1 Change log file for Exim from version 4.21
2 -------------------------------------------
8 TF/01 Correctly close the server side of TLS when forking for delivery.
10 When a message was received over SMTP with TLS, Exim failed to clear up
11 the incoming connection properly after forking off the child process to
12 deliver the message. In some situations the subsequent outgoing
13 delivery connection happened to have the same fd number as the incoming
14 connection previously had. Exim would try to use TLS and fail, logging
15 a "Bad file descriptor" error.
17 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
18 utilities have not been installed.
20 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
21 temporary space as the ACL may create new global variables.
23 TL/01 LDAP support uses per connection or global context settings, depending
24 upon the detected version of the libraries at build time.
26 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
27 to extract and use the src ip:port in logging and expansions as if it
28 were a direct connection from the outside internet.
30 JH/02 Add ${listextract {number}{list}{success}{fail}}.
32 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
33 Properly escape header and check for NULL return.
35 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
38 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
40 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
41 characters in header names, implemented as a verify condition.
42 Contributed by Michael Fischer v. Mollard.
44 TL/05 Rename SPF condition results err_perm and err_temp to standardized
45 results permerror and temperror. Is a backward incompatibility if
46 the ACL tests for either of these two results. Patch contributed by
47 user bes-internal on the mailing list.
49 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
51 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
52 selectors, in both main and reject logs.
54 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
57 JH/07 Add malware type "sock" for talking to simple daemon.
59 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
61 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
62 routers/transports under cutthrough routing.
64 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
65 numbers. Touch up "bool" conditional to keep the same definition.
67 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
69 JH/12 Expand items in router/transport headers_add or headers_remove lists
70 individually rather than the list as a whole. Bug 1452.
72 Required for reasonable handling of multiple headers_ options when
73 they may be empty; requires that headers_remove items with embedded
74 colons must have them doubled (or the list-separator changed).
80 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
82 PP/02 Make -n do something, by making it not do something.
83 When combined with -bP, the name of an option is not output.
85 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
88 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
89 $sender_host_name and config options to manage this, and basic check
92 PP/05 DSCP support for outbound connections and control modifier for inbound.
94 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
95 (Only plugin which currently uses this is kerberos4, which nobody should
96 be using, but we should make it available and other future plugins might
97 conceivably use it, even though it would break NAT; stuff *should* be
98 using channel bindings instead).
100 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
101 name; added for Sendmail compatibility; requires admin caller.
102 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
103 just ignore it); requires trusted caller.
104 Also parse but ignore: -Ac -Am -X<logfile>
107 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
109 TL/02 Add +smtp_confirmation as a default logging option.
111 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
112 Patch by Magnus Holmgren from 2007-02-20.
114 TL/04 Bugzilla 1281 - Spec typo.
115 Bugzilla 1283 - Spec typo.
116 Bugzilla 1290 - Spec grammar fixes.
118 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
120 TL/06 Add Experimental DMARC support using libopendmarc libraries.
122 TL/07 Fix an out of order global option causing a segfault. Reported to dev
123 mailing list by by Dmitry Isaikin.
125 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
127 JH/02 Support "G" suffix to numbers in ${if comparisons.
129 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
131 NM/01 Bugzilla 1197 - Spec typo
132 Bugzilla 1196 - Spec examples corrections
134 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
136 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
137 gnutls_enable_pkcs11, but renamed to more accurately indicate its
140 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
141 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
143 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
144 "acl {{name}{arg}...}", and optional args on acl condition
147 JH/05 Permit multiple router/transport headers_add/remove lines.
149 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
151 JH/07 Avoid using a waiting database for a single-message-only transport.
152 Performance patch from Paul Fisher. Bugzilla 1262.
154 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
157 JH/09 Add $headers_added variable, with content from use of ACL modifier
158 add_header (but not yet added to the message). Bugzilla 199.
160 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
161 Pulled from Bugzilla 817 by Wolfgang Breyha.
163 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
165 (nb: this is the same fix as in Exim 4.80.1)
167 JH/11 Add A= logging on delivery lines, and a client_set_id option on
170 JH/12 Add optional authenticated_sender logging to A= and a log_selector
173 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
175 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
176 advertise SMTP AUTH mechanism to us, instead of a generic
177 protocol violation error. Also, make Exim more robust to bad
178 data from the Dovecot auth socket.
180 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
182 When a queue runner is handling a message, Exim first routes the
183 recipient addresses, during which it prunes them based on the retry
184 hints database. After that it attempts to deliver the message to
185 any remaining recipients. It then updates the hints database using
188 So if a recipient address works intermittently, it can get repeatedly
189 deferred at routing time. The retry hints record remains fresh so the
190 address never reaches the final cutoff time.
192 This is a fairly common occurrence when a user is bumping up against
193 their storage quota. Exim had some logic in its local delivery code
194 to deal with this. However it did not apply to per-recipient defers
195 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
197 This change adds a proper retry rule check during routing so that the
198 final cutoff time is checked against the message's age. We only do
199 this check if there is an address retry record and there is not a
200 domain retry record; this implies that previous attempts to handle
201 the address had the retry_use_local_parts option turned on. We use
202 this as an approximation for the destination being like a local
203 delivery, as in LMTP.
205 I suspect this new check makes the old local delivery cutoff check
206 redundant, but I have not verified this so I left the code in place.
208 TF/02 Correct gecos expansion when From: is a prefix of the username.
210 Test 0254 submits a message to Exim with the header
214 When I ran the test suite under the user fanf2, Exim expanded
215 the header to contain my full name, whereas it should have added
216 a Resent-Sender: header. It erroneously treats any prefix of the
217 username as equal to the username.
219 This change corrects that bug.
221 GF/01 DCC debug and logging tidyup
222 Error conditions log to paniclog rather than rejectlog.
223 Debug lines prefixed by "DCC: " to remove any ambiguity.
225 TF/03 Avoid unnecessary rebuilds of lookup-related code.
227 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
228 Bug spotted by Jeremy Harris; was flawed since initial commit.
229 Would have resulted in OCSP responses post-SNI triggering an Exim
230 NULL dereference and crash.
232 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
234 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
235 Bug detection, analysis and fix by Samuel Thibault.
236 Bugzilla 1331, Debian bug #698092.
238 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
240 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
241 Server implementation by Todd Lyons, client by JH.
242 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
243 config variable "prdr_enable" controls whether the server
244 advertises the facility. If the client requests PRDR a new
245 acl_data_smtp_prdr ACL is called once for each recipient, after
246 the body content is received and before the acl_smtp_data ACL.
247 The client is controlled by bolth of: a hosts_try_prdr option
248 on the smtp transport, and the server advertisement.
249 Default client logging of deliveries and rejections involving
250 PRDR are flagged with the string "PRDR".
252 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
253 fclose(). Diagnosis by Todd Lyons.
255 PP/17 Update configure.default to handle IPv6 localhost better.
256 Patch by Alain Williams (plus minor tweaks).
259 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
260 This is now consistent with GnuTLS, and is now documented: the
261 previous undocumented portable approach to treating the option as
262 unset was to force an expansion failure. That still works, and
263 an empty string is now equivalent.
265 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
266 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
267 not performing validation itself.
269 PP/20 Added force_command boolean option to pipe transport.
270 Patch from Nick Koston, of cPanel Inc.
272 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
275 TF/04 Added udpsend ACL modifer and hexquote expansion operator
277 PP/21 Fix eximon continuous updating with timestamped log-files.
278 Broken in a format-string cleanup in 4.80, missed when I repaired the
279 other false fix of the same issue.
280 Report and fix from Heiko Schlichting.
283 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
284 Report from Prashanth Katuri.
286 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
287 It's SecureTransport, so affects any MacOS clients which use the
288 system-integrated TLS libraries, including email clients.
290 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
291 using a MIME ACL for non-SMTP local injection.
292 Report and assistance in diagnosis by Warren Baker.
294 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
296 JH/16 Fix comparisons for 64b. Bugzilla 1385.
298 TL/09 Add expansion variable $authenticated_fail_id to keep track of
299 last id that failed so it may be referenced in subsequent ACL's.
301 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
304 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
305 ldap library initialization, allowing self-signed CA's to be
306 used. Also properly sets require_cert option later in code by
307 using NULL (global ldap config) instead of ldap handle (per
308 session). Bug diagnosis and testing by alxgomz.
310 TL/12 Enhanced documentation in the ratelimit.pl script provided in
311 the src/util/ subdirectory.
313 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
314 renamed to Transport Post Delivery Action by Jeremy Harris, as
317 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
318 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
319 redis_servers = needs to be configured which will be used by the redis
320 lookup. Patch from Warren Baker, of The Packet Hub.
322 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
324 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
325 hostname or reverse DNS when processing a host list. Used suggestions
326 from multiple comments on this bug.
328 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
330 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
331 Missed a few lines, added it to make the runtest require no keyboard
334 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
335 contains upper case chars. Make router use caseful_local_part.
337 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
338 support when GnuTLS has been built with p11-kit.
344 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
346 This, or similar/improved, will also be change PP/11 of 4.82.
352 PP/01 Handle short writes when writing local log-files.
353 In practice, only affects FreeBSD (8 onwards).
354 Bugzilla 1053, with thanks to Dmitry Isaikin.
356 NM/01 Bugzilla 949 - Documentation tweak
358 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
361 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
363 PP/02 Implemented gsasl authenticator.
365 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
367 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
368 `pkg-config foo` for cflags/libs.
370 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
371 with rest of GSASL and with heimdal_gssapi.
373 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
374 `pkg-config foo` for cflags/libs for the TLS implementation.
376 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
377 properties get this fed in as external SSF. A number of robustness
378 and debugging improvements to the cyrus_sasl authenticator.
380 PP/08 cyrus_sasl server now expands the server_realm option.
382 PP/09 Bugzilla 1214 - Log authentication information in reject log.
383 Patch by Jeremy Harris.
385 PP/10 Added dbmjz lookup type.
387 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
389 PP/12 MAIL args handles TAB as well as SP, for better interop with
390 non-compliant senders.
391 Analysis and variant patch by Todd Lyons.
393 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
394 Bug report from Lars Müller <lars@samba.org> (via SUSE),
395 Patch from Dirk Mueller <dmueller@suse.com>
397 PP/13 tls_peerdn now print-escaped for spool files.
398 Observed some $tls_peerdn in wild which contained \n, which resulted
399 in spool file corruption.
401 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
402 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
403 or write after TLS renegotiation, which otherwise led to messages
406 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
407 as a tracking header (ie: a signed header comes before the signature).
408 Patch from Wolfgang Breyha.
410 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
411 comma-sep list; embedded commas doubled.
413 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
415 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
417 Report and patch from Dmitry Banschikov.
419 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
420 Removed SSL_clear() after SSL_new() which led to protocol negotiation
421 failures. We appear to now support TLS1.1+ with Exim.
423 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
424 lets Exim select keys and certificates based upon TLS SNI from client.
425 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
426 before an outbound SMTP session. New log_selector, +tls_sni.
428 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
429 NULL dereference. Report and patch from Alun Jones.
431 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
432 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
433 Not seeing resolver debug output on NetBSD, but suspect this is a
434 resolver implementation change.
436 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
437 Left warnings. Added "eximon gdb" invocation mode.
439 PP/21 Defaulting "accept_8bitmime" to true, not false.
441 PP/22 Added -bw for inetd wait mode support.
443 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
444 locate the relevant includes and libraries. Made this the default.
446 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
447 Bugzilla 1246, report and most of solution from Tomasz Kusy.
449 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
450 This may cause build issues on older platforms.
452 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
453 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
454 gnutls_require_mac & gnutls_require_protocols (no longer supported).
455 Added SNI support via GnuTLS too.
456 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
458 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
460 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
461 Applied second patch from Janne, implementing suggestion to default
462 multiple-strings-in-record handling to match SPF spec.
464 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
466 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
467 read-only, out of scope).
468 Patch from Wolfgang Breyha, report from Stuart Northfield.
470 PP/29 Fix three issues highlighted by clang analyser static analysis.
471 Only crash-plausible issue would require the Cambridge-specific
472 iplookup router and a misconfiguration.
473 Report from Marcin Mirosław.
475 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
477 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
478 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
479 As part of this, removing so much warning spew let me fix some minor
480 real issues in debug logging.
482 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
483 assignment on my part. Fixed.
485 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
486 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
487 Janne Snabb (who went above and beyond: thank you).
489 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
490 string otherwise requires a connection and a bunch more work and it's
491 relatively easy to get wrong. Should also expose TLS library linkage
494 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
495 64-bit ${eval} (JH/03).
497 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
498 GNU libc to support some of the 64-bit stuff, should not lead to
499 conflicts. Defined before os.h is pulled in, so if a given platform
500 needs to override this, it can.
502 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
503 protection layer was required, which is not implemented.
504 Bugzilla 1254, patch from Wolfgang Breyha.
506 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
507 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
508 tls_dhparam take prime identifiers. Also unbreak combination of
509 OpenSSL+DH_params+TLSSNI.
511 PP/39 Disable SSLv2 by default in OpenSSL support.
517 PP/01 Solaris build fix for Oracle's LDAP libraries.
518 Bugzilla 1109, patch from Stephen Usher.
520 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
522 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
525 TF/02 Fix a couple more cases where we did not log the error message
526 when unlink() failed. See also change 4.74-TF/03.
528 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
529 lock up or crash if it happened to be inside a call to libc when it
530 got a SIGUSR1 from exiwhat.
532 The SIGUSR1 handler appends the current process status to the process
533 log which is later printed by exiwhat. It used to use the general
534 purpose logging code to do this, but several functions it calls are
535 not safe for signals.
537 The new output code in the SIGUSR1 handler is specific to the process
538 log, and simple enough that it's easy to inspect for signal safety.
539 Removing some special cases also simplifies the general logging code.
540 Removing the spurious timestamps from the process log simplifies
543 TF/04 Improved ratelimit ACL condition.
545 The /noupdate option has been deprecated in favour of /readonly which
546 has clearer semantics. The /leaky, /strict, and /readonly update modes
547 are mutually exclusive. The update mode is no longer included in the
548 database key; it just determines when the database is updated. (This
549 means that when you upgrde Exim will forget old rate measurements.)
551 Exim now checks that the per_* options are used with an update mode that
552 makes sense for the current ACL. For example, when Exim is processing a
553 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
554 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
555 must specify per_mail/readonly. If you omit the update mode it defaults to
556 /leaky where that makes sense (as before) or /readonly where required.
558 The /noupdate option is now undocumented but still supported for
559 backwards compatibility. It is equivalent to /readonly except that in
560 ACLs where /readonly is required you may specify /leaky/noupdate or
561 /strict/noupdate which are treated the same as /readonly.
563 A useful new feature is the /count= option. This is a generalization
564 of the per_byte option, so that you can measure the throughput of other
565 aggregate values. For example, the per_byte option is now equivalent
566 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
568 The per_rcpt option has been generalized using the /count= mechanism
569 (though it's more complicated than the per_byte equivalence). When it is
570 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
571 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
572 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
573 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
574 Note that using per_rcpt with a non-readonly update mode in more than
575 one ACL will cause the recipients to be double-counted. (The per_mail
576 and per_byte options don't have this problem.)
578 The handling of very low rates has changed slightly. If the computed rate
579 is less than the event's count (usually one) then this event is the first
580 after a long gap. In this case the rate is set to the same as this event's
581 count, so that the first message of a spam run is counted properly.
583 The major new feature is a mechanism for counting the rate of unique
584 events. The new per_addr option counts the number of different
585 recipients that someone has sent messages to in the last time period. It
586 behaves like per_rcpt if all the recipient addresses are different, but
587 duplicate recipient addresses do not increase the measured rate. Like
588 the /count= option this is a general mechanism, so the per_addr option
589 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
590 example, measure the rate that a client uses different sender addresses
591 with the options per_mail/unique=$sender_address. There are further
592 details in the main documentation.
594 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
596 TF/06 Removed a few PCRE remnants.
598 TF/07 Automatically extract Exim's version number from tags in the git
599 repository when doing development or release builds.
601 PP/02 Raise smtp_cmd_buffer_size to 16kB.
602 Bugzilla 879. Patch from Paul Fisher.
604 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
605 Heavily based on revision 40f9a89a from Simon Arlott's tree.
608 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
610 PP/05 Variable $av_failed, true if the AV scanner deferred.
611 Bugzilla 1078. Patch from John Horne.
613 PP/06 Stop make process more reliably on build failure.
614 Bugzilla 1087. Patch from Heiko Schlittermann.
616 PP/07 Make maildir_use_size_file an _expandable_ boolean.
617 Bugzilla 1089. Patch from Heiko Schlittermann.
619 PP/08 Handle ${run} returning more data than OS pipe buffer size.
620 Bugzilla 1131. Patch from Holger Weiß.
622 PP/09 Handle IPv6 addresses with SPF.
623 Bugzilla 860. Patch from Wolfgang Breyha.
625 PP/10 GnuTLS: support TLS 1.2 & 1.1.
627 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
630 PP/11 match_* no longer expand right-hand-side by default.
631 New compile-time build option, EXPAND_LISTMATCH_RHS.
632 New expansion conditions, "inlist", "inlisti".
634 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
636 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
638 PP/14 fix log_write() format string regression from TF/03.
639 Bugzilla 1152. Patch from Dmitry Isaikin.
645 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
647 PP/02 Harmonised TLS library version reporting; only show if debugging.
648 Layout now matches that introduced for other libraries in 4.74 PP/03.
650 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
652 PP/04 New "dns_use_edns0" global option.
654 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
657 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
658 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
660 TK/01 Updated PolarSSL code to 0.14.2.
661 Bugzilla 1097. Patch from Andreas Metzler.
663 PP/07 Catch divide-by-zero in ${eval:...}.
666 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
669 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
670 format-string attack -- SECURITY: remote arbitrary code execution.
672 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
673 time unintentionally subject to list matching rules, letting the header
674 cause arbitrary Exim lookups (of items which can occur in lists, *not*
675 arbitrary string expansion). This allowed for information disclosure.
677 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
678 INT_MIN/-1 -- value coerced to INT_MAX.
684 NM/01 Workround for PCRE version dependency in version reporting
687 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
688 This fixes portability to compilers other than gcc, notably
689 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
691 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
692 makefiles for portability to HP-UX and POSIX correctness.
694 PP/01 Permit LOOKUP_foo enabling on the make command-line.
695 Also via indented variable definition in the Makefile.
696 (Debugging by Oliver Heesakkers).
698 PP/02 Restore caching of spamd results with expanded spamd_address.
699 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
701 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
702 Improves build reliability. Fix from: Frank Elsner
704 NM/02 Fix wide character breakage in the rfc2047 coding
705 Fixes bug 1064. Patch from Andrey N. Oktyabrski
707 NM/03 Allow underscore in dnslist lookups
708 Fixes bug 1026. Patch from Graeme Fowler
710 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
711 Code patches from Adam Ciarcinski of NetBSD.
713 NM/04 Fixed exiqgrep to cope with mailq missing size issue
716 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
717 is logged, to avoid truncation. Patch from John Horne.
719 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
720 Patch from Jakob Hirsch.
722 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
723 SQL string expansion failure details.
724 Patch from Andrey Oktyabrski.
726 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
727 Patch from Simon Arlott.
729 PP/09 New lookups functionality failed to compile on old gcc which rejects
730 extern declarations in function scope.
731 Patch from Oliver Fleischmann
733 PP/10 Use sig_atomic_t for flags set from signal handlers.
734 Check getgroups() return and improve debugging.
735 Fixed developed for diagnosis in bug 927 (which turned out to be
738 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
739 Patch from Mark Zealey.
741 PP/12 Bugzilla 1056: Improved spamd server selection.
742 Patch from Mark Zealey.
744 PP/13 Bugzilla 1086: Deal with maildir quota file races.
745 Based on patch from Heiko Schlittermann.
747 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
748 Patch from Uwe Doering, sign-off by Michael Haardt.
750 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
751 variable declaration deep within a block. Bug and patch from
754 PP/15 lookups-Makefile IRIX compatibilty coercion.
756 PP/16 Make DISABLE_DKIM build knob functional.
758 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
759 Patch by Simon Arlott
761 TF/03 Fix valgrind.h portability to C89 compilers that do not support
762 variable argument macros. Our copy now differs from upstream.
768 TF/01 Failure to get a lock on a hints database can have serious
769 consequences so log it to the panic log.
771 TF/02 Log LMTP confirmation messages in the same way as SMTP,
772 controlled using the smtp_confirmation log selector.
774 TF/03 Include the error message when we fail to unlink a spool file.
776 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
777 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
778 for maintaining out-of-tree patches for some time.
780 PP/01 Bugzilla 139: Documentation and portability issues.
781 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
782 Handle per-OS dynamic-module compilation flags.
784 PP/02 Let /dev/null have normal permissions.
785 The 4.73 fixes were a little too stringent and complained about the
786 permissions on /dev/null. Exempt it from some checks.
787 Reported by Andreas M. Kirchwitz.
789 PP/03 Report version information for many libraries, including
790 Exim version information for dynamically loaded libraries. Created
791 version.h, now support a version extension string for distributors
792 who patch heavily. Dynamic module ABI change.
794 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
795 privilege escalation vulnerability whereby the Exim run-time user
796 can cause root to append content of the attacker's choosing to
799 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
802 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
803 If dropping privileges for untrusted macros, we disabled normal logging
804 on the basis that it would fail; for the Exim run-time user, this is not
805 the case, and it resulted in successful deliveries going unlogged.
806 Fixed. Reported by Andreas Metzler.
812 PP/01 Date: & Message-Id: revert to normally being appended to a message,
813 only prepend for the Resent-* case. Fixes regression introduced in
814 Exim 4.70 by NM/22 for Bugzilla 607.
816 PP/02 Include check_rfc2047_length in configure.default because we're seeing
817 increasing numbers of administrators be bitten by this.
819 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
821 PP/03 Bugzilla 994: added openssl_options main configuration option.
823 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
825 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
827 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
829 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
830 without a peer certificate, leading to a segfault because of an
831 assumption that peers always have certificates. Be a little more
832 paranoid. Problem reported by Martin Tscholak.
834 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
835 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
836 NB: ClamAV planning to remove STREAM in "middle of 2010".
837 CL also introduces -bmalware, various -d+acl logging additions and
838 more caution in buffer sizes.
840 PP/09 Implemented reverse_ip expansion operator.
842 PP/10 Bugzilla 937: provide a "debug" ACL control.
844 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
846 PP/12 Bugzilla 973: Implement --version.
848 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
850 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
852 PP/15 Bugzilla 816: support multiple condition rules on Routers.
854 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
855 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
856 ignore trailing whitespace.
858 JJ/02 prevent non-panic DKIM error from being sent to paniclog
860 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
863 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
864 Notification from Dr Andrew Aitchison.
866 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
867 ExtendedDetectionInfo response format.
868 Notification from John Horne.
870 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
873 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
874 XSL and documented dependency on system catalogs, with examples of how
877 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
880 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
881 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
882 configuration file which is writeable by the Exim user or group.
884 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
885 of configuration files to cover files specified with the -C option if
886 they are going to be used with root privileges, not just the default
889 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
890 option (effectively making it always true).
892 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
893 files to be used while preserving root privileges.
895 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
896 that rogue child processes cannot use them.
898 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
899 run-time user, instead of root.
901 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
902 Exim run-time user without dropping privileges.
904 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
905 result string, instead of calling string_vformat() twice with the same
908 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
909 for other users. Others should always drop root privileges if they use
910 -C on the command line, even for a whitelisted configure file.
912 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
914 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
920 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
921 $data_path, and $header_path variables; fixed documentation bugs and
924 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
925 exipick to access non-standard spools, including the "frozen" queue
928 NM/01 Bugzilla 965: Support mysql stored procedures.
929 Patch from Alain Williams
931 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
933 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
934 Patch from Andreas Metzler
936 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
937 Patch from Kirill Miazine
939 NM/05 Bugzilla 671: Added umask to procmail example.
941 JJ/03 installed exipick 20100323.0, fixing doc bug
943 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
944 directory. Notification and patch from Dan Rosenberg.
946 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
948 TK/02 Improve log output when DKIM signing operation fails.
950 MH/01 Treat the transport option dkim_domain as a colon separated
951 list, not as a single string, and sign the message with each element,
952 omitting multiple occurences of the same signer.
954 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
955 Bugzilla 985, 986. Patch by Simon Arlott
957 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
958 Patch by Simon Arlott
960 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
961 MBX locking. Notification from Dan Rosenberg.
967 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
969 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
971 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
973 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
975 NM/04 Bugzilla 915: Flag broken perl installation during build.
981 TK/01 Added patch by Johannes Berg that expands the main option
982 "spamd_address" if it starts with a dollar sign.
984 TK/02 Write list of recipients to X-Envelope-Sender header when building
985 the mbox-format spool file for content scanning (suggested by Jakob
988 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
989 (http://www.dcc-servers.net/) support via dccifd. Activated by
990 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
992 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
993 by Mark Daniel Reidel <mr@df.eu>.
995 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
996 When building exim an external PCRE library is now needed -
997 PCRE is a system library on the majority of modern systems.
998 See entry on PCRE_LIBS in EDITME file.
1000 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1001 conversation. Added nologin parameter to request.
1002 Patch contributed by Kirill Miazine.
1004 TF/01 Do not log submission mode rewrites if they do not change the address.
1006 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1008 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1009 log files in place. Contributed by Roberto Lima.
1011 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1013 TF/03 Bugzilla 615: When checking the local_parts router precondition
1014 after a local_part_suffix or local_part_prefix option, Exim now
1015 does not use the address's named list lookup cache, since this
1016 contains cached lookups for the whole local part.
1018 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1019 Robert Millan. Documentation is in experimental-spec.txt.
1021 TF/04 Bugzilla 668: Fix parallel build (make -j).
1023 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1025 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1026 Patch provided by Jan Srzednicki.
1028 TF/05 Leading white space used to be stripped from $spam_report which
1029 wrecked the formatting. Now it is preserved.
1031 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1032 that they are available at delivery time.
1034 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1036 TF/08 TLS error reporting now respects the incoming_interface and
1037 incoming_port log selectors.
1039 TF/09 Produce a more useful error message if an SMTP transport's hosts
1040 setting expands to an empty string.
1042 NM/06 Bugzilla 744: EXPN did not work under TLS.
1043 Patch provided by Phil Pennock.
1045 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1046 Patch provided by Richard Godbee.
1048 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1049 acl_smtp_notquit, added index entry.
1051 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1052 Patch provided by Eugene Bujak.
1054 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1055 accept(). Patch provided by Maxim Dounin.
1057 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1058 Patch provided by Phil Pennock.
1060 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1062 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1063 Patch provided by Brad "anomie" Jorsch.
1065 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1066 Patch provided by Dean Brooks.
1068 TK/05 Add native DKIM support (does not depend on external libraries).
1070 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1071 Patch provided by Graeme Fowler.
1073 NM/16 Bugzilla 851: Documentation example syntax fix.
1075 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1077 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1080 NM/19 Bugzilla 745: TLS version reporting.
1081 Patch provided by Phil Pennock.
1083 NM/20 Bugzilla 167: bool: condition support.
1084 Patch provided by Phil Pennock.
1086 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1087 clients. Patch provided by Phil Pennock.
1089 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1090 Patch provided by Brad "anomie" Jorsch.
1092 NM/23 Bugzilla 687: Fix misparses in eximstats.
1093 Patch provided by Heiko Schlittermann.
1095 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1096 Patch provided by Heiko Schlittermann.
1098 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1099 plus update to original patch.
1101 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1103 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1104 Patch provided by David Brownlee.
1106 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1108 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1110 NM/30 Bugzilla 888: TLS documentation bugfixes.
1112 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1114 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1115 Unlike the original bugzilla I have changed all shell scripts in src tree.
1117 NM/33 Bugzilla 898: Transport filter timeout fix.
1118 Patch by Todd Rinaldo.
1120 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1121 Patch by Serge Demonchaux.
1123 NM/35 Bugzilla 39: Base64 decode bug fixes.
1124 Patch by Jakob Hirsch.
1126 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1128 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1130 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1132 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1138 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1139 ALT-N's libdkim that I have put here:
1140 http://duncanthrax.net/exim-experimental/
1142 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1143 were called 'true' and it seems that C99 defines that as a reserved
1144 keyword to be used with 'bool' variable types. That means you could
1145 not include C99-style headers which use bools without triggering
1146 build errors in sieve.c.
1148 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1149 as mailq or other aliases. Changed the --help handling significantly
1150 to do whats expected. exim_usage() emits usage/help information.
1152 SC/01 Added the -bylocaldomain option to eximstats.
1154 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1156 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1158 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1164 PH/01 Another patch from the Sieve maintainer.
1166 PH/02 When an IPv6 address is converted to a string for single-key lookup
1167 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1168 dots are used instead of colons so that keys in lsearch files need not
1169 contain colons. This was done some time before quoting was made available
1170 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1171 (notated using the quote facility) so as to distinguish them from IPv4
1172 keys. This meant that lookups for IP addresses in host lists did not work
1173 for iplsearch lookups.
1175 This has been fixed by arranging for IPv6 addresses to be expressed with
1176 colons if the lookup type is iplsearch. This is not incompatible, because
1177 previously such lookups could never work.
1179 The situation is now rather anomolous, since one *can* have colons in
1180 ordinary lsearch keys. However, making the change in all cases is
1181 incompatible and would probably break a number of configurations.
1183 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1186 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1187 conversion specification without a maximum field width, thereby enabling
1188 a rogue spamd server to cause a buffer overflow. While nobody in their
1189 right mind would setup Exim to query an untrusted spamd server, an
1190 attacker that gains access to a server running spamd could potentially
1191 exploit this vulnerability to run arbitrary code as the Exim user.
1193 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1194 $primary_hostname instead of what libspf2 thinks the hosts name is.
1196 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1197 a directory entry by the name of the lookup key. Previously, if a
1198 symlink pointed to a non-existing file or a file in a directory that
1199 Exim lacked permissions to read, a lookup for a key matching that
1200 symlink would fail. Now it is enough that a matching directory entry
1201 exists, symlink or not. (Bugzilla 503.)
1203 PH/03 The body_linecount and body_zerocount variables are now exported in the
1206 PH/04 Added the $dnslist_matched variable.
1208 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1209 This means they are set thereafter only if the connection becomes
1212 PH/06 Added the client_condition to authenticators so that some can be skipped
1213 by clients under certain conditions.
1215 PH/07 The error message for a badly-placed control=no_multiline_responses left
1216 "_responses" off the end of the name.
1218 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1220 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1221 (without spaces) instead of just copying the configuration text.
1223 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1225 PH/11 Added $max_received_linelength.
1227 PH/12 Added +ignore_defer and +include_defer to host lists.
1229 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1230 way in which PCRE > 7.0 is built.
1232 PH/14 Implemented queue_only_load_latch.
1234 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1235 MAIL command. The effect was to mangle the value on 64-bit systems.
1237 PH/16 Another patch from the Sieve maintainer.
1239 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1241 PH/18 If a system quota error occurred while trying to create the file for
1242 a maildir delivery, the message "Mailbox is full" was not appended to the
1243 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1244 only to a quota excession during the actual writing of the file.
1246 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1247 characters?) which causes problems in log lines. The DN values are now
1248 passed through string_printing() before being added to log lines.
1250 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1251 and InterBase are left for another time.)
1253 PH/21 Added message_body_newlines option.
1255 PH/22 Guard against possible overflow in moan_check_errorcopy().
1257 PH/23 POSIX allows open() to be a macro; guard against that.
1259 PH/24 If the recipient of an error message contained an @ in the local part
1260 (suitably quoted, of course), incorrect values were put in $domain and
1261 $local_part during the evaluation of errors_copy.
1267 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1268 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1271 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1272 issue a MAIL command.
1274 PH/02 In an ACL statement such as
1276 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1278 if a client was not listed at all, or was listed with a value other than
1279 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1280 the condition was not true (as it should be), so access was not denied.
1281 The bug was that the ! inversion was incorrectly passed on to the second
1282 item. This has been fixed.
1284 PH/03 Added additional dnslists conditions == and =& which are different from
1285 = and & when the dns lookup returns more than one IP address.
1287 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1288 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1290 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1291 FSYNC, which compiles an option called disable_fsync that allows for
1292 bypassing fsync(). The documentation is heavily laced with warnings.
1294 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1296 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1297 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1298 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1299 including adding "make clean"; (3) Added -fPIC when compiling the test
1300 dynamically loaded module, to get rid of a warning.
1302 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1303 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1304 = 0s. The bug is otherwise harmless.
1306 PH/07 There was a bug in the dovecot authenticator such that the value of
1307 $auth1 could be overwritten, and so not correctly preserved, after a
1308 successful authentication. This usually meant that the value preserved by
1309 the server_setid option was incorrect.
1311 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1313 PH/09 Installed PCRE release 7.0.
1315 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1316 run for batched SMTP input. It is now run at the start of every message
1317 in the batch. While fixing this I discovered that the process information
1318 (output by running exiwhat) was not always getting set for -bs and -bS
1319 input. This is fixed, and it now also says "batched" for BSMTP.
1321 PH/11 Added control=no_pipelining.
1323 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1324 patch, slightly modified), and move the expansion of helo_data till after
1325 the connection is made in the smtp transport (so it can use these
1328 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1330 PH/14 Added log_selector = +pid.
1332 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1334 PH/16 Add ${if forany and ${if forall.
1336 PH/17 Added dsn_from option to vary the From: line in DSNs.
1338 PH/18 Flush SMTP output before performing a callout, unless control =
1339 no_callout_flush is set.
1341 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1342 was true (the default) a successful delivery failed to delete the retry
1343 item, thus causing premature timeout of the address. The bug is now
1346 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1348 PH/21 Long custom messages for fakedefer and fakereject are now split up
1349 into multiline reponses in the same way that messages for "deny" and
1350 other ACL rejections are.
1352 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1353 with slight modification.
1355 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1356 draft, changing the syntax and factoring some duplicate code".
1358 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1359 for deliveries of the second and subsequent messages over the same SMTP
1362 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1363 ${reduce, with only minor "tidies".
1365 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1367 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1368 expansion side effects.
1370 PH/27 When a message times out after an over-quota error from an Exim-imposed
1371 quota, the bounce message says "mailbox is full". This message was not
1372 being given when it was a system quota that was exceeded. It now should
1375 MH/03 Made $recipients available in local_scan(). local_scan() already has
1376 better access to the recipient list through recipients_list[], but
1377 $recipients can be useful in postmaster-provided expansion strings.
1379 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1380 in the case of a MAIL command with additional options following the
1381 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1382 were accidentally chopped off.
1384 PH/29 SMTP synchronization checks are implemented when a command is read -
1385 there is a check that no more input is waiting when there shouldn't be
1386 any. However, for some commands, a delay in an ACL can mean that it is
1387 some time before the response is written. In this time, more input might
1388 arrive, invalidly. So now there are extra checks after an ACL has run for
1389 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1390 pipelining has not been advertised.
1392 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1394 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1395 returned in the SMTP response when smtp_return_error_details was false.
1396 This has been fixed.
1398 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1399 instead of the C I/O that was originally supplied, because problems were
1400 reported on Solaris.
1402 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1403 Exim which did not show up earlier: it was assuming that a call to
1404 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1405 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1406 was a macro that became an assignment, so it seemed to work. This has
1407 changed to a proper function call with a void return, hence the compile
1408 error. Exim's code has been fixed.
1410 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1413 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1415 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1417 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1418 because it turns out to be too big or there is a timeout) there is no
1419 "Completed" line in the log. When some messages of this type were
1420 selected by exigrep, they were listed as "not completed". Others were
1421 picked up by some special patterns. I have improved the selection
1422 criteria to be more general.
1424 PH/38 The host_find_failed option in the manualroute router can now be set
1425 to "ignore", to completely ignore a host whose IP address cannot be
1426 found. If all hosts are ignored, the behaviour is controlled by the new
1427 host_all_ignored option.
1429 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1430 homing or because of multiple MX records with /mx) generated more than
1431 one IP address, and the following item turned out to be the local host,
1432 all the secondary addresses of the first item were incorrectly removed
1433 from the list, along with the local host and any following hosts (which
1434 is what is supposed to happen).
1436 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1437 whoever called Exim into the -H file. In the case of the daemon it was
1438 behaving confusingly. When first started, it used values for whoever
1439 started the daemon, but after a SIGHUP it used the Exim user (because it
1440 calls itself on a restart). I have changed the code so that it now always
1443 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1444 message are rejected with the same error (e.g. no authentication or bad
1445 sender address), and a DATA command is nevertheless sent (as can happen
1446 with PIPELINING or a stupid MUA), the error message that was given to the
1447 RCPT commands is included in the rejection of the DATA command. This is
1448 intended to be helpful for MUAs that show only the final error to their
1451 PH/42 Another patch from the Sieve maintainer.
1453 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1454 Eximstats - Fixed some broken HTML links and added missing column headers
1456 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1457 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1459 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1461 PH/43 Yet another patch from the Sieve maintainer.
1463 PH/44 I found a way to check for a TCP/IP connection going away before sending
1464 the response to the final '.' that terminates a message, but only in the
1465 case where the client has not sent further data following the '.'
1466 (unfortunately, this is allowed). However, in many cases there won't be
1467 any further data because there won't be any more messages to send. A call
1468 to select() can be used: if it shows that the input is "ready", there is
1469 either input waiting, or the socket has been closed. An attempt to read
1470 the next input character can distinguish the two cases. Previously, Exim
1471 would have sent an OK response which the client would never have see.
1472 This could lead to message repetition. This fix should cure that, at
1473 least in a lot of common cases.
1475 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1476 advertised in response to EHLO.
1482 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1483 fixed by 4.65/MH/01 (is this a record?) are fixed:
1485 (i) An empty string was always treated as zero by the numeric comparison
1486 operators. This behaviour has been restored.
1488 (ii) It is documented that the numeric comparison operators always treat
1489 their arguments as decimal numbers. This was broken in that numbers
1490 starting with 0 were being interpreted as octal.
1492 While fixing these problems I realized that there was another issue that
1493 hadn't been noticed. Values of message_size_limit (both the global option
1494 and the transport option) were treated as octal if they started with 0.
1495 The documentation was vague. These values are now always treated as
1496 decimal, and I will make that clear in the documentation.
1502 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1503 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1506 MH/01 Don't check that the operands of numeric comparison operators are
1507 integers when their expansion is in "skipping" mode (fixes bug
1508 introduced by 4.64-PH/07).
1510 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1511 child addresses, Exim now panics and dies. Previously, because the count
1512 is held in a short int, deliveries were likely to be lost. As such a
1513 large number of recipients for a single message is ridiculous
1514 (performance will be very, very poor), I have chosen to impose a limit
1515 rather than extend the field.
1521 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1522 leftover -K file (the existence of which was triggered by #402).
1523 While we were at it, introduced process PID as part of the -K
1524 filename. This should rule out race conditions when creating
1527 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1528 processing considerably. Previous code took too long for large mails,
1529 triggering a timeout which in turn triggers #401.
1531 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1532 in the DK code in transports.c. sendfile() is not really portable,
1533 hence the _LINUX specificness.
1535 TF/01 In the add_headers option to the mail command in an Exim filter,
1536 there was a bug that Exim would claim a syntax error in any
1537 header after the first one which had an odd number of characters
1540 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1541 callout verification, Exim cached a "reject" for the entire domain. This
1542 is correct for most verifications, but it is not correct for a recipient
1543 verification with use_sender or use_postmaster set, because in that case
1544 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1545 case of MAIL FROM:<> rejection from other early rejections (e.g.
1546 rejection of HELO). When verifying a recipient using a non-null MAIL
1547 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1548 Whatever the result of the callout, the value of the domain cache is
1549 left unchanged (for any other kind of callout, getting as far as trying
1550 RCPT means that the domain itself is ok).
1552 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1555 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1556 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1557 session, the daemon ignores ECONNECT errors and logs others; it now
1558 ignores EPIPE as well.
1560 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1561 (quoted-printable decoding).
1563 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1564 later the small subsequent patch to fix an introduced bug.
1566 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1568 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1570 PH/08 An error is now given if message_size_limit is specified negative.
1572 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1573 to be given (somewhat) arbitrary names.
1575 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1578 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1579 miscellaneous code fixes
1581 PH/10 Added the log_reject_target ACL modifier to specify where to log
1584 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1585 hostname. This is wrong, because it relates to the incoming message (and
1586 probably the interface on which it is arriving) and not to the outgoing
1587 callout (which could be using a different interface). This has been
1588 changed to use the value of the helo_data option from the smtp transport
1589 instead - this is what is used when a message is actually being sent. If
1590 there is no remote transport (possible with a router that sets up host
1591 addresses), $smtp_active_hostname is used.
1593 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1594 tweaks were necessary in order to get it to work (see also 21 below):
1595 (a) The code assumed that strncpy() returns a negative number on buffer
1596 overflow, which isn't the case. Replaced with Exim's string_format()
1598 (b) There were several signed/unsigned issues. I just did the minimum
1599 hacking in of casts. There is scope for a larger refactoring.
1600 (c) The code used strcasecmp() which is not a standard C function.
1601 Replaced with Exim's strcmpic() function.
1602 (d) The code set only $1; it now sets $auth1 as well.
1603 (e) A simple test gave the error "authentication client didn't specify
1604 service in request". It would seem that Dovecot has changed its
1605 interface. Fortunately there's a specification; I followed it and
1606 changed what the client sends and it appears to be working now.
1608 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1611 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1612 address A is aliased to B and C, where B exists and C does not. Without
1613 -v the output is "A verified" because verification stops after a
1614 successful redirection if more than one address is generated. However,
1615 with -v the child addresses are also verified. Exim was outputting "A
1616 failed to verify" and then showing the successful verification for C,
1617 with its parentage. It now outputs "B failed to verify", showing B's
1618 parentage before showing the successful verification of C.
1620 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1621 look up a TXT record in a specific list after matching in a combined
1624 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1625 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1626 they consult the DNS. I had assumed they would set it the way they
1627 wanted; and indeed my experiments on Linux seem to show that in some
1628 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1629 To be on the safe side, however, I have now made the interface to
1630 host_find_byname() similar to host_find_bydns(), with an argument
1631 containing the DNS resolver options. The host_find_byname() function now
1632 sets these options at its start, just as host_find_bydns() does. The smtp
1633 transport options dns_qualify_single and dns_search_parents are passed to
1634 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1635 of host_find_byname() use the default settings of RES_DEFNAMES
1636 (qualify_single) but not RES_DNSRCH (search_parents).
1638 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1639 spool_read_header() do less string testing, by means of a preliminary
1640 switch on the second character of optional "-foo" lines. (This is
1641 overdue, caused by the large number of possibilities that now exist.
1642 Originally there were few.) While I was there, I also converted the
1643 str(n)cmp tests so they don't re-test the leading "-" and the first
1644 character, in the hope this might squeeze out yet more improvement.
1646 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1647 flag allowing group syntax was set by the header_syntax check but not
1648 turned off, possible causing trouble later; (2) The flag was not being
1649 set at all for the header_verify test, causing "group"-style headers to
1650 be rejected. I have now set it in this case, and also caused header_
1651 verify to ignore an empty address taken from a group. While doing this, I
1652 came across some other cases where the code for allowing group syntax
1653 while scanning a header line wasn't quite right (mostly, not resetting
1654 the flag correctly in the right place). These bugs could have caused
1655 trouble for malformed header lines. I hope it is now all correct.
1657 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1658 with the "reply" argument non-NULL. The code, however (which originally
1659 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1660 but it didn't always do it. This confused somebody who was copying the
1661 code for some other use. I have removed all the tests.
1663 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1664 feature that was used to support insecure browsers during the U.S. crypto
1665 embargo. It requires special client support, and Exim is probably the
1666 only MTA that supported it -- and would never use it because real RSA is
1667 always available. This code has been removed, because it had the bad
1668 effect of slowing Exim down by computing (never used) parameters for the
1669 RSA_EXPORT functionality.
1671 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1672 authenticator to fail if there's a tab character in the incoming data
1673 (there should never be unless someone is messing about, as it's supposed
1674 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1675 if the connection is using TLS or if the remote IP is the same as the
1676 local IP, and the "valid-client-cert option" if a client certificate has
1679 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1680 authenticators. This can be used for authorization after authentication
1681 succeeds. (In the case of plaintext, it servers for both authentication
1684 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1685 if any retry times were supplied.
1687 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1688 connection, where there is no client IP address to check. In this
1689 situation, the verify now always succeeds.
1691 PH/25 Applied John Jetmore's -Mset patch.
1693 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1695 PH/27 In a string expansion for a processed (not raw) header when multiple
1696 headers of the same name were present, leading whitespace was being
1697 removed from all of them, but trailing whitespace was being removed only
1698 from the last one. Now trailing whitespace is removed from each header
1699 before concatenation. Completely empty headers in a concatenation (as
1700 before) are ignored.
1702 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1703 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1705 PH/29 [Removed. This was a change that I later backed out, and forgot to
1706 correct the ChangeLog entry (that I had efficiently created) before
1707 committing the later change.]
1709 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1710 address errors (4xx response to RCPT) over the same connection as other
1711 messages routed to the same hosts. Such deliveries are always "forced",
1712 so retry times are not inspected. This resulted in far too many retries
1713 for the affected addresses. The effect occurred only when there were more
1714 hosts than the hosts_max_try setting in the smtp transport when it had
1715 the 4xx errors. Those hosts that it had tried were not added to the list
1716 of hosts for which the message was waiting, so if all were tried, there
1717 was no problem. Two fixes have been applied:
1719 (i) If there are any address or message errors in an SMTP delivery, none
1720 of the hosts (tried or untried) are now added to the list of hosts
1721 for which the message is waiting, so the message should not be a
1722 candidate for sending over the same connection that was used for a
1723 successful delivery of some other message. This seems entirely
1724 reasonable: after all the message is NOT "waiting for some host".
1725 This is so "obvious" that I'm not sure why it wasn't done
1726 previously. Hope I haven't missed anything, but it can't do any
1727 harm, as the worst effect is to miss an optimization.
1729 (ii) If, despite (i), such a delivery is accidentally attempted, the
1730 routing retry time is respected, so at least it doesn't keep
1731 hammering the server.
1733 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1734 in ${readsocket because some servers need this prod.
1736 PH/32 Added some extra debug output when updating a wait-xxx database.
1738 PH/33 The hint "could be header name not terminated by colon", which has been
1739 given for certain expansion errors for a long time, was not being given
1740 for the ${if def:h_colon_omitted{... case.
1742 PH/34 The spec says: "With one important exception, whenever a domain list is
1743 being scanned, $domain contains the subject domain." There was at least
1744 one case where this was not true.
1746 PH/35 The error "getsockname() failed: connection reset by peer" was being
1747 written to the panic log as well as the main log, but it isn't really
1748 panic-worthy as it just means the connection died rather early on. I have
1749 removed the panic log writing for the ECONNRESET error when getsockname()
1752 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1753 runs only) independently of the message's sender address. This meant
1754 that, if the 4xx error was in fact related to the sender, a different
1755 message to the same recipient with a different sender could confuse
1756 things. In particualar, this can happen when sending to a greylisting
1757 server, but other circumstances could also provoke similar problems.
1758 I have changed the default so that the retry time for these errors is now
1759 based a combination of the sender and recipient addresses. This change
1760 can be overridden by setting address_retry_include_sender=false in the
1763 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1764 remote server are returned as part of bounce messages. This was not
1765 happening for LMTP over a pipe (the lmtp transport), but now it is the
1766 same for both kinds of LMTP.
1768 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1769 in header lines that were in fact CNAMEs. This is no longer the case.
1771 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1772 and queue runs started by the daemon processed all messages. This has
1773 been fixed so that -R and -S can now usefully be given with -q<time>.
1775 PH/40 Import PCRE release 6.7 (fixes some bugs).
1777 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1779 PH/42 Give an error if -q is specified more than once.
1781 PH/43 Renamed the variables $interface_address and $interface_port as
1782 $received_ip_address and $received_port, to make it clear that these
1783 values apply to message reception, and not to the outgoing interface when
1784 a message is delivered. (The old names remain recognized, of course.)
1786 PH/44 There was no timeout on the connect() call when using a Unix domain
1787 socket in the ${readsocket expansion. There now is.
1789 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1790 be meaningful with "accept".
1792 SC/01 Eximstats V1.43
1793 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1795 SC/02 Eximstats V1.44
1796 Use a glob alias rather than an array ref in the generated
1797 parser. This improves both readability and performance.
1799 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1800 Collect SpamAssassin and rejection statistics.
1801 Don't display local sender or destination tables unless
1802 there is data to show.
1803 Added average volumes into the top table text output.
1805 SC/04 Eximstats V1.46
1806 Collect data on the number of addresses (recipients)
1807 as well as the number of messages.
1809 SC/05 Eximstats V1.47
1810 Added 'Message too big' to the list of mail rejection
1811 reasons (thanks to Marco Gaiarin).
1813 SC/06 Eximstats V1.48
1814 Mainlog lines which have GMT offsets and are too short to
1815 have a flag are now skipped.
1817 SC/07 Eximstats V1.49 (Alain Williams)
1818 Added the -emptyok flag.
1820 SC/08 Eximstats V1.50
1821 Fixes for obtaining the IP address from reject messages.
1823 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1824 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1825 whitesspace changes from 4.64-PH/27
1827 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1830 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1831 are found, allow negative numbers in numeric criteria)
1833 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1835 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1836 to match changes made in 4.64-PH/43
1838 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1840 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1841 "domain" is actually an IP address, and force a failure. This locks out
1842 those revolvers/nameservers that support "A-for-A" lookups, in
1843 contravention of the specifications.
1845 PH/48 When a host name was looked up from an IP address, and the subsequent
1846 forward lookup of the name timed out, the host name was left in
1847 $sender_host_name, contrary to the specification.
1849 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1850 restricted to single-key lookups, Exim was not diagnosing an error if
1851 * or *@ was used with a query-style lookup.
1853 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1855 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1856 long ago, but noone interested enough thought of it. Let's just say that
1857 the "1.1" means that there are some new functions that weren't there at
1858 some point in the past.
1860 PH/51 Error processing for expansion failure of helo_data from an smtp
1861 transport during callout processing was broken.
1863 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1864 tested/used via the -bh/-bhc/-bs options.
1866 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1867 bug, fixed in subsequent PCRE releases).
1869 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1870 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1872 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1878 SC/01 Use a glob alias rather than an array ref in eximstats generated
1879 parser. This improves both readability and performance.
1881 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1882 Don't display local sender or destination tables in eximstats unless
1883 there is data to show.
1884 Added average volumes into the eximstats top table text output.
1886 SC/03 Collect data on the number of addresses (recipients) as well
1887 as the number of messages in eximstats.
1889 TF/01 Correct an error in the documentation for the redirect router. Exim
1890 does (usually) call initgroups() when daemonizing.
1892 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1893 with consistent privilege compared to when running as a daemon.
1895 TF/03 Note in the spec that $authenticated_id is not set for local
1896 submissions from trusted users.
1898 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1899 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1901 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1902 by adding some example configuration directives to the default
1903 configuration file. A little bit of work is required to uncomment the
1904 directives and define how usernames and passwords are checked, but
1905 there is now a framework to start from.
1907 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1908 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1909 without this. I don't know how relevant this is to other LDAP libraries.
1911 PH/02 Add the verb name to the "unknown ACL verb" error.
1913 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1915 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1917 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1918 directory not expanded when it should be if an expanded home directory
1919 was set for the address (which is overridden by the transport).
1921 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1924 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1925 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1926 because it is too late at that time, and has no effect.
1928 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1929 security issue with \' (bugzilla #107). I could not use the
1930 PQescapeStringConn() function, because it needs a PGconn value as one of
1933 PH/08 When testing addresses using -bt, indicate those final addresses that
1934 are duplicates that would not cause an additional delivery. At least one
1935 person was confused, thinking that -bt output corresponded to deliveries.
1936 (Suppressing duplicates isn't a good idea as you lose the information
1937 about possibly different redirections that led to the duplicates.)
1939 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1940 systems where poll() doesn't work, in particular OS X.
1942 PH/10 Added more information to debugging output for retry time not reached.
1944 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1945 operations in malware.c.
1947 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1950 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1951 syslog_duplication set false, log lines that would normally be written
1952 both the the main log and to the reject log were not written to syslog at
1955 PH/14 In the default configuration, change the use of "message" in ACL warn
1956 statements to "add_header".
1958 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1959 not followed by a command (e.g. "seen endif").
1961 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1962 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1965 PH/17 Added extra conditions to the default value of delay_warning_condition
1969 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1970 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1971 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1974 The Auto-Submitted: and various List- headers are standardised, whereas I
1975 don't think Precedence: ever was.
1977 PH/18 Refactored debugging code in route_finduser() to show more information,
1978 in particular, the error code if getpwnam() issues one.
1980 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1981 This is apparently needed in addition to the PH/07 change above to avoid
1982 any possible encoding problems.
1984 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1985 but not after initializing Perl.
1987 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1988 output them only if debugging. By default they are written stderr,
1989 apparently, which is not desirable.
1991 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1994 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1997 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1999 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2000 authenticated or an ident call has been made. Suppress the default
2001 values for $authenticated_id and $authenticated_sender (but permit -oMai
2002 and -oMas) when testing with -bh.
2004 PH/24 Re-jigged the order of the tests in the default configuration so that the
2005 tests for valid domains and recipients precede the DNS black list and CSA
2006 tests, on the grounds that those ones are more expensive.
2008 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2009 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2010 This bug exists in every version of Exim that I still have, right back to
2013 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2014 However, an attempt to turn on case-sensitivity in a regex key by
2015 including (?-i) didn't work because the subject string was already
2016 lowercased, and the effects were non-intuitive. It turns out that a
2017 one-line patch can be used to allow (?-i) to work as expected.
2023 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2024 other effects) broke the use of negated acl sub-conditions.
2026 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2029 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2030 "Deny" causes Exim to reject the incoming connection with a 554 error.
2031 Unfortunately, if there is a major crisis, such as a disk failure,
2032 tcp-wrappers gives "deny", whereas what one would like would be some
2033 kind of temporary error. A kludge has been added to help with this.
2034 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2035 554 error is used if errno is still zero or contains ENOENT (which occurs
2036 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2039 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2041 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2042 errors. Otherwise a message that provokes a temporary error (when other
2043 messages do not) can cause a whole host to time out.
2045 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2046 addresses were routed directly to files or pipes from a redirect router.
2047 File deliveries just didn't batch; pipe deliveries might have suffered
2050 PH/06 A failure to get a lock for a hints database would erroneously always say
2051 "Failed to get write lock", even when it was really a read lock.
2053 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2054 of 0600. This has been changed to use the value of the lockfile_mode
2055 option (which defaults to 0600).
2057 PH/08 Applied small patch from the Sieve maintainer.
2059 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2060 folder from quota calculations, a direct delivery into this folder messed
2061 up the contents of the maildirsize file. This was because the regex was
2062 used only to exclude .Trash (or whatever) when the size of the mailbox
2063 was calculated. There was no check that a delivery was happening into an
2064 excluded directory. This bug has been fixed by ignoring all quota
2065 processing for deliveries into excluded directories.
2067 PH/10 Added the maildirfolder_create_regex option to appendfile.
2073 PH/01 The code for finding all the local interface addresses on a FreeBSD
2074 system running IPv6 was broken. This may well have applied to all BSD
2075 systems, as well as to others that have similar system calls. The broken
2076 code found IPv4 interfaces correctly, but gave incorrect values for the
2077 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2078 that it would not match correctly against @[] and not recognize the IPv6
2081 PH/02 The ipliteral router was not recognizing addresses of the form user@
2082 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2084 PH/03 Added disable_ipv6.
2086 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2087 decoded form, because it is most often used to construct To: headers
2088 lines in autoreplies, and the decoded form may well be syntactically
2089 invalid. However, $reply_address has leading white space removed, and all
2090 newlines turned into spaces so that the autoreply transport does not
2093 PH/05 If group was specified without a user on a router, and no group or user
2094 was specified on a transport, the group from the router was ignored.
2096 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2097 for visible compile-time settings that can be used to change these
2098 numbers, for those that want even more. Backwards compatibility with old
2099 spool files has been maintained. However, going back to a previous Exim
2100 release will lost any variables that are in spool files.
2102 PH/07 Two small changes when running in the test harness: increase delay when
2103 passing a TCP/IP connection to a new process, in case the original
2104 process has to generate a bounce, and remove special handling of
2105 127.0.0.2 (sic), which is no longer necessary.
2107 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2108 be the same on different OS.
2110 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2113 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2114 whether --show-vars was specified or not
2116 JJ/02 exipick: Added support for new ACL variable spool format introduced
2119 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2120 syntactically invalid From: or Reply-to: line, and a filter used this to
2121 generate an autoreply, and therefore failed to obtain an address for the
2122 autoreply, Exim could try to deliver to a non-existent relative file
2123 name, causing unrelated and misleading errors. What now happens is that
2124 it logs this as a hard delivery error, but does not attempt to create a
2127 PH/11 The exinext utility has a -C option for testing purposes, but although
2128 the given file was scanned by exinext itself; it wasn't being passed on
2129 when Exim was called.
2131 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2132 an end-of-file indication when reading a command response.
2134 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2135 compiled. In many other places in Exim, IPv6 addresses are always
2136 recognized, so I have changed this. It also means that IPv4 domain
2137 literals of the form [IPV4:n.n.n.n] are now always recognized.
2139 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2140 used if the router is not running as root, for example, when verifying at
2141 ACL time, or when using -bh. The debugging output from this situation was
2142 non-existent - all you got was a failure to exec. I have made two
2145 (a) Failures to set uid/gid, the current directory, or a process leader
2146 in a subprocess such as that created by queryprogram now generate
2147 suitable debugging ouput when -d is set.
2149 (b) The queryprogram router detects when it is not running as root,
2150 outputs suitable debugging information if -d is set, and then runs
2151 the subprocess without attempting to change uid/gid.
2153 PH/15 Minor change to Makefile for building test_host (undocumented testing
2156 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2157 additional section of a DNS packet that returns MX or SRV records.
2158 Instead, it always explicitly searches for A/AAAA records. This avoids
2159 major problems that occur when a DNS server includes only records of one
2160 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2161 fixed another bug: if SRV records were looked up and the corresponding
2162 address records were *not* found in the additional section, the port
2163 values from the SRV records were lost.
2165 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2166 using the correct key (the original address) when searching the retry
2167 rules in order to find which one to use for generating the retry hint.
2169 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2170 adding the default one. Similarly, if it contains a Reply-To: header, the
2171 errors_reply_to option, if set, is not used.
2173 PH/19 When calculating a retry time, Exim used to measure the "time since
2174 failure" by looking at the "first failed" field in the retry record. Now
2175 it does not use this if it is later than than the arrival time of the
2176 message. Instead it uses the arrival time. This makes for better
2177 behaviour in cases where some deliveries succeed, thus re-setting the
2178 "first failed" field. An example is a quota failure for a huge message
2179 when small messages continue to be delivered. Without this change, the
2180 "time since failure" will always be short, possible causing more frequent
2181 delivery attempts for the huge message than are intended.
2182 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2184 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2185 $1, $2, $3) because the numerical variables can be reset during some
2186 expansion items (e.g. "match"), thereby losing the authentication data.
2188 PH/21 Make -bV show the size of off_t variables so that the test suite can
2189 decide whether to run tests for quotas > 2G.
2191 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2192 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2193 filecount value is greater than 2G or if a quota value is greater than 2G
2194 on a system where the size of off_t is not greater than 4, a panic error
2197 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2198 never match. The debug and -bh output now contains an explicit error
2199 message indicating a malformed IPv4 address or mask.
2201 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2202 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2203 PH/23 above applies.
2205 PH/25 Do not write to syslog when running in the test harness. The only
2206 occasion when this arises is a failure to open the main or panic logs
2207 (for which there is an explicit test).
2209 PH/26 Added the /no_tell option to "control=freeze".
2211 PH/27 If a host name lookup failed very early in a connection, for example, if
2212 the IP address matched host_lookup and the reverse lookup yielded a name
2213 that did not have a forward lookup, an error message of the form "no IP
2214 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2215 could be logged. Now it outputs the IP address instead of "NULL".
2217 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2218 allows the sender and the authenticated sender to be set when
2219 submitting a message from within Exim. Since child_open_exim() is
2220 documented for local_scan(), the new function should be too.
2222 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2223 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2224 results in an empty string is now treated as unset.
2226 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2228 PH/31 Added sender_verify_fail logging option.
2230 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2231 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2232 tidied the source and removed it altogether.
2234 PH/33 When a queue run was abandoned because the load average was too high, a
2235 log line was always written; now it is written only if the queue_run log
2236 selector is set. In addition, the log line for abandonment now contains
2237 information about the queue run such as the pid. This is always present
2238 in "start" and "stop" lines but was omitted from the "abandon" line.
2240 PH/34 Omit spaces between a header name and the colon in the error message that
2241 is given when verify = headers_syntax fails (if there are lots of them,
2242 the message gets confusing).
2244 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2245 names, as there are now some PTR records that contain slashes. This check
2246 is only to protect against broken name servers that fall over on strange
2247 characters, so the fact that it applies to all lookups doesn't matter.
2249 PH/36 Now that the new test suite is complete, we can remove some of the
2250 special code in Exim that was needed for the old test suite. For example,
2251 sorting DNS records because real resolvers return them in an arbitrary
2252 order. The new test suite's fake resolver always returns records in the
2255 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2256 bounces) except when queue_only is set, to avoid logging races between
2257 the different processes.
2259 PH/38 Panic-die if .include specifies a non-absolute path.
2261 PH/39 A tweak to the "H" retry rule from its user.
2263 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2264 a label. They prevented compilation on older perls.
2266 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2267 a warning to be raised on newish perls.
2269 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2270 on queue. Changes to match documented behaviour of showing count of
2271 messages matching specified criteria.
2273 PH/40 Changed the default ident timeout from 30s to 5s.
2275 PH/41 Added support for the use of login_cap features, on those BSD systems
2276 that have them, for controlling the resources used by pipe deliveries.
2278 PH/42 The content-scanning code uses fopen() to create files in which to put
2279 message data. Previously it was not paying any attention to the mode of
2280 the files. Exim runs with umask(0) because the rest of the code creates
2281 files with open(), and sets the required mode explicitly. Thus, these
2282 files were ending up world-writeable. This was not a big issue, because,
2283 being within the spool directory, they were not world-accessible. I have
2284 created a function called modefopen, which takes an additional mode
2285 argument. It sets umask(777), creates the file, chmods it to the required
2286 mode, then resets the umask. All the relevant calls to fopen() in the
2287 content scanning code have been changed to use this function.
2289 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2290 to 24 hours. This avoids potential overflow problems when processing G
2291 and H retry rules. I suspect nobody ever tinkers with this value.
2293 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2295 PH/45 When the plaintext authenticator is running as a client, the server's
2296 challenges are checked to ensure they are valid base64 strings. By
2297 default, the authentication attempt is cancelled if an invalid string is
2298 received. Setting client_ignore_invalid_base64 true ignores these errors.
2299 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2300 they are received. Thus, the responses can be made to depend on the
2301 challenges. If an invalid string is ignored, an empty string is placed in
2304 PH/46 Messages that are created by the autoreply transport now contains a
2305 References: header, in accordance with RFCs 2822 and 3834.
2307 PH/47 Added authenticated_sender_force to the smtp transport.
2309 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2311 PH/49 Installed latest patch from the Sieve maintainer.
2313 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2314 was also set, the appendfile transport was unnecessarily scanning a
2315 directory of message files (e.g. for maildir delivery) to find the count
2316 of files (along with the size), even though it did not need this
2317 information. It now does the scan only if it needs to find either the
2318 size of the count of files.
2320 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2322 PH/52 Two bugs concerned with error handling when the smtp transport is
2325 (i) Exim was not creating retry information for temporary errors given
2326 for individual recipients after the DATA command when the smtp transport
2327 was used in LMTP mode. This meant that they could be retried too
2328 frequently, and not timed out correctly.
2330 (ii) Exim was setting the flag that allows error details to be returned
2331 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2332 recipients that were returned after the DATA command.
2334 PH/53 This is related to PH/52, but is more general: for any failing address,
2335 when detailed error information was permitted to be returned to the
2336 sender, but the error was temporary, then after the final timeout, only
2337 "retry timeout exceeded" was returned. Now it returns the full error as
2338 well as "retry timeout exceeded".
2340 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2341 do this, and (what is worse) MTAs that accept it.
2343 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2344 will now be deprecated.
2346 PH/56 New os.c-cygwin from the Cygwin maintainer.
2348 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2349 formats (previously only available in exim formats via -bpr, -bpru,
2350 and -bpra. Now also available in native and exiqgrep formats)
2352 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2353 with very large, slow to parse queues
2355 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2357 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2359 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2360 responses to authentication challenges, though it was showing the
2361 challenges; (ii) I've removed the CR characters from the debug output for
2364 PH/58 Allow for the insertion of a newline as well as a space when a string
2365 is turned into more than one encoded-word during RFC 2047 encoding. The
2366 Sieve code now uses this.
2368 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2369 data_4xx, lost_connection, tls_required.
2371 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2372 message was being sent as an SMTP response.
2374 PH/61 Add -l and -k options to exicyclog.
2376 PH/62 When verifying, if an address was redirected to one new address, so that
2377 verification continued, and the new address failed or deferred after
2378 having set something in $address_data, the value of $address_data was not
2379 passed back to the ACL. This was different to the case when no
2380 redirection occurred. The value is now passed back in both cases.
2382 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2383 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2384 use login_cap.h, so on its own it isn't the distinguishing feature. The
2385 new name refers directly to the setclassresources() function.
2387 PH/65 Added configuration files for NetBSD3.
2389 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2391 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2392 is preferred over IPv4.
2394 PH/68 The bounce_return_message and bounce_return_body options were not being
2395 honoured for bounces generated during the reception of non-SMTP messages.
2396 In particular, this applied to messages rejected by the ACL. This bug has
2397 been fixed. However, if bounce_return_message is true and bounce_return_
2398 body is false, the headers that are returned for a non-SMTP message
2399 include only those that have been read before the error was detected.
2400 (In the case of an ACL rejection, they have all been read.)
2402 PH/69 The HTML version of the specification is now built in a directory called
2403 spec_html instead of spec.html, because the latter looks like a path with
2404 a MIME-type, and this confuses some software.
2406 PH/70 Catch two compiler warnings in sieve.c.
2408 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2409 function verify_get_ident() calls ip_connect() to connect a socket, but
2410 if the "connect()" function timed out, ip_connect() used to close the
2411 socket. However, verify_get_ident() also closes the socket later, and in
2412 between Exim writes to the log, which may get opened at this point. When
2413 the socket was closed in ip_connect(), the log could get the same file
2414 descriptor number as the socket. This naturally causes chaos. The fix is
2415 not to close the socket in ip_connect(); the socket should be closed by
2416 the function that creates it. There was only one place in the code where
2417 this was missing, in the iplookup router, which I don't think anybody now
2418 uses, but I've fixed it anyway.
2420 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2421 well as to direct DNS lookups. Otherwise the handling of names in host
2422 lists is inconsistent and therefore confusing.
2428 PH/01 Two changes to the default runtime configuration:
2430 (1) Move the checks for relay_from_hosts and authenticated clients from
2431 after to before the (commented out) DNS black list checks.
2433 (2) Add control=submission to the relay_from_hosts and authenticated
2434 clients checks, on the grounds that messages accepted by these
2435 statements are most likely to be submissions.
2437 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2439 (1) Generate an error if the third argument for the ${prvs expansion is
2442 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2445 (3) Reset the variables that are obtained from the first argument of
2446 ${prvscheck and used in the second argument before leaving the code,
2447 because their memory is reclaimed, so using them afterwards may do
2450 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2451 one (it's much easier than Tom thought :-).
2453 (5) Because of (4), we can now allow for the use of $prvscheck_result
2454 inside the third argument.
2456 PH/03 For some reason, the default setting of PATH when running a command from
2457 a pipe transport was just "/usr/bin". I have changed it to
2460 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2461 anything to be listed in the output from -bV.
2463 PH/05 When a filter generated an autoreply, the entire To: header line was
2464 quoted in the delivery log line, like this:
2466 => >A.N.Other <ano@some.domain> <original@ddress> ...
2468 This has been changed so that it extracts the operative address. There
2469 may be more than one such address. If so, they are comma-separated, like
2472 => >ano@some.domain,ona@other.domain <original@ddress> ...
2474 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2475 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2476 not being looked up in the rDNS to get a host name, Exim was showing the
2477 IP address twice in Received: lines, even though the IP addresses were
2478 identical. For example:
2480 Received: from [1.2.3.4] (helo=[1.2.3.4])
2482 However, if the real host name was known, it was omitting the HELO data
2483 if it matched the actual IP address. This has been tidied up so that it
2484 doesn't show the same IP address twice.
2486 PH/07 When both +timestamp and +memory debugging was on, the value given by
2487 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2488 called by the debug printing, thereby overwriting the timestamp buffer.
2489 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2491 PH/08 When the original message was included in an autoreply transport, it
2492 always said "this is a copy of the message, including all the headers",
2493 even if body_only or headers_only was set. It now gives an appropriate
2496 PH/09 Applied a patch from the Sieve maintainer which:
2498 o fixes some comments
2499 o adds the (disabled) notify extension core
2500 o adds some debug output for the result of if/elsif tests
2501 o points to the current vacation draft in the documentation
2502 and documents the missing references header update
2506 o fixes a bug in processing the envelope test (when testing
2507 multiple envelope elements, the last element determinted the
2510 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2511 Electronic Mail") by including:
2513 Auto-submitted: auto-generated
2515 in the messages that it generates (bounce messages and others, such as
2516 warnings). In the case of bounce messages for non-SMTP mesages, there was
2517 also a typo: it was using "Auto_submitted" (underscore instead of
2518 hyphen). Since every message generated by Exim is necessarily in response
2519 to another message, thes have all been changed to:
2521 Auto-Submitted: auto-replied
2523 in accordance with these statements in the RFC:
2525 The auto-replied keyword:
2527 - SHOULD be used on messages sent in direct response to another
2528 message by an automatic process,
2530 - MUST NOT be used on manually-generated messages,
2532 - MAY be used on Delivery Status Notifications (DSNs) and Message
2533 Disposition Notifications (MDNs),
2535 - MUST NOT be used on messages generated by automatic or periodic
2536 processes, except for messages which are automatic responses to
2539 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2540 to the default Received: header definition.
2542 PH/12 Added log selector acl_warn_skipped (default on).
2544 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2545 variables because (a) they are in the wrong storage pool and (b) even if
2546 they were copied, it wouldn't work properly because of the caching.
2548 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2549 checking when decoding. Apparently there are clients that generate
2550 overlong encoded strings. Why am I not surprised?
2552 PH/15 If the first argument of "${if match_address" was not empty, but did not
2553 contain an "@" character, Exim crashed. Now it writes a panic log message
2554 and treats the condition as false.
2556 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2558 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2559 extension "envelope-auth". The code is finished and in agreement with
2560 other implementations, but there is no documentation so far and in fact,
2561 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2562 not changing the active code.
2564 Print executed "if" and "elsif" statements when debugging is used. This
2565 helps a great deal to understand what a filter does.
2567 Document more things not specified clearly in RFC3028. I had all this
2568 sorted out, when out of a sudden new issues came to my mind. Oops."
2570 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2573 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2574 canonical form (as documented). However, after a host name lookup from
2575 the IP address, check_host() was doing a simple string comparison with
2576 addresses acquired from the DNS when checking that the found name did
2577 have the original IP as one of its addresses. Since any found IPv6
2578 addresses are likely to be in abbreviated form, the comparison could
2579 fail. Luckily, there already exists a function for doing the comparison
2580 by converting both addresses to binary, so now that is used instead of
2581 the text comparison.
2583 PH/20 There was another similar case to PH/19, when a complete host name was
2584 given in a host list; looking up its IP address could give an abbreviated
2585 form, whereas the current host's name might or might not be abbreviated.
2586 The same fix has been applied.
2592 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2593 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2596 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2597 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2599 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2601 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2602 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2603 Also, for "Auto-Submitted", treat anything other than "no" as
2604 non-personal, in accordance with RFC 3834. (Previously it treated
2605 anything starting "auto-" as non-personal.)
2607 TF/01 The control=submission/name=... option had a problem with syntax
2608 errors if the name included a slash character. The /name= option
2609 now slurps the rest of the string, so it can include any characters
2610 but it must come last in the list of options (after /sender_retain
2613 PH/05 Some modifications to the interface to the fake nameserver for the new
2621 TK/01 Added the "success_on_redirect" address verification option. See
2622 NewStuff for rationale and an example.
2624 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2626 PH/02 Patch to exigrep to allow it to work on syslog lines.
2628 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2629 fread() to skip over the body file's header line, because in Cygwin the
2630 header line is locked and is inaccessible.
2632 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2633 co-exist for some time) to make it clear that it is the Exim ID that is
2634 referenced, not the Message-ID: header line.
2636 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2637 string_format() function, because snprintf() does not exist on all
2640 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2641 ${stat: expansion item.
2643 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2644 protocol synchronization error", to keep the pedants happy.
2646 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2647 well as for IRIX systems, when gcc is being used. See the host.c source
2650 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2652 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2655 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2656 transport and to the smtp transport in LMTP mode.
2658 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2660 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2661 run a filter in a subprocess. This could lead to confusion in subsequent
2662 lookups in the parent process. There should also be a search_tidyup() at
2663 the end of the subprocess.
2665 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2666 only if the host matched helo_try_verify_hosts, which caused the
2667 verification to occur when the EHLO/HELO command was issued. The ACL just
2668 tested the remembered result. Now, if a previous verification attempt has
2669 not happened, "verify = helo" does it there and then.
2671 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2673 TK/03 Fix log output including CR from clamd.
2675 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2676 exist provoked a memory error which could cause a segfault.
2678 PH/15 Installed PCRE 6.2
2680 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2682 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2683 of the problem. Specifically, suggested +O2 rather than +O1 for the
2686 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2688 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2689 record, the port was not correctly logged when the outgoing_port log
2690 selector was set (it logged the transort's default port).
2692 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2693 fallback_hosts, and "hosts" in the smtp transport.
2695 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2696 host errors such as "Connection refused".
2698 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2699 authentication with radiusclient 0.4.9:
2701 - Error returned from rc_read_config was caught wrongly
2702 - Username/password not passed on to radius server due to wrong length.
2704 The presumption is that some radiusclient API changes for 4.51/PH/17
2705 were not taken care of correctly. The code is still untested by me (my
2706 Linux distribution still has 0.3.2 of radiusclient), but it was
2707 contributed by a Radius user.
2709 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2710 expanding the "port" option of the smtp transport.
2712 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2713 while reading a MIME header. Thanks to Tom Hughes for a patch.
2715 PH/24 Include config.h inside local_scan.h so that configuration settings are
2718 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2719 that in an ACL for RCPT (for example), you can examine exactly what was
2722 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2723 commands, but it was not correctly comparing the address with the actual
2724 client host address. Thus, it would show the EHLO address in Received:
2725 header lines when this was not necessary.
2727 PH/27 Added the % operator to ${eval:}.
2729 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2730 it should be ignoring failures (because with -C, for example, it has lost
2731 privilege). It wasn't ignoring creation failures other than "already
2734 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2737 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2738 because an input line was too long, either on its own, or by virtue of
2739 too many continuations, the temporary file was not being removed, and the
2740 return code was incorrect.
2742 PH/31 Missing "BOOL" in function definition in filtertest.c.
2744 PH/32 Applied Sieve patches from the maintainer.
2746 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2748 PH/33 Added "verify = not_blind".
2750 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2751 Local/Makefile (with some defaults set). These are used in built scripts
2752 such as exicyclog, but they have never been used in the exim_install
2753 script (though there are many overriding facilities there). I have
2754 arranged that the exim_install script now takes note of these two
2757 PH/35 Installed configuration files for Dragonfly.
2759 PH/36 When a locally submitted message by a trusted user did not contain a
2760 From: header, and the sender address was obtained from -f or from an SMTP
2761 MAIL command, and the trusted user did not use -F to supply a sender
2762 name, $originator_name was incorrectly used when constructing a From:
2763 header. Furthermore, $originator_name was used for submission mode
2764 messages from external hosts without From: headers in a similar way,
2765 which is clearly wrong.
2767 PH/37 Added control=suppress_local_fixups.
2769 PH/38 When log_selector = +received_sender was set, and the addition of the
2770 sender made the log line's construction buffer exactly full, or one byte
2771 less than full, an overflow happened when the terminating "\n" was
2774 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2775 when the result of a list match is failure because a DNS lookup failed.
2777 PH/40 RM_COMMAND is now used in the building process.
2779 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2780 the "build-* directories that it finds.
2782 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2783 address in a domain literal was a prefix of an interface address.
2785 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2786 when verifying a sender address, unless rewrite_headers is false.
2788 PH/44 Wrote a long comment about why errors_to addresses are verified as
2789 recipients, not senders.
2791 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2792 the ratelimit ACL was added.
2794 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2796 PH/46 Added extra information about PostgreSQL errors to the error string.
2798 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2799 avoiding the need to install special zones in a real server. This is
2800 backwards compatible; if it can't find the fake resolver, it drops back.
2801 Thus, both old and new test suites can be run.
2803 TF/02 Added util/ratelimit.pl
2805 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2806 clock is set back in time.
2808 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2809 Candler <B.Candler@pobox.com>.
2811 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2812 OK for addresses that are the result of redirections.
2814 PH/48 A number of further additions for the benefit of the new test suite,
2815 including a fake gethostbyname() that interfaces to the fake DNS resolver
2818 TF/06 The fix for widen_domains has also been applied to qualify_single and
2819 search_parents which are the other dnslookup options that can cause
2822 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2825 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2827 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2828 service" if it encountered an MX record with an empty target hostname.
2829 The message is now "an MX or SRV record indicated no SMTP service".
2831 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2832 if the DNS of the sending site is misconfigured. This is quite a
2833 common situation. This change restores the behaviour of treating a
2834 helo verification defer as a failure.
2836 PH/51 If self=fail was set on a router, the bounce message did not include the
2837 actual error message.
2843 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2845 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2846 command itself ended in error, the underlying message about the transport
2847 filter timeout was being overwritten with the pipe command error. Now the
2848 underlying error message should be appended to the second error message.
2850 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2852 PH/02 Reduce the amount of output that "make" produces by default. Full output
2853 can still be requested.
2855 PH/03 The warning log line about a condition test deferring for a "warn" verb
2856 was being output only once per connection, rather than after each
2857 occurrence (because it was using the same function as for successful
2858 "warn" verbs). This seems wrong, so I have changed it.
2860 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2861 it should not have, which might have caused a crash in the right
2862 circumstances, but probably never did.
2864 PH/04 Installed a modified version of Tony Finch's patch to make submission
2865 mode fix the return path as well as the Sender: header line, and to
2866 add a /name= option so that you can make the user's friendly name appear
2869 TF/03 Added the control = fakedefer ACL modifier.
2871 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2872 Mark Lowes for thorough testing.
2874 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2876 TK/03 Merged latest SRS patch from Miles Wilton.
2878 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2879 with the definition in sysexits.h (which is #included earlier).
2880 Fortunately, Exim does not actually use EX_OK. The code used to try to
2881 preserve the sysexits.h value, by assumimg that macro definitions were
2882 scanned for macro replacements. I have been disabused of this notion,
2883 so now the code just undefines EX_OK before #including unistd.h.
2885 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2886 in the smtp transport. When a block could not be written in a single
2887 write() function, the timeout was being re-applied to each part-write.
2888 This seems wrong - if the receiver was accepting one byte at a time it
2889 would take for ever. The timeout is now adjusted when this happens. It
2890 doesn't have to be particularly precise.
2892 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2893 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2895 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2896 without a domain if the check to <postmaster@domain> fails.
2898 SC/01 Eximstats: added -xls and the ability to specify output files
2899 (patch written by Frank Heydlauf).
2901 SC/02 Eximstats: use FileHandles for outputing results.
2903 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2905 SC/04 Eximstats: fixed display of large numbers with -nvr option
2907 SC/05 Eximstats: fixed merging of reports with empty tables.
2909 SC/06 Eximstats: added the -include_original_destination flag
2911 SC/07 Eximstats: removed tabs and trailing whitespace.
2913 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2915 TK/06 MBOX spool code: Add real "From " MBOX separator line
2916 so the .eml file is really in mbox format (even though
2917 most programs do not really care). Patch from Alex Miller.
2919 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2920 The latter is generated from $received_to and is only set if the
2921 message has one envelope recipient. SA can use these headers,
2922 obviously out-of-the-box. Patch from Alex Miller.
2924 PH/08 The ${def test on a variable was returning false if the variable's
2925 value was "0", contrary to what the specification has always said!
2926 The result should be true unless the variable is empty.
2928 PH/09 The syntax error of a character other than { following "${if
2929 def:variable_name" (after optional whitespace) was not being diagnosed.
2930 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2931 accidental colon was present, for example, could give incorrect results.
2933 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2934 result is used (not including appendfile, where other changes are about
2937 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2938 This involved changing a lot of size variables from int to off_t. It
2939 should work with maildirs and everything.
2941 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2942 spamd dying while we are connected to it.
2944 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2947 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2948 function for BDB 4.3.
2950 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2952 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2953 the sources for PCRE have been re-arranged and more files are now
2956 PH/15 The code I had for printing potentially long long variables in PH/11
2957 above was not the best (it lost precision). The length of off_t variables
2958 is now inspected at build time, and an appropriate printing format (%ld
2959 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2960 to be "long long int" or "long int". This is needed for the internal
2961 formatting function string_vformat().
2963 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2964 the configuration file to be ":syslog", then the script "guesses" where
2965 the logs files are, rather than using the compiled in default. In our
2966 case the guess is not the same as the compiled default, so the script
2967 suddenly stopped working when I started to use syslog. The patch checks
2968 to see if log_file_path is "". If so, it attempts to read it from exim
2969 with no configuration file to get the compiled in version, before it
2970 falls back to the previous guessing code."
2972 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2973 implementing BATV in an Exim configuration. See NewStuff for the gory
2976 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2977 Makefile that are specific to HP-UX.
2979 PH/18 If the "use_postmaster" option was set for a recipient callout together
2980 with the "random" option, the postmaster address was used as the MAIL
2981 FROM address for the random test, but not for the subsequent recipient
2982 test. It is now used for both.
2984 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2985 patch removes a few documentation additions to RFC 3028, because the
2986 latest draft now contains them. It adds the new en;ascii-case comparator
2987 and a new error check for 8bit text in MIME parts. Comparator and
2988 require names are now matched exactly. I enabled the subaddress
2989 extension, but it is not well tested yet (read: it works for me)."
2991 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2992 rework some of the code of TK/09 above to avoid the hardwired use of
2993 "%lld" and "long long". Replaced the call to snprintf() with a call to
2996 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2997 records point to non-existent hosts", "retry timeout exceeded", and
2998 "retry time not reached for any host after a long failure period".
3000 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3001 experimental DomainKeys support:
3003 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3004 (2) On an error such as an illegally used "control", the wrong name for
3005 the control was given.
3007 These problems did NOT occur unless DomainKeys support was compiled.
3009 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3011 PH/24 Added ${if match_ip condition.
3013 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3014 databases so that it will be absolutely obvious if a crash occurs in the
3015 DB library. This is a regular occurrence (often caused by mis-matched
3018 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3019 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3020 picked up on a user's system that detects such things. There doesn't seem
3021 to be a gcc warning option for this - only an attribute that has to be
3022 put on the function's prototype. It seems that in Fedora Core 4 they have
3023 set this on a number of new functions. No doubt there will be more in due
3026 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3027 specify a transport. However, if an address that was verified by such a
3028 router was the subject of a callout, Exim crashed because it tried to
3029 read the rcpt_include_affixes from the non-existent transport. Now it
3030 just assumes that the setting of that option is false. This bug was
3031 introduced by 4.51/PH/31.
3033 PH/28 Changed -d+all to exclude +memory, because that information is very
3034 rarely of interest, but it makes the output a lot bigger. People tend to
3035 do -d+all out of habit.
3037 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3038 code in os-type was giving problems when libc.so lives in lib64, like on
3041 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3042 aren't the modern standard, and it seems that some systems' include files
3043 don't always have them. Exim was already checking for some of the newer
3044 ones like T_AAAA, and defining it itself. I've added checks for all the
3045 record types that Exim uses.
3047 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3048 not automatically generating a new one, as it is supposed to. This
3049 prevented TLS from working. If the file did exist, but contained invalid
3050 data, a new version was generated, as expected. It was only the case of a
3051 non-existent file that was broken.
3053 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3054 with a change in libdomainkeys > 0.64.
3056 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3057 from DNS. If the selector record carries the flag, it now has
3058 precedence over the domain-wide flag.
3060 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3062 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3063 the use of a port name that isn't defined in /etc/services) occurred, the
3064 message was deferred as in a normal delivery, and thus remained on the
3065 spool, instead of being failed because of the mua_wrapper setting. This
3066 is now fixed, and I tidied up some of the mua_wrapper messages at the
3069 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3070 the messages in a hash of arrays rather than using individual hashes.
3071 This is a bit cleaner and results in dramatic memory savings, albeit
3072 at a slight CPU cost.
3074 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3075 as requested by Marc Sherman.
3077 SC/10 Eximstats: added histograms for user specified patterns as requested
3080 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3082 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3083 fopen() in the content-scanning modules that did not already have it.
3089 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3090 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3092 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3094 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3096 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3097 to test Sieve filters that use "vacation".
3099 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3100 that changes the way the GnuTLS parameters are stored in the cache file.
3101 The new format can be generated externally. For backward compatibility,
3102 if the data in the cache doesn't make sense, Exim assumes it has read an
3103 old-format file, and it generates new data and writes a new file. This
3104 means that you can't go back to an older release without removing the
3107 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3108 work if there are any delivery delays because "one_time" forces the
3109 parent to be marked "delivered", so its unseen clone is never tried
3110 again. For this reason, Exim now forbids the simultaneous setting of
3113 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3114 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3115 entry if you want to know the details. The fix, however, neglected to
3116 consider the case when local delivery batching is involved. The test for
3117 "previously delivered" was not happening when checking to see if an
3118 address could be batched with a previous (undelivered) one; under
3119 certain circumstances this could lead to multiple deliveries to the same
3122 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3123 in its include files, and this causes problems building Exim.
3125 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3126 header_syntax) but Exim was just ignoring anything given after a slash.
3127 In particular, this caused confusion with an attempt to use "verify =
3128 reverse_host_lookup/defer_ok". An error is now given when options are
3129 supplied for verify items that do not have them. (Maybe reverse_host_
3130 lookup should have a defer_ok option, but that's a different point.)
3132 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3133 defined by RFC 821) to 2048, because there were problems with some AUTH
3134 commands, and RFC 1869 says the size should be increased for extended
3135 SMTP commands that take arguments.
3137 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3140 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3141 "unknown" error; now it says that the functionality isn't in the binary.
3143 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3144 an address' error message when a string expansion fails (syntax or
3145 whatever). Otherwise the password may appear in the log. Following change
3146 PH/42 below, there is no longer a chance of it appearing in a bounce
3149 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3151 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3152 of its addresses was ever tried. (Bugzilla bug #2.)
3154 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3155 the result incorrectly in the debug output. (It correctly added a newline
3156 to what was transported.)
3158 TF/01 Added $received_time.
3160 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3161 commented out examples of how to interface to a virus scanner and to
3162 SpamAssassin. Also added commented examples of av_scanner and
3163 spamd_address settings.
3165 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3166 and controls are allowed in which ACLs. There were a couple of minor
3167 errors. Some of the entries in the conditions table (which is a table of
3168 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3169 negation of where the condition IS allowed.
3171 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3173 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3174 header file does not have a version number, so I've had to invent a new
3175 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3176 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3177 radiusclient), but it was contributed by a Radius user.
3179 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3180 files or directories, for OpenSSL.
3182 PH/19 When an Exim process that is running as root has to create an Exim log
3183 file, it does so in a subprocess that runs as exim:exim so as to get the
3184 ownership right at creation (otherwise, other Exim processes might see
3185 the file with the wrong ownership). There was no test for failure of this
3186 fork() call, which would lead to the process getting stuck as it waited
3187 for a non-existent subprocess. Forks do occasionally fail when resources
3188 run out. I reviewed all the other calls to fork(); they all seem to check
3191 PH/20 When checking for unexpected SMTP input at connect time (before writing
3192 the banner), Exim was not dealing correctly with a non-positive return
3193 from the read() function. If the client had disconnected by this time,
3194 the result was a log entry for a synchronization error with an empty
3195 string after "input=" when read() returned zero. If read() returned -1
3196 (an event I could not check), uninitialized data bytes were printed.
3197 There were reports of junk text (parts of files, etc) appearing after
3200 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3202 PH/22 Added support for macro redefinition, and (re)definition in between
3203 driver and ACL definitions.
3205 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3206 forgetting to use the resulting value; it was using the unexpanded value.
3208 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3209 hadn't been configured. The fix is from Juergen Kreileder, who
3210 understands it better than I do:
3212 "Here's what I see happening with three configured cyrus_sasl
3213 authenticators configured (plain, login, cram-md5):
3215 On startup auth_cyrus_sasl_init() gets called for each of these.
3216 This means three calls to sasl_listmech() without a specified mech_list.
3217 => SASL tests which mechs of all available mechs actually work
3218 => three warnings about OTP not working
3219 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3221 With the patch, sasl_listmech() also gets called three times. But now
3222 SASL's mech_list option is set to the server_mech specified in the the
3223 authenticator. Or in other words, the answer from sasl_listmech()
3224 gets limited to just the mech you're testing for (which is different
3226 => the return list contains just 'plain' or 'login', 'cram-md5' or
3227 nothing depending on the value of ob->server_mech.
3229 I've just tested the patch: Authentication still works fine,
3230 unavailable mechs specified in the exim configuration are still
3231 caught, and the auth.log warnings about OTP are gone."
3233 PH/25 When debugging is enabled, the contents of the command line are added
3234 to the debugging output, even when log_selector=+arguments is not
3237 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3238 answer is "GNU", and only if the return is "GNU/something" is the answer
3241 PH/27 $acl_verify_message is now set immediately after the failure of a
3242 verification in an ACL, and so is available in subsequent modifiers. In
3243 particular, the message can be preserved by coding like this:
3245 warn !verify = sender
3246 set acl_m0 = $acl_verify_message
3248 Previously, $acl_verify_message was set only while expanding "message"
3249 and "log_message" when a very denied access.
3251 PH/28 Modified OS/os.c-Linux with
3253 -#ifndef OS_LOAD_AVERAGE
3254 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3256 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3257 nomenclature these days.)
3259 PH/29 Installed patch from the Sieve maintainer that adds the options
3260 sieve_useraddress and sieve_subaddress to the redirect router.
3262 PH/30 In these circumstances:
3263 . Two addresses routed to the same list of hosts;
3264 . First host does not offer TLS;
3265 . First host accepts first address;
3266 . First host gives temporary error to second address;
3267 . Second host offers TLS and a TLS session is established;
3268 . Second host accepts second address.
3269 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3270 and peerdn, if requested) that were in fact used only for the second
3273 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3274 attention to any local part prefix or suffix that was matched by the
3275 router that accepted the address. It now behaves in the same way as it
3276 does for delivery: the affixes are removed from the local part unless
3277 rcpt_include_affixes is set on the transport.
3279 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3280 timeout during the DATA phase of an incoming message.
3282 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3283 applied a patch sanctioned by the Sieve maintainer.
3285 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3286 the uid or gid is negative. A case of a negative gid caused this to be
3287 noticed. The fix allows for either to be negative.
3289 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3290 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3293 PH/36 The change PH/12 above was broken. Fixed it.
3295 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3296 the grounds that routing the same address twice would always produce the
3297 same answer. This might have been true once, but it is certainly no
3298 longer true now. Routing a child address may depend on the previous
3299 routing that produced that child. Some complicated redirection strategies
3300 went wrong when messages had multiple recipients, and made Exim's
3301 behaviour dependent on the order in which the addresses were given.
3303 I have moved the duplicate checking until after the routing is complete.
3304 Exim scans the addresses that are assigned to local and remote
3305 transports, and removes any duplicates. This means that more work will be
3306 done, as duplicates will always all be routed, but duplicates are
3307 presumably rare, so I don't expect this is of any significance.
3309 For deliveries to pipes, files, and autoreplies, the duplicate checking
3310 still happens during the routing process, since they are not going to be
3313 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3314 It corrects a timeout issue with spamd. This is Ian's comment: "The
3315 background is that sometimes spamd either never reads data from a
3316 connection it has accepted, or it never writes response data. The exiscan
3317 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3318 blindly assumes that writes won't block so it may never time out."
3320 PH/39 Allow G after quota size as well as K and M.
3322 PH/40 The value set for $authenticated_id in an authenticator may not contain
3323 binary zeroes or newlines because the value is written to log lines and
3324 to spool files. There was no check on this. Now the value is run through
3325 the string_printing() function so that such characters are converted to
3326 printable escape sequences.
3328 PH/41 $message_linecount is a new variable that contains the total number of
3329 lines in the message. Compare $body_linecount, which is the count for the
3332 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3333 bounce and delay warning messages, except in certain special cases, which
3336 (a) An SMTP error message from a remote host;
3337 (b) A message specified in a :fail: redirection;
3338 (c) A message specified in a "fail" command in a system filter;
3339 (d) A message specified in a FAIL return from the queryprogram router;
3340 (e) A message specified by the cannot_route_message router option.
3342 In these cases only, Exim does include the error details in bounce and
3343 warning messages. There are also a few cases where bland messages such
3344 as "unrouteable address" or "local delivery error" are given.
3346 PH/43 $value is now also set for the "else" part of a ${run expansion.
3348 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3349 being worked on, but at least Exim now implements the latest version to
3352 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3353 process to complete was treated as a delivery failure, a timeout while
3354 writing the message to the pipe was logged, but erroneously treated as a
3355 successful delivery. Such timeouts include transport filter timeouts. For
3356 consistency with the overall process timeout, these timeouts are now
3357 treated as errors, giving rise to delivery failures by default. However,
3358 there is now a new Boolean option for the pipe transport called
3359 timeout_defer, which, if set TRUE, converts the failures into defers for
3360 both kinds of timeout. A transport filter timeout is now identified in
3363 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3364 systems where "make" and "gmake" are different, calling "gmake" at top
3365 level broke things. I've arranged for the value of $(MAKE) to be passed
3366 from the Makefile to this script so that it can call the same version of
3370 A note about Exim versions 4.44 and 4.50
3371 ----------------------------------------
3373 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3374 changes of various kinds. As a consequence, a big documentation update was
3375 needed. This delayed the release for rather longer than seemed good, especially
3376 in the light of a couple of (minor) security issues. Therefore, the changes
3377 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3378 release. So 4.44 and 4.50 are in effect two different branches that both start
3381 I have left the 4.50 change log unchanged; it contains all the changes since
3382 4.43. The change log for 4.44 is below; many of its items are identical to
3383 those for 4.50. This seems to be the most sensible way to preserve the
3384 historical information.
3390 1. Minor wording change to the doc/README.SIEVE file.
3392 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3393 computation of the current number of files was incorrect.
3395 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3396 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3399 4. Give more explanation in the error message when the command for a transport
3400 filter fails to execute.
3402 5. There are several places where Exim runs a non-Exim command in a
3403 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3404 was being done only for the command run by the queryprogram router. It is
3405 now done for all such subprocesses. The other cases are: ${run, transport
3406 filters, and the commands run by the lmtp and pipe transports.
3408 6. Added CONFIGURE_GROUP build-time option.
3410 7. Some older OS have a limit of 256 on the maximum number of file
3411 descriptors. Exim was using setrlimit() to set 1000 as a large value
3412 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3413 systems. I've change it so that if it can't get 1000, it tries for 256.
3415 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3416 was an oversight, and furthermore, ever since the addition of extra
3417 controls (e.g. 4.43/32), the checks on when to allow different forms of
3418 "control" were broken. There should now be diagnostics for all cases when a
3419 control that does not make sense is encountered.
3421 9. Added the /retain_sender option to "control=submission".
3423 10. $recipients is now available in the predata ACL (oversight).
3425 11. Tidy the search cache before the fork to do a delivery from a message
3426 received from the command line. Otherwise the child will trigger a lookup
3427 failure and thereby defer the delivery if it tries to use (for example) a
3428 cached ldap connection that the parent has called unbind on.
3430 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3431 of $address_data from the recipient verification was clobbered by the
3432 sender verification.
3434 13. The value of address_data from a sender verification is now available in
3435 $sender_address_data in subsequent conditions in the ACL statement.
3437 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3439 15. Added a new option "connect=<time>" to callout options, to set a different
3442 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3443 was its contents. (It was OK if the option was not defined at all.)
3445 17. A "Completed" log line is now written for messages that are removed from
3446 the spool by the -Mrm option.
3448 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3449 information about exactly what failed.
3451 19. Added -dd to debug only the daemon process.
3453 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3454 handles timeouts, both on the server side and network timeouts. Renamed the
3455 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3457 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3458 It is now set to "smtps".
3460 22. $host_address is now set to the target address during the checking of
3461 ignore_target_hosts.
3463 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3464 being passed; this would have caused $sender_host_name to have been used if
3465 matching the list had actually called for a host name (not very likely,
3466 since this list is usually IP addresses). A host name is now passed as
3469 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3470 code that specifies a non-restarting handler (typically sigaction() in
3471 modern systems) in an attempt to fix a rare and obscure crash bug.
3473 25. Narrowed the window for a race in the daemon that could cause it to ignore
3474 SIGCHLD signals. This is not a major problem, because they are used only to
3475 wake it up if nothing else does.
3477 26. A malformed maildirsize file could cause Exim to calculate negative values
3478 for the mailbox size or file count. Odd effects could occur as a result.
3479 The maildirsize information is now recalculated if the size or filecount
3482 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3483 support for a long time. Removed HAVE_SYS_VFS_H.
3485 28. Installed the latest version of exipick from John Jetmore.
3487 29. In an address list, if the pattern was not a regular expression, an empty
3488 subject address (from a bounce message) matched only if the pattern was an
3489 empty string. Non-empty patterns were not even tested. This was the wrong
3490 because it is perfectly reasonable to use an empty address as part of a
3491 database query. An empty address is now tested by patterns that are
3492 lookups. However, all the other forms of pattern expect the subject to
3493 contain a local part and a domain, and therefore, for them, an empty
3494 address still always fails if the pattern is not itself empty.
3496 30. Exim went into a mad DNS loop when attempting to do a callout where the
3497 host was specified on an smtp transport, and looking it up yielded more
3498 than one IP address.
3500 31. Re-factored the code for checking spool and log partition space into a
3501 function that finds that data and another that does the check. The former
3502 is then used to implement four new variables: $spool_space, $log_space,
3503 $spool_inodes, and $log_inodes.
3505 32. The RFC2047 encoding function was originally intended for short strings
3506 such as real names; it was not keeping to the 75-character limit for
3507 encoded words that the RFC imposes. It now respects the limit, and
3508 generates multiple encoded words if necessary. To be on the safe side, I
3509 have increased the buffer size for the ${rfc2047: expansion operator from
3512 33. It is now permitted to omit both strings after an "if" condition; if the
3513 condition is true, the result is "true". As before, when the second string
3514 is omitted, a false condition yields an empty string. This makes it less
3515 cumbersome to write custom ACL and router conditions.
3517 34. Failure to deliver a bounce message always caused it to be frozen, even if
3518 there was an errors_to setting on the router. The errors_to setting is now
3521 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3522 canonical form (fully expanded) before being placed in
3523 $sender_host_address.
3525 36. The table in the code that translates DNS record types into text (T_A to
3526 "A" for instance) was missing entries for NS and CNAME. It is just possible
3527 that this could have caused confusion if both these types were looked up
3528 for the same domain, because the text type is used as part of Exim's
3529 per-process caching. But the chance of anyone hitting this buglet seems
3532 37. The dnsdb lookup has been extended in a number of ways.
3534 (1) There is a new type, "zns", which walks up the domain tree until it
3535 finds some nameserver records. It should be used with care.
3537 (2) There is a new type, "mxh", which is like "mx" except that it returns
3538 just the host names, not the priorities.
3540 (3) It is now possible to give a list of domains (or IP addresses) to be
3541 looked up. The behaviour when one of the lookups defers can be
3542 controlled by a keyword.
3544 (4) It is now possible to specify the separator character for use when
3545 multiple records are returned.
3547 38. The dnslists ACL condition has been extended: it is now possible to supply
3548 a list of IP addresses and/or domains to be looked up in a particular DNS
3551 39. Added log_selector=+queue_time_overall.
3553 40. When running the queue in the test harness, wait just a tad after forking a
3554 delivery process, to get repeatability of debugging output.
3556 41. Include certificate and key file names in error message when GnuTLS fails
3557 to set them up, because the GnuTLS error message doesn't include the name
3558 of the failing file when there is a problem reading it.
3560 42. Allow both -bf and -bF in the same test run.
3562 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3564 44. The "Exiscan patch" is now merged into the mainline Exim source.
3566 45. Sometimes the final signoff response after QUIT could fail to get
3567 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3568 < 0 before doing a fflush(). This bug looks as though it goes back to the
3569 introduction of TLS in release 3.20, but "sometimes" must have been rare
3570 because the tests only now provoked it.
3572 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3573 (this can affect the format of dates).
3575 47. exim_tidydb, when checking for the continued existence of a message for
3576 which it has found a message-specific retry record, was not finding
3577 messages that were in split spool directories. Consequently, it was
3578 deleting retry records that should have stayed in existence.
3580 48. Steve fixed some bugs in eximstats.
3582 49. The SPA authentication driver was not abandoning authentication and moving
3583 on to the next authenticator when an expansion was forced to fail,
3584 contradicting the general specification for all authenticators. Instead it
3585 was generating a temporary error. It now behaves as specified.
3587 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3588 (the order specifies the preference for clients). The order is now AES256,
3589 AES128, 3DES, ARCFOUR128.
3591 51. Small patch to Sieve code - explicitly set From: when generating an
3594 52. Exim crashed if a remote delivery caused a very long error message to be
3595 recorded - for instance if somebody sent an entire SpamAssassin report back
3596 as a large number of 550 error lines. This bug was coincidentally fixed by
3597 increasing the size of one of Exim's internal buffers (big_buffer) that
3598 happened as part of the Exiscan merge. However, to be on the safe side, I
3599 have made the code more robust (and fixed the comments that describe what
3602 53. Now that there can be additional text after "Completed" in log lines (if
3603 the queue_time_overall log selector is set), a one-byte patch to exigrep
3604 was needed to allow it to recognize "Completed" as not the last thing in
3607 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3608 patch that reportedly fixes this has been added. I am not expert enough to
3609 create a test for it. This is what the patch creator wrote:
3611 "I found a little strange behaviour of ldap code when working with
3612 Windows 2003 AD Domain, where users was placed in more than one
3613 Organization Units. When I tried to give exim partial DN, the exit code
3614 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3615 But simultaneously result of request was absolutely normal ldap result,
3616 so I produce this patch..."
3618 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3619 REFERENCE, so I have modified the code to exclude the patch when that macro
3622 55. Some experimental protocols are using DNS PTR records for new purposes. The
3623 keys for these records are domain names, not reversed IP addresses. The
3624 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3625 leaves it alone. Component reversal etc. now happens only for IP addresses.
3628 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3630 57. Double the size of the debug message buffer (to 2048) so that more of very
3631 long debug lines gets shown.
3633 58. The exicyclog utility now does better if the number of log files to keep
3634 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3636 59. Two changes related to the smtp_active_hostname option:
3638 (1) $smtp_active_hostname is now available as a variable.
3639 (2) The default for smtp_banner uses $smtp_active_hostname instead
3640 of $primary_hostname.
3642 60. The host_aton() function is supposed to be passed a string that is known
3643 to be a valid IP address. However, in the case of IPv6 addresses, it was
3644 not checking this. This is a hostage to fortune. Exim now panics and dies
3645 if the condition is not met. A case was found where this could be provoked
3646 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3647 components; fortuitously, this particular loophole had already been fixed
3648 by change 4.50/55 above.
3650 If there are any other similar loopholes, the new check in host_aton()
3651 itself should stop them being exploited. The report I received stated that
3652 data on the command line could provoke the exploit when Exim was running as
3653 exim, but did not say which command line option was involved. All I could
3654 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3655 running as the user.
3658 61. There was a buffer overflow vulnerability in the SPA authentication code
3659 (which came originally from the Samba project). I have added a test to the
3660 spa_base64_to_bits() function which I hope fixes it.
3663 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3664 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3666 63. The daemon start-up calls getloadavg() while still root for those OS that
3667 need the first call to be done as root, but it missed one case: when
3668 deliver_queue_load_max is set with deliver_drop_privilege. This is
3669 necessary for the benefit of the queue runner, because there is no re-exec
3670 when deliver_drop_privilege is set.
3672 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3673 This has been fixed.
3675 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3676 was in use, was not putting the data itself into the right store pool;
3677 consequently, it could be overwritten for a subsequent message in the same
3678 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3681 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3683 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3684 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3685 the function were treating the return as a boolean value, which happened to
3686 work because 0=false and not-0=true, but is not correct code.
3688 68. The host_aton() function was not handling scoped IPv6 addresses (those
3689 with, for example, "%eth0" on the end) correctly.
3691 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3692 negated items (that is, ~something) in unsigned ints. Some compilers
3693 apparently mutter when there is no cast.
3695 70. If an address verification called from an ACL failed, and did not produce a
3696 user-specific message (i.e. there was only a "system" message), nothing was
3697 put in $acl_verify_message. In this situation, it now puts the system
3700 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3701 session; change 4.31/43 added the unwanted input to the log line - except
3702 that it did not do this in the start of session case. It now does.
3704 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3705 This is wrong and can cause the other end to generate a synchronization
3706 error if it is another Exim or anything else that does the synchronization
3707 check. A QUIT command is no longer sent after a timeout.
3709 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3710 during host lookups.
3712 74. The defer_ok option of callout verification was not working if it was used
3713 when verifying addresses in header lines, that is, for this case:
3715 verify = header_sender/callout=defer_ok
3717 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3718 those file descriptors could be used for SMTP connections. If anything
3719 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3720 could be sent to the SMTP client, causing chaos. The daemon now opens
3721 stdin, stdout, and stderr to /dev/null when it puts itself into the
3724 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3725 log by default. The user can override this with suitable Perl magic.
3727 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3728 the log message when discard triggers, was not working for the DATA ACL or
3729 for the non-SMTP ACL.
3731 78. Error message wording change in sieve.c.
3733 79. If smtp_accept_max_per_host was set, the number of connections could be
3734 restricted to fewer than expected, because the daemon was trying to set up
3735 a new connection before checking whether the processes handling previous
3736 connections had finished. The check for completed processes is now done
3737 earlier. On busy systems, this bug wouldn't be noticed because something
3738 else would have woken the daemon, and it would have reaped the completed
3741 80. If a message was submitted locally by a user whose login name contained one
3742 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3743 It caused a spool format error. I have fixed the spool reading code. A
3744 related problem was that the "from" clause in the Received: line became
3745 illegal because of the space(s). It is now covered by ${quote_local_part.
3747 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3750 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3752 83. Patch from Sieve maintainer to fix unterminated string problem in
3753 "vacation" handling.
3755 84. Some minor changes to the Linux configuration files to help with other
3756 OS variants using glibc.
3758 85. One more patch for Sieve to update vacation handling to latest spec.
3761 ----------------------------------------------------
3762 See the note above about the 4.44 and 4.50 releases.
3763 ----------------------------------------------------
3769 1. Change 4.43/35 introduced a bug that caused file counts to be
3770 incorrectly computed when quota_filecount was set in an appendfile
3773 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3774 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3777 3. Give more explanation in the error message when the command for a transport
3778 filter fails to execute.
3780 4. There are several places where Exim runs a non-Exim command in a
3781 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3782 was being done only for the command run by the queryprogram router. It is
3783 now done for all such subprocesses. The other cases are: ${run, transport
3784 filters, and the commands run by the lmtp and pipe transports.
3786 5. Some older OS have a limit of 256 on the maximum number of file
3787 descriptors. Exim was using setrlimit() to set 1000 as a large value
3788 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3789 systems. I've change it so that if it can't get 1000, it tries for 256.
3791 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3792 was an oversight, and furthermore, ever since the addition of extra
3793 controls (e.g. 4.43/32), the checks on when to allow different forms of
3794 "control" were broken. There should now be diagnostics for all cases when a
3795 control that does not make sense is encountered.
3797 7. $recipients is now available in the predata ACL (oversight).
3799 8. Tidy the search cache before the fork to do a delivery from a message
3800 received from the command line. Otherwise the child will trigger a lookup
3801 failure and thereby defer the delivery if it tries to use (for example) a
3802 cached ldap connection that the parent has called unbind on.
3804 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3805 of $address_data from the recipient verification was clobbered by the
3806 sender verification.
3808 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3809 was its contents. (It was OK if the option was not defined at all.)
3811 11. A "Completed" log line is now written for messages that are removed from
3812 the spool by the -Mrm option.
3814 12. $host_address is now set to the target address during the checking of
3815 ignore_target_hosts.
3817 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3818 being passed; this would have caused $sender_host_name to have been used if
3819 matching the list had actually called for a host name (not very likely,
3820 since this list is usually IP addresses). A host name is now passed as
3823 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3824 code that specifies a non-restarting handler (typically sigaction() in
3825 modern systems) in an attempt to fix a rare and obscure crash bug.
3827 15. Narrowed the window for a race in the daemon that could cause it to ignore
3828 SIGCHLD signals. This is not a major problem, because they are used only to
3829 wake it up if nothing else does.
3831 16. A malformed maildirsize file could cause Exim to calculate negative values
3832 for the mailbox size or file count. Odd effects could occur as a result.
3833 The maildirsize information is now recalculated if the size or filecount
3836 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3837 support for a long time. Removed HAVE_SYS_VFS_H.
3839 18. Updated exipick to current release from John Jetmore.
3841 19. Allow an empty sender to be matched against a lookup in an address list.
3842 Previously the only cases considered were a regular expression, or an
3845 20. Exim went into a mad DNS lookup loop when doing a callout where the
3846 host was specified on the transport, if the DNS lookup yielded more than
3849 21. The RFC2047 encoding function was originally intended for short strings
3850 such as real names; it was not keeping to the 75-character limit for
3851 encoded words that the RFC imposes. It now respects the limit, and
3852 generates multiple encoded words if necessary. To be on the safe side, I
3853 have increased the buffer size for the ${rfc2047: expansion operator from
3856 22. Failure to deliver a bounce message always caused it to be frozen, even if
3857 there was an errors_to setting on the router. The errors_to setting is now
3860 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3861 canonical form (fully expanded) before being placed in
3862 $sender_host_address.
3864 24. Updated eximstats to version 1.33
3866 25. Include certificate and key file names in error message when GnuTLS fails
3867 to set them up, because the GnuTLS error message doesn't include the name
3868 of the failing file when there is a problem reading it.
3870 26. Expand error message when OpenSSL has problems setting up cert/key files.
3873 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3874 (this can affect the format of dates).
3876 28. exim_tidydb, when checking for the continued existence of a message for
3877 which it has found a message-specific retry record, was not finding
3878 messages that were in split spool directories. Consequently, it was
3879 deleting retry records that should have stayed in existence.
3881 29. eximstats updated to version 1.35
3882 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3883 1.35 - bugfix such that pie charts by volume are generated correctly
3885 30. The SPA authentication driver was not abandoning authentication and moving
3886 on to the next authenticator when an expansion was forced to fail,
3887 contradicting the general specification for all authenticators. Instead it
3888 was generating a temporary error. It now behaves as specified.
3890 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3891 (the order specifies the preference for clients). The order is now AES256,
3892 AES128, 3DES, ARCFOUR128.
3894 31. Small patch to Sieve code - explicitly set From: when generating an
3897 32. Exim crashed if a remote delivery caused a very long error message to be
3898 recorded - for instance if somebody sent an entire SpamAssassin report back
3899 as a large number of 550 error lines. This bug was coincidentally fixed by
3900 increasing the size of one of Exim's internal buffers (big_buffer) that
3901 happened as part of the Exiscan merge. However, to be on the safe side, I
3902 have made the code more robust (and fixed the comments that describe what
3905 33. Some experimental protocols are using DNS PTR records for new purposes. The
3906 keys for these records are domain names, not reversed IP addresses. The
3907 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3908 leaves it alone. Component reversal etc. now happens only for IP addresses.
3911 34. The host_aton() function is supposed to be passed a string that is known
3912 to be a valid IP address. However, in the case of IPv6 addresses, it was
3913 not checking this. This is a hostage to fortune. Exim now panics and dies
3914 if the condition is not met. A case was found where this could be provoked
3915 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3916 components; fortuitously, this particular loophole had already been fixed
3917 by change 4.50/55 or 4.44/33 above.
3919 If there are any other similar loopholes, the new check in host_aton()
3920 itself should stop them being exploited. The report I received stated that
3921 data on the command line could provoke the exploit when Exim was running as
3922 exim, but did not say which command line option was involved. All I could
3923 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3924 running as the user.
3927 35. There was a buffer overflow vulnerability in the SPA authentication code
3928 (which came originally from the Samba project). I have added a test to the
3929 spa_base64_to_bits() function which I hope fixes it.
3932 36. The daemon start-up calls getloadavg() while still root for those OS that
3933 need the first call to be done as root, but it missed one case: when
3934 deliver_queue_load_max is set with deliver_drop_privilege. This is
3935 necessary for the benefit of the queue runner, because there is no re-exec
3936 when deliver_drop_privilege is set.
3938 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3939 was in use, was not putting the data itself into the right store pool;
3940 consequently, it could be overwritten for a subsequent message in the same
3941 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3944 38. Sometimes the final signoff response after QUIT could fail to get
3945 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3946 < 0 before doing a fflush(). This bug looks as though it goes back to the
3947 introduction of TLS in release 3.20, but "sometimes" must have been rare
3948 because the tests only now provoked it.
3954 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3955 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3956 Now it returns an integer. A number of calls were still expecting a T/F
3957 return. Fortuitously, in all cases, the tests worked in OK situations,
3958 which is the norm. However, things would have gone wrong on any write
3959 failures on the smtp file descriptor. This function is used when sending
3960 messages over SMTP and also when doing verify callouts.
3962 2. When Exim is called to do synchronous delivery of a locally submitted
3963 message (the -odf or -odi options), it no longer closes stderr before doing
3966 3. Implemented the mua_wrapper option.
3968 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3970 5. Implemented the functions header_remove(), header_testname(),
3971 header_add_at_position(), and receive_remove_recipient(), and exported them
3974 6. If an ACL "warn" statement specified the addition of headers, Exim already
3975 inserted X-ACL-Warn: at the start if there was no header name. However, it
3976 was not making this test for the second and subsequent header lines if
3977 there were newlines in the string. This meant that an invalid header could
3978 be inserted if Exim was badly configured.
3980 7. Allow an ACL "warn" statement to add header lines at the start or after all
3981 the Received: headers, as well as at the end.
3983 8. Added the rcpt_4xx retry error code.
3985 9. Added postmaster_mailfrom=xxx to callout verification option.
3987 10. Added mailfrom=xxxx to the callout verification option, for verify=
3990 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3991 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3993 12. Inserted some casts to stop certain compilers warning when using pointer
3994 differences as field lengths or precisions in printf-type calls (mostly
3995 affecting debugging statements).
3997 13. Added optional readline() support for -be (dynamically loaded).
3999 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4000 same clock tick as a message's arrival, so that its received time was the
4001 same as the "first fail" time on the retry record, and that message
4002 remained on the queue past the ultimate address timeout, every queue runner
4003 would try a delivery (because it was past the ultimate address timeout) but
4004 after another failure, the ultimate address timeout, which should have then
4005 bounced the address, did not kick in. This was a "< instead of <=" error;
4006 in most cases the first failure would have been in the next clock tick
4007 after the received time, and all would be well.
4009 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4010 being recognized when the domain list was tested by the match_domain
4011 condition in an expansion string.
4013 16. Added the ${str2b64: operator.
4015 17. Exim was always calling setrlimit() to set a large limit for the number of
4016 processes, without checking whether the existing limit was already
4017 adequate. (It did check for the limit on file descriptors.) Furthermore,
4018 errors from getrlimit() and setrlimit() were being ignored. Now they are
4019 logged to the main and panic logs, but Exim does carry on, to try to do its
4020 job under whatever limits there are.
4022 18. Imported PCRE 5.0.
4024 19. Trivial typo in log message " temporarily refused connection" (the leading
4027 20. If the log selector return_path_on_delivery was set and an address was
4028 redirected to /dev/null, the delivery process crashed because it assumed
4029 that a return path would always be set for a "successful" delivery. In this
4030 case, the whole delivery is bypassed as an optimization, and therefore no
4033 21. Internal re-arrangement: the function for sending a challenge and reading
4034 a response while authentication was assuming a zero-terminated challenge
4035 string. It's now changed to take a pointer and a length, to allow for
4036 binary data in such strings.
4038 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4040 23. Exim was not respecting finduser_retries when seeking the login of the
4041 uid under which it was called; it was always trying 10 times. (The default
4042 setting of finduser_retries is zero.) Also, it was sleeping after the final
4043 failure, which is pointless.
4045 24. Implemented tls_on_connect_ports.
4047 25. Implemented acl_smtp_predata.
4049 26. If the domain in control=submission is set empty, Exim assumes that the
4050 authenticated id is a complete email address when it generates From: or
4051 Sender: header lines.
4053 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4054 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4055 chown and chgrp in /bin and hostname in /usr/bin.
4057 28. Exim was keeping the "process log" file open after each use, just as it
4058 does for the main log. This opens the possibility of it remaining open for
4059 long periods when the USR1 signal hits a daemon. Occasional processlog
4060 errors were reported, that could have been caused by this. Anyway, it seems
4061 much more sensible not to leave this file open at all, so that is what now
4064 29. The long-running daemon process does not normally write to the log once it
4065 has entered its main loop, and it closes the log before doing so. This is
4066 so that log files can straightforwardly be renamed and moved. However,
4067 there are a couple of unusual error situations where the daemon does write
4068 log entries, and I had neglected to close the log afterwards.
4070 30. The text of an SMTP error response that was received during a remote
4071 delivery was being truncated at 512 bytes. This is too short for some of
4072 the long messages that one sometimes sees. I've increased the limit to
4075 31. It is now possible to make retry rules that apply only when a message has a
4076 specific sender, in particular, an empty sender.
4078 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4079 it possible to be selective about when SMTP synchronization is enforced.
4081 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4083 32. Implemented hosts_connection_nolog.
4085 33. Added an ACL for QUIT.
4087 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4090 35. Added mailbox_size and mailbox_filecount to appendfile.
4092 36. Added control = no_multiline_responses to ACLs.
4094 37. There was a bug in the logic of the code that waits for the clock to tick
4095 in the case where the clock went backwards by a substantial amount such
4096 that the microsecond fraction of "now" was more than the microsecond
4097 fraction of "then" (but the whole seconds number was less).
4099 38. Added support for the libradius Radius client library this is found on
4100 FreeBSD (previously only the radiusclient library was supported).
4106 1. When certain lookups returned multiple values in the form name=value, the
4107 quoting of the values was not always being done properly. Specifically:
4108 (a) If the value started with a double quote, but contained no whitespace,
4110 (b) If the value contained whitespace other than a space character (i.e.
4111 tabs or newlines or carriage returns) it was not quoted.
4112 This fix has been applied to the mysql and pgsql lookups by writing a
4113 separate quoting function and calling it from the lookup code. The fix
4114 should probably also be applied to nisplus, ibase and oracle lookups, but
4115 since I cannot test any of those, I have not disturbed their existing code.
4117 2. A hit in the callout cache for a specific address caused a log line with no
4118 reason for rejecting RCPT. Now it says "Previous (cached) callout
4119 verification failure".
4121 3. There was an off-by-one bug in the queryprogram router. An over-long
4122 return line was truncated at 256 instead of 255 characters, thereby
4123 overflowing its buffer with the terminating zero. As well as fixing this, I
4124 have increased the buffer size to 1024 (and made a note to document this).
4126 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4127 when Exim is waiting for an SMTP response from a remote server, Exim
4128 restarts its select() call on the socket, thereby resetting its timeout.
4129 This is not a problem when such interrupts are rare. Somebody set up a cron
4130 job to run exiwhat every 2 minutes, which is less than the normal select()
4131 timeout (5 or 10 minutes). This meant that the select() timeout never
4132 kicked in because it was always reset. I have fixed this by comparing the
4133 time when an interrupt arrives with the time at the start of the first call
4134 to select(). If more time than the timeout has elapsed, the interrupt is
4135 treated as a timeout.
4137 5. Some internal re-factoring in preparation for the addition of Sieve
4138 extensions (by MH). In particular, the "personal" test is moved to a
4139 separate function, and given an option for scanning Cc: and Bcc: (which is
4140 not set for Exim filters).
4142 6. When Exim created an email address using the login of the caller as the
4143 local part (e.g. when creating a From: or Sender: header line), it was not
4144 quoting the local part when it contained special characters such as @.
4146 7. Installed new OpenBSD configuration files.
4148 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4149 try to make them clearer.
4151 9. Callout options, other than the timeout value, were being ignored when
4152 verifying sender addresses in header lines. For example, when using
4154 verify = header_sender/callout=no_cache
4156 the cache was (incorrectly) being used.
4158 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4159 only the Cygwin environment.
4161 11. When return_path_on_delivery was set as a log selector, if different remote
4162 addresses in the same message used different return paths and parallel
4163 remote delivery occurred, the wrong values would sometimes be logged.
4164 (Whenever a remote delivery process finished, the return path value from
4165 the most recently started remote delivery process was logged.)
4167 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4168 lines when AUTH and/or TLS are in use. This is the "received protocol"
4169 field. Exim used to use "asmtp" for authenticated SMTP, without any
4170 indication (in the protocol name) for TLS use. Now it follows the RFC and
4171 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4172 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4173 names appear in log lines as well as in Received: header lines.
4175 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4176 extensions, and comparison tests, and to fix some bugs.
4178 14. Changes to the "personal" filter test:
4180 (1) The test was buggy in that it was just doing the equivalent of
4181 "contains" tests on header lines. For example, if a user's address was
4182 anne@some.where, the "personal" test would incorrectly be true for
4184 To: susanne@some.where
4186 This test is now done by extracting each address from the header in turn,
4187 and checking the entire address. Other tests that are part of "personal"
4188 are now done using regular expressions (for example, to check local parts
4189 of addresses in From: header lines).
4191 (2) The list of non-personal local parts in From: addresses has been
4192 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4193 taken from the Sieve specification recommendations.
4195 (3) If the message contains any header line starting with "List-" it is
4196 treated as non-personal.
4198 (4) The test for "circular" in the Subject: header line has been removed
4199 because it now seems ill-conceived.
4201 15. Minor typos in src/EDITME comments corrected.
4203 16. Installed latest exipick from John Jetmore.
4205 17. If headers_add on a router specified a text string that was too long for
4206 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4207 of string_sprintf() is now avoided.
4209 18. $message_body_size was not set (it was always zero) when running the DATA
4210 ACL and the local_scan() function.
4212 19. For the "mail" command in an Exim filter, no default was being set for
4213 the once_repeat time, causing a random time value to be used if "once" was
4214 specified. (If the value happened to be <= 0, no repeat happened.) The
4215 default is now 0s, meaning "never repeat". The "vacation" command was OK
4216 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4217 (I found it when inspecting the code).
4219 20. There is now an overall timeout for performing a callout verification. It
4220 defaults to 4 times the callout timeout, which applies to individual SMTP
4221 commands during the callout. The overall timeout applies when there is more
4222 than one host that can be tried. The timeout is checked before trying the
4223 next host. This prevents very long delays if there are a large number of
4224 hosts and all are timing out (e.g. when the network connections are timing
4225 out). The value of the overall timeout can be changed by specifying an
4226 additional sub-option for "callout", called "maxwait". For example:
4228 verify = sender/callout=5s,maxwait=20s
4230 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4231 to the end before writing, but this should make it even safer).
4233 22. Exim was forgetting that it had advertised PIPELINING for the second and
4234 subsequent messages on an SMTP connection. It was also not resetting its
4235 memory on STARTTLS and an internal HELO.
4237 23. When Exim logs an SMTP synchronization error within a session, it now
4238 records whether PIPELINING has been advertised or not.
4240 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4241 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4242 rather than long int.
4244 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4246 26. Added the never_mail option to autoreply.
4252 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4253 crash if the getsockname() call failed; this can happen if a connection is
4254 closed very soon after it is established. The problem was simply in the
4255 order in which certain operations were done, causing Exim to try to write
4256 to the SMTP stream before it had set up the file descriptor. The bug has
4257 been fixed by making things happen in the correct order.
4263 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4264 before the connection was closed, thus losing the rejection response.
4266 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4267 some early Solaris releases, but causes trouble in current releases where
4268 socklen_t is defined.
4270 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4273 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4276 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4277 that releasing the top of it at the end releases what was used for sub-
4278 expansions (unless the block got too big). However, discard this block if
4279 the first thing is a variable or header, so that we can use its block when
4280 it is dynamic (useful for very large $message_headers, for example).
4282 6. Lookups now cache *every* query, not just the most recent. A new, separate
4283 store pool is used for this. It can be recovered when all lookup caches are
4284 flushed. Lookups now release memory at the end of their result strings.
4285 This has involved some general refactoring of the lookup sources.
4287 7. Some code has been added to the store_xxx() functions to reduce the amount
4288 of flapping under certain conditions.
4290 8. log_incoming_interface used to affect only the <= reception log lines. Now
4291 it causes the local interface and port to be added to several more SMTP log
4292 lines, for example "SMTP connection from", and rejection lines.
4294 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4296 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4298 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4299 could be overwritten at the end of the current message (or the start of a
4300 new message if it was set in a HELO ACL). The value is now preserved for
4301 the duration of the SMTP connection.
4303 12. If a transport had a headers_rewrite setting, and a matching header line
4304 contained an unqualified address, that address was qualified, even if it
4305 did not match any rewriting rules. The underlying bug was that the values
4306 of the flags that permit the existence of unqualified sender and recipient
4307 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4308 non-local messages, and by -bnq for local messages) were not being
4309 preserved with the message after it was received.
4311 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4312 "next input=" as part of the text comprising the host identity instead of
4313 the correct text. The code was using the same buffer for two different
4314 strings. However, depending on which order the printing function evaluated
4315 its arguments, the bug did not always show up. Under Linux, for example, my
4316 test suite worked just fine.
4318 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4319 doesn't work with some older versions of Perl. It has been changed to "my",
4320 which in any case is probably the better facility to use.
4322 15. A really picky compiler found some instances of statements for creating
4323 error messages that either had too many or two few arguments for the format
4326 16. The size of the buffer for calls to the DNS resolver has been increased
4327 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4328 for addresses that have a lot of PTR records. This alleviates a problem; it
4329 does not fully solve it.
4331 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4332 buffer now truncates the list and logs the incident, which is the same
4333 action as happens when Exim is looking up a host name and its aliases.
4334 Previously in this situation something unpredictable would happen;
4335 sometimes it was "internal error: store_reset failed".
4337 18. If a server dropped the connection unexpectedly when an Exim client was
4338 using GnuTLS and trying to read a response, the client delivery process
4339 crashed while trying to generate an error log message.
4341 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4342 string, for example:
4344 warn message = H1: something\nH2: something
4346 the text was added as a single header line from Exim's point of view
4347 though it ended up OK in the delivered message. However, searching for the
4348 second and subsequent header lines using $h_h2: did not work. This has been
4349 fixed. Similarly, if a system filter added multiple headers in this way,
4350 the routers could not see them.
4352 20. Expanded the error message when iplsearch is called with an invalid key to
4353 suggest using net-iplsearch in a host list.
4355 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4356 ACLs are no longer actually imposed (and a message to that effect is
4359 22. If a "gecos" field in a passwd entry contained escaped characters, in
4360 particular, if it contained a \" sequence, Exim got it wrong when building
4361 a From: or a Sender: header from that name. A second bug also caused
4362 incorrect handling when an unquoted " was present following a character
4363 that needed quoting.
4365 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4366 was not being matched caselessly.
4368 24. Arranged for all hyphens in the exim.8 source to be escaped with
4371 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4372 properly. Recipient callout cache records were still being keyed to include
4373 the sender, even when use_sender was set false. This led to far more
4374 callouts that were necessary. The sender is no longer included in the key
4375 when use_sender is false.
4377 26. Added "control = submission" modifier to ACLs.
4379 27. Added the ${base62d: operator to decode base 62 numbers.
4381 28. dnsdb lookups can now access SRV records.
4383 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4384 the configuration file.
4386 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4387 (-v) mode. This makes the output for a verbose queue run more intelligible.
4389 31. Added a use_postmaster feature to recipient callouts.
4391 32. Added the $body_zerocount variable, containing the number of binary zero
4392 bytes in the message body.
4394 33. The time of last modification of the "new" subdirectory is now used as the
4395 "mailbox time last read" when there is a quota error for a maildir
4398 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4400 35. Added +ignore_unknown as a special item in host lists.
4402 36. Code for decoding IPv6 addresses in host lists is now included, even if
4403 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4404 address was recognized as an IP address, but was then not correctly decoded
4405 into binary, causing unexpected and incorrect effects when compared with
4412 1. Very minor rewording of debugging text in manualroute to say "list of
4413 hosts" instead of "hostlist".
4415 2. If verify=header_syntax was set, and a header line with an unqualified
4416 address (no domain) and a large number of spaces between the end of the
4417 name and the colon was received, the reception process suffered a buffer
4418 overflow, and (when I tested it) crashed. This was caused by some obsolete
4419 code that should have been removed. The fix is to remove it!
4421 3. When running in the test harness, delay a bit after writing a bounce
4422 message to get a bit more predictability in the log output.
4424 4. Added a call to search_tidyup() just before forking a reception process. In
4425 theory, someone could use a lookup in the expansion of smtp_accept_max_
4426 per_host which, without the tidyup, could leave open a database connection.
4428 5. Added the variables $recipient_data and $sender_data which get set from a
4429 lookup success in an ACL "recipients" or "senders" condition, or a router
4430 "senders" option, similar to $domain_data and $local_part_data.
4432 6. Moved the writing of debug_print from before to after the "senders" test
4435 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4436 problems for message scanning, either using a data ACL, or using
4437 local_scan() because the Received: header was not generated till after they
4438 were called (in order to set the time as the time of reception completion).
4439 I have revised the way this works. The header is now generated after the
4440 body is received, but before the ACL or local_scan() are called. After they
4441 are run, the timestamp in the header is updated.
4447 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4448 before starting a queue runner without re-exec. This happened only when
4449 deliver_drop_privilege was set or when the Exim user was set to root. The
4450 effect of the bug was that timeouts during subsequent deliveries caused
4451 crashes instead of being properly handled. The handler is now left at its
4452 default (and expected) setting.
4454 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4455 message, again when deliver_drop_privilege is set or Exim is run as root.
4456 The bug described in (1) was not present in this case, but the tidying up
4457 of the other signals was missing. I have made the two cases consistent.
4459 3. The ignore_target_hosts setting on a manualroute router was being ignored
4460 for hosts that were looked up using the /MX notation.
4462 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4465 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4466 operated on the sender address. After changing the $sender_address to <>
4467 for the sender address verify, Exim was re-instated it as the original
4468 (before rewriting) address, but remembering that it had rewritten it, so it
4469 wasn't rewriting it again. This bug also had the effect of breaking the
4470 sender address verification caching when the sender address was rewritten.
4472 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4473 This has been changed so that if the ip literal address matches
4474 ignore_target_hosts, the router declines.
4476 7. Added expansion conditions match_domain, match_address, and match_local_
4477 part (NOT match_host).
4479 8. The placeholder for the Received: header didn't have a length field set.
4481 9. Added code to Exim itself and to exim_lock to test for a specific race
4482 condition that could lead to file corruption when using MBX delivery. The
4483 issue is with the lockfile that is created in /tmp. If this file is removed
4484 after a process has opened it but before that process has acquired a lock,
4485 there is the potential for a second process to recreate the file and also
4486 acquire a lock. This could lead to two Exim processes writing to the file
4487 at the same time. The added code performs the same test as UW imapd; it
4488 checks after acquiring the lock that its file descriptor still refers to
4489 the same named file.
4491 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4492 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4495 11. Added the smtp_active_hostname option. If used, this will typically be made
4496 to depend on the incoming interface address. Because $interface_address is
4497 not set up until the daemon has forked a reception process, error responses
4498 that can happen earlier (such as "too many connections") no longer contain
4501 12. If an expansion in a condition on a "warn" statement fails because a lookup
4502 defers, the "warn" statement is abandoned, and the next ACL statement is
4503 processed. Previously this caused the whole ACL to be aborted.
4505 13. Added the iplsearch lookup type.
4507 14. Added ident_timeout as a log selector.
4509 15. Added tls_certificate_verified as a log selector.
4511 16. Added a global option tls_require_ciphers (compare the smtp transport
4512 option of the same name). This controls incoming TLS connections.
4514 17. I finally figured out how to make tls_require_ciphers do a similar thing
4515 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4516 before starting the TLS session.
4518 18. Tabs are now shown as \t in -bP output.
4520 19. If the log selector return_path_on_delivery was set, Exim crashed when
4521 bouncing a message because it had too many Received: header lines.
4523 20. If two routers both had headers_remove settings, and the first one included
4524 a superfluous trailing colon, the final name in the first list and the
4525 first name in the second list were incorrectly joined into one item (with a
4526 colon in the middle).
4532 1. Added -C and -D options to the exinext utility, mainly to make it easier
4533 to include in the automated testing, but these could be helpful when
4534 multiple configurations are in use.
4536 2. The exinext utility was not formatting the output nicely when there was
4537 an alternate port involved in the retry record key, nor when there was a
4538 message id as well (for retries that were specific to a specific message
4539 and a specific host). It was also confused by IPv6 addresses, because of
4540 the additional colons they contain. I have fixed the IPv4 problem, and
4541 patched it up to do a reasonable job for IPv6.
4543 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4544 delivery, the log line now contains "pipelined" if PIPELINING was used.
4546 4. An SMTP transport process used to panic and die if the bind() call to set
4547 an explicit outgoing interface failed. This has been changed; it is now
4548 treated in the same way as a connect() failure.
4550 5. A reference to $sender_host_name in the part of a conditional expansion
4551 that was being skipped was still causing a DNS lookup. This no longer
4554 6. The def: expansion condition was not recognizing references to header lines
4555 that used bh_ and bheader_.
4557 7. Added the _cache feature to named lists.
4559 8. The code for checking quota_filecount in the appendfile transport was
4560 allowing one more file than it should have been.
4562 9. For compatibility with Sendmail, the command line option
4570 and sets the incoming protocol and host name (for trusted callers). The
4571 host name and its colon can be omitted when only the protocol is to be set.
4572 Note the Exim already has two private options, -pd and -ps, that refer to
4573 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4574 "s", but I don't think that's a major issue.
4576 10. A number of refactoring changes to the code, none of which should affect
4579 (a) The number of logging options was getting close to filling up the
4580 32-bit word that was used as a bit map. I have split them into two classes:
4581 those that are passed in the argument to log_write(), and those that are
4582 only ever tested independently outside of that function. These are now in
4583 separate 32-bit words, so there is plenty of room for expansion again.
4584 There is no change in the user interface or the logging behaviour.
4586 (b) When building, for example, log lines, the code previously used a
4587 macro that called string_cat() twice, in order to add two strings. This is
4588 not really sufficiently general. Furthermore, there was one instance where
4589 it was actually wrong because one of the argument was used twice, and in
4590 one call a function was used. (As it happened, calling the function twice
4591 did not affect the overall behaviour.) The macro has been replaced by a
4592 function that can join an arbitrary number of extra strings onto a growing
4595 (c) The code for expansion conditions now uses a table and a binary chop
4596 instead of a serial search (which was left over from when there were very
4597 few conditions). Also, it now recognizes conditions like "pam" even when
4598 the relevant support is not compiled in: a suitably worded error message is
4599 given if an attempt is made to use such a condition.
4601 11. Added ${time_interval:xxxxx}.
4603 12. A bug was causing one of the ddress fields not to be passed back correctly
4604 from remote delivery subprocesses. The field in question was not being
4605 subsequently used, so this caused to problems in practice.
4607 13. Added new log selectors queue_time and deliver_time.
4609 14. Might have fixed a bug in maildirsizefile handling that threw up
4610 "unexpected character" debug warnings, and recalculated the data
4611 unnecessarily. In any case, I expanded the warning message to give more
4614 15. Added the message "Restricted characters in address" to the statements in
4615 the default ACL that block characters like @ and % in local parts.
4617 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4618 Three changes have been made:
4620 (a) There was a serious bug; a negative response to MAIL caused the whole
4621 recipient domain to be cached as invalid, thereby blocking all messages
4622 to all local parts at the same domain, from all senders. This bug has
4623 been fixed. The domain is no longer cached after a negative response to
4624 MAIL if the sender used is not empty.
4626 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4629 (c) A new callout option, "use_sender" has been added for people who want
4630 the modified behaviour.
4636 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4639 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4640 indeed breaks things for older releases.
4642 3. Added additional logging to the case where there is a problem reading data
4643 from a filter that is running in a subprocess using a pipe, in order to
4644 try to track down a specific problem.
4646 4. Testing facility fudge: when running in the test harness and attempting
4647 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4648 getting "No route to host". Convert this to a timeout.
4650 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4653 6. Some OS don't have socklen_t but use size_t instead. This affects the
4654 fifth argument of getsockopt() amongst other things. This is now
4655 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4656 can be set for individual OS. I have set it for SunOS5, OSF1, and
4657 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4658 some earlier ones do not.
4660 7. Change 4.30/15 was not doing the test caselessly.
4662 8. The standard form for an IPv6 address literal was being rejected by address
4663 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4664 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4665 this, as well as the form without the "IPv6" on the front (but only when
4666 address literals are enabled, of course).
4668 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4670 10. Exim crashed if a message with an empty sender address specified by -f
4671 encountered a router with an errors_to setting. This could be provoked only
4672 by a command such as
4676 where an empty string was supplied; "<>" did not hit this bug.
4678 11. Installed PCRE release 4.5.
4680 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4681 remained set. It is now erased.
4683 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4684 times from message ids (which are base 36 rather than the normal 62).
4686 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4687 were being counted as actual protocol errors, and logged if the log
4688 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4689 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4690 and DATA following a set of rejected RCPTs do not count as protocol errors.
4691 In other words, Exim assumes they were pipelined, though this may not
4692 actually be the case. Of course, in all cases the client gets an
4693 appropriate error code.
4695 15. If a lookup fails in an ACL condition, a message about the failure may
4696 be available; it is used if testing the ACL cannot continue, because most
4697 such messages specify what the cause of the deferral is. However, some
4698 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4699 that caused an old message to be retained and used if a later statement
4700 caused a defer, replacing the real cause of the deferral.
4702 16. If an IP address had so many PTR records that the DNS lookup buffer
4703 was not large enough to hold them, Exim could crash while trying to process
4704 the truncated data. It now detects and logs this case.
4706 17. Further to 4.21/58, another change has been made: if (and only if) the
4707 first line of a message (the first header line) ends with CRLF, a bare LF
4708 in a subsequent header line has a space inserted after it, so as not to
4709 terminate the header.
4711 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4712 unnecessarily, used atoi() instead of strtol(), and didn't check the
4713 termination when getting file sizes from file names by regex.
4715 19. Completely re-implemented the support for maildirsize files, in the light
4716 of a number of problems with the previous contributed implementation
4717 (4.30/29). In particular:
4719 . If the quota is zero, the maildirsize file is maintained, but no quota is
4722 . If the maildir directory does not exist, it is created before any attempt
4723 to write a maildirsize file.
4725 . The quota value in the file is just a cache; if the quota is changed in
4726 the transport, the new value overrides.
4728 . A regular expression is available for excluding directories from the
4731 20. The autoreply transport checks the characters in options that define the
4732 message's headers; it allows continued headers, but it was checking with
4733 isspace() after an embedded newline instead of explicitly looking for a
4736 21. If all the "regular" hosts to which an address was routed had passed their
4737 expiry times, and had not reached their retry times, the address was
4738 bounced, even if fallback hosts were defined. Now Exim should go on to try
4741 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4742 equivalent code in the SMTP transport. Some hosts send humungous responses
4743 to HELO/EHLO, more than 1024 it seems.
4745 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4746 gives compiler warnings in some environments. I've now done it "properly",
4749 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4750 (because of problems with the built-in one) was declared to return uschar *
4751 instead of char *, causing compiler failure.
4753 25. Fixed a file descriptor leak when processing alias/forward files.
4755 26. Fixed a minor format string issue in dbfn.c.
4757 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4759 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4760 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4763 29. When a sender address is verified, it is cached, to save repeating the test
4764 when there is more than one recipient in a message. However, when the
4765 verification involves a callout, it is possible for different callout
4766 options to be set for different recipients. It is too complicated to keep
4767 track of this in the cache, so now Exim always runs a verification when a
4768 callout is required, relying on the callout cache for the optimization.
4769 The overhead is duplication of the address routing, but this should not be
4772 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4773 to be verified with callout=postmaster, and the main callout worked but the
4774 postmaster check failed, the verification correctly failed. However, if a
4775 subsequent RCPT command asked for sender verification *without* the
4776 postmaster check, incorrect caching caused this verification also to fail,
4779 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4780 it was not caching the DNS options (qualify_single, search_parents) that
4781 were used when the lookup failed. A subsequent lookup with different
4782 options therefore always gave the same answer, though there were cases
4783 where it should not have. (Example: a "domains = !$mx_any" option on a
4784 dnslookup router: the "domains" option is always processed without any
4785 widening, but the router might have qualify_single set.) Now Exim uses the
4786 cached value only when the same options are set.
4788 32. Added John Jetmore's "exipick" utility to the distribution.
4790 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4791 than a timeout (e.g. a certificate is required, and is not provided), an
4792 Exim server now closes the connection immediately. Previously it waited for
4793 the client to close - but if the client is SSL, it seems that they each
4794 wait for each other, leading to a delay before one of them times out.
4796 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4797 maintained 0.8.x compatibility because I don't think many are using it, and
4798 it is clearly obsolete.
4800 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4803 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4804 client certificate was expired. A simple patch fixes this, though I don't
4805 understand the full logic of why the verify callback is called multiple
4808 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4809 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4810 which causes problems with some clients (such as the Certicom SSL Plus
4811 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4812 disables the coutermeasure allowing Eudora to connect."
4814 38. Exim was not checking that a write() to a log file succeeded. This could
4815 lead to Bad Things if a log got too big, in particular if it hit a file
4816 size limit. Exim now panics and dies if it cannot write to a log file, just
4817 as it does if it cannot open a log file.
4819 39. Modified OS/Makefile-Linux so that it now contains
4821 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4823 The two -D definitions ensure that Exim is compiled with large file
4824 support, which makes it possible to handle log files that are bigger than
4827 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4828 instance) a domain was checked against a named list that involved a lookup,
4829 causing $domain_data to be set, then another domain was checked against the
4830 same list, then the first domain was re-checked, the value of $domain_data
4831 after the final check could be wrong. In particular, if the second check
4832 failed, it could be set empty. This bug probably also applied to
4835 41. The strip_trailing_dot option was not being applied to the address given
4836 with the -f command-line option.
4838 42. The code for reading a message's header from the spool was incrementing
4839 $received_count, but never initializing it. This meant that the value was
4840 incorrect (doubled) while delivering a message in the same process in which
4841 it was received. In the most common configuration of Exim, this never
4842 happens - a fresh exec is done - but it can happen when
4843 deliver_drop_privilege is set.
4845 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4846 it now includes up to 150 characters of the unexpected data in the log
4849 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4850 and building data strings. The size of both of these buffers was 10 000
4851 bytes - far larger than anybody would *ever* want, thought I. Needless to
4852 say, somebody hit the limit. I have increased the maximum line length to
4853 20 000 and the maximum data length of concatenated lines to 100 000. I have
4854 also fixed two bugs, because there was no checking on these buffers. Tsk,
4855 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4856 buffer is too small.
4858 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4859 lsearch lookups. Now it does.
4861 46. When parsing a route_list item in a manualroute router, a fixed-length
4862 buffer was used for the list of hosts. I made this 1024 bytes long,
4863 thinking that nobody would ever have a list of hosts that long. Wrong.
4864 Somebody had a whole pile of complicated expansion conditions, and the
4865 string was silently truncated, leading to an expansion error. It turns out
4866 that it is easier to change to an unlimited length (owing to other changes
4867 that have happened since this code was originally written) than to build
4868 structure for giving a limitation error. The length of the item that
4869 expands into the list of hosts is now unlimited.
4871 47. The lsearch lookup could not handle data where the length of text line was
4872 more than 4095 characters. Such lines were truncated, leading to shortened
4873 data being returned. It should now handle lines of any length.
4875 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4876 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4879 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4880 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4881 no longer happens in comment lines. A list of replacements is now placed
4882 at the head of all of the source files, except those whose only change is
4883 to replace PERL_COMMAND in the very first #! line.
4885 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4886 messages on the queue, with a bottom-up merge sort, using code contributed
4887 by Michael Haardt. This should make operations like -bp somewhat faster on
4888 large queues. It won't affect queue runners, except when queue_run_in_order
4891 51. Installed eximstats 1.31 in the distribution.
4893 52. Added support for SRV lookups to the dnslookup router.
4895 53. If an ACL referred to $message_body or $message_body_end, the value was not
4896 reset for any messages that followed in the same SMTP session.
4898 54. The store-handling optimization for building very long strings was not
4899 differentiating between the different store pools. I don't think this
4900 actually made any difference in practice, but I've tidied it.
4902 55. While running the routers to verify a sender address, $sender_address
4903 was still set to the sender address. This is wrong, because when routing to
4904 send a bounce to the sender, it would be empty. Therefore, I have changed
4905 it so that, while verifying a sender address, $sender_address is set to <>.
4906 (There is no change to what happens when verifying a recipient address.)
4908 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4909 target A or AAAA records (if not already returned) without resetting the
4910 qualify_single or search_parents options of the DNS resolver. These are
4911 inappropriate in this case because the targets of MX and SRV records must
4912 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4913 target that, when qualified, matched something in the local domain. These
4914 two options are now turned off when doing these lookups.
4916 57. It seems that at least some releases of Reiserfs (which does not have the
4917 concept of a fixed number of inodes) returns zero and not -1 for the
4918 number of available inodes. This interacted badly with check_spool_inodes,
4919 which assumed that -1 was the "no such thing" setting. What I have done is
4920 to check that the total number of inodes is greater than zero before doing
4921 the test of how many are available.
4923 58. When a "warn" ACL statement has a log_message modifier, the message is
4924 remembered, and not repeated. This is to avoid a lot of repetition when a
4925 message has many recipients that cause the same warning to be written.
4926 Howewer, Exim was preserving the list of already written lines for an
4927 entire SMTP session, which doesn't seem right. The memory is now reset if a
4928 new message is started.
4930 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4931 debugging output unless log_rewrite was also set.
4933 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4934 of (int)(handle) when we know that handle contains (void *)(-1).
4936 61. The Exim daemon panic-logs an error return when it closes the incoming
4937 connection. However "connection reset by peer" seems to be common, and
4938 isn't really an error worthy of noting specially, so that particular error
4941 62. When Exim is trying to find all the local interfaces, it used to panic and
4942 die if the ioctl to get the interface flags failed. However, it seems that
4943 on at least one OS (Solaris 9) it is possible to have an interface that is
4944 included in the list of interfaces, but for which you get a failure error
4945 for this call. This happens when the interface is not "plumbed" into a
4946 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4947 failure of the "get flags" call assumes that the interface is down.
4949 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4950 makes life easier for people who are doing arithmetic on fields extracted
4951 from dates, where you often get leading zeros that should not be
4952 interpreted as octal.
4954 64. Added qualify_domain to the redirect router, to override the global
4957 65. If a pathologically long header line contained very many addresses (the
4958 report of this problem mentioned 10 000) and each of them was rewritten,
4959 Exim could use up a very large amount of memory. (It kept on making new
4960 copies of the header line as it rewrote, and never released the old ones.)
4961 At the expense of a bit more processing, the header rewriting function has
4962 been changed so that it no longer eats memory in this way.
4964 66. The generation of the Received: header has been moved from the time that a
4965 message starts to be received, to the time that it finishes. The timestamp
4966 in the Received: header should now be very close to that of the <= log
4967 line. There are two side-effects of this change:
4969 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4970 logged header lines no longer include the local Received: line, because
4971 it has not yet been created. The same applies to a copy of the message
4972 that is returned to a non-SMTP sender when a message is rejected.
4974 (b) When a filter file is tested using -bf, no additional Received: header
4975 is added to the test message. After some thought, I decided that this
4978 This change does not affect the value of $received_for. It is still set
4979 after address rewriting, but before local_scan() is called.
4981 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4983 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4984 gave an unhelpful panic error message, and a defer error. I have managed to
4985 change this behaviour so that it now rejects any supplied certificate,
4986 which seems right, as the list of acceptable certificates is empty.
4988 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4989 gave an unhelpful defer error. I have not managed to make this reject any
4990 supplied certificates, but the error message it gives is "no certificate
4991 supplied", which is not helpful.
4993 70. exigrep's output now also includes lines that are not associated with any
4994 message, but which match the given pattern. Implemented by a patch from
4995 Martin Sluka, which also tidied up the Perl a bit.
4997 71. Recipient callout verification, like sender verification, was using <> in
4998 the MAIL FROM command. This isn't really the right thing, since the actual
4999 sender may affect whether the remote host accepts the recipient or not. I
5000 have changed it to use the actual sender in the callout; this means that
5001 the cache record is now keyed on a recipient/sender pair, not just the
5002 recipient address. There doesn't seem to be a real danger of callout loops,
5003 since a callout by the remote host to check the sender would use <>.
5004 [SEE ABOVE: changed after hitting problems.]
5006 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5007 temporary errors. However, in the case of such a code being given after
5008 the end of a data transmission (i.e. after ".") Exim was failing to write
5009 a retry record for the message. (Yes, there was some broken host that was
5010 actually sending 8xx at this point.)
5012 73. An unknown lookup type in a host list could cause Exim to panic-die when
5013 the list was checked. (An example that provoked this was putting <; in the
5014 middle of a list instead of at the start.) If this happened during a DATA
5015 ACL check, a -D file could be left lying around. This kind of configuration
5016 error no longer causes Exim to die; instead it causes a defer errror. The
5017 incident is still logged to the main and panic logs.
5019 74. Buglet left over from Exim 3 conversion. The message "too many messages
5020 in one connection" was written to the rejectlog but not the mainlog, except
5021 when address rewriting (yes!) was being logged.
5023 75. Added write_rejectlog option.
5025 76. When a system filter was run not as root (that is, when system_filter_user
5026 was set), the values of the $n variables were not being returned to the
5027 main process; thus, they were not subsequently available in the $sn
5030 77. Added +return_path_on_delivery log selector.
5032 78. A connection timeout was being treated differently from recipients deferred
5033 when testing hosts_max_try with a message that was older than the host's
5034 retry timeout. (The host should not be counted, thus allowing all hosts to
5035 be tried at least once before bouncing.) This may have been the cause of an
5036 occasionally reported bug whereby a message would remain on the queue
5037 longer than the retry timeout, but would be bounced if a delivery was
5038 forced. I say "may" because I never totally pinned down the problem;
5039 setting up timeout/retry tests is difficult. See also the next item.
5041 79. The ultimate address timeout was not being applied to errors that involved
5042 a combination of host plus message (for example, a timeout on a MAIL
5043 command). When an address resolved to a number of possible hosts, and they
5044 were not all tried for each delivery (e.g. because of hosts_max_try), a
5045 message could remain on the queue longer than the retry timeout.
5047 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5050 81. Fixed an obscure SMTP outgoing bug which required at least the following
5051 conditions: (a) there was another message waiting for the same server;
5052 (b) the server returned 5xx to all RCPT commands in the first message so
5053 that the message was not completed; (c) the server dropped the connection
5054 or gave a negative response to the RSET that Exim sends to abort the
5055 transaction. The observed case was a dropped connection after DATA that had
5056 been sent in pipelining mode. That is, the server had advertised PIPELINING
5057 but was not implementing it correctly. The effect of the bug was incorrect
5058 behaviour, such as trying another host, and this could lead to a crash.
5064 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5065 and daemon.c were passed as pointers to ints; they should have been
5066 pointers to socklen_t variables (which are typically unsigned ints).
5068 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5071 3. Fixed a really odd bug that affected only the testing scheme; patching a
5072 certain fixed string in the binary changed the value of another string that
5073 happened to be identical to the end of the original first string.
5075 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5076 name", it returns that address as the IP address. On some operating
5077 systems (e.g. Solaris), it also passes back the IP address string as the
5078 "host name". However, on others (e.g. Linux), it passes back an empty
5079 string. Exim wasn't checking for this, and was changing the host name to an
5080 empty string, assuming it had been canonicized.
5082 5. Although rare, it is permitted to have more than one PTR record for a given
5083 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5084 all the names associated with an address, because they do in Solaris.
5085 However, it seems that they do not in Linux for data that comes from the
5086 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5087 I found this out when I moved to a new Linux workstation and tried to run
5088 the Exim test suite.
5090 To get round this problem I have changed the code so that it now does its
5091 own call to the DNS to look up PTR records when searching for a host name.
5092 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5093 addresses that are only in /etc/hosts are still found.
5095 This behaviour is, however, controlled by an option called host_lookup_
5096 order, which defaults to "bydns:byaddr". If people want to use the other
5097 order, or indeed, just use one or the other means of lookup, they can
5098 specify it in this variable.
5100 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5101 some operating systems, this comes back from gethostbyaddr() as an empty
5102 string, and this is what Exim used to test for. However, it seems that in
5103 other systems, "." is yielded. Exim now tests for this case too.
5105 7. The values of check_spool_space and check_log_space are now held internally
5106 as a number of kilobytes instead of an absolute number of bytes. If a
5107 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5108 kilobyte. This means that much larger values can be stored.
5110 8. Exim monitor: an attempt to get the action menu when not actually pointing
5111 at a message produces an empty menu entitled "No message selected". This
5112 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5113 no entries in it ("Shell widget menu has zero width and/or height"). So I
5114 have added a single, blank menu entry in this case.
5116 9. Added ${quote_local_part.
5118 10. MIME decoding is now applied to the contents of Subject: header lines when
5121 11. Now that a reference to $sender_host_address automatically causes a reverse
5122 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5123 host lookup before query-style lookups in lists that might use this
5124 variable. This has therefore been abolished, and the "net-" prefix is no
5125 longer necessary for query-style lookups.
5127 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5128 have been a typo for LFLAGS, so it has been changed.
5130 13. The install script calls Exim with "-C /dev/null" in order to find the
5131 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5132 to be output. Howeve, since Exim outputs its version number before the
5133 error, it didn't break the script. It just looked ugly. I fixed this by
5134 always allowing "-C /dev/null" if the caller is root.
5136 14. Ignore overlarge ACL variable number when reading spool file - insurance
5137 against a later release with more variables having written the file.
5139 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5140 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5141 this, as well as the form without the "IPv6" on the front.
5143 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5144 OS/Makefile-Darwin file.
5146 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5147 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5149 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5150 variable. However, some OS use ints for inodes. I've added cast to long int
5151 to get rid of the compiler warning.
5153 19. I had forgotten to lock out "/../" in configuration file names when
5154 ALT_CONFIG_PREFIX was set.
5156 20. Routers used for verification do not need to specify transports. However,
5157 if such a router generated a host list, and callout was configured, Exim
5158 crashed, because it could not find a port number from the (non-existent)
5159 transport. It now assumes port 25 in this circumstance.
5161 21. Added the -t option to exigrep.
5163 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5164 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5165 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5166 them is set, LOOKUP_LSEARCH is forced.
5168 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5170 24. Added sender and host information to the "rejected by local_scan()" log
5171 line; previously there was no indication of these.
5173 25. Added .include_if_exists.
5175 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5176 Linux. It turns out that not all file systems support this. Apparently some
5177 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5178 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5179 not supported on the file descriptor, is now ignored when Exim is trying to
5180 sync a directory. This applies only to Linux.
5182 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5184 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5185 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5188 29. Added support for maildirsize files from supplied patch (modified a bit).
5190 30. The use of :fail: followed by an empty string could lead Exim to respond to
5191 sender verification failures with (e.g.):
5193 550 Verification failed for <xxx>
5194 550 Sender verify failed
5196 where the first response line was missing the '-' that indicates it is not
5197 the final line of the response.
5199 31. The loop for finding the name of the user that called Exim had a hardwired
5200 limit of 10; it now uses the value of finduser_retries, which is used for
5201 all other user lookups.
5203 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5206 33. Exim was neglecting to zero errno before one call of strtol() when
5207 expanding a string and expecting an integer value. On some systems this
5208 resulted in spurious "integer overflow" errors. Also, it was casting the
5209 result into an int without checking.
5211 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5212 did not work. The code looks as if it has *never* worked, though it appears
5213 to have been documented since at least releast 1.62. I have made it work.
5215 35. The "timeout_DNS" error in retry rules, also documented since at least
5216 1.62, also never worked. As it isn't clear exactly what this means, and
5217 clearly it isn't a major issue, I have abolished the feature by treating it
5218 as "timeout", and writing a warning to the main and panic logs.
5220 36. The display of retry rules for -brt wasn't always showing the error code
5223 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5224 timeout_connect_A, timeout_connect_MX.
5226 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5227 to the empty sender.
5229 39. The daemon was not analysing the content of -oX till after it had closed
5230 stderr and disconnected from the controlling terminal. This meant that any
5231 syntax errors were only noted on the panic log, and the return code from
5232 the command was 0. By re-arranging the code a little, I've made the
5233 decoding happen first, so such errors now appear on stderr, and the return
5234 code is 1. However, the actual setting up of the sockets still happens in
5235 the disconnected process, so errors there are still only recorded on the
5238 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5239 connections (as happens on some IP stacks) was logged at start up time as
5240 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5241 it from "IPv6 and IPv4", which means that two separate sockets are being
5244 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5245 says whether AF_INET or AF_INET6 was passed as an argument.
5247 42. Exiwhat output was messed up when time zones were included in log
5250 43. Exiwhat now gives more information about the daemon's listening ports,
5251 and whether -tls-on-connect was used.
5253 44. The "port" option of the smtp transport is now expanded.
5255 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5256 silently ignored. Now an error message is written to the main and panic
5259 46. There's a new ACL modifier called "logwrite" which writes to a log file
5260 as soon as it is encountered.
5262 47. Added $local_user_uid and $local_user_gid at routing time.
5264 48. Exim crashed when trying to verify a sender address that was being
5267 49. Exim was recognizing only a space character after ".include". It now also
5268 recognizes a tab character.
5270 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5271 extracting the relevant information from the specification. The man page no
5272 longer contains scrambled data for the -d option, and I've added a section
5273 at the front about calling Exim under different names.
5275 51. Added "extra_headers" argument to the "mail" command in filter files.
5277 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5280 53. Installed eximstats 1.29.
5282 54. Added transport_filter_timeout as a generic transport option.
5284 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5285 Cc: header lines. This was required by RFC 822, but it not required by RFC
5288 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5289 incoming messages that did not have them. Now it does so only if the
5290 message originates locally, that is, if there is no associated remote host
5291 address. When Resent- header lines are present, this applies to the Resent-
5292 lines rather than the non-Resent- lines.
5294 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5295 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5297 58. Messages for configuration errors now include the name of the main
5298 configuration file - useful now that there may be more than one file in a
5299 list (.included file names were always shown).
5301 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5302 for those rare installations that do not start the daemon as root or run it
5303 setuid root. I've cut out the call to initgroups() if the daemon is not
5306 60. The Exim user and group can now be bound into the binary as text strings
5307 that are looked up at the start of Exim's processing.
5309 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5311 62. Added $mailstore_basename variable.
5313 63. Installed patch to sieve.c from Michael Haardt.
5315 64. When Exim failed to open the panic log after failing to open the main log,
5316 the original message it was trying to log was written to stderr and debug
5317 output, but if they were not available (the usual case in production), it
5318 was lost. Now it is written to syslog before the two lines that record the
5319 failures to open the logs.
5321 65. Users' Exim filters run in subprocesses under the user's uid. It is
5322 possible for a "deliver" command or an alias in a "personal" command to
5323 provoke an address rewrite. If logging of address rewriting is configured,
5324 this fails because the process is not running as root or exim. There may be
5325 a better way of dealing with this, but for the moment (because 4.30 needs
5326 to be released), I have disabled address rewrite logging when running a
5327 filter in a non-root, non-exim process.
5333 1. The buildconfig auxiliary program wasn't quoting the value set for
5334 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5335 not defined. This bug was masked in 4.22 by the effect that was fixed in
5338 2. Some messages that were rejected after a message id was allocated were
5339 shown as "incomplete" by exigrep. It no longer does this for messages that
5340 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5342 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5343 have allow_domain_literals set, the ID did not get logged in the <= line.
5344 Domain literals are now always recognized in Message-ID: header lines.
5346 4. The first argument for a ${extract expansion item is the key name or field
5347 number. Leading and trailing spaces in this item were not being ignored,
5348 causing some misleading effects.
5350 5. When deliver_drop_privilege was set, single queue runner processes started
5351 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5352 same command in the process it spins off) were not dropping privilege.
5354 6. When the daemon running as "exim" started a queue runner, it always
5355 re-executed Exim in the spun-off process. This is a waste of effort when
5356 deliver_drop_privilege is set. The new process now just calls the
5357 queue-runner function directly.
5363 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5366 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5367 ignored. Though the use of -oP was forcing the writing of a pid file, it
5368 was always written to the default place.
5370 3. If the message "no IP address found for host xxxx" is generated during
5371 incoming verification, it is now followed by identification of the incoming
5372 connection (so you can more easily find what provoked it).
5374 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5376 5. Added some features to "harden" Exim a bit more against certain attacks:
5378 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5379 be put in Local/Makefile. This is like the never_users runtime option,
5380 but it cannot be overridden. The default setting is "root".
5382 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5383 prefix string with which any file named in a -C command line option
5386 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5387 is retained for -C and -D only if the caller of Exim is root. Without
5388 it, the exim user may also use -C and -D and retain privilege.
5390 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5391 command line option is disabled.
5393 6. Macro names set by the -D option must start with an upper case letter, just
5394 like macro names defined in the configuration file.
5396 7. Added "dereference=" facility to LDAP.
5398 8. Two instances of the typo "uknown" in the source files are fixed.
5400 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5401 the Configure-Makefile script screwed up while processing it.
5403 10. Incorporated PCRE 4.4.
5405 11. The SMTP synchronization check was not operating right at the start of an
5406 SMTP session. For example, it could not catch a HELO sent before the client
5407 waited for the greeting. There is now a check for outstanding input at the
5408 point when the greeting is written. Because of the duplex, asynchronous
5409 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5410 way, but not yet received, when the check is performed.
5412 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5413 on TCP/IP sockets, because this apparently causes some broken clients to
5416 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5417 unchanged) from the Cygwin maintainer.
5419 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5420 was supported instead of testing for mbx. Effectively a typo.
5422 15. The spa authenticator server code was not checking that the input it
5423 received was valid base64.
5425 16. The debug output line for the "set" modifier in ACLs was not showing the
5426 name of the variable that was being set.
5428 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5430 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5431 lookup on its own. Something else was needed to trigger the lookup. For
5432 example, a match in host_lookup or the need for a host name in a host list.
5433 Now, if $sender_host_name is referenced and the host name has not yet been
5434 looked up, a lookup is performed. If the lookup fails, the variable remains
5435 empty, and $host_lookup_failed is set to "1".
5437 19. Added "eqi" as a case-independent comparison operator.
5439 20. The saslauthd authentication condition could segfault if neither service
5440 nor realm was specified.
5442 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5443 error message that was logged was misleading, and incoming SMTP
5444 connections were dropped. The message is now more accurate, and temporary
5445 errors are given to SMTP connections.
5447 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5448 (except RSET) with a 421 error, until QUIT is received. However, it was
5449 failing to send a response to QUIT.
5451 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5452 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5453 ACL is now run after the helo_try_verify_hosts code.
5455 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5456 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5457 case-independent, so other case variants are also recognized). Apparently
5458 some systems use these upper case variants.
5460 25. If more than two messages were waiting for the same host, and a transport
5461 filter was specified for the transport, Exim sent two messages over the
5462 same TCP/IP connection, and then failed with "socket operation on non-
5463 socket" when it tried to send the third.
5465 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5467 27. The extern definition of crypt16() in expand.c was not being excluded when
5468 the OS had its own crypt16() function.
5470 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5471 as a preferred synonym for return_size_limit, both as an option and as an
5474 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5476 30. Changed the default configuration ACL to relax the local part checking rule
5477 for addresses that are not in any local domains. For these addresses,
5478 slashes and pipe symbols are allowed within local parts, but the sequence
5479 /../ is explicitly forbidden.
5481 31. SPA server authentication was not clearing the challenge buffer before
5484 32. log_message in a "warn" ACL statement was writing to the reject log as
5485 well as to the main log, which contradicts the documentation and doesn't
5486 seem right (because no rejection is happening). So I have stopped it.
5488 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5489 However, I am unable to do any testing of this.
5491 34. Fixed an infelicity in the appendfile transport. When checking directories
5492 for a mailbox, to see if any needed to be created, it was accidentally
5493 using path names with one or more superfluous leading slashes; tracing
5494 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5496 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5497 contents are added to the log line that is written for every discarded
5498 recipient. (Previously a log_message setting was ignored.)
5500 36. The ${quote: operator now quotes the string if it is empty.
5502 37. The install script runs exim in order to find its version number. If for
5503 some reason other than non-existence or emptiness, which it checks, it
5504 could not run './exim', it was installing it with an empty version number,
5505 i.e. as "exim-". This error state is now caught, and the installation is
5508 38. An argument was missing from the function that creates an error message
5509 when Exim fails to connect to the socket for saslauthd authentication.
5510 This could cause Exim to crash, or give a corrupted message.
5512 39. Added isip, isip4, and isip6 to ${if conditions.
5514 40. The ACL variables $acl_xx are now saved with the message, and can be
5515 accessed later in routers, transports, and filters.
5517 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5518 strings in the file are not string-expanded.
5520 42. If a MAIL command specified a SIZE value that was too large to fit into an
5521 int variable, the check against message_size_limit failed. Such values are
5522 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5523 day this will have to be increased, but I don't think I want to be around
5524 when emails are that large.
5531 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5532 iconv() is not standard in FreeBSD.
5534 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5535 IPv6 enabled. The observed symptom was a segmentation fault on return from
5536 the function os_common_find_running_interfaces() in src/os.c.
5538 3. In the check_special_case() function in daemon.c I had used "errno" as an
5539 argument name, which causes warnings on some systems. This was basically a
5540 typo, since it was named "eno" in the comments!
5542 4. The code that waits for the clock to tick (at a resolution of some fraction
5543 of a second) so as to ensure message-id uniqueness was always waiting for
5544 at least one whole tick, when it could have waited for less. [This is
5545 almost certainly not relevant at current processor speeds, where it is
5546 unlikely to ever wait at all. But we try to future-proof.]
5548 5. The function that sleeps for a time interval that includes fractions of a
5549 second contained a race. It did not block SIGALRM between setting the
5550 timer, and suspending (a couple of lines later). If the interval was short
5551 and the sigsuspend() was delayed until after it had expired, the suspension
5552 never ended. On busy systems this could lead to processes getting stuck for
5555 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5556 process, before it forks any delivery processes. The open lookup caching
5557 mechanism meant that the open file or database connection was passed into
5558 the delivery process. The problem was that delivery processes always tidy
5559 up cached lookup data. This could cause a problem for the next delivery
5560 process started by the queue runner, because the external queue runner
5561 process does not know about the closure. So the next delivery process
5562 still has data in the lookup cache. In the case of a file lookup, there was
5563 no problem because closing a file descriptor in a subprocess doesn't affect
5564 the parent. However, if the lookup was caching a connection to a database,
5565 the connection was closed, and the second delivery process was likely to
5566 see errors such as "PGSQL: query failed: server closed the connection
5567 unexpectedly". The problem has been fixed by closing all cached lookups
5568 in a queue runner before running a delivery process.
5570 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5571 seem to have the "const" qualifier which it has on other OS. I've
5574 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5575 *of the same type* (client or server) with the same public name that an
5576 error should be diagnosed.
5578 9. When Exim looked up a host name for an IP address, but failed to find the
5579 original IP address when looking up the host name (a safety check), it
5580 output the message "<ip address> does not match any IP for NULL", which was
5581 confusing, to say the least. The bug was that the host name should have
5582 appeared instead of "NULL".
5584 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5585 user that is built into the binary, and the -C or -D options is used, root
5586 privilege is dropped before the configuration file is read. In addition,
5587 logging is switched to stderr instead of the normal log files. If the
5588 configuration then re-defines the Exim user, the unprivileged environment
5589 is probably not what is expected, so Exim logs a panic warning message (but
5592 However, if deliver_drop_privilege is set, the unprivileged state may well
5593 be exactly what is intended, so the warning has been cut out in that case,
5594 and Exim is allowed to try to write to its normal log files.
5600 1. smtp_return_error_details was not giving details for temporary sender
5601 or receiver verification errors.
5603 2. Diagnose a configuration error if two authenticators have the same public
5606 3. Exim used not to create the message log file for a message until the first
5607 delivery attempt. This could be confusing when incoming messages were held
5608 for policy or load reasons. The message log file is now created at the time
5609 the message is received, and an initial "Received" line is written to it.
5611 4. The automatically generated man page for command line options had a minor
5612 bug that caused no ill effects; however, a more serious problem was that
5613 the procedure for building the man page automatically didn't always
5614 operate. Consequently, release 4.20 contains an out-of-date version. This
5615 shouldn't happen again.
5617 5. When building Exim with embedded Perl support, the script that builds the
5618 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5619 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5621 6. The freeze_tell option was not being used for messages that were frozen on
5622 arrival, either by an ACL or by local_scan().
5624 7. Added the smtp_incomplete_transaction log selector.
5626 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5627 was accepting AUTH without a new EHLO.
5629 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5630 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5633 10. Logging of TCP/IP connections (when configured) now happens in the main
5634 daemon process instead of the child process, so that the TCP/IP connection
5635 count is more accurate (but it can never be perfect).
5637 11. The use of "drop" in a nested ACL was not being handled correctly in the
5638 outer ACL. Now, if condition failure induced by the nested "drop" causes
5639 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5640 or "require"), the connection is dropped.
5642 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5643 that yield "discard" can now be used with an "accept" or a "discard" verb,
5644 but an error is generated for any others (because I can't see a useful way
5645 to define what should happen).
5647 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5648 are now processed in the same way as lines in the Exim configuration file.
5649 In particular, continuation lines are supported.
5651 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5653 15. Added -ti meaning -t -i.
5655 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5656 domains, and warn by logging if others are found.
5658 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5659 returned from GIFCONF when Exim is trying to find the list of interfaces on
5660 a host. The code in os.c has been modified to copy each ifreq to an aligned
5661 structure in all cases.
5663 Also, in some cases, the returned ifreq's were being copied to a 'struct
5664 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5665 means the last couple of bytes of an IPv6 address could be chopped if the
5666 ifreq contained only a normal sockaddr (14 bytes storage).
5668 18. Named domain lists were not supported in the hosts_treat_as_local option.
5669 An entry such as +xxxx was not recognized, and was treated as a literal
5672 19. Ensure that header lines added by a DATA ACL are included in the reject log
5673 if the ACL subsequently rejects the message.
5675 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5676 MD5 (which is deprecated).
5678 21. When testing a filter file using -bf, Exim was writing a message when it
5679 took the sender from a "From " line in the message, but it was not doing so
5680 when it took $return_path from a Return-Path: header line. It now does.
5682 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5683 with a valid header line field name (a series of printing characters
5684 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5686 23. Changed "disc" in the source to "disk" to conform to the documentation and
5687 the book and for uniformity.
5689 24. Ignore Sendmail's -Ooption=value command line item.
5691 25. When execve() failed while trying to run a command in a pipe transport,
5692 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5693 could be confused with a return value of 69 from the command itself. This
5694 has been changed to 127, the value the shell returns if it is asked to run
5695 a non-existent command. The wording for the related log line suggests a
5696 non-existent command as the problem.
5698 26. If received_header_text expands to an empty string, do not add a Received:
5699 header line to the message. (Well, it adds a token one on the spool, but
5700 marks it "old" so that it doesn't get used or transmitted.)
5702 27. Installed eximstats 1.28 (addition of -nt option).
5704 28. There was no check for failure on the call to getsockname() in the daemon
5705 code. This can fail if there is a shortage of resources on the system, with
5706 ENOMEM, for example. A temporary error is now given on failure.
5708 29. Contrary to the C standard, it seems that in some environments, the
5709 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5710 program. Exim now does this explicitly; it affects the formatting of
5711 timestamps using strftime().
5713 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5714 complaints. I've now initialized all the variables, to avoid this.
5716 32. Header lines added by a system filter were not being "seen" during
5717 transport-time rewrites.
5719 33. The info_callback() function passed to OpenSSL is set up with type void
5720 (*)(SSL *, int, int), as described somewhere. However, when calling the
5721 function (actually a macro) that sets it up, the type void(*)() is
5722 expected. I've put in a cast to prevent warnings from picky compilers.
5724 34. If a DNS black list lookup found a CNAME record, but there were no A
5725 records associated with the domain it pointed at, Exim crashed.
5727 35. If a DNS black list lookup returned more than one A record, Exim ignored
5728 all but the first. It now scans all returned addresses if a particular IP
5729 value is being sought. In this situation, the contents of the
5730 $dnslist_value variable are a list of all the addresses, separated by a
5733 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5734 to accept a host name and all its aliases if the forward lookup for any of
5735 them yielded the IP address of the incoming connection. Now it accepts only
5736 those names whose forward lookup yields the correct IP address. Any other
5737 names are discarded. This closes a loophole whereby a rogue DNS
5738 administrator could create reverse DNS records to break through a
5739 wildcarded host restriction in an ACL.
5741 37. If a user filter or a system filter that ran in a subprocess used any of
5742 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5743 the wrong values were passed to the pipe command ($thisaddress had the
5744 value of $0, $0 had the value of $1, etc). This bug was introduced by
5745 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5747 38. Improved the line breaking for long SMTP error messages from ACLs.
5748 Previously, if there was no break point between 40 and 75 characters, Exim
5749 left the rest of the message alone. Two changes have been made: (a) I've
5750 reduced the minimum length to 35 characters; (b) if it can't find a break
5751 point between 35 and 75 characters, it looks ahead and uses the first one
5752 that it finds. This may give the occasional overlong line, but at least the
5753 remaining text gets split now.
5755 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5756 file descriptors might be low, and that setting 1000 would always raise it.
5757 It turns out that in some environments, the limit is already over 1000 and
5758 that lowering it causes trouble. So now Exim takes care not to decrease it.
5760 40. When delivering a message, the value of $return_path is set to $sender_
5761 address at the start of routing (routers may change the value). By an
5762 oversight, this default was not being set up when an address was tested by
5763 -bt or -bv, which affected the outcome if any router or filter referred to
5766 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5767 log only when "message" or "log_message" are set. However, if one of the
5768 conditions was an address verification, or a call to a nested ACL, the
5769 messages generated by the underlying test were being passed through. This
5770 no longer happens. The underlying message is available in $acl_verify_
5771 message for both "message" and "log_message" expansions, so it can be
5772 passed through if needed.
5774 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5775 new expansion $bh_ to give the encoded byte string without charset
5776 translation. Translation happens only if iconv() is available; HAVE_ICONV
5777 indicates this at build time. HEADERS_CHARSET gives the charset to
5778 translate to; headers_charset can change it in the configuration, and
5779 "headers charset" can change it in an individual filter file.
5781 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5782 that creates RFC 2047 encoded "words" labels them as that charset instead
5783 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5784 expansion operator; (ii) when Exim creates a From: line for a local
5785 message; (iii) when a header line is rewritten to include a "phrase" part.
5787 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5788 buggy, causing it to skip the first lines of messages whose message ID
5789 ended in 'D'. This would not have bitten before Exim release 4.14, because
5790 message IDs were unlikely to end in 'D' before then. The effect was to have
5791 incorrect size information for certain domains.
5793 45. #include "config.h" was missing at the start of the crypt16.c module. This
5794 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5797 46. If there was a timeout during a "random" callout check, Exim treated it as
5798 a failure of the random address, and carried on sending RSET and the real
5799 address. If the delay was just some slowness somewhere, the response to the
5800 original RCPT would be taken as a response to RSET and so on, causing
5801 mayhem of various kinds.
5803 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5804 when I implemented it. It didn't allow for the fact that some option values
5805 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5806 the right test for positive values.
5808 48. Domain names in DNS records are case-independent. Exim always looks them up
5809 in lower case. Some resolvers return domain names in exactly the case they
5810 appear in the zone file, that is, they may contain uppercase letters. Not
5811 all resolvers do this - some return always lower case. Exim was treating a
5812 change of case by a resolver as a change of domain, similar to a widening
5813 of a domain abbreviation. This triggered its re-routing code and so it was
5814 trying to route what was effectively the same domain again. This normally
5815 caused routing to fail (because the router wouldn't handle the domain
5816 twice). Now Exim checks for this case specially, and just changes the
5817 casing of the domain that it ultimately uses when it transmits the message
5820 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5823 50. If a filter generated a file delivery with a non-absolute name (possible if
5824 no home directory exists for the router), the forbid_file option was not
5827 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5828 the existing equality matching.
5830 52. Exim was using ints instead of ino_t variables in some places where it was
5831 dealing with inode numbers.
5833 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5834 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5835 called TMPDIR, and if it finds it is different, it changes its value.
5837 54. The smtp_printf() function is now made available to local_scan() so
5838 additional output lines can be written before returning. There is also an
5839 smtp_fflush() function to enable the detection of a dropped connection.
5840 The variables smtp_input and smtp_batched_input are exported to
5843 55. Changed the default runtime configuration: the message "Unknown user"
5844 has been removed from the ACL, and instead placed on the localuser router,
5845 using the cannot_route_message feature. This means that any verification
5846 failures that generate their own messages won't get overridden. Similarly,
5847 the "Unrouteable address" message that was in the ACL for unverifiable
5848 relay addresses has also been removed.
5850 56. Added hosts_avoid_esmtp to the smtp transport.
5852 57. The exicyclog script was not checking for the esoteric option
5853 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5854 will work only if exicyclog is run under the appropriate euid.
5856 58. Following a discussion on the list, the rules by which Exim recognises line
5857 endings on incoming messages have been changed. The -dropcr and drop_cr
5858 options are now no-ops, retained only for backwards compatibility. The
5859 following line terminators are recognized: LF CRLF CR. However, special
5860 processing applies to CR:
5862 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5863 nor a local message in the state where . is a terminator.
5865 (ii) If a bare CR is encountered in a header line, an extra space is added
5866 after the line terminator so as not to end the header. The reasoning
5867 behind this is that bare CRs in header lines are most likely either
5868 to be mistakes, or people trying to play silly games.
5870 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5871 was being incorrectly given as 18 bytes larger than it should have been.
5872 This is a VOB (very old bug).
5874 60. This may never have affected anything current, but just in case it has:
5875 When the local host is found other than at the start of a list of hosts,
5876 the local host, those with the same MX, and any that follow, are discarded.
5877 When the list in question was part of a longer list of hosts, the following
5878 hosts (not currently being processed) were also being discarded. This no
5879 longer happens. I'm not sure if this situation could ever has previously
5882 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5885 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5886 header. This is something that is recommended in a new Internet Draft, and
5887 is something that is documented as being done by Sendmail. There are two
5888 possible values. For messages generated by the autoreply transport, Exim
5891 Auto-Submitted: auto-replied
5893 whereas for all other generated messages (e.g. bounces) it adds
5895 Auto-Submitted: auto-generated
5897 63. The "personal" condition in filters now includes a test for the
5898 Auto-Submitted: header. If it contains the string "auto-" the message it
5899 not considered personal.
5901 64. Added rcpt_include_affixes as a generic transport option.
5903 65. Added queue_only_override (default true).
5905 66. Added the syslog_duplication option.
5907 67. If what should have been the first header line of a message consisted of
5908 a space followed by a colon, Exim was mis-interpreting it as a header line.
5909 It isn't of course - it is syntactically invalid and should therefore be
5910 treated as the start of the message body. The misbehaviour could have
5911 caused a number of strange effects, including loss of data in subsequent
5912 header lines, and spool format errors.
5914 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5915 client host had authenticated. This control can now be exercised by an ACL
5916 for more flexibility.
5918 69. By default, callouts do not happen when testing with -bh. There is now a
5919 variant, -bhc, which does actually run the callout code, including
5920 consulting and updating the callout cache.
5922 70. Added support for saslauthd authentication, courtesy of Alexander
5925 71. If statvfs() failed on the spool or log directories while checking their
5926 size for availability, Exim confusingly gave the error "space shortage".
5927 Furthermore, in debugging mode it crashed with a floating point exception.
5928 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5929 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5930 serious problem, Exim now writes to the main and panic logs when this
5931 happens, with details of the failure. It then refuses to accept the
5932 incoming message, giving the message "spool directory problem" or "log
5933 directory problem" with a 421 code for SMTP messages.
5935 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5936 0, 1, and 2 exist, because some OS complain for execs without them (see
5937 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5938 descriptors. However, the code omitted to check that the open succeeded,
5939 causing mysterious errors if for some reason the permissions on /dev/null
5940 got screwed. Now Exim writes a message to the main and panic logs, and
5941 bombs out if it can't open /dev/null.
5943 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5944 interact so that it is all more flexible. It is supposed to remain
5945 backwards compatible. Also added extra_local_interfaces.
5947 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5948 to bomb out with an assertion failure - to the client this appears as a
5949 connection drop. This problem occurs in the part of the code that was taken
5950 from the Samba project. Fortunately, the assertion is in a very simple
5951 function, so I have fixed this by reproducing the function inline in the
5952 one place where it is called, and arranging for authentication to fail
5953 instead of killing the process with assert().
5955 75. The SPA client code was not working when the server requested OEM rather
5956 than Unicode encoding.
5958 76. Added code to make require_files with a specific uid setting more usable in
5959 the case where statting the file as root fails - usually a non-root-mounted
5960 NFS file system. When this happens and the failure is EACCES, Exim now
5961 forks a subprocess and does the per-uid checking as the relevant uid.
5963 77. Added process_log_path.
5965 78. If log_file_path was not explicitly set, a setting of check_log_space or
5966 check_log_inodes was ignored.
5968 79. If a space check for the spool or log partitions fails, the incident is now
5969 logged. Of course, in the latter case the data may get lost...
5971 80. Added the %p formatting code to string_format() so that it can be used to
5972 print addresses in debug_print(). Adjusted all the address printing in the
5973 debugging in store.c to use %p rather than %d.
5975 81. There was a concern that a line of code in smtp_in.c could overflow a
5976 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5977 initially expressed, the concern was not well-founded, because trailing
5978 spaces are removed early. However, if the trailing spaces were followed by
5979 a NULL, they did not get removed, so the overflow was possible. Two fixes
5982 (a) I re-wrote the offending code in a cleaner fashion.
5983 (b) If an incoming SMTP command contains a NULL character, it is rejected
5986 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5987 runs initgroups(), so that if the Exim user is in any additional groups,
5988 they will be used during message reception.
5994 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob