name
3520 of the failing file when there is a problem reading it.
3522 42. Allow both -bf and -bF in the same test run.
3524 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3526 44. The "Exiscan patch" is now merged into the mainline Exim source.
3528 45. Sometimes the final signoff response after QUIT could fail to get
3529 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3530 < 0 before doing a fflush(). This bug looks as though it goes back to the
3531 introduction of TLS in release 3.20, but "sometimes" must have been rare
3532 because the tests only now provoked it.
3534 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3535 (this can affect the format of dates).
3537 47. exim_tidydb, when checking for the continued existence of a message for
3538 which it has found a message-specific retry record, was not finding
3539 messages that were in split spool directories. Consequently, it was
3540 deleting retry records that should have stayed in existence.
3542 48. Steve fixed some bugs in eximstats.
3544 49. The SPA authentication driver was not abandoning authentication and moving
3545 on to the next authenticator when an expansion was forced to fail,
3546 contradicting the general specification for all authenticators. Instead it
3547 was generating a temporary error. It now behaves as specified.
3549 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3550 (the order specifies the preference for clients). The order is now AES256,
3551 AES128, 3DES, ARCFOUR128.
3553 51. Small patch to Sieve code - explicitly set From: when generating an
3556 52. Exim crashed if a remote delivery caused a very long error message to be
3557 recorded - for instance if somebody sent an entire SpamAssassin report back
3558 as a large number of 550 error lines. This bug was coincidentally fixed by
3559 increasing the size of one of Exim's internal buffers (big_buffer) that
3560 happened as part of the Exiscan merge. However, to be on the safe side, I
3561 have made the code more robust (and fixed the comments that describe what
3564 53. Now that there can be additional text after "Completed" in log lines (if
3565 the queue_time_overall log selector is set), a one-byte patch to exigrep
3566 was needed to allow it to recognize "Completed" as not the last thing in
3569 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3570 patch that reportedly fixes this has been added. I am not expert enough to
3571 create a test for it. This is what the patch creator wrote:
3573 "I found a little strange behaviour of ldap code when working with
3574 Windows 2003 AD Domain, where users was placed in more than one
3575 Organization Units. When I tried to give exim partial DN, the exit code
3576 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3577 But simultaneously result of request was absolutely normal ldap result,
3578 so I produce this patch..."
3580 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3581 REFERENCE, so I have modified the code to exclude the patch when that macro
3584 55. Some experimental protocols are using DNS PTR records for new purposes. The
3585 keys for these records are domain names, not reversed IP addresses. The
3586 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3587 leaves it alone. Component reversal etc. now happens only for IP addresses.
3590 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3592 57. Double the size of the debug message buffer (to 2048) so that more of very
3593 long debug lines gets shown.
3595 58. The exicyclog utility now does better if the number of log files to keep
3596 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3598 59. Two changes related to the smtp_active_hostname option:
3600 (1) $smtp_active_hostname is now available as a variable.
3601 (2) The default for smtp_banner uses $smtp_active_hostname instead
3602 of $primary_hostname.
3604 60. The host_aton() function is supposed to be passed a string that is known
3605 to be a valid IP address. However, in the case of IPv6 addresses, it was
3606 not checking this. This is a hostage to fortune. Exim now panics and dies
3607 if the condition is not met. A case was found where this could be provoked
3608 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3609 components; fortuitously, this particular loophole had already been fixed
3610 by change 4.50/55 above.
3612 If there are any other similar loopholes, the new check in host_aton()
3613 itself should stop them being exploited. The report I received stated that
3614 data on the command line could provoke the exploit when Exim was running as
3615 exim, but did not say which command line option was involved. All I could
3616 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3617 running as the user.
3620 61. There was a buffer overflow vulnerability in the SPA authentication code
3621 (which came originally from the Samba project). I have added a test to the
3622 spa_base64_to_bits() function which I hope fixes it.
3625 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3626 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3628 63. The daemon start-up calls getloadavg() while still root for those OS that
3629 need the first call to be done as root, but it missed one case: when
3630 deliver_queue_load_max is set with deliver_drop_privilege. This is
3631 necessary for the benefit of the queue runner, because there is no re-exec
3632 when deliver_drop_privilege is set.
3634 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3635 This has been fixed.
3637 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3638 was in use, was not putting the data itself into the right store pool;
3639 consequently, it could be overwritten for a subsequent message in the same
3640 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3643 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3645 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3646 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3647 the function were treating the return as a boolean value, which happened to
3648 work because 0=false and not-0=true, but is not correct code.
3650 68. The host_aton() function was not handling scoped IPv6 addresses (those
3651 with, for example, "%eth0" on the end) correctly.
3653 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3654 negated items (that is, ~something) in unsigned ints. Some compilers
3655 apparently mutter when there is no cast.
3657 70. If an address verification called from an ACL failed, and did not produce a
3658 user-specific message (i.e. there was only a "system" message), nothing was
3659 put in $acl_verify_message. In this situation, it now puts the system
3662 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3663 session; change 4.31/43 added the unwanted input to the log line - except
3664 that it did not do this in the start of session case. It now does.
3666 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3667 This is wrong and can cause the other end to generate a synchronization
3668 error if it is another Exim or anything else that does the synchronization
3669 check. A QUIT command is no longer sent after a timeout.
3671 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3672 during host lookups.
3674 74. The defer_ok option of callout verification was not working if it was used
3675 when verifying addresses in header lines, that is, for this case:
3677 verify = header_sender/callout=defer_ok
3679 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3680 those file descriptors could be used for SMTP connections. If anything
3681 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3682 could be sent to the SMTP client, causing chaos. The daemon now opens
3683 stdin, stdout, and stderr to /dev/null when it puts itself into the
3686 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3687 log by default. The user can override this with suitable Perl magic.
3689 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3690 the log message when discard triggers, was not working for the DATA ACL or
3691 for the non-SMTP ACL.
3693 78. Error message wording change in sieve.c.
3695 79. If smtp_accept_max_per_host was set, the number of connections could be
3696 restricted to fewer than expected, because the daemon was trying to set up
3697 a new connection before checking whether the processes handling previous
3698 connections had finished. The check for completed processes is now done
3699 earlier. On busy systems, this bug wouldn't be noticed because something
3700 else would have woken the daemon, and it would have reaped the completed
3703 80. If a message was submitted locally by a user whose login name contained one
3704 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3705 It caused a spool format error. I have fixed the spool reading code. A
3706 related problem was that the "from" clause in the Received: line became
3707 illegal because of the space(s). It is now covered by ${quote_local_part.
3709 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3712 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3714 83. Patch from Sieve maintainer to fix unterminated string problem in
3715 "vacation" handling.
3717 84. Some minor changes to the Linux configuration files to help with other
3718 OS variants using glibc.
3720 85. One more patch for Sieve to update vacation handling to latest spec.
3723 ----------------------------------------------------
3724 See the note above about the 4.44 and 4.50 releases.
3725 ----------------------------------------------------
3731 1. Change 4.43/35 introduced a bug that caused file counts to be
3732 incorrectly computed when quota_filecount was set in an appendfile
3735 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3736 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3739 3. Give more explanation in the error message when the command for a transport
3740 filter fails to execute.
3742 4. There are several places where Exim runs a non-Exim command in a
3743 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3744 was being done only for the command run by the queryprogram router. It is
3745 now done for all such subprocesses. The other cases are: ${run, transport
3746 filters, and the commands run by the lmtp and pipe transports.
3748 5. Some older OS have a limit of 256 on the maximum number of file
3749 descriptors. Exim was using setrlimit() to set 1000 as a large value
3750 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3751 systems. I've change it so that if it can't get 1000, it tries for 256.
3753 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3754 was an oversight, and furthermore, ever since the addition of extra
3755 controls (e.g. 4.43/32), the checks on when to allow different forms of
3756 "control" were broken. There should now be diagnostics for all cases when a
3757 control that does not make sense is encountered.
3759 7. $recipients is now available in the predata ACL (oversight).
3761 8. Tidy the search cache before the fork to do a delivery from a message
3762 received from the command line. Otherwise the child will trigger a lookup
3763 failure and thereby defer the delivery if it tries to use (for example) a
3764 cached ldap connection that the parent has called unbind on.
3766 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3767 of $address_data from the recipient verification was clobbered by the
3768 sender verification.
3770 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3771 was its contents. (It was OK if the option was not defined at all.)
3773 11. A "Completed" log line is now written for messages that are removed from
3774 the spool by the -Mrm option.
3776 12. $host_address is now set to the target address during the checking of
3777 ignore_target_hosts.
3779 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3780 being passed; this would have caused $sender_host_name to have been used if
3781 matching the list had actually called for a host name (not very likely,
3782 since this list is usually IP addresses). A host name is now passed as
3785 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3786 code that specifies a non-restarting handler (typically sigaction() in
3787 modern systems) in an attempt to fix a rare and obscure crash bug.
3789 15. Narrowed the window for a race in the daemon that could cause it to ignore
3790 SIGCHLD signals. This is not a major problem, because they are used only to
3791 wake it up if nothing else does.
3793 16. A malformed maildirsize file could cause Exim to calculate negative values
3794 for the mailbox size or file count. Odd effects could occur as a result.
3795 The maildirsize information is now recalculated if the size or filecount
3798 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3799 support for a long time. Removed HAVE_SYS_VFS_H.
3801 18. Updated exipick to current release from John Jetmore.
3803 19. Allow an empty sender to be matched against a lookup in an address list.
3804 Previously the only cases considered were a regular expression, or an
3807 20. Exim went into a mad DNS lookup loop when doing a callout where the
3808 host was specified on the transport, if the DNS lookup yielded more than
3811 21. The RFC2047 encoding function was originally intended for short strings
3812 such as real names; it was not keeping to the 75-character limit for
3813 encoded words that the RFC imposes. It now respects the limit, and
3814 generates multiple encoded words if necessary. To be on the safe side, I
3815 have increased the buffer size for the ${rfc2047: expansion operator from
3818 22. Failure to deliver a bounce message always caused it to be frozen, even if
3819 there was an errors_to setting on the router. The errors_to setting is now
3822 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3823 canonical form (fully expanded) before being placed in
3824 $sender_host_address.
3826 24. Updated eximstats to version 1.33
3828 25. Include certificate and key file names in error message when GnuTLS fails
3829 to set them up, because the GnuTLS error message doesn't include the name
3830 of the failing file when there is a problem reading it.
3832 26. Expand error message when OpenSSL has problems setting up cert/key files.
3835 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3836 (this can affect the format of dates).
3838 28. exim_tidydb, when checking for the continued existence of a message for
3839 which it has found a message-specific retry record, was not finding
3840 messages that were in split spool directories. Consequently, it was
3841 deleting retry records that should have stayed in existence.
3843 29. eximstats updated to version 1.35
3844 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3845 1.35 - bugfix such that pie charts by volume are generated correctly
3847 30. The SPA authentication driver was not abandoning authentication and moving
3848 on to the next authenticator when an expansion was forced to fail,
3849 contradicting the general specification for all authenticators. Instead it
3850 was generating a temporary error. It now behaves as specified.
3852 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3853 (the order specifies the preference for clients). The order is now AES256,
3854 AES128, 3DES, ARCFOUR128.
3856 31. Small patch to Sieve code - explicitly set From: when generating an
3859 32. Exim crashed if a remote delivery caused a very long error message to be
3860 recorded - for instance if somebody sent an entire SpamAssassin report back
3861 as a large number of 550 error lines. This bug was coincidentally fixed by
3862 increasing the size of one of Exim's internal buffers (big_buffer) that
3863 happened as part of the Exiscan merge. However, to be on the safe side, I
3864 have made the code more robust (and fixed the comments that describe what
3867 33. Some experimental protocols are using DNS PTR records for new purposes. The
3868 keys for these records are domain names, not reversed IP addresses. The
3869 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3870 leaves it alone. Component reversal etc. now happens only for IP addresses.
3873 34. The host_aton() function is supposed to be passed a string that is known
3874 to be a valid IP address. However, in the case of IPv6 addresses, it was
3875 not checking this. This is a hostage to fortune. Exim now panics and dies
3876 if the condition is not met. A case was found where this could be provoked
3877 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3878 components; fortuitously, this particular loophole had already been fixed
3879 by change 4.50/55 or 4.44/33 above.
3881 If there are any other similar loopholes, the new check in host_aton()
3882 itself should stop them being exploited. The report I received stated that
3883 data on the command line could provoke the exploit when Exim was running as
3884 exim, but did not say which command line option was involved. All I could
3885 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3886 running as the user.
3889 35. There was a buffer overflow vulnerability in the SPA authentication code
3890 (which came originally from the Samba project). I have added a test to the
3891 spa_base64_to_bits() function which I hope fixes it.
3894 36. The daemon start-up calls getloadavg() while still root for those OS that
3895 need the first call to be done as root, but it missed one case: when
3896 deliver_queue_load_max is set with deliver_drop_privilege. This is
3897 necessary for the benefit of the queue runner, because there is no re-exec
3898 when deliver_drop_privilege is set.
3900 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3901 was in use, was not putting the data itself into the right store pool;
3902 consequently, it could be overwritten for a subsequent message in the same
3903 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3906 38. Sometimes the final signoff response after QUIT could fail to get
3907 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3908 < 0 before doing a fflush(). This bug looks as though it goes back to the
3909 introduction of TLS in release 3.20, but "sometimes" must have been rare
3910 because the tests only now provoked it.
3916 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3917 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3918 Now it returns an integer. A number of calls were still expecting a T/F
3919 return. Fortuitously, in all cases, the tests worked in OK situations,
3920 which is the norm. However, things would have gone wrong on any write
3921 failures on the smtp file descriptor. This function is used when sending
3922 messages over SMTP and also when doing verify callouts.
3924 2. When Exim is called to do synchronous delivery of a locally submitted
3925 message (the -odf or -odi options), it no longer closes stderr before doing
3928 3. Implemented the mua_wrapper option.
3930 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3932 5. Implemented the functions header_remove(), header_testname(),
3933 header_add_at_position(), and receive_remove_recipient(), and exported them
3936 6. If an ACL "warn" statement specified the addition of headers, Exim already
3937 inserted X-ACL-Warn: at the start if there was no header name. However, it
3938 was not making this test for the second and subsequent header lines if
3939 there were newlines in the string. This meant that an invalid header could
3940 be inserted if Exim was badly configured.
3942 7. Allow an ACL "warn" statement to add header lines at the start or after all
3943 the Received: headers, as well as at the end.
3945 8. Added the rcpt_4xx retry error code.
3947 9. Added postmaster_mailfrom=xxx to callout verification option.
3949 10. Added mailfrom=xxxx to the callout verification option, for verify=
3952 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3953 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3955 12. Inserted some casts to stop certain compilers warning when using pointer
3956 differences as field lengths or precisions in printf-type calls (mostly
3957 affecting debugging statements).
3959 13. Added optional readline() support for -be (dynamically loaded).
3961 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3962 same clock tick as a message's arrival, so that its received time was the
3963 same as the "first fail" time on the retry record, and that message
3964 remained on the queue past the ultimate address timeout, every queue runner
3965 would try a delivery (because it was past the ultimate address timeout) but
3966 after another failure, the ultimate address timeout, which should have then
3967 bounced the address, did not kick in. This was a "< instead of <=" error;
3968 in most cases the first failure would have been in the next clock tick
3969 after the received time, and all would be well.
3971 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3972 being recognized when the domain list was tested by the match_domain
3973 condition in an expansion string.
3975 16. Added the ${str2b64: operator.
3977 17. Exim was always calling setrlimit() to set a large limit for the number of
3978 processes, without checking whether the existing limit was already
3979 adequate. (It did check for the limit on file descriptors.) Furthermore,
3980 errors from getrlimit() and setrlimit() were being ignored. Now they are
3981 logged to the main and panic logs, but Exim does carry on, to try to do its
3982 job under whatever limits there are.
3984 18. Imported PCRE 5.0.
3986 19. Trivial typo in log message " temporarily refused connection" (the leading
3989 20. If the log selector return_path_on_delivery was set and an address was
3990 redirected to /dev/null, the delivery process crashed because it assumed
3991 that a return path would always be set for a "successful" delivery. In this
3992 case, the whole delivery is bypassed as an optimization, and therefore no
3995 21. Internal re-arrangement: the function for sending a challenge and reading
3996 a response while authentication was assuming a zero-terminated challenge
3997 string. It's now changed to take a pointer and a length, to allow for
3998 binary data in such strings.
4000 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4002 23. Exim was not respecting finduser_retries when seeking the login of the
4003 uid under which it was called; it was always trying 10 times. (The default
4004 setting of finduser_retries is zero.) Also, it was sleeping after the final
4005 failure, which is pointless.
4007 24. Implemented tls_on_connect_ports.
4009 25. Implemented acl_smtp_predata.
4011 26. If the domain in control=submission is set empty, Exim assumes that the
4012 authenticated id is a complete email address when it generates From: or
4013 Sender: header lines.
4015 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4016 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4017 chown and chgrp in /bin and hostname in /usr/bin.
4019 28. Exim was keeping the "process log" file open after each use, just as it
4020 does for the main log. This opens the possibility of it remaining open for
4021 long periods when the USR1 signal hits a daemon. Occasional processlog
4022 errors were reported, that could have been caused by this. Anyway, it seems
4023 much more sensible not to leave this file open at all, so that is what now
4026 29. The long-running daemon process does not normally write to the log once it
4027 has entered its main loop, and it closes the log before doing so. This is
4028 so that log files can straightforwardly be renamed and moved. However,
4029 there are a couple of unusual error situations where the daemon does write
4030 log entries, and I had neglected to close the log afterwards.
4032 30. The text of an SMTP error response that was received during a remote
4033 delivery was being truncated at 512 bytes. This is too short for some of
4034 the long messages that one sometimes sees. I've increased the limit to
4037 31. It is now possible to make retry rules that apply only when a message has a
4038 specific sender, in particular, an empty sender.
4040 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4041 it possible to be selective about when SMTP synchronization is enforced.
4043 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4045 32. Implemented hosts_connection_nolog.
4047 33. Added an ACL for QUIT.
4049 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4052 35. Added mailbox_size and mailbox_filecount to appendfile.
4054 36. Added control = no_multiline_responses to ACLs.
4056 37. There was a bug in the logic of the code that waits for the clock to tick
4057 in the case where the clock went backwards by a substantial amount such
4058 that the microsecond fraction of "now" was more than the microsecond
4059 fraction of "then" (but the whole seconds number was less).
4061 38. Added support for the libradius Radius client library this is found on
4062 FreeBSD (previously only the radiusclient library was supported).
4068 1. When certain lookups returned multiple values in the form name=value, the
4069 quoting of the values was not always being done properly. Specifically:
4070 (a) If the value started with a double quote, but contained no whitespace,
4072 (b) If the value contained whitespace other than a space character (i.e.
4073 tabs or newlines or carriage returns) it was not quoted.
4074 This fix has been applied to the mysql and pgsql lookups by writing a
4075 separate quoting function and calling it from the lookup code. The fix
4076 should probably also be applied to nisplus, ibase and oracle lookups, but
4077 since I cannot test any of those, I have not disturbed their existing code.
4079 2. A hit in the callout cache for a specific address caused a log line with no
4080 reason for rejecting RCPT. Now it says "Previous (cached) callout
4081 verification failure".
4083 3. There was an off-by-one bug in the queryprogram router. An over-long
4084 return line was truncated at 256 instead of 255 characters, thereby
4085 overflowing its buffer with the terminating zero. As well as fixing this, I
4086 have increased the buffer size to 1024 (and made a note to document this).
4088 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4089 when Exim is waiting for an SMTP response from a remote server, Exim
4090 restarts its select() call on the socket, thereby resetting its timeout.
4091 This is not a problem when such interrupts are rare. Somebody set up a cron
4092 job to run exiwhat every 2 minutes, which is less than the normal select()
4093 timeout (5 or 10 minutes). This meant that the select() timeout never
4094 kicked in because it was always reset. I have fixed this by comparing the
4095 time when an interrupt arrives with the time at the start of the first call
4096 to select(). If more time than the timeout has elapsed, the interrupt is
4097 treated as a timeout.
4099 5. Some internal re-factoring in preparation for the addition of Sieve
4100 extensions (by MH). In particular, the "personal" test is moved to a
4101 separate function, and given an option for scanning Cc: and Bcc: (which is
4102 not set for Exim filters).
4104 6. When Exim created an email address using the login of the caller as the
4105 local part (e.g. when creating a From: or Sender: header line), it was not
4106 quoting the local part when it contained special characters such as @.
4108 7. Installed new OpenBSD configuration files.
4110 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4111 try to make them clearer.
4113 9. Callout options, other than the timeout value, were being ignored when
4114 verifying sender addresses in header lines. For example, when using
4116 verify = header_sender/callout=no_cache
4118 the cache was (incorrectly) being used.
4120 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4121 only the Cygwin environment.
4123 11. When return_path_on_delivery was set as a log selector, if different remote
4124 addresses in the same message used different return paths and parallel
4125 remote delivery occurred, the wrong values would sometimes be logged.
4126 (Whenever a remote delivery process finished, the return path value from
4127 the most recently started remote delivery process was logged.)
4129 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4130 lines when AUTH and/or TLS are in use. This is the "received protocol"
4131 field. Exim used to use "asmtp" for authenticated SMTP, without any
4132 indication (in the protocol name) for TLS use. Now it follows the RFC and
4133 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4134 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4135 names appear in log lines as well as in Received: header lines.
4137 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4138 extensions, and comparison tests, and to fix some bugs.
4140 14. Changes to the "personal" filter test:
4142 (1) The test was buggy in that it was just doing the equivalent of
4143 "contains" tests on header lines. For example, if a user's address was
4144 anne@some.where, the "personal" test would incorrectly be true for
4146 To: susanne@some.where
4148 This test is now done by extracting each address from the header in turn,
4149 and checking the entire address. Other tests that are part of "personal"
4150 are now done using regular expressions (for example, to check local parts
4151 of addresses in From: header lines).
4153 (2) The list of non-personal local parts in From: addresses has been
4154 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4155 taken from the Sieve specification recommendations.
4157 (3) If the message contains any header line starting with "List-" it is
4158 treated as non-personal.
4160 (4) The test for "circular" in the Subject: header line has been removed
4161 because it now seems ill-conceived.
4163 15. Minor typos in src/EDITME comments corrected.
4165 16. Installed latest exipick from John Jetmore.
4167 17. If headers_add on a router specified a text string that was too long for
4168 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4169 of string_sprintf() is now avoided.
4171 18. $message_body_size was not set (it was always zero) when running the DATA
4172 ACL and the local_scan() function.
4174 19. For the "mail" command in an Exim filter, no default was being set for
4175 the once_repeat time, causing a random time value to be used if "once" was
4176 specified. (If the value happened to be <= 0, no repeat happened.) The
4177 default is now 0s, meaning "never repeat". The "vacation" command was OK
4178 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4179 (I found it when inspecting the code).
4181 20. There is now an overall timeout for performing a callout verification. It
4182 defaults to 4 times the callout timeout, which applies to individual SMTP
4183 commands during the callout. The overall timeout applies when there is more
4184 than one host that can be tried. The timeout is checked before trying the
4185 next host. This prevents very long delays if there are a large number of
4186 hosts and all are timing out (e.g. when the network connections are timing
4187 out). The value of the overall timeout can be changed by specifying an
4188 additional sub-option for "callout", called "maxwait". For example:
4190 verify = sender/callout=5s,maxwait=20s
4192 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4193 to the end before writing, but this should make it even safer).
4195 22. Exim was forgetting that it had advertised PIPELINING for the second and
4196 subsequent messages on an SMTP connection. It was also not resetting its
4197 memory on STARTTLS and an internal HELO.
4199 23. When Exim logs an SMTP synchronization error within a session, it now
4200 records whether PIPELINING has been advertised or not.
4202 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4203 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4204 rather than long int.
4206 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4208 26. Added the never_mail option to autoreply.
4214 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4215 crash if the getsockname() call failed; this can happen if a connection is
4216 closed very soon after it is established. The problem was simply in the
4217 order in which certain operations were done, causing Exim to try to write
4218 to the SMTP stream before it had set up the file descriptor. The bug has
4219 been fixed by making things happen in the correct order.
4225 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4226 before the connection was closed, thus losing the rejection response.
4228 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4229 some early Solaris releases, but causes trouble in current releases where
4230 socklen_t is defined.
4232 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4235 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4238 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4239 that releasing the top of it at the end releases what was used for sub-
4240 expansions (unless the block got too big). However, discard this block if
4241 the first thing is a variable or header, so that we can use its block when
4242 it is dynamic (useful for very large $message_headers, for example).
4244 6. Lookups now cache *every* query, not just the most recent. A new, separate
4245 store pool is used for this. It can be recovered when all lookup caches are
4246 flushed. Lookups now release memory at the end of their result strings.
4247 This has involved some general refactoring of the lookup sources.
4249 7. Some code has been added to the store_xxx() functions to reduce the amount
4250 of flapping under certain conditions.
4252 8. log_incoming_interface used to affect only the <= reception log lines. Now
4253 it causes the local interface and port to be added to several more SMTP log
4254 lines, for example "SMTP connection from", and rejection lines.
4256 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4258 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4260 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4261 could be overwritten at the end of the current message (or the start of a
4262 new message if it was set in a HELO ACL). The value is now preserved for
4263 the duration of the SMTP connection.
4265 12. If a transport had a headers_rewrite setting, and a matching header line
4266 contained an unqualified address, that address was qualified, even if it
4267 did not match any rewriting rules. The underlying bug was that the values
4268 of the flags that permit the existence of unqualified sender and recipient
4269 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4270 non-local messages, and by -bnq for local messages) were not being
4271 preserved with the message after it was received.
4273 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4274 "next input=" as part of the text comprising the host identity instead of
4275 the correct text. The code was using the same buffer for two different
4276 strings. However, depending on which order the printing function evaluated
4277 its arguments, the bug did not always show up. Under Linux, for example, my
4278 test suite worked just fine.
4280 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4281 doesn't work with some older versions of Perl. It has been changed to "my",
4282 which in any case is probably the better facility to use.
4284 15. A really picky compiler found some instances of statements for creating
4285 error messages that either had too many or two few arguments for the format
4288 16. The size of the buffer for calls to the DNS resolver has been increased
4289 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4290 for addresses that have a lot of PTR records. This alleviates a problem; it
4291 does not fully solve it.
4293 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4294 buffer now truncates the list and logs the incident, which is the same
4295 action as happens when Exim is looking up a host name and its aliases.
4296 Previously in this situation something unpredictable would happen;
4297 sometimes it was "internal error: store_reset failed".
4299 18. If a server dropped the connection unexpectedly when an Exim client was
4300 using GnuTLS and trying to read a response, the client delivery process
4301 crashed while trying to generate an error log message.
4303 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4304 string, for example:
4306 warn message = H1: something\nH2: something
4308 the text was added as a single header line from Exim's point of view
4309 though it ended up OK in the delivered message. However, searching for the
4310 second and subsequent header lines using $h_h2: did not work. This has been
4311 fixed. Similarly, if a system filter added multiple headers in this way,
4312 the routers could not see them.
4314 20. Expanded the error message when iplsearch is called with an invalid key to
4315 suggest using net-iplsearch in a host list.
4317 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4318 ACLs are no longer actually imposed (and a message to that effect is
4321 22. If a "gecos" field in a passwd entry contained escaped characters, in
4322 particular, if it contained a \" sequence, Exim got it wrong when building
4323 a From: or a Sender: header from that name. A second bug also caused
4324 incorrect handling when an unquoted " was present following a character
4325 that needed quoting.
4327 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4328 was not being matched caselessly.
4330 24. Arranged for all hyphens in the exim.8 source to be escaped with
4333 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4334 properly. Recipient callout cache records were still being keyed to include
4335 the sender, even when use_sender was set false. This led to far more
4336 callouts that were necessary. The sender is no longer included in the key
4337 when use_sender is false.
4339 26. Added "control = submission" modifier to ACLs.
4341 27. Added the ${base62d: operator to decode base 62 numbers.
4343 28. dnsdb lookups can now access SRV records.
4345 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4346 the configuration file.
4348 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4349 (-v) mode. This makes the output for a verbose queue run more intelligible.
4351 31. Added a use_postmaster feature to recipient callouts.
4353 32. Added the $body_zerocount variable, containing the number of binary zero
4354 bytes in the message body.
4356 33. The time of last modification of the "new" subdirectory is now used as the
4357 "mailbox time last read" when there is a quota error for a maildir
4360 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4362 35. Added +ignore_unknown as a special item in host lists.
4364 36. Code for decoding IPv6 addresses in host lists is now included, even if
4365 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4366 address was recognized as an IP address, but was then not correctly decoded
4367 into binary, causing unexpected and incorrect effects when compared with
4374 1. Very minor rewording of debugging text in manualroute to say "list of
4375 hosts" instead of "hostlist".
4377 2. If verify=header_syntax was set, and a header line with an unqualified
4378 address (no domain) and a large number of spaces between the end of the
4379 name and the colon was received, the reception process suffered a buffer
4380 overflow, and (when I tested it) crashed. This was caused by some obsolete
4381 code that should have been removed. The fix is to remove it!
4383 3. When running in the test harness, delay a bit after writing a bounce
4384 message to get a bit more predictability in the log output.
4386 4. Added a call to search_tidyup() just before forking a reception process. In
4387 theory, someone could use a lookup in the expansion of smtp_accept_max_
4388 per_host which, without the tidyup, could leave open a database connection.
4390 5. Added the variables $recipient_data and $sender_data which get set from a
4391 lookup success in an ACL "recipients" or "senders" condition, or a router
4392 "senders" option, similar to $domain_data and $local_part_data.
4394 6. Moved the writing of debug_print from before to after the "senders" test
4397 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4398 problems for message scanning, either using a data ACL, or using
4399 local_scan() because the Received: header was not generated till after they
4400 were called (in order to set the time as the time of reception completion).
4401 I have revised the way this works. The header is now generated after the
4402 body is received, but before the ACL or local_scan() are called. After they
4403 are run, the timestamp in the header is updated.
4409 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4410 before starting a queue runner without re-exec. This happened only when
4411 deliver_drop_privilege was set or when the Exim user was set to root. The
4412 effect of the bug was that timeouts during subsequent deliveries caused
4413 crashes instead of being properly handled. The handler is now left at its
4414 default (and expected) setting.
4416 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4417 message, again when deliver_drop_privilege is set or Exim is run as root.
4418 The bug described in (1) was not present in this case, but the tidying up
4419 of the other signals was missing. I have made the two cases consistent.
4421 3. The ignore_target_hosts setting on a manualroute router was being ignored
4422 for hosts that were looked up using the /MX notation.
4424 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4427 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4428 operated on the sender address. After changing the $sender_address to <>
4429 for the sender address verify, Exim was re-instated it as the original
4430 (before rewriting) address, but remembering that it had rewritten it, so it
4431 wasn't rewriting it again. This bug also had the effect of breaking the
4432 sender address verification caching when the sender address was rewritten.
4434 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4435 This has been changed so that if the ip literal address matches
4436 ignore_target_hosts, the router declines.
4438 7. Added expansion conditions match_domain, match_address, and match_local_
4439 part (NOT match_host).
4441 8. The placeholder for the Received: header didn't have a length field set.
4443 9. Added code to Exim itself and to exim_lock to test for a specific race
4444 condition that could lead to file corruption when using MBX delivery. The
4445 issue is with the lockfile that is created in /tmp. If this file is removed
4446 after a process has opened it but before that process has acquired a lock,
4447 there is the potential for a second process to recreate the file and also
4448 acquire a lock. This could lead to two Exim processes writing to the file
4449 at the same time. The added code performs the same test as UW imapd; it
4450 checks after acquiring the lock that its file descriptor still refers to
4451 the same named file.
4453 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4454 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4457 11. Added the smtp_active_hostname option. If used, this will typically be made
4458 to depend on the incoming interface address. Because $interface_address is
4459 not set up until the daemon has forked a reception process, error responses
4460 that can happen earlier (such as "too many connections") no longer contain
4463 12. If an expansion in a condition on a "warn" statement fails because a lookup
4464 defers, the "warn" statement is abandoned, and the next ACL statement is
4465 processed. Previously this caused the whole ACL to be aborted.
4467 13. Added the iplsearch lookup type.
4469 14. Added ident_timeout as a log selector.
4471 15. Added tls_certificate_verified as a log selector.
4473 16. Added a global option tls_require_ciphers (compare the smtp transport
4474 option of the same name). This controls incoming TLS connections.
4476 17. I finally figured out how to make tls_require_ciphers do a similar thing
4477 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4478 before starting the TLS session.
4480 18. Tabs are now shown as \t in -bP output.
4482 19. If the log selector return_path_on_delivery was set, Exim crashed when
4483 bouncing a message because it had too many Received: header lines.
4485 20. If two routers both had headers_remove settings, and the first one included
4486 a superfluous trailing colon, the final name in the first list and the
4487 first name in the second list were incorrectly joined into one item (with a
4488 colon in the middle).
4494 1. Added -C and -D options to the exinext utility, mainly to make it easier
4495 to include in the automated testing, but these could be helpful when
4496 multiple configurations are in use.
4498 2. The exinext utility was not formatting the output nicely when there was
4499 an alternate port involved in the retry record key, nor when there was a
4500 message id as well (for retries that were specific to a specific message
4501 and a specific host). It was also confused by IPv6 addresses, because of
4502 the additional colons they contain. I have fixed the IPv4 problem, and
4503 patched it up to do a reasonable job for IPv6.
4505 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4506 delivery, the log line now contains "pipelined" if PIPELINING was used.
4508 4. An SMTP transport process used to panic and die if the bind() call to set
4509 an explicit outgoing interface failed. This has been changed; it is now
4510 treated in the same way as a connect() failure.
4512 5. A reference to $sender_host_name in the part of a conditional expansion
4513 that was being skipped was still causing a DNS lookup. This no longer
4516 6. The def: expansion condition was not recognizing references to header lines
4517 that used bh_ and bheader_.
4519 7. Added the _cache feature to named lists.
4521 8. The code for checking quota_filecount in the appendfile transport was
4522 allowing one more file than it should have been.
4524 9. For compatibility with Sendmail, the command line option
4532 and sets the incoming protocol and host name (for trusted callers). The
4533 host name and its colon can be omitted when only the protocol is to be set.
4534 Note the Exim already has two private options, -pd and -ps, that refer to
4535 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4536 "s", but I don't think that's a major issue.
4538 10. A number of refactoring changes to the code, none of which should affect
4541 (a) The number of logging options was getting close to filling up the
4542 32-bit word that was used as a bit map. I have split them into two classes:
4543 those that are passed in the argument to log_write(), and those that are
4544 only ever tested independently outside of that function. These are now in
4545 separate 32-bit words, so there is plenty of room for expansion again.
4546 There is no change in the user interface or the logging behaviour.
4548 (b) When building, for example, log lines, the code previously used a
4549 macro that called string_cat() twice, in order to add two strings. This is
4550 not really sufficiently general. Furthermore, there was one instance where
4551 it was actually wrong because one of the argument was used twice, and in
4552 one call a function was used. (As it happened, calling the function twice
4553 did not affect the overall behaviour.) The macro has been replaced by a
4554 function that can join an arbitrary number of extra strings onto a growing
4557 (c) The code for expansion conditions now uses a table and a binary chop
4558 instead of a serial search (which was left over from when there were very
4559 few conditions). Also, it now recognizes conditions like "pam" even when
4560 the relevant support is not compiled in: a suitably worded error message is
4561 given if an attempt is made to use such a condition.
4563 11. Added ${time_interval:xxxxx}.
4565 12. A bug was causing one of the ddress fields not to be passed back correctly
4566 from remote delivery subprocesses. The field in question was not being
4567 subsequently used, so this caused to problems in practice.
4569 13. Added new log selectors queue_time and deliver_time.
4571 14. Might have fixed a bug in maildirsizefile handling that threw up
4572 "unexpected character" debug warnings, and recalculated the data
4573 unnecessarily. In any case, I expanded the warning message to give more
4576 15. Added the message "Restricted characters in address" to the statements in
4577 the default ACL that block characters like @ and % in local parts.
4579 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4580 Three changes have been made:
4582 (a) There was a serious bug; a negative response to MAIL caused the whole
4583 recipient domain to be cached as invalid, thereby blocking all messages
4584 to all local parts at the same domain, from all senders. This bug has
4585 been fixed. The domain is no longer cached after a negative response to
4586 MAIL if the sender used is not empty.
4588 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4591 (c) A new callout option, "use_sender" has been added for people who want
4592 the modified behaviour.
4598 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4601 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4602 indeed breaks things for older releases.
4604 3. Added additional logging to the case where there is a problem reading data
4605 from a filter that is running in a subprocess using a pipe, in order to
4606 try to track down a specific problem.
4608 4. Testing facility fudge: when running in the test harness and attempting
4609 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4610 getting "No route to host". Convert this to a timeout.
4612 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4615 6. Some OS don't have socklen_t but use size_t instead. This affects the
4616 fifth argument of getsockopt() amongst other things. This is now
4617 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4618 can be set for individual OS. I have set it for SunOS5, OSF1, and
4619 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4620 some earlier ones do not.
4622 7. Change 4.30/15 was not doing the test caselessly.
4624 8. The standard form for an IPv6 address literal was being rejected by address
4625 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4626 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4627 this, as well as the form without the "IPv6" on the front (but only when
4628 address literals are enabled, of course).
4630 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4632 10. Exim crashed if a message with an empty sender address specified by -f
4633 encountered a router with an errors_to setting. This could be provoked only
4634 by a command such as
4638 where an empty string was supplied; "<>" did not hit this bug.
4640 11. Installed PCRE release 4.5.
4642 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4643 remained set. It is now erased.
4645 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4646 times from message ids (which are base 36 rather than the normal 62).
4648 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4649 were being counted as actual protocol errors, and logged if the log
4650 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4651 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4652 and DATA following a set of rejected RCPTs do not count as protocol errors.
4653 In other words, Exim assumes they were pipelined, though this may not
4654 actually be the case. Of course, in all cases the client gets an
4655 appropriate error code.
4657 15. If a lookup fails in an ACL condition, a message about the failure may
4658 be available; it is used if testing the ACL cannot continue, because most
4659 such messages specify what the cause of the deferral is. However, some
4660 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4661 that caused an old message to be retained and used if a later statement
4662 caused a defer, replacing the real cause of the deferral.
4664 16. If an IP address had so many PTR records that the DNS lookup buffer
4665 was not large enough to hold them, Exim could crash while trying to process
4666 the truncated data. It now detects and logs this case.
4668 17. Further to 4.21/58, another change has been made: if (and only if) the
4669 first line of a message (the first header line) ends with CRLF, a bare LF
4670 in a subsequent header line has a space inserted after it, so as not to
4671 terminate the header.
4673 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4674 unnecessarily, used atoi() instead of strtol(), and didn't check the
4675 termination when getting file sizes from file names by regex.
4677 19. Completely re-implemented the support for maildirsize files, in the light
4678 of a number of problems with the previous contributed implementation
4679 (4.30/29). In particular:
4681 . If the quota is zero, the maildirsize file is maintained, but no quota is
4684 . If the maildir directory does not exist, it is created before any attempt
4685 to write a maildirsize file.
4687 . The quota value in the file is just a cache; if the quota is changed in
4688 the transport, the new value overrides.
4690 . A regular expression is available for excluding directories from the
4693 20. The autoreply transport checks the characters in options that define the
4694 message's headers; it allows continued headers, but it was checking with
4695 isspace() after an embedded newline instead of explicitly looking for a
4698 21. If all the "regular" hosts to which an address was routed had passed their
4699 expiry times, and had not reached their retry times, the address was
4700 bounced, even if fallback hosts were defined. Now Exim should go on to try
4703 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4704 equivalent code in the SMTP transport. Some hosts send humungous responses
4705 to HELO/EHLO, more than 1024 it seems.
4707 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4708 gives compiler warnings in some environments. I've now done it "properly",
4711 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4712 (because of problems with the built-in one) was declared to return uschar *
4713 instead of char *, causing compiler failure.
4715 25. Fixed a file descriptor leak when processing alias/forward files.
4717 26. Fixed a minor format string issue in dbfn.c.
4719 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4721 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4722 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4725 29. When a sender address is verified, it is cached, to save repeating the test
4726 when there is more than one recipient in a message. However, when the
4727 verification involves a callout, it is possible for different callout
4728 options to be set for different recipients. It is too complicated to keep
4729 track of this in the cache, so now Exim always runs a verification when a
4730 callout is required, relying on the callout cache for the optimization.
4731 The overhead is duplication of the address routing, but this should not be
4734 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4735 to be verified with callout=postmaster, and the main callout worked but the
4736 postmaster check failed, the verification correctly failed. However, if a
4737 subsequent RCPT command asked for sender verification *without* the
4738 postmaster check, incorrect caching caused this verification also to fail,
4741 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4742 it was not caching the DNS options (qualify_single, search_parents) that
4743 were used when the lookup failed. A subsequent lookup with different
4744 options therefore always gave the same answer, though there were cases
4745 where it should not have. (Example: a "domains = !$mx_any" option on a
4746 dnslookup router: the "domains" option is always processed without any
4747 widening, but the router might have qualify_single set.) Now Exim uses the
4748 cached value only when the same options are set.
4750 32. Added John Jetmore's "exipick" utility to the distribution.
4752 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4753 than a timeout (e.g. a certificate is required, and is not provided), an
4754 Exim server now closes the connection immediately. Previously it waited for
4755 the client to close - but if the client is SSL, it seems that they each
4756 wait for each other, leading to a delay before one of them times out.
4758 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4759 maintained 0.8.x compatibility because I don't think many are using it, and
4760 it is clearly obsolete.
4762 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4765 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4766 client certificate was expired. A simple patch fixes this, though I don't
4767 understand the full logic of why the verify callback is called multiple
4770 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4771 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4772 which causes problems with some clients (such as the Certicom SSL Plus
4773 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4774 disables the coutermeasure allowing Eudora to connect."
4776 38. Exim was not checking that a write() to a log file succeeded. This could
4777 lead to Bad Things if a log got too big, in particular if it hit a file
4778 size limit. Exim now panics and dies if it cannot write to a log file, just
4779 as it does if it cannot open a log file.
4781 39. Modified OS/Makefile-Linux so that it now contains
4783 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4785 The two -D definitions ensure that Exim is compiled with large file
4786 support, which makes it possible to handle log files that are bigger than
4789 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4790 instance) a domain was checked against a named list that involved a lookup,
4791 causing $domain_data to be set, then another domain was checked against the
4792 same list, then the first domain was re-checked, the value of $domain_data
4793 after the final check could be wrong. In particular, if the second check
4794 failed, it could be set empty. This bug probably also applied to
4797 41. The strip_trailing_dot option was not being applied to the address given
4798 with the -f command-line option.
4800 42. The code for reading a message's header from the spool was incrementing
4801 $received_count, but never initializing it. This meant that the value was
4802 incorrect (doubled) while delivering a message in the same process in which
4803 it was received. In the most common configuration of Exim, this never
4804 happens - a fresh exec is done - but it can happen when
4805 deliver_drop_privilege is set.
4807 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4808 it now includes up to 150 characters of the unexpected data in the log
4811 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4812 and building data strings. The size of both of these buffers was 10 000
4813 bytes - far larger than anybody would *ever* want, thought I. Needless to
4814 say, somebody hit the limit. I have increased the maximum line length to
4815 20 000 and the maximum data length of concatenated lines to 100 000. I have
4816 also fixed two bugs, because there was no checking on these buffers. Tsk,
4817 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4818 buffer is too small.
4820 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4821 lsearch lookups. Now it does.
4823 46. When parsing a route_list item in a manualroute router, a fixed-length
4824 buffer was used for the list of hosts. I made this 1024 bytes long,
4825 thinking that nobody would ever have a list of hosts that long. Wrong.
4826 Somebody had a whole pile of complicated expansion conditions, and the
4827 string was silently truncated, leading to an expansion error. It turns out
4828 that it is easier to change to an unlimited length (owing to other changes
4829 that have happened since this code was originally written) than to build
4830 structure for giving a limitation error. The length of the item that
4831 expands into the list of hosts is now unlimited.
4833 47. The lsearch lookup could not handle data where the length of text line was
4834 more than 4095 characters. Such lines were truncated, leading to shortened
4835 data being returned. It should now handle lines of any length.
4837 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4838 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4841 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4842 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4843 no longer happens in comment lines. A list of replacements is now placed
4844 at the head of all of the source files, except those whose only change is
4845 to replace PERL_COMMAND in the very first #! line.
4847 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4848 messages on the queue, with a bottom-up merge sort, using code contributed
4849 by Michael Haardt. This should make operations like -bp somewhat faster on
4850 large queues. It won't affect queue runners, except when queue_run_in_order
4853 51. Installed eximstats 1.31 in the distribution.
4855 52. Added support for SRV lookups to the dnslookup router.
4857 53. If an ACL referred to $message_body or $message_body_end, the value was not
4858 reset for any messages that followed in the same SMTP session.
4860 54. The store-handling optimization for building very long strings was not
4861 differentiating between the different store pools. I don't think this
4862 actually made any difference in practice, but I've tidied it.
4864 55. While running the routers to verify a sender address, $sender_address
4865 was still set to the sender address. This is wrong, because when routing to
4866 send a bounce to the sender, it would be empty. Therefore, I have changed
4867 it so that, while verifying a sender address, $sender_address is set to <>.
4868 (There is no change to what happens when verifying a recipient address.)
4870 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4871 target A or AAAA records (if not already returned) without resetting the
4872 qualify_single or search_parents options of the DNS resolver. These are
4873 inappropriate in this case because the targets of MX and SRV records must
4874 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4875 target that, when qualified, matched something in the local domain. These
4876 two options are now turned off when doing these lookups.
4878 57. It seems that at least some releases of Reiserfs (which does not have the
4879 concept of a fixed number of inodes) returns zero and not -1 for the
4880 number of available inodes. This interacted badly with check_spool_inodes,
4881 which assumed that -1 was the "no such thing" setting. What I have done is
4882 to check that the total number of inodes is greater than zero before doing
4883 the test of how many are available.
4885 58. When a "warn" ACL statement has a log_message modifier, the message is
4886 remembered, and not repeated. This is to avoid a lot of repetition when a
4887 message has many recipients that cause the same warning to be written.
4888 Howewer, Exim was preserving the list of already written lines for an
4889 entire SMTP session, which doesn't seem right. The memory is now reset if a
4890 new message is started.
4892 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4893 debugging output unless log_rewrite was also set.
4895 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4896 of (int)(handle) when we know that handle contains (void *)(-1).
4898 61. The Exim daemon panic-logs an error return when it closes the incoming
4899 connection. However "connection reset by peer" seems to be common, and
4900 isn't really an error worthy of noting specially, so that particular error
4903 62. When Exim is trying to find all the local interfaces, it used to panic and
4904 die if the ioctl to get the interface flags failed. However, it seems that
4905 on at least one OS (Solaris 9) it is possible to have an interface that is
4906 included in the list of interfaces, but for which you get a failure error
4907 for this call. This happens when the interface is not "plumbed" into a
4908 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4909 failure of the "get flags" call assumes that the interface is down.
4911 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4912 makes life easier for people who are doing arithmetic on fields extracted
4913 from dates, where you often get leading zeros that should not be
4914 interpreted as octal.
4916 64. Added qualify_domain to the redirect router, to override the global
4919 65. If a pathologically long header line contained very many addresses (the
4920 report of this problem mentioned 10 000) and each of them was rewritten,
4921 Exim could use up a very large amount of memory. (It kept on making new
4922 copies of the header line as it rewrote, and never released the old ones.)
4923 At the expense of a bit more processing, the header rewriting function has
4924 been changed so that it no longer eats memory in this way.
4926 66. The generation of the Received: header has been moved from the time that a
4927 message starts to be received, to the time that it finishes. The timestamp
4928 in the Received: header should now be very close to that of the <= log
4929 line. There are two side-effects of this change:
4931 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4932 logged header lines no longer include the local Received: line, because
4933 it has not yet been created. The same applies to a copy of the message
4934 that is returned to a non-SMTP sender when a message is rejected.
4936 (b) When a filter file is tested using -bf, no additional Received: header
4937 is added to the test message. After some thought, I decided that this
4940 This change does not affect the value of $received_for. It is still set
4941 after address rewriting, but before local_scan() is called.
4943 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4945 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4946 gave an unhelpful panic error message, and a defer error. I have managed to
4947 change this behaviour so that it now rejects any supplied certificate,
4948 which seems right, as the list of acceptable certificates is empty.
4950 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4951 gave an unhelpful defer error. I have not managed to make this reject any
4952 supplied certificates, but the error message it gives is "no certificate
4953 supplied", which is not helpful.
4955 70. exigrep's output now also includes lines that are not associated with any
4956 message, but which match the given pattern. Implemented by a patch from
4957 Martin Sluka, which also tidied up the Perl a bit.
4959 71. Recipient callout verification, like sender verification, was using <> in
4960 the MAIL FROM command. This isn't really the right thing, since the actual
4961 sender may affect whether the remote host accepts the recipient or not. I
4962 have changed it to use the actual sender in the callout; this means that
4963 the cache record is now keyed on a recipient/sender pair, not just the
4964 recipient address. There doesn't seem to be a real danger of callout loops,
4965 since a callout by the remote host to check the sender would use <>.
4966 [SEE ABOVE: changed after hitting problems.]
4968 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4969 temporary errors. However, in the case of such a code being given after
4970 the end of a data transmission (i.e. after ".") Exim was failing to write
4971 a retry record for the message. (Yes, there was some broken host that was
4972 actually sending 8xx at this point.)
4974 73. An unknown lookup type in a host list could cause Exim to panic-die when
4975 the list was checked. (An example that provoked this was putting <; in the
4976 middle of a list instead of at the start.) If this happened during a DATA
4977 ACL check, a -D file could be left lying around. This kind of configuration
4978 error no longer causes Exim to die; instead it causes a defer errror. The
4979 incident is still logged to the main and panic logs.
4981 74. Buglet left over from Exim 3 conversion. The message "too many messages
4982 in one connection" was written to the rejectlog but not the mainlog, except
4983 when address rewriting (yes!) was being logged.
4985 75. Added write_rejectlog option.
4987 76. When a system filter was run not as root (that is, when system_filter_user
4988 was set), the values of the $n variables were not being returned to the
4989 main process; thus, they were not subsequently available in the $sn
4992 77. Added +return_path_on_delivery log selector.
4994 78. A connection timeout was being treated differently from recipients deferred
4995 when testing hosts_max_try with a message that was older than the host's
4996 retry timeout. (The host should not be counted, thus allowing all hosts to
4997 be tried at least once before bouncing.) This may have been the cause of an
4998 occasionally reported bug whereby a message would remain on the queue
4999 longer than the retry timeout, but would be bounced if a delivery was
5000 forced. I say "may" because I never totally pinned down the problem;
5001 setting up timeout/retry tests is difficult. See also the next item.
5003 79. The ultimate address timeout was not being applied to errors that involved
5004 a combination of host plus message (for example, a timeout on a MAIL
5005 command). When an address resolved to a number of possible hosts, and they
5006 were not all tried for each delivery (e.g. because of hosts_max_try), a
5007 message could remain on the queue longer than the retry timeout.
5009 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5012 81. Fixed an obscure SMTP outgoing bug which required at least the following
5013 conditions: (a) there was another message waiting for the same server;
5014 (b) the server returned 5xx to all RCPT commands in the first message so
5015 that the message was not completed; (c) the server dropped the connection
5016 or gave a negative response to the RSET that Exim sends to abort the
5017 transaction. The observed case was a dropped connection after DATA that had
5018 been sent in pipelining mode. That is, the server had advertised PIPELINING
5019 but was not implementing it correctly. The effect of the bug was incorrect
5020 behaviour, such as trying another host, and this could lead to a crash.
5026 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5027 and daemon.c were passed as pointers to ints; they should have been
5028 pointers to socklen_t variables (which are typically unsigned ints).
5030 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5033 3. Fixed a really odd bug that affected only the testing scheme; patching a
5034 certain fixed string in the binary changed the value of another string that
5035 happened to be identical to the end of the original first string.
5037 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5038 name", it returns that address as the IP address. On some operating
5039 systems (e.g. Solaris), it also passes back the IP address string as the
5040 "host name". However, on others (e.g. Linux), it passes back an empty
5041 string. Exim wasn't checking for this, and was changing the host name to an
5042 empty string, assuming it had been canonicized.
5044 5. Although rare, it is permitted to have more than one PTR record for a given
5045 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5046 all the names associated with an address, because they do in Solaris.
5047 However, it seems that they do not in Linux for data that comes from the
5048 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5049 I found this out when I moved to a new Linux workstation and tried to run
5050 the Exim test suite.
5052 To get round this problem I have changed the code so that it now does its
5053 own call to the DNS to look up PTR records when searching for a host name.
5054 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5055 addresses that are only in /etc/hosts are still found.
5057 This behaviour is, however, controlled by an option called host_lookup_
5058 order, which defaults to "bydns:byaddr". If people want to use the other
5059 order, or indeed, just use one or the other means of lookup, they can
5060 specify it in this variable.
5062 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5063 some operating systems, this comes back from gethostbyaddr() as an empty
5064 string, and this is what Exim used to test for. However, it seems that in
5065 other systems, "." is yielded. Exim now tests for this case too.
5067 7. The values of check_spool_space and check_log_space are now held internally
5068 as a number of kilobytes instead of an absolute number of bytes. If a
5069 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5070 kilobyte. This means that much larger values can be stored.
5072 8. Exim monitor: an attempt to get the action menu when not actually pointing
5073 at a message produces an empty menu entitled "No message selected". This
5074 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5075 no entries in it ("Shell widget menu has zero width and/or height"). So I
5076 have added a single, blank menu entry in this case.
5078 9. Added ${quote_local_part.
5080 10. MIME decoding is now applied to the contents of Subject: header lines when
5083 11. Now that a reference to $sender_host_address automatically causes a reverse
5084 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5085 host lookup before query-style lookups in lists that might use this
5086 variable. This has therefore been abolished, and the "net-" prefix is no
5087 longer necessary for query-style lookups.
5089 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5090 have been a typo for LFLAGS, so it has been changed.
5092 13. The install script calls Exim with "-C /dev/null" in order to find the
5093 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5094 to be output. Howeve, since Exim outputs its version number before the
5095 error, it didn't break the script. It just looked ugly. I fixed this by
5096 always allowing "-C /dev/null" if the caller is root.
5098 14. Ignore overlarge ACL variable number when reading spool file - insurance
5099 against a later release with more variables having written the file.
5101 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5102 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5103 this, as well as the form without the "IPv6" on the front.
5105 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5106 OS/Makefile-Darwin file.
5108 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5109 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5111 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5112 variable. However, some OS use ints for inodes. I've added cast to long int
5113 to get rid of the compiler warning.
5115 19. I had forgotten to lock out "/../" in configuration file names when
5116 ALT_CONFIG_PREFIX was set.
5118 20. Routers used for verification do not need to specify transports. However,
5119 if such a router generated a host list, and callout was configured, Exim
5120 crashed, because it could not find a port number from the (non-existent)
5121 transport. It now assumes port 25 in this circumstance.
5123 21. Added the -t option to exigrep.
5125 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5126 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5127 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5128 them is set, LOOKUP_LSEARCH is forced.
5130 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5132 24. Added sender and host information to the "rejected by local_scan()" log
5133 line; previously there was no indication of these.
5135 25. Added .include_if_exists.
5137 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5138 Linux. It turns out that not all file systems support this. Apparently some
5139 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5140 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5141 not supported on the file descriptor, is now ignored when Exim is trying to
5142 sync a directory. This applies only to Linux.
5144 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5146 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5147 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5150 29. Added support for maildirsize files from supplied patch (modified a bit).
5152 30. The use of :fail: followed by an empty string could lead Exim to respond to
5153 sender verification failures with (e.g.):
5155 550 Verification failed for <xxx>
5156 550 Sender verify failed
5158 where the first response line was missing the '-' that indicates it is not
5159 the final line of the response.
5161 31. The loop for finding the name of the user that called Exim had a hardwired
5162 limit of 10; it now uses the value of finduser_retries, which is used for
5163 all other user lookups.
5165 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5168 33. Exim was neglecting to zero errno before one call of strtol() when
5169 expanding a string and expecting an integer value. On some systems this
5170 resulted in spurious "integer overflow" errors. Also, it was casting the
5171 result into an int without checking.
5173 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5174 did not work. The code looks as if it has *never* worked, though it appears
5175 to have been documented since at least releast 1.62. I have made it work.
5177 35. The "timeout_DNS" error in retry rules, also documented since at least
5178 1.62, also never worked. As it isn't clear exactly what this means, and
5179 clearly it isn't a major issue, I have abolished the feature by treating it
5180 as "timeout", and writing a warning to the main and panic logs.
5182 36. The display of retry rules for -brt wasn't always showing the error code
5185 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5186 timeout_connect_A, timeout_connect_MX.
5188 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5189 to the empty sender.
5191 39. The daemon was not analysing the content of -oX till after it had closed
5192 stderr and disconnected from the controlling terminal. This meant that any
5193 syntax errors were only noted on the panic log, and the return code from
5194 the command was 0. By re-arranging the code a little, I've made the
5195 decoding happen first, so such errors now appear on stderr, and the return
5196 code is 1. However, the actual setting up of the sockets still happens in
5197 the disconnected process, so errors there are still only recorded on the
5200 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5201 connections (as happens on some IP stacks) was logged at start up time as
5202 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5203 it from "IPv6 and IPv4", which means that two separate sockets are being
5206 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5207 says whether AF_INET or AF_INET6 was passed as an argument.
5209 42. Exiwhat output was messed up when time zones were included in log
5212 43. Exiwhat now gives more information about the daemon's listening ports,
5213 and whether -tls-on-connect was used.
5215 44. The "port" option of the smtp transport is now expanded.
5217 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5218 silently ignored. Now an error message is written to the main and panic
5221 46. There's a new ACL modifier called "logwrite" which writes to a log file
5222 as soon as it is encountered.
5224 47. Added $local_user_uid and $local_user_gid at routing time.
5226 48. Exim crashed when trying to verify a sender address that was being
5229 49. Exim was recognizing only a space character after ".include". It now also
5230 recognizes a tab character.
5232 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5233 extracting the relevant information from the specification. The man page no
5234 longer contains scrambled data for the -d option, and I've added a section
5235 at the front about calling Exim under different names.
5237 51. Added "extra_headers" argument to the "mail" command in filter files.
5239 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5242 53. Installed eximstats 1.29.
5244 54. Added transport_filter_timeout as a generic transport option.
5246 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5247 Cc: header lines. This was required by RFC 822, but it not required by RFC
5250 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5251 incoming messages that did not have them. Now it does so only if the
5252 message originates locally, that is, if there is no associated remote host
5253 address. When Resent- header lines are present, this applies to the Resent-
5254 lines rather than the non-Resent- lines.
5256 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5257 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5259 58. Messages for configuration errors now include the name of the main
5260 configuration file - useful now that there may be more than one file in a
5261 list (.included file names were always shown).
5263 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5264 for those rare installations that do not start the daemon as root or run it
5265 setuid root. I've cut out the call to initgroups() if the daemon is not
5268 60. The Exim user and group can now be bound into the binary as text strings
5269 that are looked up at the start of Exim's processing.
5271 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5273 62. Added $mailstore_basename variable.
5275 63. Installed patch to sieve.c from Michael Haardt.
5277 64. When Exim failed to open the panic log after failing to open the main log,
5278 the original message it was trying to log was written to stderr and debug
5279 output, but if they were not available (the usual case in production), it
5280 was lost. Now it is written to syslog before the two lines that record the
5281 failures to open the logs.
5283 65. Users' Exim filters run in subprocesses under the user's uid. It is
5284 possible for a "deliver" command or an alias in a "personal" command to
5285 provoke an address rewrite. If logging of address rewriting is configured,
5286 this fails because the process is not running as root or exim. There may be
5287 a better way of dealing with this, but for the moment (because 4.30 needs
5288 to be released), I have disabled address rewrite logging when running a
5289 filter in a non-root, non-exim process.
5295 1. The buildconfig auxiliary program wasn't quoting the value set for
5296 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5297 not defined. This bug was masked in 4.22 by the effect that was fixed in
5300 2. Some messages that were rejected after a message id was allocated were
5301 shown as "incomplete" by exigrep. It no longer does this for messages that
5302 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5304 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5305 have allow_domain_literals set, the ID did not get logged in the <= line.
5306 Domain literals are now always recognized in Message-ID: header lines.
5308 4. The first argument for a ${extract expansion item is the key name or field
5309 number. Leading and trailing spaces in this item were not being ignored,
5310 causing some misleading effects.
5312 5. When deliver_drop_privilege was set, single queue runner processes started
5313 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5314 same command in the process it spins off) were not dropping privilege.
5316 6. When the daemon running as "exim" started a queue runner, it always
5317 re-executed Exim in the spun-off process. This is a waste of effort when
5318 deliver_drop_privilege is set. The new process now just calls the
5319 queue-runner function directly.
5325 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5328 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5329 ignored. Though the use of -oP was forcing the writing of a pid file, it
5330 was always written to the default place.
5332 3. If the message "no IP address found for host xxxx" is generated during
5333 incoming verification, it is now followed by identification of the incoming
5334 connection (so you can more easily find what provoked it).
5336 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5338 5. Added some features to "harden" Exim a bit more against certain attacks:
5340 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5341 be put in Local/Makefile. This is like the never_users runtime option,
5342 but it cannot be overridden. The default setting is "root".
5344 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5345 prefix string with which any file named in a -C command line option
5348 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5349 is retained for -C and -D only if the caller of Exim is root. Without
5350 it, the exim user may also use -C and -D and retain privilege.
5352 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5353 command line option is disabled.
5355 6. Macro names set by the -D option must start with an upper case letter, just
5356 like macro names defined in the configuration file.
5358 7. Added "dereference=" facility to LDAP.
5360 8. Two instances of the typo "uknown" in the source files are fixed.
5362 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5363 the Configure-Makefile script screwed up while processing it.
5365 10. Incorporated PCRE 4.4.
5367 11. The SMTP synchronization check was not operating right at the start of an
5368 SMTP session. For example, it could not catch a HELO sent before the client
5369 waited for the greeting. There is now a check for outstanding input at the
5370 point when the greeting is written. Because of the duplex, asynchronous
5371 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5372 way, but not yet received, when the check is performed.
5374 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5375 on TCP/IP sockets, because this apparently causes some broken clients to
5378 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5379 unchanged) from the Cygwin maintainer.
5381 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5382 was supported instead of testing for mbx. Effectively a typo.
5384 15. The spa authenticator server code was not checking that the input it
5385 received was valid base64.
5387 16. The debug output line for the "set" modifier in ACLs was not showing the
5388 name of the variable that was being set.
5390 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5392 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5393 lookup on its own. Something else was needed to trigger the lookup. For
5394 example, a match in host_lookup or the need for a host name in a host list.
5395 Now, if $sender_host_name is referenced and the host name has not yet been
5396 looked up, a lookup is performed. If the lookup fails, the variable remains
5397 empty, and $host_lookup_failed is set to "1".
5399 19. Added "eqi" as a case-independent comparison operator.
5401 20. The saslauthd authentication condition could segfault if neither service
5402 nor realm was specified.
5404 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5405 error message that was logged was misleading, and incoming SMTP
5406 connections were dropped. The message is now more accurate, and temporary
5407 errors are given to SMTP connections.
5409 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5410 (except RSET) with a 421 error, until QUIT is received. However, it was
5411 failing to send a response to QUIT.
5413 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5414 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5415 ACL is now run after the helo_try_verify_hosts code.
5417 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5418 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5419 case-independent, so other case variants are also recognized). Apparently
5420 some systems use these upper case variants.
5422 25. If more than two messages were waiting for the same host, and a transport
5423 filter was specified for the transport, Exim sent two messages over the
5424 same TCP/IP connection, and then failed with "socket operation on non-
5425 socket" when it tried to send the third.
5427 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5429 27. The extern definition of crypt16() in expand.c was not being excluded when
5430 the OS had its own crypt16() function.
5432 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5433 as a preferred synonym for return_size_limit, both as an option and as an
5436 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5438 30. Changed the default configuration ACL to relax the local part checking rule
5439 for addresses that are not in any local domains. For these addresses,
5440 slashes and pipe symbols are allowed within local parts, but the sequence
5441 /../ is explicitly forbidden.
5443 31. SPA server authentication was not clearing the challenge buffer before
5446 32. log_message in a "warn" ACL statement was writing to the reject log as
5447 well as to the main log, which contradicts the documentation and doesn't
5448 seem right (because no rejection is happening). So I have stopped it.
5450 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5451 However, I am unable to do any testing of this.
5453 34. Fixed an infelicity in the appendfile transport. When checking directories
5454 for a mailbox, to see if any needed to be created, it was accidentally
5455 using path names with one or more superfluous leading slashes; tracing
5456 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5458 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5459 contents are added to the log line that is written for every discarded
5460 recipient. (Previously a log_message setting was ignored.)
5462 36. The ${quote: operator now quotes the string if it is empty.
5464 37. The install script runs exim in order to find its version number. If for
5465 some reason other than non-existence or emptiness, which it checks, it
5466 could not run './exim', it was installing it with an empty version number,
5467 i.e. as "exim-". This error state is now caught, and the installation is
5470 38. An argument was missing from the function that creates an error message
5471 when Exim fails to connect to the socket for saslauthd authentication.
5472 This could cause Exim to crash, or give a corrupted message.
5474 39. Added isip, isip4, and isip6 to ${if conditions.
5476 40. The ACL variables $acl_xx are now saved with the message, and can be
5477 accessed later in routers, transports, and filters.
5479 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5480 strings in the file are not string-expanded.
5482 42. If a MAIL command specified a SIZE value that was too large to fit into an
5483 int variable, the check against message_size_limit failed. Such values are
5484 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5485 day this will have to be increased, but I don't think I want to be around
5486 when emails are that large.
5493 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5494 iconv() is not standard in FreeBSD.
5496 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5497 IPv6 enabled. The observed symptom was a segmentation fault on return from
5498 the function os_common_find_running_interfaces() in src/os.c.
5500 3. In the check_special_case() function in daemon.c I had used "errno" as an
5501 argument name, which causes warnings on some systems. This was basically a
5502 typo, since it was named "eno" in the comments!
5504 4. The code that waits for the clock to tick (at a resolution of some fraction
5505 of a second) so as to ensure message-id uniqueness was always waiting for
5506 at least one whole tick, when it could have waited for less. [This is
5507 almost certainly not relevant at current processor speeds, where it is
5508 unlikely to ever wait at all. But we try to future-proof.]
5510 5. The function that sleeps for a time interval that includes fractions of a
5511 second contained a race. It did not block SIGALRM between setting the
5512 timer, and suspending (a couple of lines later). If the interval was short
5513 and the sigsuspend() was delayed until after it had expired, the suspension
5514 never ended. On busy systems this could lead to processes getting stuck for
5517 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5518 process, before it forks any delivery processes. The open lookup caching
5519 mechanism meant that the open file or database connection was passed into
5520 the delivery process. The problem was that delivery processes always tidy
5521 up cached lookup data. This could cause a problem for the next delivery
5522 process started by the queue runner, because the external queue runner
5523 process does not know about the closure. So the next delivery process
5524 still has data in the lookup cache. In the case of a file lookup, there was
5525 no problem because closing a file descriptor in a subprocess doesn't affect
5526 the parent. However, if the lookup was caching a connection to a database,
5527 the connection was closed, and the second delivery process was likely to
5528 see errors such as "PGSQL: query failed: server closed the connection
5529 unexpectedly". The problem has been fixed by closing all cached lookups
5530 in a queue runner before running a delivery process.
5532 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5533 seem to have the "const" qualifier which it has on other OS. I've
5536 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5537 *of the same type* (client or server) with the same public name that an
5538 error should be diagnosed.
5540 9. When Exim looked up a host name for an IP address, but failed to find the
5541 original IP address when looking up the host name (a safety check), it
5542 output the message "<ip address> does not match any IP for NULL", which was
5543 confusing, to say the least. The bug was that the host name should have
5544 appeared instead of "NULL".
5546 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5547 user that is built into the binary, and the -C or -D options is used, root
5548 privilege is dropped before the configuration file is read. In addition,
5549 logging is switched to stderr instead of the normal log files. If the
5550 configuration then re-defines the Exim user, the unprivileged environment
5551 is probably not what is expected, so Exim logs a panic warning message (but
5554 However, if deliver_drop_privilege is set, the unprivileged state may well
5555 be exactly what is intended, so the warning has been cut out in that case,
5556 and Exim is allowed to try to write to its normal log files.
5562 1. smtp_return_error_details was not giving details for temporary sender
5563 or receiver verification errors.
5565 2. Diagnose a configuration error if two authenticators have the same public
5568 3. Exim used not to create the message log file for a message until the first
5569 delivery attempt. This could be confusing when incoming messages were held
5570 for policy or load reasons. The message log file is now created at the time
5571 the message is received, and an initial "Received" line is written to it.
5573 4. The automatically generated man page for command line options had a minor
5574 bug that caused no ill effects; however, a more serious problem was that
5575 the procedure for building the man page automatically didn't always
5576 operate. Consequently, release 4.20 contains an out-of-date version. This
5577 shouldn't happen again.
5579 5. When building Exim with embedded Perl support, the script that builds the
5580 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5581 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5583 6. The freeze_tell option was not being used for messages that were frozen on
5584 arrival, either by an ACL or by local_scan().
5586 7. Added the smtp_incomplete_transaction log selector.
5588 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5589 was accepting AUTH without a new EHLO.
5591 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5592 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5595 10. Logging of TCP/IP connections (when configured) now happens in the main
5596 daemon process instead of the child process, so that the TCP/IP connection
5597 count is more accurate (but it can never be perfect).
5599 11. The use of "drop" in a nested ACL was not being handled correctly in the
5600 outer ACL. Now, if condition failure induced by the nested "drop" causes
5601 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5602 or "require"), the connection is dropped.
5604 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5605 that yield "discard" can now be used with an "accept" or a "discard" verb,
5606 but an error is generated for any others (because I can't see a useful way
5607 to define what should happen).
5609 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5610 are now processed in the same way as lines in the Exim configuration file.
5611 In particular, continuation lines are supported.
5613 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5615 15. Added -ti meaning -t -i.
5617 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5618 domains, and warn by logging if others are found.
5620 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5621 returned from GIFCONF when Exim is trying to find the list of interfaces on
5622 a host. The code in os.c has been modified to copy each ifreq to an aligned
5623 structure in all cases.
5625 Also, in some cases, the returned ifreq's were being copied to a 'struct
5626 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5627 means the last couple of bytes of an IPv6 address could be chopped if the
5628 ifreq contained only a normal sockaddr (14 bytes storage).
5630 18. Named domain lists were not supported in the hosts_treat_as_local option.
5631 An entry such as +xxxx was not recognized, and was treated as a literal
5634 19. Ensure that header lines added by a DATA ACL are included in the reject log
5635 if the ACL subsequently rejects the message.
5637 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5638 MD5 (which is deprecated).
5640 21. When testing a filter file using -bf, Exim was writing a message when it
5641 took the sender from a "From " line in the message, but it was not doing so
5642 when it took $return_path from a Return-Path: header line. It now does.
5644 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5645 with a valid header line field name (a series of printing characters
5646 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5648 23. Changed "disc" in the source to "disk" to conform to the documentation and
5649 the book and for uniformity.
5651 24. Ignore Sendmail's -Ooption=value command line item.
5653 25. When execve() failed while trying to run a command in a pipe transport,
5654 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5655 could be confused with a return value of 69 from the command itself. This
5656 has been changed to 127, the value the shell returns if it is asked to run
5657 a non-existent command. The wording for the related log line suggests a
5658 non-existent command as the problem.
5660 26. If received_header_text expands to an empty string, do not add a Received:
5661 header line to the message. (Well, it adds a token one on the spool, but
5662 marks it "old" so that it doesn't get used or transmitted.)
5664 27. Installed eximstats 1.28 (addition of -nt option).
5666 28. There was no check for failure on the call to getsockname() in the daemon
5667 code. This can fail if there is a shortage of resources on the system, with
5668 ENOMEM, for example. A temporary error is now given on failure.
5670 29. Contrary to the C standard, it seems that in some environments, the
5671 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5672 program. Exim now does this explicitly; it affects the formatting of
5673 timestamps using strftime().
5675 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5676 complaints. I've now initialized all the variables, to avoid this.
5678 32. Header lines added by a system filter were not being "seen" during
5679 transport-time rewrites.
5681 33. The info_callback() function passed to OpenSSL is set up with type void
5682 (*)(SSL *, int, int), as described somewhere. However, when calling the
5683 function (actually a macro) that sets it up, the type void(*)() is
5684 expected. I've put in a cast to prevent warnings from picky compilers.
5686 34. If a DNS black list lookup found a CNAME record, but there were no A
5687 records associated with the domain it pointed at, Exim crashed.
5689 35. If a DNS black list lookup returned more than one A record, Exim ignored
5690 all but the first. It now scans all returned addresses if a particular IP
5691 value is being sought. In this situation, the contents of the
5692 $dnslist_value variable are a list of all the addresses, separated by a
5695 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5696 to accept a host name and all its aliases if the forward lookup for any of
5697 them yielded the IP address of the incoming connection. Now it accepts only
5698 those names whose forward lookup yields the correct IP address. Any other
5699 names are discarded. This closes a loophole whereby a rogue DNS
5700 administrator could create reverse DNS records to break through a
5701 wildcarded host restriction in an ACL.
5703 37. If a user filter or a system filter that ran in a subprocess used any of
5704 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5705 the wrong values were passed to the pipe command ($thisaddress had the
5706 value of $0, $0 had the value of $1, etc). This bug was introduced by
5707 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5709 38. Improved the line breaking for long SMTP error messages from ACLs.
5710 Previously, if there was no break point between 40 and 75 characters, Exim
5711 left the rest of the message alone. Two changes have been made: (a) I've
5712 reduced the minimum length to 35 characters; (b) if it can't find a break
5713 point between 35 and 75 characters, it looks ahead and uses the first one
5714 that it finds. This may give the occasional overlong line, but at least the
5715 remaining text gets split now.
5717 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5718 file descriptors might be low, and that setting 1000 would always raise it.
5719 It turns out that in some environments, the limit is already over 1000 and
5720 that lowering it causes trouble. So now Exim takes care not to decrease it.
5722 40. When delivering a message, the value of $return_path is set to $sender_
5723 address at the start of routing (routers may change the value). By an
5724 oversight, this default was not being set up when an address was tested by
5725 -bt or -bv, which affected the outcome if any router or filter referred to
5728 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5729 log only when "message" or "log_message" are set. However, if one of the
5730 conditions was an address verification, or a call to a nested ACL, the
5731 messages generated by the underlying test were being passed through. This
5732 no longer happens. The underlying message is available in $acl_verify_
5733 message for both "message" and "log_message" expansions, so it can be
5734 passed through if needed.
5736 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5737 new expansion $bh_ to give the encoded byte string without charset
5738 translation. Translation happens only if iconv() is available; HAVE_ICONV
5739 indicates this at build time. HEADERS_CHARSET gives the charset to
5740 translate to; headers_charset can change it in the configuration, and
5741 "headers charset" can change it in an individual filter file.
5743 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5744 that creates RFC 2047 encoded "words" labels them as that charset instead
5745 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5746 expansion operator; (ii) when Exim creates a From: line for a local
5747 message; (iii) when a header line is rewritten to include a "phrase" part.
5749 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5750 buggy, causing it to skip the first lines of messages whose message ID
5751 ended in 'D'. This would not have bitten before Exim release 4.14, because
5752 message IDs were unlikely to end in 'D' before then. The effect was to have
5753 incorrect size information for certain domains.
5755 45. #include "config.h" was missing at the start of the crypt16.c module. This
5756 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5759 46. If there was a timeout during a "random" callout check, Exim treated it as
5760 a failure of the random address, and carried on sending RSET and the real
5761 address. If the delay was just some slowness somewhere, the response to the
5762 original RCPT would be taken as a response to RSET and so on, causing
5763 mayhem of various kinds.
5765 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5766 when I implemented it. It didn't allow for the fact that some option values
5767 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5768 the right test for positive values.
5770 48. Domain names in DNS records are case-independent. Exim always looks them up
5771 in lower case. Some resolvers return domain names in exactly the case they
5772 appear in the zone file, that is, they may contain uppercase letters. Not
5773 all resolvers do this - some return always lower case. Exim was treating a
5774 change of case by a resolver as a change of domain, similar to a widening
5775 of a domain abbreviation. This triggered its re-routing code and so it was
5776 trying to route what was effectively the same domain again. This normally
5777 caused routing to fail (because the router wouldn't handle the domain
5778 twice). Now Exim checks for this case specially, and just changes the
5779 casing of the domain that it ultimately uses when it transmits the message
5782 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5785 50. If a filter generated a file delivery with a non-absolute name (possible if
5786 no home directory exists for the router), the forbid_file option was not
5789 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5790 the existing equality matching.
5792 52. Exim was using ints instead of ino_t variables in some places where it was
5793 dealing with inode numbers.
5795 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5796 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5797 called TMPDIR, and if it finds it is different, it changes its value.
5799 54. The smtp_printf() function is now made available to local_scan() so
5800 additional output lines can be written before returning. There is also an
5801 smtp_fflush() function to enable the detection of a dropped connection.
5802 The variables smtp_input and smtp_batched_input are exported to
5805 55. Changed the default runtime configuration: the message "Unknown user"
5806 has been removed from the ACL, and instead placed on the localuser router,
5807 using the cannot_route_message feature. This means that any verification
5808 failures that generate their own messages won't get overridden. Similarly,
5809 the "Unrouteable address" message that was in the ACL for unverifiable
5810 relay addresses has also been removed.
5812 56. Added hosts_avoid_esmtp to the smtp transport.
5814 57. The exicyclog script was not checking for the esoteric option
5815 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5816 will work only if exicyclog is run under the appropriate euid.
5818 58. Following a discussion on the list, the rules by which Exim recognises line
5819 endings on incoming messages have been changed. The -dropcr and drop_cr
5820 options are now no-ops, retained only for backwards compatibility. The
5821 following line terminators are recognized: LF CRLF CR. However, special
5822 processing applies to CR:
5824 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5825 nor a local message in the state where . is a terminator.
5827 (ii) If a bare CR is encountered in a header line, an extra space is added
5828 after the line terminator so as not to end the header. The reasoning
5829 behind this is that bare CRs in header lines are most likely either
5830 to be mistakes, or people trying to play silly games.
5832 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5833 was being incorrectly given as 18 bytes larger than it should have been.
5834 This is a VOB (very old bug).
5836 60. This may never have affected anything current, but just in case it has:
5837 When the local host is found other than at the start of a list of hosts,
5838 the local host, those with the same MX, and any that follow, are discarded.
5839 When the list in question was part of a longer list of hosts, the following
5840 hosts (not currently being processed) were also being discarded. This no
5841 longer happens. I'm not sure if this situation could ever has previously
5844 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5847 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5848 header. This is something that is recommended in a new Internet Draft, and
5849 is something that is documented as being done by Sendmail. There are two
5850 possible values. For messages generated by the autoreply transport, Exim
5853 Auto-Submitted: auto-replied
5855 whereas for all other generated messages (e.g. bounces) it adds
5857 Auto-Submitted: auto-generated
5859 63. The "personal" condition in filters now includes a test for the
5860 Auto-Submitted: header. If it contains the string "auto-" the message it
5861 not considered personal.
5863 64. Added rcpt_include_affixes as a generic transport option.
5865 65. Added queue_only_override (default true).
5867 66. Added the syslog_duplication option.
5869 67. If what should have been the first header line of a message consisted of
5870 a space followed by a colon, Exim was mis-interpreting it as a header line.
5871 It isn't of course - it is syntactically invalid and should therefore be
5872 treated as the start of the message body. The misbehaviour could have
5873 caused a number of strange effects, including loss of data in subsequent
5874 header lines, and spool format errors.
5876 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5877 client host had authenticated. This control can now be exercised by an ACL
5878 for more flexibility.
5880 69. By default, callouts do not happen when testing with -bh. There is now a
5881 variant, -bhc, which does actually run the callout code, including
5882 consulting and updating the callout cache.
5884 70. Added support for saslauthd authentication, courtesy of Alexander
5887 71. If statvfs() failed on the spool or log directories while checking their
5888 size for availability, Exim confusingly gave the error "space shortage".
5889 Furthermore, in debugging mode it crashed with a floating point exception.
5890 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5891 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5892 serious problem, Exim now writes to the main and panic logs when this
5893 happens, with details of the failure. It then refuses to accept the
5894 incoming message, giving the message "spool directory problem" or "log
5895 directory problem" with a 421 code for SMTP messages.
5897 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5898 0, 1, and 2 exist, because some OS complain for execs without them (see
5899 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5900 descriptors. However, the code omitted to check that the open succeeded,
5901 causing mysterious errors if for some reason the permissions on /dev/null
5902 got screwed. Now Exim writes a message to the main and panic logs, and
5903 bombs out if it can't open /dev/null.
5905 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5906 interact so that it is all more flexible. It is supposed to remain
5907 backwards compatible. Also added extra_local_interfaces.
5909 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5910 to bomb out with an assertion failure - to the client this appears as a
5911 connection drop. This problem occurs in the part of the code that was taken
5912 from the Samba project. Fortunately, the assertion is in a very simple
5913 function, so I have fixed this by reproducing the function inline in the
5914 one place where it is called, and arranging for authentication to fail
5915 instead of killing the process with assert().
5917 75. The SPA client code was not working when the server requested OEM rather
5918 than Unicode encoding.
5920 76. Added code to make require_files with a specific uid setting more usable in
5921 the case where statting the file as root fails - usually a non-root-mounted
5922 NFS file system. When this happens and the failure is EACCES, Exim now
5923 forks a subprocess and does the per-uid checking as the relevant uid.
5925 77. Added process_log_path.
5927 78. If log_file_path was not explicitly set, a setting of check_log_space or
5928 check_log_inodes was ignored.
5930 79. If a space check for the spool or log partitions fails, the incident is now
5931 logged. Of course, in the latter case the data may get lost...
5933 80. Added the %p formatting code to string_format() so that it can be used to
5934 print addresses in debug_print(). Adjusted all the address printing in the
5935 debugging in store.c to use %p rather than %d.
5937 81. There was a concern that a line of code in smtp_in.c could overflow a
5938 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5939 initially expressed, the concern was not well-founded, because trailing
5940 spaces are removed early. However, if the trailing spaces were followed by
5941 a NULL, they did not get removed, so the overflow was possible. Two fixes
5944 (a) I re-wrote the offending code in a cleaner fashion.
5945 (b) If an incoming SMTP command contains a NULL character, it is rejected
5948 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5949 runs initgroups(), so that if the Exim user is in any additional groups,
5950 they will be used during message reception.
5956 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob