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
60 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
62 PP/02 Make -n do something, by making it not do something.
63 When combined with -bP, the name of an option is not output.
65 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
68 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
69 $sender_host_name and config options to manage this, and basic check
72 PP/05 DSCP support for outbound connections and control modifier for inbound.
74 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
75 (Only plugin which currently uses this is kerberos4, which nobody should
76 be using, but we should make it available and other future plugins might
77 conceivably use it, even though it would break NAT; stuff *should* be
78 using channel bindings instead).
80 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
81 name; added for Sendmail compatibility; requires admin caller.
82 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
83 just ignore it); requires trusted caller.
84 Also parse but ignore: -Ac -Am -X<logfile>
87 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
89 TL/02 Add +smtp_confirmation as a default logging option.
91 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
92 Patch by Magnus Holmgren from 2007-02-20.
94 TL/04 Bugzilla 1281 - Spec typo.
95 Bugzilla 1283 - Spec typo.
96 Bugzilla 1290 - Spec grammar fixes.
98 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
100 TL/06 Add Experimental DMARC support using libopendmarc libraries.
102 TL/07 Fix an out of order global option causing a segfault. Reported to dev
103 mailing list by by Dmitry Isaikin.
105 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
107 JH/02 Support "G" suffix to numbers in ${if comparisons.
109 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
111 NM/01 Bugzilla 1197 - Spec typo
112 Bugzilla 1196 - Spec examples corrections
114 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
116 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
117 gnutls_enable_pkcs11, but renamed to more accurately indicate its
120 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
121 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
123 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
124 "acl {{name}{arg}...}", and optional args on acl condition
127 JH/05 Permit multiple router/transport headers_add/remove lines.
129 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
131 JH/07 Avoid using a waiting database for a single-message-only transport.
132 Performance patch from Paul Fisher. Bugzilla 1262.
134 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
137 JH/09 Add $headers_added variable, with content from use of ACL modifier
138 add_header (but not yet added to the message). Bugzilla 199.
140 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
141 Pulled from Bugzilla 817 by Wolfgang Breyha.
143 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
145 (nb: this is the same fix as in Exim 4.80.1)
147 JH/11 Add A= logging on delivery lines, and a client_set_id option on
150 JH/12 Add optional authenticated_sender logging to A= and a log_selector
153 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
155 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
156 advertise SMTP AUTH mechanism to us, instead of a generic
157 protocol violation error. Also, make Exim more robust to bad
158 data from the Dovecot auth socket.
160 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
162 When a queue runner is handling a message, Exim first routes the
163 recipient addresses, during which it prunes them based on the retry
164 hints database. After that it attempts to deliver the message to
165 any remaining recipients. It then updates the hints database using
168 So if a recipient address works intermittently, it can get repeatedly
169 deferred at routing time. The retry hints record remains fresh so the
170 address never reaches the final cutoff time.
172 This is a fairly common occurrence when a user is bumping up against
173 their storage quota. Exim had some logic in its local delivery code
174 to deal with this. However it did not apply to per-recipient defers
175 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
177 This change adds a proper retry rule check during routing so that the
178 final cutoff time is checked against the message's age. We only do
179 this check if there is an address retry record and there is not a
180 domain retry record; this implies that previous attempts to handle
181 the address had the retry_use_local_parts option turned on. We use
182 this as an approximation for the destination being like a local
183 delivery, as in LMTP.
185 I suspect this new check makes the old local delivery cutoff check
186 redundant, but I have not verified this so I left the code in place.
188 TF/02 Correct gecos expansion when From: is a prefix of the username.
190 Test 0254 submits a message to Exim with the header
194 When I ran the test suite under the user fanf2, Exim expanded
195 the header to contain my full name, whereas it should have added
196 a Resent-Sender: header. It erroneously treats any prefix of the
197 username as equal to the username.
199 This change corrects that bug.
201 GF/01 DCC debug and logging tidyup
202 Error conditions log to paniclog rather than rejectlog.
203 Debug lines prefixed by "DCC: " to remove any ambiguity.
205 TF/03 Avoid unnecessary rebuilds of lookup-related code.
207 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
208 Bug spotted by Jeremy Harris; was flawed since initial commit.
209 Would have resulted in OCSP responses post-SNI triggering an Exim
210 NULL dereference and crash.
212 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
214 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
215 Bug detection, analysis and fix by Samuel Thibault.
216 Bugzilla 1331, Debian bug #698092.
218 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
220 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
221 Server implementation by Todd Lyons, client by JH.
222 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
223 config variable "prdr_enable" controls whether the server
224 advertises the facility. If the client requests PRDR a new
225 acl_data_smtp_prdr ACL is called once for each recipient, after
226 the body content is received and before the acl_smtp_data ACL.
227 The client is controlled by bolth of: a hosts_try_prdr option
228 on the smtp transport, and the server advertisement.
229 Default client logging of deliveries and rejections involving
230 PRDR are flagged with the string "PRDR".
232 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
233 fclose(). Diagnosis by Todd Lyons.
235 PP/17 Update configure.default to handle IPv6 localhost better.
236 Patch by Alain Williams (plus minor tweaks).
239 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
240 This is now consistent with GnuTLS, and is now documented: the
241 previous undocumented portable approach to treating the option as
242 unset was to force an expansion failure. That still works, and
243 an empty string is now equivalent.
245 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
246 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
247 not performing validation itself.
249 PP/20 Added force_command boolean option to pipe transport.
250 Patch from Nick Koston, of cPanel Inc.
252 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
255 TF/04 Added udpsend ACL modifer and hexquote expansion operator
257 PP/21 Fix eximon continuous updating with timestamped log-files.
258 Broken in a format-string cleanup in 4.80, missed when I repaired the
259 other false fix of the same issue.
260 Report and fix from Heiko Schlichting.
263 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
264 Report from Prashanth Katuri.
266 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
267 It's SecureTransport, so affects any MacOS clients which use the
268 system-integrated TLS libraries, including email clients.
270 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
271 using a MIME ACL for non-SMTP local injection.
272 Report and assistance in diagnosis by Warren Baker.
274 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
276 JH/16 Fix comparisons for 64b. Bugzilla 1385.
278 TL/09 Add expansion variable $authenticated_fail_id to keep track of
279 last id that failed so it may be referenced in subsequent ACL's.
281 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
284 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
285 ldap library initialization, allowing self-signed CA's to be
286 used. Also properly sets require_cert option later in code by
287 using NULL (global ldap config) instead of ldap handle (per
288 session). Bug diagnosis and testing by alxgomz.
290 TL/12 Enhanced documentation in the ratelimit.pl script provided in
291 the src/util/ subdirectory.
293 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
294 renamed to Transport Post Delivery Action by Jeremy Harris, as
297 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
298 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
299 redis_servers = needs to be configured which will be used by the redis
300 lookup. Patch from Warren Baker, of The Packet Hub.
302 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
304 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
305 hostname or reverse DNS when processing a host list. Used suggestions
306 from multiple comments on this bug.
308 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
310 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
311 Missed a few lines, added it to make the runtest require no keyboard
314 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
315 contains upper case chars. Make router use caseful_local_part.
317 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
318 support when GnuTLS has been built with p11-kit.
324 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
326 This, or similar/improved, will also be change PP/11 of 4.82.
332 PP/01 Handle short writes when writing local log-files.
333 In practice, only affects FreeBSD (8 onwards).
334 Bugzilla 1053, with thanks to Dmitry Isaikin.
336 NM/01 Bugzilla 949 - Documentation tweak
338 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
341 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
343 PP/02 Implemented gsasl authenticator.
345 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
347 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
348 `pkg-config foo` for cflags/libs.
350 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
351 with rest of GSASL and with heimdal_gssapi.
353 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
354 `pkg-config foo` for cflags/libs for the TLS implementation.
356 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
357 properties get this fed in as external SSF. A number of robustness
358 and debugging improvements to the cyrus_sasl authenticator.
360 PP/08 cyrus_sasl server now expands the server_realm option.
362 PP/09 Bugzilla 1214 - Log authentication information in reject log.
363 Patch by Jeremy Harris.
365 PP/10 Added dbmjz lookup type.
367 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
369 PP/12 MAIL args handles TAB as well as SP, for better interop with
370 non-compliant senders.
371 Analysis and variant patch by Todd Lyons.
373 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
374 Bug report from Lars Müller <lars@samba.org> (via SUSE),
375 Patch from Dirk Mueller <dmueller@suse.com>
377 PP/13 tls_peerdn now print-escaped for spool files.
378 Observed some $tls_peerdn in wild which contained \n, which resulted
379 in spool file corruption.
381 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
382 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
383 or write after TLS renegotiation, which otherwise led to messages
386 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
387 as a tracking header (ie: a signed header comes before the signature).
388 Patch from Wolfgang Breyha.
390 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
391 comma-sep list; embedded commas doubled.
393 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
395 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
397 Report and patch from Dmitry Banschikov.
399 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
400 Removed SSL_clear() after SSL_new() which led to protocol negotiation
401 failures. We appear to now support TLS1.1+ with Exim.
403 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
404 lets Exim select keys and certificates based upon TLS SNI from client.
405 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
406 before an outbound SMTP session. New log_selector, +tls_sni.
408 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
409 NULL dereference. Report and patch from Alun Jones.
411 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
412 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
413 Not seeing resolver debug output on NetBSD, but suspect this is a
414 resolver implementation change.
416 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
417 Left warnings. Added "eximon gdb" invocation mode.
419 PP/21 Defaulting "accept_8bitmime" to true, not false.
421 PP/22 Added -bw for inetd wait mode support.
423 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
424 locate the relevant includes and libraries. Made this the default.
426 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
427 Bugzilla 1246, report and most of solution from Tomasz Kusy.
429 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
430 This may cause build issues on older platforms.
432 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
433 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
434 gnutls_require_mac & gnutls_require_protocols (no longer supported).
435 Added SNI support via GnuTLS too.
436 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
438 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
440 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
441 Applied second patch from Janne, implementing suggestion to default
442 multiple-strings-in-record handling to match SPF spec.
444 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
446 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
447 read-only, out of scope).
448 Patch from Wolfgang Breyha, report from Stuart Northfield.
450 PP/29 Fix three issues highlighted by clang analyser static analysis.
451 Only crash-plausible issue would require the Cambridge-specific
452 iplookup router and a misconfiguration.
453 Report from Marcin Mirosław.
455 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
457 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
458 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
459 As part of this, removing so much warning spew let me fix some minor
460 real issues in debug logging.
462 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
463 assignment on my part. Fixed.
465 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
466 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
467 Janne Snabb (who went above and beyond: thank you).
469 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
470 string otherwise requires a connection and a bunch more work and it's
471 relatively easy to get wrong. Should also expose TLS library linkage
474 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
475 64-bit ${eval} (JH/03).
477 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
478 GNU libc to support some of the 64-bit stuff, should not lead to
479 conflicts. Defined before os.h is pulled in, so if a given platform
480 needs to override this, it can.
482 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
483 protection layer was required, which is not implemented.
484 Bugzilla 1254, patch from Wolfgang Breyha.
486 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
487 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
488 tls_dhparam take prime identifiers. Also unbreak combination of
489 OpenSSL+DH_params+TLSSNI.
491 PP/39 Disable SSLv2 by default in OpenSSL support.
497 PP/01 Solaris build fix for Oracle's LDAP libraries.
498 Bugzilla 1109, patch from Stephen Usher.
500 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
502 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
505 TF/02 Fix a couple more cases where we did not log the error message
506 when unlink() failed. See also change 4.74-TF/03.
508 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
509 lock up or crash if it happened to be inside a call to libc when it
510 got a SIGUSR1 from exiwhat.
512 The SIGUSR1 handler appends the current process status to the process
513 log which is later printed by exiwhat. It used to use the general
514 purpose logging code to do this, but several functions it calls are
515 not safe for signals.
517 The new output code in the SIGUSR1 handler is specific to the process
518 log, and simple enough that it's easy to inspect for signal safety.
519 Removing some special cases also simplifies the general logging code.
520 Removing the spurious timestamps from the process log simplifies
523 TF/04 Improved ratelimit ACL condition.
525 The /noupdate option has been deprecated in favour of /readonly which
526 has clearer semantics. The /leaky, /strict, and /readonly update modes
527 are mutually exclusive. The update mode is no longer included in the
528 database key; it just determines when the database is updated. (This
529 means that when you upgrde Exim will forget old rate measurements.)
531 Exim now checks that the per_* options are used with an update mode that
532 makes sense for the current ACL. For example, when Exim is processing a
533 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
534 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
535 must specify per_mail/readonly. If you omit the update mode it defaults to
536 /leaky where that makes sense (as before) or /readonly where required.
538 The /noupdate option is now undocumented but still supported for
539 backwards compatibility. It is equivalent to /readonly except that in
540 ACLs where /readonly is required you may specify /leaky/noupdate or
541 /strict/noupdate which are treated the same as /readonly.
543 A useful new feature is the /count= option. This is a generalization
544 of the per_byte option, so that you can measure the throughput of other
545 aggregate values. For example, the per_byte option is now equivalent
546 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
548 The per_rcpt option has been generalized using the /count= mechanism
549 (though it's more complicated than the per_byte equivalence). When it is
550 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
551 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
552 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
553 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
554 Note that using per_rcpt with a non-readonly update mode in more than
555 one ACL will cause the recipients to be double-counted. (The per_mail
556 and per_byte options don't have this problem.)
558 The handling of very low rates has changed slightly. If the computed rate
559 is less than the event's count (usually one) then this event is the first
560 after a long gap. In this case the rate is set to the same as this event's
561 count, so that the first message of a spam run is counted properly.
563 The major new feature is a mechanism for counting the rate of unique
564 events. The new per_addr option counts the number of different
565 recipients that someone has sent messages to in the last time period. It
566 behaves like per_rcpt if all the recipient addresses are different, but
567 duplicate recipient addresses do not increase the measured rate. Like
568 the /count= option this is a general mechanism, so the per_addr option
569 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
570 example, measure the rate that a client uses different sender addresses
571 with the options per_mail/unique=$sender_address. There are further
572 details in the main documentation.
574 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
576 TF/06 Removed a few PCRE remnants.
578 TF/07 Automatically extract Exim's version number from tags in the git
579 repository when doing development or release builds.
581 PP/02 Raise smtp_cmd_buffer_size to 16kB.
582 Bugzilla 879. Patch from Paul Fisher.
584 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
585 Heavily based on revision 40f9a89a from Simon Arlott's tree.
588 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
590 PP/05 Variable $av_failed, true if the AV scanner deferred.
591 Bugzilla 1078. Patch from John Horne.
593 PP/06 Stop make process more reliably on build failure.
594 Bugzilla 1087. Patch from Heiko Schlittermann.
596 PP/07 Make maildir_use_size_file an _expandable_ boolean.
597 Bugzilla 1089. Patch from Heiko Schlittermann.
599 PP/08 Handle ${run} returning more data than OS pipe buffer size.
600 Bugzilla 1131. Patch from Holger Weiß.
602 PP/09 Handle IPv6 addresses with SPF.
603 Bugzilla 860. Patch from Wolfgang Breyha.
605 PP/10 GnuTLS: support TLS 1.2 & 1.1.
607 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
610 PP/11 match_* no longer expand right-hand-side by default.
611 New compile-time build option, EXPAND_LISTMATCH_RHS.
612 New expansion conditions, "inlist", "inlisti".
614 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
616 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
618 PP/14 fix log_write() format string regression from TF/03.
619 Bugzilla 1152. Patch from Dmitry Isaikin.
625 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
627 PP/02 Harmonised TLS library version reporting; only show if debugging.
628 Layout now matches that introduced for other libraries in 4.74 PP/03.
630 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
632 PP/04 New "dns_use_edns0" global option.
634 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
637 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
638 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
640 TK/01 Updated PolarSSL code to 0.14.2.
641 Bugzilla 1097. Patch from Andreas Metzler.
643 PP/07 Catch divide-by-zero in ${eval:...}.
646 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
649 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
650 format-string attack -- SECURITY: remote arbitrary code execution.
652 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
653 time unintentionally subject to list matching rules, letting the header
654 cause arbitrary Exim lookups (of items which can occur in lists, *not*
655 arbitrary string expansion). This allowed for information disclosure.
657 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
658 INT_MIN/-1 -- value coerced to INT_MAX.
664 NM/01 Workround for PCRE version dependency in version reporting
667 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
668 This fixes portability to compilers other than gcc, notably
669 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
671 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
672 makefiles for portability to HP-UX and POSIX correctness.
674 PP/01 Permit LOOKUP_foo enabling on the make command-line.
675 Also via indented variable definition in the Makefile.
676 (Debugging by Oliver Heesakkers).
678 PP/02 Restore caching of spamd results with expanded spamd_address.
679 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
681 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
682 Improves build reliability. Fix from: Frank Elsner
684 NM/02 Fix wide character breakage in the rfc2047 coding
685 Fixes bug 1064. Patch from Andrey N. Oktyabrski
687 NM/03 Allow underscore in dnslist lookups
688 Fixes bug 1026. Patch from Graeme Fowler
690 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
691 Code patches from Adam Ciarcinski of NetBSD.
693 NM/04 Fixed exiqgrep to cope with mailq missing size issue
696 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
697 is logged, to avoid truncation. Patch from John Horne.
699 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
700 Patch from Jakob Hirsch.
702 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
703 SQL string expansion failure details.
704 Patch from Andrey Oktyabrski.
706 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
707 Patch from Simon Arlott.
709 PP/09 New lookups functionality failed to compile on old gcc which rejects
710 extern declarations in function scope.
711 Patch from Oliver Fleischmann
713 PP/10 Use sig_atomic_t for flags set from signal handlers.
714 Check getgroups() return and improve debugging.
715 Fixed developed for diagnosis in bug 927 (which turned out to be
718 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
719 Patch from Mark Zealey.
721 PP/12 Bugzilla 1056: Improved spamd server selection.
722 Patch from Mark Zealey.
724 PP/13 Bugzilla 1086: Deal with maildir quota file races.
725 Based on patch from Heiko Schlittermann.
727 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
728 Patch from Uwe Doering, sign-off by Michael Haardt.
730 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
731 variable declaration deep within a block. Bug and patch from
734 PP/15 lookups-Makefile IRIX compatibilty coercion.
736 PP/16 Make DISABLE_DKIM build knob functional.
738 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
739 Patch by Simon Arlott
741 TF/03 Fix valgrind.h portability to C89 compilers that do not support
742 variable argument macros. Our copy now differs from upstream.
748 TF/01 Failure to get a lock on a hints database can have serious
749 consequences so log it to the panic log.
751 TF/02 Log LMTP confirmation messages in the same way as SMTP,
752 controlled using the smtp_confirmation log selector.
754 TF/03 Include the error message when we fail to unlink a spool file.
756 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
757 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
758 for maintaining out-of-tree patches for some time.
760 PP/01 Bugzilla 139: Documentation and portability issues.
761 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
762 Handle per-OS dynamic-module compilation flags.
764 PP/02 Let /dev/null have normal permissions.
765 The 4.73 fixes were a little too stringent and complained about the
766 permissions on /dev/null. Exempt it from some checks.
767 Reported by Andreas M. Kirchwitz.
769 PP/03 Report version information for many libraries, including
770 Exim version information for dynamically loaded libraries. Created
771 version.h, now support a version extension string for distributors
772 who patch heavily. Dynamic module ABI change.
774 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
775 privilege escalation vulnerability whereby the Exim run-time user
776 can cause root to append content of the attacker's choosing to
779 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
782 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
783 If dropping privileges for untrusted macros, we disabled normal logging
784 on the basis that it would fail; for the Exim run-time user, this is not
785 the case, and it resulted in successful deliveries going unlogged.
786 Fixed. Reported by Andreas Metzler.
792 PP/01 Date: & Message-Id: revert to normally being appended to a message,
793 only prepend for the Resent-* case. Fixes regression introduced in
794 Exim 4.70 by NM/22 for Bugzilla 607.
796 PP/02 Include check_rfc2047_length in configure.default because we're seeing
797 increasing numbers of administrators be bitten by this.
799 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
801 PP/03 Bugzilla 994: added openssl_options main configuration option.
803 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
805 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
807 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
809 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
810 without a peer certificate, leading to a segfault because of an
811 assumption that peers always have certificates. Be a little more
812 paranoid. Problem reported by Martin Tscholak.
814 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
815 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
816 NB: ClamAV planning to remove STREAM in "middle of 2010".
817 CL also introduces -bmalware, various -d+acl logging additions and
818 more caution in buffer sizes.
820 PP/09 Implemented reverse_ip expansion operator.
822 PP/10 Bugzilla 937: provide a "debug" ACL control.
824 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
826 PP/12 Bugzilla 973: Implement --version.
828 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
830 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
832 PP/15 Bugzilla 816: support multiple condition rules on Routers.
834 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
835 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
836 ignore trailing whitespace.
838 JJ/02 prevent non-panic DKIM error from being sent to paniclog
840 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
843 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
844 Notification from Dr Andrew Aitchison.
846 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
847 ExtendedDetectionInfo response format.
848 Notification from John Horne.
850 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
853 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
854 XSL and documented dependency on system catalogs, with examples of how
857 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
860 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
861 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
862 configuration file which is writeable by the Exim user or group.
864 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
865 of configuration files to cover files specified with the -C option if
866 they are going to be used with root privileges, not just the default
869 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
870 option (effectively making it always true).
872 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
873 files to be used while preserving root privileges.
875 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
876 that rogue child processes cannot use them.
878 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
879 run-time user, instead of root.
881 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
882 Exim run-time user without dropping privileges.
884 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
885 result string, instead of calling string_vformat() twice with the same
888 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
889 for other users. Others should always drop root privileges if they use
890 -C on the command line, even for a whitelisted configure file.
892 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
894 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
900 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
901 $data_path, and $header_path variables; fixed documentation bugs and
904 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
905 exipick to access non-standard spools, including the "frozen" queue
908 NM/01 Bugzilla 965: Support mysql stored procedures.
909 Patch from Alain Williams
911 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
913 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
914 Patch from Andreas Metzler
916 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
917 Patch from Kirill Miazine
919 NM/05 Bugzilla 671: Added umask to procmail example.
921 JJ/03 installed exipick 20100323.0, fixing doc bug
923 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
924 directory. Notification and patch from Dan Rosenberg.
926 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
928 TK/02 Improve log output when DKIM signing operation fails.
930 MH/01 Treat the transport option dkim_domain as a colon separated
931 list, not as a single string, and sign the message with each element,
932 omitting multiple occurences of the same signer.
934 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
935 Bugzilla 985, 986. Patch by Simon Arlott
937 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
938 Patch by Simon Arlott
940 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
941 MBX locking. Notification from Dan Rosenberg.
947 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
949 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
951 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
953 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
955 NM/04 Bugzilla 915: Flag broken perl installation during build.
961 TK/01 Added patch by Johannes Berg that expands the main option
962 "spamd_address" if it starts with a dollar sign.
964 TK/02 Write list of recipients to X-Envelope-Sender header when building
965 the mbox-format spool file for content scanning (suggested by Jakob
968 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
969 (http://www.dcc-servers.net/) support via dccifd. Activated by
970 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
972 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
973 by Mark Daniel Reidel <mr@df.eu>.
975 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
976 When building exim an external PCRE library is now needed -
977 PCRE is a system library on the majority of modern systems.
978 See entry on PCRE_LIBS in EDITME file.
980 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
981 conversation. Added nologin parameter to request.
982 Patch contributed by Kirill Miazine.
984 TF/01 Do not log submission mode rewrites if they do not change the address.
986 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
988 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
989 log files in place. Contributed by Roberto Lima.
991 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
993 TF/03 Bugzilla 615: When checking the local_parts router precondition
994 after a local_part_suffix or local_part_prefix option, Exim now
995 does not use the address's named list lookup cache, since this
996 contains cached lookups for the whole local part.
998 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
999 Robert Millan. Documentation is in experimental-spec.txt.
1001 TF/04 Bugzilla 668: Fix parallel build (make -j).
1003 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1005 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1006 Patch provided by Jan Srzednicki.
1008 TF/05 Leading white space used to be stripped from $spam_report which
1009 wrecked the formatting. Now it is preserved.
1011 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1012 that they are available at delivery time.
1014 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1016 TF/08 TLS error reporting now respects the incoming_interface and
1017 incoming_port log selectors.
1019 TF/09 Produce a more useful error message if an SMTP transport's hosts
1020 setting expands to an empty string.
1022 NM/06 Bugzilla 744: EXPN did not work under TLS.
1023 Patch provided by Phil Pennock.
1025 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1026 Patch provided by Richard Godbee.
1028 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1029 acl_smtp_notquit, added index entry.
1031 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1032 Patch provided by Eugene Bujak.
1034 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1035 accept(). Patch provided by Maxim Dounin.
1037 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1038 Patch provided by Phil Pennock.
1040 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1042 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1043 Patch provided by Brad "anomie" Jorsch.
1045 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1046 Patch provided by Dean Brooks.
1048 TK/05 Add native DKIM support (does not depend on external libraries).
1050 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1051 Patch provided by Graeme Fowler.
1053 NM/16 Bugzilla 851: Documentation example syntax fix.
1055 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1057 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1060 NM/19 Bugzilla 745: TLS version reporting.
1061 Patch provided by Phil Pennock.
1063 NM/20 Bugzilla 167: bool: condition support.
1064 Patch provided by Phil Pennock.
1066 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1067 clients. Patch provided by Phil Pennock.
1069 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1070 Patch provided by Brad "anomie" Jorsch.
1072 NM/23 Bugzilla 687: Fix misparses in eximstats.
1073 Patch provided by Heiko Schlittermann.
1075 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1076 Patch provided by Heiko Schlittermann.
1078 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1079 plus update to original patch.
1081 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1083 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1084 Patch provided by David Brownlee.
1086 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1088 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1090 NM/30 Bugzilla 888: TLS documentation bugfixes.
1092 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1094 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1095 Unlike the original bugzilla I have changed all shell scripts in src tree.
1097 NM/33 Bugzilla 898: Transport filter timeout fix.
1098 Patch by Todd Rinaldo.
1100 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1101 Patch by Serge Demonchaux.
1103 NM/35 Bugzilla 39: Base64 decode bug fixes.
1104 Patch by Jakob Hirsch.
1106 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1108 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1110 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1112 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1118 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1119 ALT-N's libdkim that I have put here:
1120 http://duncanthrax.net/exim-experimental/
1122 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1123 were called 'true' and it seems that C99 defines that as a reserved
1124 keyword to be used with 'bool' variable types. That means you could
1125 not include C99-style headers which use bools without triggering
1126 build errors in sieve.c.
1128 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1129 as mailq or other aliases. Changed the --help handling significantly
1130 to do whats expected. exim_usage() emits usage/help information.
1132 SC/01 Added the -bylocaldomain option to eximstats.
1134 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1136 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1138 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1144 PH/01 Another patch from the Sieve maintainer.
1146 PH/02 When an IPv6 address is converted to a string for single-key lookup
1147 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1148 dots are used instead of colons so that keys in lsearch files need not
1149 contain colons. This was done some time before quoting was made available
1150 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1151 (notated using the quote facility) so as to distinguish them from IPv4
1152 keys. This meant that lookups for IP addresses in host lists did not work
1153 for iplsearch lookups.
1155 This has been fixed by arranging for IPv6 addresses to be expressed with
1156 colons if the lookup type is iplsearch. This is not incompatible, because
1157 previously such lookups could never work.
1159 The situation is now rather anomolous, since one *can* have colons in
1160 ordinary lsearch keys. However, making the change in all cases is
1161 incompatible and would probably break a number of configurations.
1163 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1166 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1167 conversion specification without a maximum field width, thereby enabling
1168 a rogue spamd server to cause a buffer overflow. While nobody in their
1169 right mind would setup Exim to query an untrusted spamd server, an
1170 attacker that gains access to a server running spamd could potentially
1171 exploit this vulnerability to run arbitrary code as the Exim user.
1173 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1174 $primary_hostname instead of what libspf2 thinks the hosts name is.
1176 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1177 a directory entry by the name of the lookup key. Previously, if a
1178 symlink pointed to a non-existing file or a file in a directory that
1179 Exim lacked permissions to read, a lookup for a key matching that
1180 symlink would fail. Now it is enough that a matching directory entry
1181 exists, symlink or not. (Bugzilla 503.)
1183 PH/03 The body_linecount and body_zerocount variables are now exported in the
1186 PH/04 Added the $dnslist_matched variable.
1188 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1189 This means they are set thereafter only if the connection becomes
1192 PH/06 Added the client_condition to authenticators so that some can be skipped
1193 by clients under certain conditions.
1195 PH/07 The error message for a badly-placed control=no_multiline_responses left
1196 "_responses" off the end of the name.
1198 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1200 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1201 (without spaces) instead of just copying the configuration text.
1203 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1205 PH/11 Added $max_received_linelength.
1207 PH/12 Added +ignore_defer and +include_defer to host lists.
1209 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1210 way in which PCRE > 7.0 is built.
1212 PH/14 Implemented queue_only_load_latch.
1214 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1215 MAIL command. The effect was to mangle the value on 64-bit systems.
1217 PH/16 Another patch from the Sieve maintainer.
1219 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1221 PH/18 If a system quota error occurred while trying to create the file for
1222 a maildir delivery, the message "Mailbox is full" was not appended to the
1223 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1224 only to a quota excession during the actual writing of the file.
1226 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1227 characters?) which causes problems in log lines. The DN values are now
1228 passed through string_printing() before being added to log lines.
1230 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1231 and InterBase are left for another time.)
1233 PH/21 Added message_body_newlines option.
1235 PH/22 Guard against possible overflow in moan_check_errorcopy().
1237 PH/23 POSIX allows open() to be a macro; guard against that.
1239 PH/24 If the recipient of an error message contained an @ in the local part
1240 (suitably quoted, of course), incorrect values were put in $domain and
1241 $local_part during the evaluation of errors_copy.
1247 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1248 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1251 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1252 issue a MAIL command.
1254 PH/02 In an ACL statement such as
1256 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1258 if a client was not listed at all, or was listed with a value other than
1259 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1260 the condition was not true (as it should be), so access was not denied.
1261 The bug was that the ! inversion was incorrectly passed on to the second
1262 item. This has been fixed.
1264 PH/03 Added additional dnslists conditions == and =& which are different from
1265 = and & when the dns lookup returns more than one IP address.
1267 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1268 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1270 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1271 FSYNC, which compiles an option called disable_fsync that allows for
1272 bypassing fsync(). The documentation is heavily laced with warnings.
1274 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1276 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1277 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1278 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1279 including adding "make clean"; (3) Added -fPIC when compiling the test
1280 dynamically loaded module, to get rid of a warning.
1282 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1283 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1284 = 0s. The bug is otherwise harmless.
1286 PH/07 There was a bug in the dovecot authenticator such that the value of
1287 $auth1 could be overwritten, and so not correctly preserved, after a
1288 successful authentication. This usually meant that the value preserved by
1289 the server_setid option was incorrect.
1291 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1293 PH/09 Installed PCRE release 7.0.
1295 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1296 run for batched SMTP input. It is now run at the start of every message
1297 in the batch. While fixing this I discovered that the process information
1298 (output by running exiwhat) was not always getting set for -bs and -bS
1299 input. This is fixed, and it now also says "batched" for BSMTP.
1301 PH/11 Added control=no_pipelining.
1303 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1304 patch, slightly modified), and move the expansion of helo_data till after
1305 the connection is made in the smtp transport (so it can use these
1308 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1310 PH/14 Added log_selector = +pid.
1312 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1314 PH/16 Add ${if forany and ${if forall.
1316 PH/17 Added dsn_from option to vary the From: line in DSNs.
1318 PH/18 Flush SMTP output before performing a callout, unless control =
1319 no_callout_flush is set.
1321 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1322 was true (the default) a successful delivery failed to delete the retry
1323 item, thus causing premature timeout of the address. The bug is now
1326 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1328 PH/21 Long custom messages for fakedefer and fakereject are now split up
1329 into multiline reponses in the same way that messages for "deny" and
1330 other ACL rejections are.
1332 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1333 with slight modification.
1335 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1336 draft, changing the syntax and factoring some duplicate code".
1338 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1339 for deliveries of the second and subsequent messages over the same SMTP
1342 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1343 ${reduce, with only minor "tidies".
1345 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1347 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1348 expansion side effects.
1350 PH/27 When a message times out after an over-quota error from an Exim-imposed
1351 quota, the bounce message says "mailbox is full". This message was not
1352 being given when it was a system quota that was exceeded. It now should
1355 MH/03 Made $recipients available in local_scan(). local_scan() already has
1356 better access to the recipient list through recipients_list[], but
1357 $recipients can be useful in postmaster-provided expansion strings.
1359 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1360 in the case of a MAIL command with additional options following the
1361 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1362 were accidentally chopped off.
1364 PH/29 SMTP synchronization checks are implemented when a command is read -
1365 there is a check that no more input is waiting when there shouldn't be
1366 any. However, for some commands, a delay in an ACL can mean that it is
1367 some time before the response is written. In this time, more input might
1368 arrive, invalidly. So now there are extra checks after an ACL has run for
1369 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1370 pipelining has not been advertised.
1372 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1374 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1375 returned in the SMTP response when smtp_return_error_details was false.
1376 This has been fixed.
1378 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1379 instead of the C I/O that was originally supplied, because problems were
1380 reported on Solaris.
1382 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1383 Exim which did not show up earlier: it was assuming that a call to
1384 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1385 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1386 was a macro that became an assignment, so it seemed to work. This has
1387 changed to a proper function call with a void return, hence the compile
1388 error. Exim's code has been fixed.
1390 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1393 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1395 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1397 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1398 because it turns out to be too big or there is a timeout) there is no
1399 "Completed" line in the log. When some messages of this type were
1400 selected by exigrep, they were listed as "not completed". Others were
1401 picked up by some special patterns. I have improved the selection
1402 criteria to be more general.
1404 PH/38 The host_find_failed option in the manualroute router can now be set
1405 to "ignore", to completely ignore a host whose IP address cannot be
1406 found. If all hosts are ignored, the behaviour is controlled by the new
1407 host_all_ignored option.
1409 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1410 homing or because of multiple MX records with /mx) generated more than
1411 one IP address, and the following item turned out to be the local host,
1412 all the secondary addresses of the first item were incorrectly removed
1413 from the list, along with the local host and any following hosts (which
1414 is what is supposed to happen).
1416 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1417 whoever called Exim into the -H file. In the case of the daemon it was
1418 behaving confusingly. When first started, it used values for whoever
1419 started the daemon, but after a SIGHUP it used the Exim user (because it
1420 calls itself on a restart). I have changed the code so that it now always
1423 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1424 message are rejected with the same error (e.g. no authentication or bad
1425 sender address), and a DATA command is nevertheless sent (as can happen
1426 with PIPELINING or a stupid MUA), the error message that was given to the
1427 RCPT commands is included in the rejection of the DATA command. This is
1428 intended to be helpful for MUAs that show only the final error to their
1431 PH/42 Another patch from the Sieve maintainer.
1433 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1434 Eximstats - Fixed some broken HTML links and added missing column headers
1436 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1437 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1439 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1441 PH/43 Yet another patch from the Sieve maintainer.
1443 PH/44 I found a way to check for a TCP/IP connection going away before sending
1444 the response to the final '.' that terminates a message, but only in the
1445 case where the client has not sent further data following the '.'
1446 (unfortunately, this is allowed). However, in many cases there won't be
1447 any further data because there won't be any more messages to send. A call
1448 to select() can be used: if it shows that the input is "ready", there is
1449 either input waiting, or the socket has been closed. An attempt to read
1450 the next input character can distinguish the two cases. Previously, Exim
1451 would have sent an OK response which the client would never have see.
1452 This could lead to message repetition. This fix should cure that, at
1453 least in a lot of common cases.
1455 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1456 advertised in response to EHLO.
1462 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1463 fixed by 4.65/MH/01 (is this a record?) are fixed:
1465 (i) An empty string was always treated as zero by the numeric comparison
1466 operators. This behaviour has been restored.
1468 (ii) It is documented that the numeric comparison operators always treat
1469 their arguments as decimal numbers. This was broken in that numbers
1470 starting with 0 were being interpreted as octal.
1472 While fixing these problems I realized that there was another issue that
1473 hadn't been noticed. Values of message_size_limit (both the global option
1474 and the transport option) were treated as octal if they started with 0.
1475 The documentation was vague. These values are now always treated as
1476 decimal, and I will make that clear in the documentation.
1482 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1483 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1486 MH/01 Don't check that the operands of numeric comparison operators are
1487 integers when their expansion is in "skipping" mode (fixes bug
1488 introduced by 4.64-PH/07).
1490 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1491 child addresses, Exim now panics and dies. Previously, because the count
1492 is held in a short int, deliveries were likely to be lost. As such a
1493 large number of recipients for a single message is ridiculous
1494 (performance will be very, very poor), I have chosen to impose a limit
1495 rather than extend the field.
1501 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1502 leftover -K file (the existence of which was triggered by #402).
1503 While we were at it, introduced process PID as part of the -K
1504 filename. This should rule out race conditions when creating
1507 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1508 processing considerably. Previous code took too long for large mails,
1509 triggering a timeout which in turn triggers #401.
1511 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1512 in the DK code in transports.c. sendfile() is not really portable,
1513 hence the _LINUX specificness.
1515 TF/01 In the add_headers option to the mail command in an Exim filter,
1516 there was a bug that Exim would claim a syntax error in any
1517 header after the first one which had an odd number of characters
1520 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1521 callout verification, Exim cached a "reject" for the entire domain. This
1522 is correct for most verifications, but it is not correct for a recipient
1523 verification with use_sender or use_postmaster set, because in that case
1524 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1525 case of MAIL FROM:<> rejection from other early rejections (e.g.
1526 rejection of HELO). When verifying a recipient using a non-null MAIL
1527 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1528 Whatever the result of the callout, the value of the domain cache is
1529 left unchanged (for any other kind of callout, getting as far as trying
1530 RCPT means that the domain itself is ok).
1532 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1535 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1536 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1537 session, the daemon ignores ECONNECT errors and logs others; it now
1538 ignores EPIPE as well.
1540 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1541 (quoted-printable decoding).
1543 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1544 later the small subsequent patch to fix an introduced bug.
1546 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1548 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1550 PH/08 An error is now given if message_size_limit is specified negative.
1552 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1553 to be given (somewhat) arbitrary names.
1555 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1558 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1559 miscellaneous code fixes
1561 PH/10 Added the log_reject_target ACL modifier to specify where to log
1564 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1565 hostname. This is wrong, because it relates to the incoming message (and
1566 probably the interface on which it is arriving) and not to the outgoing
1567 callout (which could be using a different interface). This has been
1568 changed to use the value of the helo_data option from the smtp transport
1569 instead - this is what is used when a message is actually being sent. If
1570 there is no remote transport (possible with a router that sets up host
1571 addresses), $smtp_active_hostname is used.
1573 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1574 tweaks were necessary in order to get it to work (see also 21 below):
1575 (a) The code assumed that strncpy() returns a negative number on buffer
1576 overflow, which isn't the case. Replaced with Exim's string_format()
1578 (b) There were several signed/unsigned issues. I just did the minimum
1579 hacking in of casts. There is scope for a larger refactoring.
1580 (c) The code used strcasecmp() which is not a standard C function.
1581 Replaced with Exim's strcmpic() function.
1582 (d) The code set only $1; it now sets $auth1 as well.
1583 (e) A simple test gave the error "authentication client didn't specify
1584 service in request". It would seem that Dovecot has changed its
1585 interface. Fortunately there's a specification; I followed it and
1586 changed what the client sends and it appears to be working now.
1588 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1591 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1592 address A is aliased to B and C, where B exists and C does not. Without
1593 -v the output is "A verified" because verification stops after a
1594 successful redirection if more than one address is generated. However,
1595 with -v the child addresses are also verified. Exim was outputting "A
1596 failed to verify" and then showing the successful verification for C,
1597 with its parentage. It now outputs "B failed to verify", showing B's
1598 parentage before showing the successful verification of C.
1600 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1601 look up a TXT record in a specific list after matching in a combined
1604 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1605 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1606 they consult the DNS. I had assumed they would set it the way they
1607 wanted; and indeed my experiments on Linux seem to show that in some
1608 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1609 To be on the safe side, however, I have now made the interface to
1610 host_find_byname() similar to host_find_bydns(), with an argument
1611 containing the DNS resolver options. The host_find_byname() function now
1612 sets these options at its start, just as host_find_bydns() does. The smtp
1613 transport options dns_qualify_single and dns_search_parents are passed to
1614 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1615 of host_find_byname() use the default settings of RES_DEFNAMES
1616 (qualify_single) but not RES_DNSRCH (search_parents).
1618 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1619 spool_read_header() do less string testing, by means of a preliminary
1620 switch on the second character of optional "-foo" lines. (This is
1621 overdue, caused by the large number of possibilities that now exist.
1622 Originally there were few.) While I was there, I also converted the
1623 str(n)cmp tests so they don't re-test the leading "-" and the first
1624 character, in the hope this might squeeze out yet more improvement.
1626 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1627 flag allowing group syntax was set by the header_syntax check but not
1628 turned off, possible causing trouble later; (2) The flag was not being
1629 set at all for the header_verify test, causing "group"-style headers to
1630 be rejected. I have now set it in this case, and also caused header_
1631 verify to ignore an empty address taken from a group. While doing this, I
1632 came across some other cases where the code for allowing group syntax
1633 while scanning a header line wasn't quite right (mostly, not resetting
1634 the flag correctly in the right place). These bugs could have caused
1635 trouble for malformed header lines. I hope it is now all correct.
1637 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1638 with the "reply" argument non-NULL. The code, however (which originally
1639 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1640 but it didn't always do it. This confused somebody who was copying the
1641 code for some other use. I have removed all the tests.
1643 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1644 feature that was used to support insecure browsers during the U.S. crypto
1645 embargo. It requires special client support, and Exim is probably the
1646 only MTA that supported it -- and would never use it because real RSA is
1647 always available. This code has been removed, because it had the bad
1648 effect of slowing Exim down by computing (never used) parameters for the
1649 RSA_EXPORT functionality.
1651 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1652 authenticator to fail if there's a tab character in the incoming data
1653 (there should never be unless someone is messing about, as it's supposed
1654 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1655 if the connection is using TLS or if the remote IP is the same as the
1656 local IP, and the "valid-client-cert option" if a client certificate has
1659 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1660 authenticators. This can be used for authorization after authentication
1661 succeeds. (In the case of plaintext, it servers for both authentication
1664 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1665 if any retry times were supplied.
1667 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1668 connection, where there is no client IP address to check. In this
1669 situation, the verify now always succeeds.
1671 PH/25 Applied John Jetmore's -Mset patch.
1673 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1675 PH/27 In a string expansion for a processed (not raw) header when multiple
1676 headers of the same name were present, leading whitespace was being
1677 removed from all of them, but trailing whitespace was being removed only
1678 from the last one. Now trailing whitespace is removed from each header
1679 before concatenation. Completely empty headers in a concatenation (as
1680 before) are ignored.
1682 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1683 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1685 PH/29 [Removed. This was a change that I later backed out, and forgot to
1686 correct the ChangeLog entry (that I had efficiently created) before
1687 committing the later change.]
1689 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1690 address errors (4xx response to RCPT) over the same connection as other
1691 messages routed to the same hosts. Such deliveries are always "forced",
1692 so retry times are not inspected. This resulted in far too many retries
1693 for the affected addresses. The effect occurred only when there were more
1694 hosts than the hosts_max_try setting in the smtp transport when it had
1695 the 4xx errors. Those hosts that it had tried were not added to the list
1696 of hosts for which the message was waiting, so if all were tried, there
1697 was no problem. Two fixes have been applied:
1699 (i) If there are any address or message errors in an SMTP delivery, none
1700 of the hosts (tried or untried) are now added to the list of hosts
1701 for which the message is waiting, so the message should not be a
1702 candidate for sending over the same connection that was used for a
1703 successful delivery of some other message. This seems entirely
1704 reasonable: after all the message is NOT "waiting for some host".
1705 This is so "obvious" that I'm not sure why it wasn't done
1706 previously. Hope I haven't missed anything, but it can't do any
1707 harm, as the worst effect is to miss an optimization.
1709 (ii) If, despite (i), such a delivery is accidentally attempted, the
1710 routing retry time is respected, so at least it doesn't keep
1711 hammering the server.
1713 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1714 in ${readsocket because some servers need this prod.
1716 PH/32 Added some extra debug output when updating a wait-xxx database.
1718 PH/33 The hint "could be header name not terminated by colon", which has been
1719 given for certain expansion errors for a long time, was not being given
1720 for the ${if def:h_colon_omitted{... case.
1722 PH/34 The spec says: "With one important exception, whenever a domain list is
1723 being scanned, $domain contains the subject domain." There was at least
1724 one case where this was not true.
1726 PH/35 The error "getsockname() failed: connection reset by peer" was being
1727 written to the panic log as well as the main log, but it isn't really
1728 panic-worthy as it just means the connection died rather early on. I have
1729 removed the panic log writing for the ECONNRESET error when getsockname()
1732 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1733 runs only) independently of the message's sender address. This meant
1734 that, if the 4xx error was in fact related to the sender, a different
1735 message to the same recipient with a different sender could confuse
1736 things. In particualar, this can happen when sending to a greylisting
1737 server, but other circumstances could also provoke similar problems.
1738 I have changed the default so that the retry time for these errors is now
1739 based a combination of the sender and recipient addresses. This change
1740 can be overridden by setting address_retry_include_sender=false in the
1743 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1744 remote server are returned as part of bounce messages. This was not
1745 happening for LMTP over a pipe (the lmtp transport), but now it is the
1746 same for both kinds of LMTP.
1748 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1749 in header lines that were in fact CNAMEs. This is no longer the case.
1751 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1752 and queue runs started by the daemon processed all messages. This has
1753 been fixed so that -R and -S can now usefully be given with -q<time>.
1755 PH/40 Import PCRE release 6.7 (fixes some bugs).
1757 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1759 PH/42 Give an error if -q is specified more than once.
1761 PH/43 Renamed the variables $interface_address and $interface_port as
1762 $received_ip_address and $received_port, to make it clear that these
1763 values apply to message reception, and not to the outgoing interface when
1764 a message is delivered. (The old names remain recognized, of course.)
1766 PH/44 There was no timeout on the connect() call when using a Unix domain
1767 socket in the ${readsocket expansion. There now is.
1769 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1770 be meaningful with "accept".
1772 SC/01 Eximstats V1.43
1773 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1775 SC/02 Eximstats V1.44
1776 Use a glob alias rather than an array ref in the generated
1777 parser. This improves both readability and performance.
1779 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1780 Collect SpamAssassin and rejection statistics.
1781 Don't display local sender or destination tables unless
1782 there is data to show.
1783 Added average volumes into the top table text output.
1785 SC/04 Eximstats V1.46
1786 Collect data on the number of addresses (recipients)
1787 as well as the number of messages.
1789 SC/05 Eximstats V1.47
1790 Added 'Message too big' to the list of mail rejection
1791 reasons (thanks to Marco Gaiarin).
1793 SC/06 Eximstats V1.48
1794 Mainlog lines which have GMT offsets and are too short to
1795 have a flag are now skipped.
1797 SC/07 Eximstats V1.49 (Alain Williams)
1798 Added the -emptyok flag.
1800 SC/08 Eximstats V1.50
1801 Fixes for obtaining the IP address from reject messages.
1803 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1804 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1805 whitesspace changes from 4.64-PH/27
1807 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1810 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1811 are found, allow negative numbers in numeric criteria)
1813 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1815 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1816 to match changes made in 4.64-PH/43
1818 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1820 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1821 "domain" is actually an IP address, and force a failure. This locks out
1822 those revolvers/nameservers that support "A-for-A" lookups, in
1823 contravention of the specifications.
1825 PH/48 When a host name was looked up from an IP address, and the subsequent
1826 forward lookup of the name timed out, the host name was left in
1827 $sender_host_name, contrary to the specification.
1829 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1830 restricted to single-key lookups, Exim was not diagnosing an error if
1831 * or *@ was used with a query-style lookup.
1833 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1835 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1836 long ago, but noone interested enough thought of it. Let's just say that
1837 the "1.1" means that there are some new functions that weren't there at
1838 some point in the past.
1840 PH/51 Error processing for expansion failure of helo_data from an smtp
1841 transport during callout processing was broken.
1843 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1844 tested/used via the -bh/-bhc/-bs options.
1846 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1847 bug, fixed in subsequent PCRE releases).
1849 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1850 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1852 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1858 SC/01 Use a glob alias rather than an array ref in eximstats generated
1859 parser. This improves both readability and performance.
1861 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1862 Don't display local sender or destination tables in eximstats unless
1863 there is data to show.
1864 Added average volumes into the eximstats top table text output.
1866 SC/03 Collect data on the number of addresses (recipients) as well
1867 as the number of messages in eximstats.
1869 TF/01 Correct an error in the documentation for the redirect router. Exim
1870 does (usually) call initgroups() when daemonizing.
1872 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1873 with consistent privilege compared to when running as a daemon.
1875 TF/03 Note in the spec that $authenticated_id is not set for local
1876 submissions from trusted users.
1878 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1879 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1881 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1882 by adding some example configuration directives to the default
1883 configuration file. A little bit of work is required to uncomment the
1884 directives and define how usernames and passwords are checked, but
1885 there is now a framework to start from.
1887 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1888 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1889 without this. I don't know how relevant this is to other LDAP libraries.
1891 PH/02 Add the verb name to the "unknown ACL verb" error.
1893 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1895 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1897 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1898 directory not expanded when it should be if an expanded home directory
1899 was set for the address (which is overridden by the transport).
1901 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1904 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1905 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1906 because it is too late at that time, and has no effect.
1908 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1909 security issue with \' (bugzilla #107). I could not use the
1910 PQescapeStringConn() function, because it needs a PGconn value as one of
1913 PH/08 When testing addresses using -bt, indicate those final addresses that
1914 are duplicates that would not cause an additional delivery. At least one
1915 person was confused, thinking that -bt output corresponded to deliveries.
1916 (Suppressing duplicates isn't a good idea as you lose the information
1917 about possibly different redirections that led to the duplicates.)
1919 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1920 systems where poll() doesn't work, in particular OS X.
1922 PH/10 Added more information to debugging output for retry time not reached.
1924 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1925 operations in malware.c.
1927 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1930 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1931 syslog_duplication set false, log lines that would normally be written
1932 both the the main log and to the reject log were not written to syslog at
1935 PH/14 In the default configuration, change the use of "message" in ACL warn
1936 statements to "add_header".
1938 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1939 not followed by a command (e.g. "seen endif").
1941 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1942 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1945 PH/17 Added extra conditions to the default value of delay_warning_condition
1949 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1950 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1951 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1954 The Auto-Submitted: and various List- headers are standardised, whereas I
1955 don't think Precedence: ever was.
1957 PH/18 Refactored debugging code in route_finduser() to show more information,
1958 in particular, the error code if getpwnam() issues one.
1960 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1961 This is apparently needed in addition to the PH/07 change above to avoid
1962 any possible encoding problems.
1964 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1965 but not after initializing Perl.
1967 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1968 output them only if debugging. By default they are written stderr,
1969 apparently, which is not desirable.
1971 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1974 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1977 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1979 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1980 authenticated or an ident call has been made. Suppress the default
1981 values for $authenticated_id and $authenticated_sender (but permit -oMai
1982 and -oMas) when testing with -bh.
1984 PH/24 Re-jigged the order of the tests in the default configuration so that the
1985 tests for valid domains and recipients precede the DNS black list and CSA
1986 tests, on the grounds that those ones are more expensive.
1988 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1989 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1990 This bug exists in every version of Exim that I still have, right back to
1993 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1994 However, an attempt to turn on case-sensitivity in a regex key by
1995 including (?-i) didn't work because the subject string was already
1996 lowercased, and the effects were non-intuitive. It turns out that a
1997 one-line patch can be used to allow (?-i) to work as expected.
2003 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2004 other effects) broke the use of negated acl sub-conditions.
2006 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2009 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2010 "Deny" causes Exim to reject the incoming connection with a 554 error.
2011 Unfortunately, if there is a major crisis, such as a disk failure,
2012 tcp-wrappers gives "deny", whereas what one would like would be some
2013 kind of temporary error. A kludge has been added to help with this.
2014 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2015 554 error is used if errno is still zero or contains ENOENT (which occurs
2016 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2019 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2021 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2022 errors. Otherwise a message that provokes a temporary error (when other
2023 messages do not) can cause a whole host to time out.
2025 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2026 addresses were routed directly to files or pipes from a redirect router.
2027 File deliveries just didn't batch; pipe deliveries might have suffered
2030 PH/06 A failure to get a lock for a hints database would erroneously always say
2031 "Failed to get write lock", even when it was really a read lock.
2033 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2034 of 0600. This has been changed to use the value of the lockfile_mode
2035 option (which defaults to 0600).
2037 PH/08 Applied small patch from the Sieve maintainer.
2039 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2040 folder from quota calculations, a direct delivery into this folder messed
2041 up the contents of the maildirsize file. This was because the regex was
2042 used only to exclude .Trash (or whatever) when the size of the mailbox
2043 was calculated. There was no check that a delivery was happening into an
2044 excluded directory. This bug has been fixed by ignoring all quota
2045 processing for deliveries into excluded directories.
2047 PH/10 Added the maildirfolder_create_regex option to appendfile.
2053 PH/01 The code for finding all the local interface addresses on a FreeBSD
2054 system running IPv6 was broken. This may well have applied to all BSD
2055 systems, as well as to others that have similar system calls. The broken
2056 code found IPv4 interfaces correctly, but gave incorrect values for the
2057 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2058 that it would not match correctly against @[] and not recognize the IPv6
2061 PH/02 The ipliteral router was not recognizing addresses of the form user@
2062 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2064 PH/03 Added disable_ipv6.
2066 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2067 decoded form, because it is most often used to construct To: headers
2068 lines in autoreplies, and the decoded form may well be syntactically
2069 invalid. However, $reply_address has leading white space removed, and all
2070 newlines turned into spaces so that the autoreply transport does not
2073 PH/05 If group was specified without a user on a router, and no group or user
2074 was specified on a transport, the group from the router was ignored.
2076 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2077 for visible compile-time settings that can be used to change these
2078 numbers, for those that want even more. Backwards compatibility with old
2079 spool files has been maintained. However, going back to a previous Exim
2080 release will lost any variables that are in spool files.
2082 PH/07 Two small changes when running in the test harness: increase delay when
2083 passing a TCP/IP connection to a new process, in case the original
2084 process has to generate a bounce, and remove special handling of
2085 127.0.0.2 (sic), which is no longer necessary.
2087 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2088 be the same on different OS.
2090 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2093 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2094 whether --show-vars was specified or not
2096 JJ/02 exipick: Added support for new ACL variable spool format introduced
2099 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2100 syntactically invalid From: or Reply-to: line, and a filter used this to
2101 generate an autoreply, and therefore failed to obtain an address for the
2102 autoreply, Exim could try to deliver to a non-existent relative file
2103 name, causing unrelated and misleading errors. What now happens is that
2104 it logs this as a hard delivery error, but does not attempt to create a
2107 PH/11 The exinext utility has a -C option for testing purposes, but although
2108 the given file was scanned by exinext itself; it wasn't being passed on
2109 when Exim was called.
2111 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2112 an end-of-file indication when reading a command response.
2114 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2115 compiled. In many other places in Exim, IPv6 addresses are always
2116 recognized, so I have changed this. It also means that IPv4 domain
2117 literals of the form [IPV4:n.n.n.n] are now always recognized.
2119 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2120 used if the router is not running as root, for example, when verifying at
2121 ACL time, or when using -bh. The debugging output from this situation was
2122 non-existent - all you got was a failure to exec. I have made two
2125 (a) Failures to set uid/gid, the current directory, or a process leader
2126 in a subprocess such as that created by queryprogram now generate
2127 suitable debugging ouput when -d is set.
2129 (b) The queryprogram router detects when it is not running as root,
2130 outputs suitable debugging information if -d is set, and then runs
2131 the subprocess without attempting to change uid/gid.
2133 PH/15 Minor change to Makefile for building test_host (undocumented testing
2136 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2137 additional section of a DNS packet that returns MX or SRV records.
2138 Instead, it always explicitly searches for A/AAAA records. This avoids
2139 major problems that occur when a DNS server includes only records of one
2140 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2141 fixed another bug: if SRV records were looked up and the corresponding
2142 address records were *not* found in the additional section, the port
2143 values from the SRV records were lost.
2145 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2146 using the correct key (the original address) when searching the retry
2147 rules in order to find which one to use for generating the retry hint.
2149 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2150 adding the default one. Similarly, if it contains a Reply-To: header, the
2151 errors_reply_to option, if set, is not used.
2153 PH/19 When calculating a retry time, Exim used to measure the "time since
2154 failure" by looking at the "first failed" field in the retry record. Now
2155 it does not use this if it is later than than the arrival time of the
2156 message. Instead it uses the arrival time. This makes for better
2157 behaviour in cases where some deliveries succeed, thus re-setting the
2158 "first failed" field. An example is a quota failure for a huge message
2159 when small messages continue to be delivered. Without this change, the
2160 "time since failure" will always be short, possible causing more frequent
2161 delivery attempts for the huge message than are intended.
2162 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2164 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2165 $1, $2, $3) because the numerical variables can be reset during some
2166 expansion items (e.g. "match"), thereby losing the authentication data.
2168 PH/21 Make -bV show the size of off_t variables so that the test suite can
2169 decide whether to run tests for quotas > 2G.
2171 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2172 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2173 filecount value is greater than 2G or if a quota value is greater than 2G
2174 on a system where the size of off_t is not greater than 4, a panic error
2177 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2178 never match. The debug and -bh output now contains an explicit error
2179 message indicating a malformed IPv4 address or mask.
2181 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2182 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2183 PH/23 above applies.
2185 PH/25 Do not write to syslog when running in the test harness. The only
2186 occasion when this arises is a failure to open the main or panic logs
2187 (for which there is an explicit test).
2189 PH/26 Added the /no_tell option to "control=freeze".
2191 PH/27 If a host name lookup failed very early in a connection, for example, if
2192 the IP address matched host_lookup and the reverse lookup yielded a name
2193 that did not have a forward lookup, an error message of the form "no IP
2194 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2195 could be logged. Now it outputs the IP address instead of "NULL".
2197 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2198 allows the sender and the authenticated sender to be set when
2199 submitting a message from within Exim. Since child_open_exim() is
2200 documented for local_scan(), the new function should be too.
2202 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2203 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2204 results in an empty string is now treated as unset.
2206 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2208 PH/31 Added sender_verify_fail logging option.
2210 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2211 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2212 tidied the source and removed it altogether.
2214 PH/33 When a queue run was abandoned because the load average was too high, a
2215 log line was always written; now it is written only if the queue_run log
2216 selector is set. In addition, the log line for abandonment now contains
2217 information about the queue run such as the pid. This is always present
2218 in "start" and "stop" lines but was omitted from the "abandon" line.
2220 PH/34 Omit spaces between a header name and the colon in the error message that
2221 is given when verify = headers_syntax fails (if there are lots of them,
2222 the message gets confusing).
2224 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2225 names, as there are now some PTR records that contain slashes. This check
2226 is only to protect against broken name servers that fall over on strange
2227 characters, so the fact that it applies to all lookups doesn't matter.
2229 PH/36 Now that the new test suite is complete, we can remove some of the
2230 special code in Exim that was needed for the old test suite. For example,
2231 sorting DNS records because real resolvers return them in an arbitrary
2232 order. The new test suite's fake resolver always returns records in the
2235 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2236 bounces) except when queue_only is set, to avoid logging races between
2237 the different processes.
2239 PH/38 Panic-die if .include specifies a non-absolute path.
2241 PH/39 A tweak to the "H" retry rule from its user.
2243 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2244 a label. They prevented compilation on older perls.
2246 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2247 a warning to be raised on newish perls.
2249 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2250 on queue. Changes to match documented behaviour of showing count of
2251 messages matching specified criteria.
2253 PH/40 Changed the default ident timeout from 30s to 5s.
2255 PH/41 Added support for the use of login_cap features, on those BSD systems
2256 that have them, for controlling the resources used by pipe deliveries.
2258 PH/42 The content-scanning code uses fopen() to create files in which to put
2259 message data. Previously it was not paying any attention to the mode of
2260 the files. Exim runs with umask(0) because the rest of the code creates
2261 files with open(), and sets the required mode explicitly. Thus, these
2262 files were ending up world-writeable. This was not a big issue, because,
2263 being within the spool directory, they were not world-accessible. I have
2264 created a function called modefopen, which takes an additional mode
2265 argument. It sets umask(777), creates the file, chmods it to the required
2266 mode, then resets the umask. All the relevant calls to fopen() in the
2267 content scanning code have been changed to use this function.
2269 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2270 to 24 hours. This avoids potential overflow problems when processing G
2271 and H retry rules. I suspect nobody ever tinkers with this value.
2273 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2275 PH/45 When the plaintext authenticator is running as a client, the server's
2276 challenges are checked to ensure they are valid base64 strings. By
2277 default, the authentication attempt is cancelled if an invalid string is
2278 received. Setting client_ignore_invalid_base64 true ignores these errors.
2279 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2280 they are received. Thus, the responses can be made to depend on the
2281 challenges. If an invalid string is ignored, an empty string is placed in
2284 PH/46 Messages that are created by the autoreply transport now contains a
2285 References: header, in accordance with RFCs 2822 and 3834.
2287 PH/47 Added authenticated_sender_force to the smtp transport.
2289 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2291 PH/49 Installed latest patch from the Sieve maintainer.
2293 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2294 was also set, the appendfile transport was unnecessarily scanning a
2295 directory of message files (e.g. for maildir delivery) to find the count
2296 of files (along with the size), even though it did not need this
2297 information. It now does the scan only if it needs to find either the
2298 size of the count of files.
2300 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2302 PH/52 Two bugs concerned with error handling when the smtp transport is
2305 (i) Exim was not creating retry information for temporary errors given
2306 for individual recipients after the DATA command when the smtp transport
2307 was used in LMTP mode. This meant that they could be retried too
2308 frequently, and not timed out correctly.
2310 (ii) Exim was setting the flag that allows error details to be returned
2311 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2312 recipients that were returned after the DATA command.
2314 PH/53 This is related to PH/52, but is more general: for any failing address,
2315 when detailed error information was permitted to be returned to the
2316 sender, but the error was temporary, then after the final timeout, only
2317 "retry timeout exceeded" was returned. Now it returns the full error as
2318 well as "retry timeout exceeded".
2320 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2321 do this, and (what is worse) MTAs that accept it.
2323 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2324 will now be deprecated.
2326 PH/56 New os.c-cygwin from the Cygwin maintainer.
2328 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2329 formats (previously only available in exim formats via -bpr, -bpru,
2330 and -bpra. Now also available in native and exiqgrep formats)
2332 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2333 with very large, slow to parse queues
2335 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2337 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2339 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2340 responses to authentication challenges, though it was showing the
2341 challenges; (ii) I've removed the CR characters from the debug output for
2344 PH/58 Allow for the insertion of a newline as well as a space when a string
2345 is turned into more than one encoded-word during RFC 2047 encoding. The
2346 Sieve code now uses this.
2348 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2349 data_4xx, lost_connection, tls_required.
2351 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2352 message was being sent as an SMTP response.
2354 PH/61 Add -l and -k options to exicyclog.
2356 PH/62 When verifying, if an address was redirected to one new address, so that
2357 verification continued, and the new address failed or deferred after
2358 having set something in $address_data, the value of $address_data was not
2359 passed back to the ACL. This was different to the case when no
2360 redirection occurred. The value is now passed back in both cases.
2362 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2363 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2364 use login_cap.h, so on its own it isn't the distinguishing feature. The
2365 new name refers directly to the setclassresources() function.
2367 PH/65 Added configuration files for NetBSD3.
2369 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2371 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2372 is preferred over IPv4.
2374 PH/68 The bounce_return_message and bounce_return_body options were not being
2375 honoured for bounces generated during the reception of non-SMTP messages.
2376 In particular, this applied to messages rejected by the ACL. This bug has
2377 been fixed. However, if bounce_return_message is true and bounce_return_
2378 body is false, the headers that are returned for a non-SMTP message
2379 include only those that have been read before the error was detected.
2380 (In the case of an ACL rejection, they have all been read.)
2382 PH/69 The HTML version of the specification is now built in a directory called
2383 spec_html instead of spec.html, because the latter looks like a path with
2384 a MIME-type, and this confuses some software.
2386 PH/70 Catch two compiler warnings in sieve.c.
2388 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2389 function verify_get_ident() calls ip_connect() to connect a socket, but
2390 if the "connect()" function timed out, ip_connect() used to close the
2391 socket. However, verify_get_ident() also closes the socket later, and in
2392 between Exim writes to the log, which may get opened at this point. When
2393 the socket was closed in ip_connect(), the log could get the same file
2394 descriptor number as the socket. This naturally causes chaos. The fix is
2395 not to close the socket in ip_connect(); the socket should be closed by
2396 the function that creates it. There was only one place in the code where
2397 this was missing, in the iplookup router, which I don't think anybody now
2398 uses, but I've fixed it anyway.
2400 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2401 well as to direct DNS lookups. Otherwise the handling of names in host
2402 lists is inconsistent and therefore confusing.
2408 PH/01 Two changes to the default runtime configuration:
2410 (1) Move the checks for relay_from_hosts and authenticated clients from
2411 after to before the (commented out) DNS black list checks.
2413 (2) Add control=submission to the relay_from_hosts and authenticated
2414 clients checks, on the grounds that messages accepted by these
2415 statements are most likely to be submissions.
2417 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2419 (1) Generate an error if the third argument for the ${prvs expansion is
2422 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2425 (3) Reset the variables that are obtained from the first argument of
2426 ${prvscheck and used in the second argument before leaving the code,
2427 because their memory is reclaimed, so using them afterwards may do
2430 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2431 one (it's much easier than Tom thought :-).
2433 (5) Because of (4), we can now allow for the use of $prvscheck_result
2434 inside the third argument.
2436 PH/03 For some reason, the default setting of PATH when running a command from
2437 a pipe transport was just "/usr/bin". I have changed it to
2440 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2441 anything to be listed in the output from -bV.
2443 PH/05 When a filter generated an autoreply, the entire To: header line was
2444 quoted in the delivery log line, like this:
2446 => >A.N.Other <ano@some.domain> <original@ddress> ...
2448 This has been changed so that it extracts the operative address. There
2449 may be more than one such address. If so, they are comma-separated, like
2452 => >ano@some.domain,ona@other.domain <original@ddress> ...
2454 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2455 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2456 not being looked up in the rDNS to get a host name, Exim was showing the
2457 IP address twice in Received: lines, even though the IP addresses were
2458 identical. For example:
2460 Received: from [1.2.3.4] (helo=[1.2.3.4])
2462 However, if the real host name was known, it was omitting the HELO data
2463 if it matched the actual IP address. This has been tidied up so that it
2464 doesn't show the same IP address twice.
2466 PH/07 When both +timestamp and +memory debugging was on, the value given by
2467 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2468 called by the debug printing, thereby overwriting the timestamp buffer.
2469 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2471 PH/08 When the original message was included in an autoreply transport, it
2472 always said "this is a copy of the message, including all the headers",
2473 even if body_only or headers_only was set. It now gives an appropriate
2476 PH/09 Applied a patch from the Sieve maintainer which:
2478 o fixes some comments
2479 o adds the (disabled) notify extension core
2480 o adds some debug output for the result of if/elsif tests
2481 o points to the current vacation draft in the documentation
2482 and documents the missing references header update
2486 o fixes a bug in processing the envelope test (when testing
2487 multiple envelope elements, the last element determinted the
2490 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2491 Electronic Mail") by including:
2493 Auto-submitted: auto-generated
2495 in the messages that it generates (bounce messages and others, such as
2496 warnings). In the case of bounce messages for non-SMTP mesages, there was
2497 also a typo: it was using "Auto_submitted" (underscore instead of
2498 hyphen). Since every message generated by Exim is necessarily in response
2499 to another message, thes have all been changed to:
2501 Auto-Submitted: auto-replied
2503 in accordance with these statements in the RFC:
2505 The auto-replied keyword:
2507 - SHOULD be used on messages sent in direct response to another
2508 message by an automatic process,
2510 - MUST NOT be used on manually-generated messages,
2512 - MAY be used on Delivery Status Notifications (DSNs) and Message
2513 Disposition Notifications (MDNs),
2515 - MUST NOT be used on messages generated by automatic or periodic
2516 processes, except for messages which are automatic responses to
2519 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2520 to the default Received: header definition.
2522 PH/12 Added log selector acl_warn_skipped (default on).
2524 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2525 variables because (a) they are in the wrong storage pool and (b) even if
2526 they were copied, it wouldn't work properly because of the caching.
2528 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2529 checking when decoding. Apparently there are clients that generate
2530 overlong encoded strings. Why am I not surprised?
2532 PH/15 If the first argument of "${if match_address" was not empty, but did not
2533 contain an "@" character, Exim crashed. Now it writes a panic log message
2534 and treats the condition as false.
2536 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2538 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2539 extension "envelope-auth". The code is finished and in agreement with
2540 other implementations, but there is no documentation so far and in fact,
2541 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2542 not changing the active code.
2544 Print executed "if" and "elsif" statements when debugging is used. This
2545 helps a great deal to understand what a filter does.
2547 Document more things not specified clearly in RFC3028. I had all this
2548 sorted out, when out of a sudden new issues came to my mind. Oops."
2550 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2553 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2554 canonical form (as documented). However, after a host name lookup from
2555 the IP address, check_host() was doing a simple string comparison with
2556 addresses acquired from the DNS when checking that the found name did
2557 have the original IP as one of its addresses. Since any found IPv6
2558 addresses are likely to be in abbreviated form, the comparison could
2559 fail. Luckily, there already exists a function for doing the comparison
2560 by converting both addresses to binary, so now that is used instead of
2561 the text comparison.
2563 PH/20 There was another similar case to PH/19, when a complete host name was
2564 given in a host list; looking up its IP address could give an abbreviated
2565 form, whereas the current host's name might or might not be abbreviated.
2566 The same fix has been applied.
2572 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2573 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2576 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2577 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2579 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2581 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2582 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2583 Also, for "Auto-Submitted", treat anything other than "no" as
2584 non-personal, in accordance with RFC 3834. (Previously it treated
2585 anything starting "auto-" as non-personal.)
2587 TF/01 The control=submission/name=... option had a problem with syntax
2588 errors if the name included a slash character. The /name= option
2589 now slurps the rest of the string, so it can include any characters
2590 but it must come last in the list of options (after /sender_retain
2593 PH/05 Some modifications to the interface to the fake nameserver for the new
2601 TK/01 Added the "success_on_redirect" address verification option. See
2602 NewStuff for rationale and an example.
2604 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2606 PH/02 Patch to exigrep to allow it to work on syslog lines.
2608 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2609 fread() to skip over the body file's header line, because in Cygwin the
2610 header line is locked and is inaccessible.
2612 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2613 co-exist for some time) to make it clear that it is the Exim ID that is
2614 referenced, not the Message-ID: header line.
2616 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2617 string_format() function, because snprintf() does not exist on all
2620 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2621 ${stat: expansion item.
2623 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2624 protocol synchronization error", to keep the pedants happy.
2626 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2627 well as for IRIX systems, when gcc is being used. See the host.c source
2630 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2632 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2635 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2636 transport and to the smtp transport in LMTP mode.
2638 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2640 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2641 run a filter in a subprocess. This could lead to confusion in subsequent
2642 lookups in the parent process. There should also be a search_tidyup() at
2643 the end of the subprocess.
2645 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2646 only if the host matched helo_try_verify_hosts, which caused the
2647 verification to occur when the EHLO/HELO command was issued. The ACL just
2648 tested the remembered result. Now, if a previous verification attempt has
2649 not happened, "verify = helo" does it there and then.
2651 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2653 TK/03 Fix log output including CR from clamd.
2655 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2656 exist provoked a memory error which could cause a segfault.
2658 PH/15 Installed PCRE 6.2
2660 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2662 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2663 of the problem. Specifically, suggested +O2 rather than +O1 for the
2666 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2668 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2669 record, the port was not correctly logged when the outgoing_port log
2670 selector was set (it logged the transort's default port).
2672 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2673 fallback_hosts, and "hosts" in the smtp transport.
2675 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2676 host errors such as "Connection refused".
2678 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2679 authentication with radiusclient 0.4.9:
2681 - Error returned from rc_read_config was caught wrongly
2682 - Username/password not passed on to radius server due to wrong length.
2684 The presumption is that some radiusclient API changes for 4.51/PH/17
2685 were not taken care of correctly. The code is still untested by me (my
2686 Linux distribution still has 0.3.2 of radiusclient), but it was
2687 contributed by a Radius user.
2689 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2690 expanding the "port" option of the smtp transport.
2692 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2693 while reading a MIME header. Thanks to Tom Hughes for a patch.
2695 PH/24 Include config.h inside local_scan.h so that configuration settings are
2698 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2699 that in an ACL for RCPT (for example), you can examine exactly what was
2702 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2703 commands, but it was not correctly comparing the address with the actual
2704 client host address. Thus, it would show the EHLO address in Received:
2705 header lines when this was not necessary.
2707 PH/27 Added the % operator to ${eval:}.
2709 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2710 it should be ignoring failures (because with -C, for example, it has lost
2711 privilege). It wasn't ignoring creation failures other than "already
2714 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2717 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2718 because an input line was too long, either on its own, or by virtue of
2719 too many continuations, the temporary file was not being removed, and the
2720 return code was incorrect.
2722 PH/31 Missing "BOOL" in function definition in filtertest.c.
2724 PH/32 Applied Sieve patches from the maintainer.
2726 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2728 PH/33 Added "verify = not_blind".
2730 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2731 Local/Makefile (with some defaults set). These are used in built scripts
2732 such as exicyclog, but they have never been used in the exim_install
2733 script (though there are many overriding facilities there). I have
2734 arranged that the exim_install script now takes note of these two
2737 PH/35 Installed configuration files for Dragonfly.
2739 PH/36 When a locally submitted message by a trusted user did not contain a
2740 From: header, and the sender address was obtained from -f or from an SMTP
2741 MAIL command, and the trusted user did not use -F to supply a sender
2742 name, $originator_name was incorrectly used when constructing a From:
2743 header. Furthermore, $originator_name was used for submission mode
2744 messages from external hosts without From: headers in a similar way,
2745 which is clearly wrong.
2747 PH/37 Added control=suppress_local_fixups.
2749 PH/38 When log_selector = +received_sender was set, and the addition of the
2750 sender made the log line's construction buffer exactly full, or one byte
2751 less than full, an overflow happened when the terminating "\n" was
2754 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2755 when the result of a list match is failure because a DNS lookup failed.
2757 PH/40 RM_COMMAND is now used in the building process.
2759 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2760 the "build-* directories that it finds.
2762 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2763 address in a domain literal was a prefix of an interface address.
2765 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2766 when verifying a sender address, unless rewrite_headers is false.
2768 PH/44 Wrote a long comment about why errors_to addresses are verified as
2769 recipients, not senders.
2771 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2772 the ratelimit ACL was added.
2774 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2776 PH/46 Added extra information about PostgreSQL errors to the error string.
2778 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2779 avoiding the need to install special zones in a real server. This is
2780 backwards compatible; if it can't find the fake resolver, it drops back.
2781 Thus, both old and new test suites can be run.
2783 TF/02 Added util/ratelimit.pl
2785 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2786 clock is set back in time.
2788 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2789 Candler <B.Candler@pobox.com>.
2791 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2792 OK for addresses that are the result of redirections.
2794 PH/48 A number of further additions for the benefit of the new test suite,
2795 including a fake gethostbyname() that interfaces to the fake DNS resolver
2798 TF/06 The fix for widen_domains has also been applied to qualify_single and
2799 search_parents which are the other dnslookup options that can cause
2802 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2805 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2807 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2808 service" if it encountered an MX record with an empty target hostname.
2809 The message is now "an MX or SRV record indicated no SMTP service".
2811 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2812 if the DNS of the sending site is misconfigured. This is quite a
2813 common situation. This change restores the behaviour of treating a
2814 helo verification defer as a failure.
2816 PH/51 If self=fail was set on a router, the bounce message did not include the
2817 actual error message.
2823 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2825 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2826 command itself ended in error, the underlying message about the transport
2827 filter timeout was being overwritten with the pipe command error. Now the
2828 underlying error message should be appended to the second error message.
2830 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2832 PH/02 Reduce the amount of output that "make" produces by default. Full output
2833 can still be requested.
2835 PH/03 The warning log line about a condition test deferring for a "warn" verb
2836 was being output only once per connection, rather than after each
2837 occurrence (because it was using the same function as for successful
2838 "warn" verbs). This seems wrong, so I have changed it.
2840 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2841 it should not have, which might have caused a crash in the right
2842 circumstances, but probably never did.
2844 PH/04 Installed a modified version of Tony Finch's patch to make submission
2845 mode fix the return path as well as the Sender: header line, and to
2846 add a /name= option so that you can make the user's friendly name appear
2849 TF/03 Added the control = fakedefer ACL modifier.
2851 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2852 Mark Lowes for thorough testing.
2854 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2856 TK/03 Merged latest SRS patch from Miles Wilton.
2858 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2859 with the definition in sysexits.h (which is #included earlier).
2860 Fortunately, Exim does not actually use EX_OK. The code used to try to
2861 preserve the sysexits.h value, by assumimg that macro definitions were
2862 scanned for macro replacements. I have been disabused of this notion,
2863 so now the code just undefines EX_OK before #including unistd.h.
2865 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2866 in the smtp transport. When a block could not be written in a single
2867 write() function, the timeout was being re-applied to each part-write.
2868 This seems wrong - if the receiver was accepting one byte at a time it
2869 would take for ever. The timeout is now adjusted when this happens. It
2870 doesn't have to be particularly precise.
2872 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2873 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2875 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2876 without a domain if the check to <postmaster@domain> fails.
2878 SC/01 Eximstats: added -xls and the ability to specify output files
2879 (patch written by Frank Heydlauf).
2881 SC/02 Eximstats: use FileHandles for outputing results.
2883 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2885 SC/04 Eximstats: fixed display of large numbers with -nvr option
2887 SC/05 Eximstats: fixed merging of reports with empty tables.
2889 SC/06 Eximstats: added the -include_original_destination flag
2891 SC/07 Eximstats: removed tabs and trailing whitespace.
2893 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2895 TK/06 MBOX spool code: Add real "From " MBOX separator line
2896 so the .eml file is really in mbox format (even though
2897 most programs do not really care). Patch from Alex Miller.
2899 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2900 The latter is generated from $received_to and is only set if the
2901 message has one envelope recipient. SA can use these headers,
2902 obviously out-of-the-box. Patch from Alex Miller.
2904 PH/08 The ${def test on a variable was returning false if the variable's
2905 value was "0", contrary to what the specification has always said!
2906 The result should be true unless the variable is empty.
2908 PH/09 The syntax error of a character other than { following "${if
2909 def:variable_name" (after optional whitespace) was not being diagnosed.
2910 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2911 accidental colon was present, for example, could give incorrect results.
2913 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2914 result is used (not including appendfile, where other changes are about
2917 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2918 This involved changing a lot of size variables from int to off_t. It
2919 should work with maildirs and everything.
2921 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2922 spamd dying while we are connected to it.
2924 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2927 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2928 function for BDB 4.3.
2930 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2932 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2933 the sources for PCRE have been re-arranged and more files are now
2936 PH/15 The code I had for printing potentially long long variables in PH/11
2937 above was not the best (it lost precision). The length of off_t variables
2938 is now inspected at build time, and an appropriate printing format (%ld
2939 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2940 to be "long long int" or "long int". This is needed for the internal
2941 formatting function string_vformat().
2943 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2944 the configuration file to be ":syslog", then the script "guesses" where
2945 the logs files are, rather than using the compiled in default. In our
2946 case the guess is not the same as the compiled default, so the script
2947 suddenly stopped working when I started to use syslog. The patch checks
2948 to see if log_file_path is "". If so, it attempts to read it from exim
2949 with no configuration file to get the compiled in version, before it
2950 falls back to the previous guessing code."
2952 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2953 implementing BATV in an Exim configuration. See NewStuff for the gory
2956 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2957 Makefile that are specific to HP-UX.
2959 PH/18 If the "use_postmaster" option was set for a recipient callout together
2960 with the "random" option, the postmaster address was used as the MAIL
2961 FROM address for the random test, but not for the subsequent recipient
2962 test. It is now used for both.
2964 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2965 patch removes a few documentation additions to RFC 3028, because the
2966 latest draft now contains them. It adds the new en;ascii-case comparator
2967 and a new error check for 8bit text in MIME parts. Comparator and
2968 require names are now matched exactly. I enabled the subaddress
2969 extension, but it is not well tested yet (read: it works for me)."
2971 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2972 rework some of the code of TK/09 above to avoid the hardwired use of
2973 "%lld" and "long long". Replaced the call to snprintf() with a call to
2976 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2977 records point to non-existent hosts", "retry timeout exceeded", and
2978 "retry time not reached for any host after a long failure period".
2980 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2981 experimental DomainKeys support:
2983 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2984 (2) On an error such as an illegally used "control", the wrong name for
2985 the control was given.
2987 These problems did NOT occur unless DomainKeys support was compiled.
2989 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2991 PH/24 Added ${if match_ip condition.
2993 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2994 databases so that it will be absolutely obvious if a crash occurs in the
2995 DB library. This is a regular occurrence (often caused by mis-matched
2998 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2999 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3000 picked up on a user's system that detects such things. There doesn't seem
3001 to be a gcc warning option for this - only an attribute that has to be
3002 put on the function's prototype. It seems that in Fedora Core 4 they have
3003 set this on a number of new functions. No doubt there will be more in due
3006 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3007 specify a transport. However, if an address that was verified by such a
3008 router was the subject of a callout, Exim crashed because it tried to
3009 read the rcpt_include_affixes from the non-existent transport. Now it
3010 just assumes that the setting of that option is false. This bug was
3011 introduced by 4.51/PH/31.
3013 PH/28 Changed -d+all to exclude +memory, because that information is very
3014 rarely of interest, but it makes the output a lot bigger. People tend to
3015 do -d+all out of habit.
3017 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3018 code in os-type was giving problems when libc.so lives in lib64, like on
3021 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3022 aren't the modern standard, and it seems that some systems' include files
3023 don't always have them. Exim was already checking for some of the newer
3024 ones like T_AAAA, and defining it itself. I've added checks for all the
3025 record types that Exim uses.
3027 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3028 not automatically generating a new one, as it is supposed to. This
3029 prevented TLS from working. If the file did exist, but contained invalid
3030 data, a new version was generated, as expected. It was only the case of a
3031 non-existent file that was broken.
3033 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3034 with a change in libdomainkeys > 0.64.
3036 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3037 from DNS. If the selector record carries the flag, it now has
3038 precedence over the domain-wide flag.
3040 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3042 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3043 the use of a port name that isn't defined in /etc/services) occurred, the
3044 message was deferred as in a normal delivery, and thus remained on the
3045 spool, instead of being failed because of the mua_wrapper setting. This
3046 is now fixed, and I tidied up some of the mua_wrapper messages at the
3049 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3050 the messages in a hash of arrays rather than using individual hashes.
3051 This is a bit cleaner and results in dramatic memory savings, albeit
3052 at a slight CPU cost.
3054 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3055 as requested by Marc Sherman.
3057 SC/10 Eximstats: added histograms for user specified patterns as requested
3060 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3062 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3063 fopen() in the content-scanning modules that did not already have it.
3069 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3070 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3072 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3074 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3076 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3077 to test Sieve filters that use "vacation".
3079 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3080 that changes the way the GnuTLS parameters are stored in the cache file.
3081 The new format can be generated externally. For backward compatibility,
3082 if the data in the cache doesn't make sense, Exim assumes it has read an
3083 old-format file, and it generates new data and writes a new file. This
3084 means that you can't go back to an older release without removing the
3087 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3088 work if there are any delivery delays because "one_time" forces the
3089 parent to be marked "delivered", so its unseen clone is never tried
3090 again. For this reason, Exim now forbids the simultaneous setting of
3093 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3094 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3095 entry if you want to know the details. The fix, however, neglected to
3096 consider the case when local delivery batching is involved. The test for
3097 "previously delivered" was not happening when checking to see if an
3098 address could be batched with a previous (undelivered) one; under
3099 certain circumstances this could lead to multiple deliveries to the same
3102 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3103 in its include files, and this causes problems building Exim.
3105 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3106 header_syntax) but Exim was just ignoring anything given after a slash.
3107 In particular, this caused confusion with an attempt to use "verify =
3108 reverse_host_lookup/defer_ok". An error is now given when options are
3109 supplied for verify items that do not have them. (Maybe reverse_host_
3110 lookup should have a defer_ok option, but that's a different point.)
3112 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3113 defined by RFC 821) to 2048, because there were problems with some AUTH
3114 commands, and RFC 1869 says the size should be increased for extended
3115 SMTP commands that take arguments.
3117 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3120 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3121 "unknown" error; now it says that the functionality isn't in the binary.
3123 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3124 an address' error message when a string expansion fails (syntax or
3125 whatever). Otherwise the password may appear in the log. Following change
3126 PH/42 below, there is no longer a chance of it appearing in a bounce
3129 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3131 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3132 of its addresses was ever tried. (Bugzilla bug #2.)
3134 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3135 the result incorrectly in the debug output. (It correctly added a newline
3136 to what was transported.)
3138 TF/01 Added $received_time.
3140 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3141 commented out examples of how to interface to a virus scanner and to
3142 SpamAssassin. Also added commented examples of av_scanner and
3143 spamd_address settings.
3145 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3146 and controls are allowed in which ACLs. There were a couple of minor
3147 errors. Some of the entries in the conditions table (which is a table of
3148 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3149 negation of where the condition IS allowed.
3151 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3153 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3154 header file does not have a version number, so I've had to invent a new
3155 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3156 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3157 radiusclient), but it was contributed by a Radius user.
3159 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3160 files or directories, for OpenSSL.
3162 PH/19 When an Exim process that is running as root has to create an Exim log
3163 file, it does so in a subprocess that runs as exim:exim so as to get the
3164 ownership right at creation (otherwise, other Exim processes might see
3165 the file with the wrong ownership). There was no test for failure of this
3166 fork() call, which would lead to the process getting stuck as it waited
3167 for a non-existent subprocess. Forks do occasionally fail when resources
3168 run out. I reviewed all the other calls to fork(); they all seem to check
3171 PH/20 When checking for unexpected SMTP input at connect time (before writing
3172 the banner), Exim was not dealing correctly with a non-positive return
3173 from the read() function. If the client had disconnected by this time,
3174 the result was a log entry for a synchronization error with an empty
3175 string after "input=" when read() returned zero. If read() returned -1
3176 (an event I could not check), uninitialized data bytes were printed.
3177 There were reports of junk text (parts of files, etc) appearing after
3180 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3182 PH/22 Added support for macro redefinition, and (re)definition in between
3183 driver and ACL definitions.
3185 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3186 forgetting to use the resulting value; it was using the unexpanded value.
3188 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3189 hadn't been configured. The fix is from Juergen Kreileder, who
3190 understands it better than I do:
3192 "Here's what I see happening with three configured cyrus_sasl
3193 authenticators configured (plain, login, cram-md5):
3195 On startup auth_cyrus_sasl_init() gets called for each of these.
3196 This means three calls to sasl_listmech() without a specified mech_list.
3197 => SASL tests which mechs of all available mechs actually work
3198 => three warnings about OTP not working
3199 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3201 With the patch, sasl_listmech() also gets called three times. But now
3202 SASL's mech_list option is set to the server_mech specified in the the
3203 authenticator. Or in other words, the answer from sasl_listmech()
3204 gets limited to just the mech you're testing for (which is different
3206 => the return list contains just 'plain' or 'login', 'cram-md5' or
3207 nothing depending on the value of ob->server_mech.
3209 I've just tested the patch: Authentication still works fine,
3210 unavailable mechs specified in the exim configuration are still
3211 caught, and the auth.log warnings about OTP are gone."
3213 PH/25 When debugging is enabled, the contents of the command line are added
3214 to the debugging output, even when log_selector=+arguments is not
3217 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3218 answer is "GNU", and only if the return is "GNU/something" is the answer
3221 PH/27 $acl_verify_message is now set immediately after the failure of a
3222 verification in an ACL, and so is available in subsequent modifiers. In
3223 particular, the message can be preserved by coding like this:
3225 warn !verify = sender
3226 set acl_m0 = $acl_verify_message
3228 Previously, $acl_verify_message was set only while expanding "message"
3229 and "log_message" when a very denied access.
3231 PH/28 Modified OS/os.c-Linux with
3233 -#ifndef OS_LOAD_AVERAGE
3234 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3236 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3237 nomenclature these days.)
3239 PH/29 Installed patch from the Sieve maintainer that adds the options
3240 sieve_useraddress and sieve_subaddress to the redirect router.
3242 PH/30 In these circumstances:
3243 . Two addresses routed to the same list of hosts;
3244 . First host does not offer TLS;
3245 . First host accepts first address;
3246 . First host gives temporary error to second address;
3247 . Second host offers TLS and a TLS session is established;
3248 . Second host accepts second address.
3249 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3250 and peerdn, if requested) that were in fact used only for the second
3253 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3254 attention to any local part prefix or suffix that was matched by the
3255 router that accepted the address. It now behaves in the same way as it
3256 does for delivery: the affixes are removed from the local part unless
3257 rcpt_include_affixes is set on the transport.
3259 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3260 timeout during the DATA phase of an incoming message.
3262 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3263 applied a patch sanctioned by the Sieve maintainer.
3265 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3266 the uid or gid is negative. A case of a negative gid caused this to be
3267 noticed. The fix allows for either to be negative.
3269 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3270 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3273 PH/36 The change PH/12 above was broken. Fixed it.
3275 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3276 the grounds that routing the same address twice would always produce the
3277 same answer. This might have been true once, but it is certainly no
3278 longer true now. Routing a child address may depend on the previous
3279 routing that produced that child. Some complicated redirection strategies
3280 went wrong when messages had multiple recipients, and made Exim's
3281 behaviour dependent on the order in which the addresses were given.
3283 I have moved the duplicate checking until after the routing is complete.
3284 Exim scans the addresses that are assigned to local and remote
3285 transports, and removes any duplicates. This means that more work will be
3286 done, as duplicates will always all be routed, but duplicates are
3287 presumably rare, so I don't expect this is of any significance.
3289 For deliveries to pipes, files, and autoreplies, the duplicate checking
3290 still happens during the routing process, since they are not going to be
3293 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3294 It corrects a timeout issue with spamd. This is Ian's comment: "The
3295 background is that sometimes spamd either never reads data from a
3296 connection it has accepted, or it never writes response data. The exiscan
3297 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3298 blindly assumes that writes won't block so it may never time out."
3300 PH/39 Allow G after quota size as well as K and M.
3302 PH/40 The value set for $authenticated_id in an authenticator may not contain
3303 binary zeroes or newlines because the value is written to log lines and
3304 to spool files. There was no check on this. Now the value is run through
3305 the string_printing() function so that such characters are converted to
3306 printable escape sequences.
3308 PH/41 $message_linecount is a new variable that contains the total number of
3309 lines in the message. Compare $body_linecount, which is the count for the
3312 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3313 bounce and delay warning messages, except in certain special cases, which
3316 (a) An SMTP error message from a remote host;
3317 (b) A message specified in a :fail: redirection;
3318 (c) A message specified in a "fail" command in a system filter;
3319 (d) A message specified in a FAIL return from the queryprogram router;
3320 (e) A message specified by the cannot_route_message router option.
3322 In these cases only, Exim does include the error details in bounce and
3323 warning messages. There are also a few cases where bland messages such
3324 as "unrouteable address" or "local delivery error" are given.
3326 PH/43 $value is now also set for the "else" part of a ${run expansion.
3328 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3329 being worked on, but at least Exim now implements the latest version to
3332 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3333 process to complete was treated as a delivery failure, a timeout while
3334 writing the message to the pipe was logged, but erroneously treated as a
3335 successful delivery. Such timeouts include transport filter timeouts. For
3336 consistency with the overall process timeout, these timeouts are now
3337 treated as errors, giving rise to delivery failures by default. However,
3338 there is now a new Boolean option for the pipe transport called
3339 timeout_defer, which, if set TRUE, converts the failures into defers for
3340 both kinds of timeout. A transport filter timeout is now identified in
3343 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3344 systems where "make" and "gmake" are different, calling "gmake" at top
3345 level broke things. I've arranged for the value of $(MAKE) to be passed
3346 from the Makefile to this script so that it can call the same version of
3350 A note about Exim versions 4.44 and 4.50
3351 ----------------------------------------
3353 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3354 changes of various kinds. As a consequence, a big documentation update was
3355 needed. This delayed the release for rather longer than seemed good, especially
3356 in the light of a couple of (minor) security issues. Therefore, the changes
3357 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3358 release. So 4.44 and 4.50 are in effect two different branches that both start
3361 I have left the 4.50 change log unchanged; it contains all the changes since
3362 4.43. The change log for 4.44 is below; many of its items are identical to
3363 those for 4.50. This seems to be the most sensible way to preserve the
3364 historical information.
3370 1. Minor wording change to the doc/README.SIEVE file.
3372 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3373 computation of the current number of files was incorrect.
3375 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3376 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3379 4. Give more explanation in the error message when the command for a transport
3380 filter fails to execute.
3382 5. There are several places where Exim runs a non-Exim command in a
3383 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3384 was being done only for the command run by the queryprogram router. It is
3385 now done for all such subprocesses. The other cases are: ${run, transport
3386 filters, and the commands run by the lmtp and pipe transports.
3388 6. Added CONFIGURE_GROUP build-time option.
3390 7. Some older OS have a limit of 256 on the maximum number of file
3391 descriptors. Exim was using setrlimit() to set 1000 as a large value
3392 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3393 systems. I've change it so that if it can't get 1000, it tries for 256.
3395 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3396 was an oversight, and furthermore, ever since the addition of extra
3397 controls (e.g. 4.43/32), the checks on when to allow different forms of
3398 "control" were broken. There should now be diagnostics for all cases when a
3399 control that does not make sense is encountered.
3401 9. Added the /retain_sender option to "control=submission".
3403 10. $recipients is now available in the predata ACL (oversight).
3405 11. Tidy the search cache before the fork to do a delivery from a message
3406 received from the command line. Otherwise the child will trigger a lookup
3407 failure and thereby defer the delivery if it tries to use (for example) a
3408 cached ldap connection that the parent has called unbind on.
3410 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3411 of $address_data from the recipient verification was clobbered by the
3412 sender verification.
3414 13. The value of address_data from a sender verification is now available in
3415 $sender_address_data in subsequent conditions in the ACL statement.
3417 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3419 15. Added a new option "connect=<time>" to callout options, to set a different
3422 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3423 was its contents. (It was OK if the option was not defined at all.)
3425 17. A "Completed" log line is now written for messages that are removed from
3426 the spool by the -Mrm option.
3428 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3429 information about exactly what failed.
3431 19. Added -dd to debug only the daemon process.
3433 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3434 handles timeouts, both on the server side and network timeouts. Renamed the
3435 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3437 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3438 It is now set to "smtps".
3440 22. $host_address is now set to the target address during the checking of
3441 ignore_target_hosts.
3443 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3444 being passed; this would have caused $sender_host_name to have been used if
3445 matching the list had actually called for a host name (not very likely,
3446 since this list is usually IP addresses). A host name is now passed as
3449 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3450 code that specifies a non-restarting handler (typically sigaction() in
3451 modern systems) in an attempt to fix a rare and obscure crash bug.
3453 25. Narrowed the window for a race in the daemon that could cause it to ignore
3454 SIGCHLD signals. This is not a major problem, because they are used only to
3455 wake it up if nothing else does.
3457 26. A malformed maildirsize file could cause Exim to calculate negative values
3458 for the mailbox size or file count. Odd effects could occur as a result.
3459 The maildirsize information is now recalculated if the size or filecount
3462 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3463 support for a long time. Removed HAVE_SYS_VFS_H.
3465 28. Installed the latest version of exipick from John Jetmore.
3467 29. In an address list, if the pattern was not a regular expression, an empty
3468 subject address (from a bounce message) matched only if the pattern was an
3469 empty string. Non-empty patterns were not even tested. This was the wrong
3470 because it is perfectly reasonable to use an empty address as part of a
3471 database query. An empty address is now tested by patterns that are
3472 lookups. However, all the other forms of pattern expect the subject to
3473 contain a local part and a domain, and therefore, for them, an empty
3474 address still always fails if the pattern is not itself empty.
3476 30. Exim went into a mad DNS loop when attempting to do a callout where the
3477 host was specified on an smtp transport, and looking it up yielded more
3478 than one IP address.
3480 31. Re-factored the code for checking spool and log partition space into a
3481 function that finds that data and another that does the check. The former
3482 is then used to implement four new variables: $spool_space, $log_space,
3483 $spool_inodes, and $log_inodes.
3485 32. The RFC2047 encoding function was originally intended for short strings
3486 such as real names; it was not keeping to the 75-character limit for
3487 encoded words that the RFC imposes. It now respects the limit, and
3488 generates multiple encoded words if necessary. To be on the safe side, I
3489 have increased the buffer size for the ${rfc2047: expansion operator from
3492 33. It is now permitted to omit both strings after an "if" condition; if the
3493 condition is true, the result is "true". As before, when the second string
3494 is omitted, a false condition yields an empty string. This makes it less
3495 cumbersome to write custom ACL and router conditions.
3497 34. Failure to deliver a bounce message always caused it to be frozen, even if
3498 there was an errors_to setting on the router. The errors_to setting is now
3501 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3502 canonical form (fully expanded) before being placed in
3503 $sender_host_address.
3505 36. The table in the code that translates DNS record types into text (T_A to
3506 "A" for instance) was missing entries for NS and CNAME. It is just possible
3507 that this could have caused confusion if both these types were looked up
3508 for the same domain, because the text type is used as part of Exim's
3509 per-process caching. But the chance of anyone hitting this buglet seems
3512 37. The dnsdb lookup has been extended in a number of ways.
3514 (1) There is a new type, "zns", which walks up the domain tree until it
3515 finds some nameserver records. It should be used with care.
3517 (2) There is a new type, "mxh", which is like "mx" except that it returns
3518 just the host names, not the priorities.
3520 (3) It is now possible to give a list of domains (or IP addresses) to be
3521 looked up. The behaviour when one of the lookups defers can be
3522 controlled by a keyword.
3524 (4) It is now possible to specify the separator character for use when
3525 multiple records are returned.
3527 38. The dnslists ACL condition has been extended: it is now possible to supply
3528 a list of IP addresses and/or domains to be looked up in a particular DNS
3531 39. Added log_selector=+queue_time_overall.
3533 40. When running the queue in the test harness, wait just a tad after forking a
3534 delivery process, to get repeatability of debugging output.
3536 41. Include certificate and key file names in error message when GnuTLS fails
3537 to set them up, because the GnuTLS error message doesn't include the name
3538 of the failing file when there is a problem reading it.
3540 42. Allow both -bf and -bF in the same test run.
3542 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3544 44. The "Exiscan patch" is now merged into the mainline Exim source.
3546 45. Sometimes the final signoff response after QUIT could fail to get
3547 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3548 < 0 before doing a fflush(). This bug looks as though it goes back to the
3549 introduction of TLS in release 3.20, but "sometimes" must have been rare
3550 because the tests only now provoked it.
3552 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3553 (this can affect the format of dates).
3555 47. exim_tidydb, when checking for the continued existence of a message for
3556 which it has found a message-specific retry record, was not finding
3557 messages that were in split spool directories. Consequently, it was
3558 deleting retry records that should have stayed in existence.
3560 48. Steve fixed some bugs in eximstats.
3562 49. The SPA authentication driver was not abandoning authentication and moving
3563 on to the next authenticator when an expansion was forced to fail,
3564 contradicting the general specification for all authenticators. Instead it
3565 was generating a temporary error. It now behaves as specified.
3567 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3568 (the order specifies the preference for clients). The order is now AES256,
3569 AES128, 3DES, ARCFOUR128.
3571 51. Small patch to Sieve code - explicitly set From: when generating an
3574 52. Exim crashed if a remote delivery caused a very long error message to be
3575 recorded - for instance if somebody sent an entire SpamAssassin report back
3576 as a large number of 550 error lines. This bug was coincidentally fixed by
3577 increasing the size of one of Exim's internal buffers (big_buffer) that
3578 happened as part of the Exiscan merge. However, to be on the safe side, I
3579 have made the code more robust (and fixed the comments that describe what
3582 53. Now that there can be additional text after "Completed" in log lines (if
3583 the queue_time_overall log selector is set), a one-byte patch to exigrep
3584 was needed to allow it to recognize "Completed" as not the last thing in
3587 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3588 patch that reportedly fixes this has been added. I am not expert enough to
3589 create a test for it. This is what the patch creator wrote:
3591 "I found a little strange behaviour of ldap code when working with
3592 Windows 2003 AD Domain, where users was placed in more than one
3593 Organization Units. When I tried to give exim partial DN, the exit code
3594 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3595 But simultaneously result of request was absolutely normal ldap result,
3596 so I produce this patch..."
3598 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3599 REFERENCE, so I have modified the code to exclude the patch when that macro
3602 55. Some experimental protocols are using DNS PTR records for new purposes. The
3603 keys for these records are domain names, not reversed IP addresses. The
3604 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3605 leaves it alone. Component reversal etc. now happens only for IP addresses.
3608 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3610 57. Double the size of the debug message buffer (to 2048) so that more of very
3611 long debug lines gets shown.
3613 58. The exicyclog utility now does better if the number of log files to keep
3614 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3616 59. Two changes related to the smtp_active_hostname option:
3618 (1) $smtp_active_hostname is now available as a variable.
3619 (2) The default for smtp_banner uses $smtp_active_hostname instead
3620 of $primary_hostname.
3622 60. The host_aton() function is supposed to be passed a string that is known
3623 to be a valid IP address. However, in the case of IPv6 addresses, it was
3624 not checking this. This is a hostage to fortune. Exim now panics and dies
3625 if the condition is not met. A case was found where this could be provoked
3626 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3627 components; fortuitously, this particular loophole had already been fixed
3628 by change 4.50/55 above.
3630 If there are any other similar loopholes, the new check in host_aton()
3631 itself should stop them being exploited. The report I received stated that
3632 data on the command line could provoke the exploit when Exim was running as
3633 exim, but did not say which command line option was involved. All I could
3634 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3635 running as the user.
3638 61. There was a buffer overflow vulnerability in the SPA authentication code
3639 (which came originally from the Samba project). I have added a test to the
3640 spa_base64_to_bits() function which I hope fixes it.
3643 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3644 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3646 63. The daemon start-up calls getloadavg() while still root for those OS that
3647 need the first call to be done as root, but it missed one case: when
3648 deliver_queue_load_max is set with deliver_drop_privilege. This is
3649 necessary for the benefit of the queue runner, because there is no re-exec
3650 when deliver_drop_privilege is set.
3652 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3653 This has been fixed.
3655 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3656 was in use, was not putting the data itself into the right store pool;
3657 consequently, it could be overwritten for a subsequent message in the same
3658 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3661 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3663 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3664 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3665 the function were treating the return as a boolean value, which happened to
3666 work because 0=false and not-0=true, but is not correct code.
3668 68. The host_aton() function was not handling scoped IPv6 addresses (those
3669 with, for example, "%eth0" on the end) correctly.
3671 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3672 negated items (that is, ~something) in unsigned ints. Some compilers
3673 apparently mutter when there is no cast.
3675 70. If an address verification called from an ACL failed, and did not produce a
3676 user-specific message (i.e. there was only a "system" message), nothing was
3677 put in $acl_verify_message. In this situation, it now puts the system
3680 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3681 session; change 4.31/43 added the unwanted input to the log line - except
3682 that it did not do this in the start of session case. It now does.
3684 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3685 This is wrong and can cause the other end to generate a synchronization
3686 error if it is another Exim or anything else that does the synchronization
3687 check. A QUIT command is no longer sent after a timeout.
3689 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3690 during host lookups.
3692 74. The defer_ok option of callout verification was not working if it was used
3693 when verifying addresses in header lines, that is, for this case:
3695 verify = header_sender/callout=defer_ok
3697 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3698 those file descriptors could be used for SMTP connections. If anything
3699 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3700 could be sent to the SMTP client, causing chaos. The daemon now opens
3701 stdin, stdout, and stderr to /dev/null when it puts itself into the
3704 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3705 log by default. The user can override this with suitable Perl magic.
3707 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3708 the log message when discard triggers, was not working for the DATA ACL or
3709 for the non-SMTP ACL.
3711 78. Error message wording change in sieve.c.
3713 79. If smtp_accept_max_per_host was set, the number of connections could be
3714 restricted to fewer than expected, because the daemon was trying to set up
3715 a new connection before checking whether the processes handling previous
3716 connections had finished. The check for completed processes is now done
3717 earlier. On busy systems, this bug wouldn't be noticed because something
3718 else would have woken the daemon, and it would have reaped the completed
3721 80. If a message was submitted locally by a user whose login name contained one
3722 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3723 It caused a spool format error. I have fixed the spool reading code. A
3724 related problem was that the "from" clause in the Received: line became
3725 illegal because of the space(s). It is now covered by ${quote_local_part.
3727 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3730 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3732 83. Patch from Sieve maintainer to fix unterminated string problem in
3733 "vacation" handling.
3735 84. Some minor changes to the Linux configuration files to help with other
3736 OS variants using glibc.
3738 85. One more patch for Sieve to update vacation handling to latest spec.
3741 ----------------------------------------------------
3742 See the note above about the 4.44 and 4.50 releases.
3743 ----------------------------------------------------
3749 1. Change 4.43/35 introduced a bug that caused file counts to be
3750 incorrectly computed when quota_filecount was set in an appendfile
3753 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3754 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3757 3. Give more explanation in the error message when the command for a transport
3758 filter fails to execute.
3760 4. There are several places where Exim runs a non-Exim command in a
3761 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3762 was being done only for the command run by the queryprogram router. It is
3763 now done for all such subprocesses. The other cases are: ${run, transport
3764 filters, and the commands run by the lmtp and pipe transports.
3766 5. Some older OS have a limit of 256 on the maximum number of file
3767 descriptors. Exim was using setrlimit() to set 1000 as a large value
3768 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3769 systems. I've change it so that if it can't get 1000, it tries for 256.
3771 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3772 was an oversight, and furthermore, ever since the addition of extra
3773 controls (e.g. 4.43/32), the checks on when to allow different forms of
3774 "control" were broken. There should now be diagnostics for all cases when a
3775 control that does not make sense is encountered.
3777 7. $recipients is now available in the predata ACL (oversight).
3779 8. Tidy the search cache before the fork to do a delivery from a message
3780 received from the command line. Otherwise the child will trigger a lookup
3781 failure and thereby defer the delivery if it tries to use (for example) a
3782 cached ldap connection that the parent has called unbind on.
3784 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3785 of $address_data from the recipient verification was clobbered by the
3786 sender verification.
3788 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3789 was its contents. (It was OK if the option was not defined at all.)
3791 11. A "Completed" log line is now written for messages that are removed from
3792 the spool by the -Mrm option.
3794 12. $host_address is now set to the target address during the checking of
3795 ignore_target_hosts.
3797 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3798 being passed; this would have caused $sender_host_name to have been used if
3799 matching the list had actually called for a host name (not very likely,
3800 since this list is usually IP addresses). A host name is now passed as
3803 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3804 code that specifies a non-restarting handler (typically sigaction() in
3805 modern systems) in an attempt to fix a rare and obscure crash bug.
3807 15. Narrowed the window for a race in the daemon that could cause it to ignore
3808 SIGCHLD signals. This is not a major problem, because they are used only to
3809 wake it up if nothing else does.
3811 16. A malformed maildirsize file could cause Exim to calculate negative values
3812 for the mailbox size or file count. Odd effects could occur as a result.
3813 The maildirsize information is now recalculated if the size or filecount
3816 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3817 support for a long time. Removed HAVE_SYS_VFS_H.
3819 18. Updated exipick to current release from John Jetmore.
3821 19. Allow an empty sender to be matched against a lookup in an address list.
3822 Previously the only cases considered were a regular expression, or an
3825 20. Exim went into a mad DNS lookup loop when doing a callout where the
3826 host was specified on the transport, if the DNS lookup yielded more than
3829 21. The RFC2047 encoding function was originally intended for short strings
3830 such as real names; it was not keeping to the 75-character limit for
3831 encoded words that the RFC imposes. It now respects the limit, and
3832 generates multiple encoded words if necessary. To be on the safe side, I
3833 have increased the buffer size for the ${rfc2047: expansion operator from
3836 22. Failure to deliver a bounce message always caused it to be frozen, even if
3837 there was an errors_to setting on the router. The errors_to setting is now
3840 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3841 canonical form (fully expanded) before being placed in
3842 $sender_host_address.
3844 24. Updated eximstats to version 1.33
3846 25. Include certificate and key file names in error message when GnuTLS fails
3847 to set them up, because the GnuTLS error message doesn't include the name
3848 of the failing file when there is a problem reading it.
3850 26. Expand error message when OpenSSL has problems setting up cert/key files.
3853 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3854 (this can affect the format of dates).
3856 28. exim_tidydb, when checking for the continued existence of a message for
3857 which it has found a message-specific retry record, was not finding
3858 messages that were in split spool directories. Consequently, it was
3859 deleting retry records that should have stayed in existence.
3861 29. eximstats updated to version 1.35
3862 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3863 1.35 - bugfix such that pie charts by volume are generated correctly
3865 30. The SPA authentication driver was not abandoning authentication and moving
3866 on to the next authenticator when an expansion was forced to fail,
3867 contradicting the general specification for all authenticators. Instead it
3868 was generating a temporary error. It now behaves as specified.
3870 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3871 (the order specifies the preference for clients). The order is now AES256,
3872 AES128, 3DES, ARCFOUR128.
3874 31. Small patch to Sieve code - explicitly set From: when generating an
3877 32. Exim crashed if a remote delivery caused a very long error message to be
3878 recorded - for instance if somebody sent an entire SpamAssassin report back
3879 as a large number of 550 error lines. This bug was coincidentally fixed by
3880 increasing the size of one of Exim's internal buffers (big_buffer) that
3881 happened as part of the Exiscan merge. However, to be on the safe side, I
3882 have made the code more robust (and fixed the comments that describe what
3885 33. Some experimental protocols are using DNS PTR records for new purposes. The
3886 keys for these records are domain names, not reversed IP addresses. The
3887 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3888 leaves it alone. Component reversal etc. now happens only for IP addresses.
3891 34. The host_aton() function is supposed to be passed a string that is known
3892 to be a valid IP address. However, in the case of IPv6 addresses, it was
3893 not checking this. This is a hostage to fortune. Exim now panics and dies
3894 if the condition is not met. A case was found where this could be provoked
3895 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3896 components; fortuitously, this particular loophole had already been fixed
3897 by change 4.50/55 or 4.44/33 above.
3899 If there are any other similar loopholes, the new check in host_aton()
3900 itself should stop them being exploited. The report I received stated that
3901 data on the command line could provoke the exploit when Exim was running as
3902 exim, but did not say which command line option was involved. All I could
3903 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3904 running as the user.
3907 35. There was a buffer overflow vulnerability in the SPA authentication code
3908 (which came originally from the Samba project). I have added a test to the
3909 spa_base64_to_bits() function which I hope fixes it.
3912 36. The daemon start-up calls getloadavg() while still root for those OS that
3913 need the first call to be done as root, but it missed one case: when
3914 deliver_queue_load_max is set with deliver_drop_privilege. This is
3915 necessary for the benefit of the queue runner, because there is no re-exec
3916 when deliver_drop_privilege is set.
3918 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3919 was in use, was not putting the data itself into the right store pool;
3920 consequently, it could be overwritten for a subsequent message in the same
3921 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3924 38. Sometimes the final signoff response after QUIT could fail to get
3925 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3926 < 0 before doing a fflush(). This bug looks as though it goes back to the
3927 introduction of TLS in release 3.20, but "sometimes" must have been rare
3928 because the tests only now provoked it.
3934 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3935 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3936 Now it returns an integer. A number of calls were still expecting a T/F
3937 return. Fortuitously, in all cases, the tests worked in OK situations,
3938 which is the norm. However, things would have gone wrong on any write
3939 failures on the smtp file descriptor. This function is used when sending
3940 messages over SMTP and also when doing verify callouts.
3942 2. When Exim is called to do synchronous delivery of a locally submitted
3943 message (the -odf or -odi options), it no longer closes stderr before doing
3946 3. Implemented the mua_wrapper option.
3948 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3950 5. Implemented the functions header_remove(), header_testname(),
3951 header_add_at_position(), and receive_remove_recipient(), and exported them
3954 6. If an ACL "warn" statement specified the addition of headers, Exim already
3955 inserted X-ACL-Warn: at the start if there was no header name. However, it
3956 was not making this test for the second and subsequent header lines if
3957 there were newlines in the string. This meant that an invalid header could
3958 be inserted if Exim was badly configured.
3960 7. Allow an ACL "warn" statement to add header lines at the start or after all
3961 the Received: headers, as well as at the end.
3963 8. Added the rcpt_4xx retry error code.
3965 9. Added postmaster_mailfrom=xxx to callout verification option.
3967 10. Added mailfrom=xxxx to the callout verification option, for verify=
3970 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3971 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3973 12. Inserted some casts to stop certain compilers warning when using pointer
3974 differences as field lengths or precisions in printf-type calls (mostly
3975 affecting debugging statements).
3977 13. Added optional readline() support for -be (dynamically loaded).
3979 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3980 same clock tick as a message's arrival, so that its received time was the
3981 same as the "first fail" time on the retry record, and that message
3982 remained on the queue past the ultimate address timeout, every queue runner
3983 would try a delivery (because it was past the ultimate address timeout) but
3984 after another failure, the ultimate address timeout, which should have then
3985 bounced the address, did not kick in. This was a "< instead of <=" error;
3986 in most cases the first failure would have been in the next clock tick
3987 after the received time, and all would be well.
3989 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3990 being recognized when the domain list was tested by the match_domain
3991 condition in an expansion string.
3993 16. Added the ${str2b64: operator.
3995 17. Exim was always calling setrlimit() to set a large limit for the number of
3996 processes, without checking whether the existing limit was already
3997 adequate. (It did check for the limit on file descriptors.) Furthermore,
3998 errors from getrlimit() and setrlimit() were being ignored. Now they are
3999 logged to the main and panic logs, but Exim does carry on, to try to do its
4000 job under whatever limits there are.
4002 18. Imported PCRE 5.0.
4004 19. Trivial typo in log message " temporarily refused connection" (the leading
4007 20. If the log selector return_path_on_delivery was set and an address was
4008 redirected to /dev/null, the delivery process crashed because it assumed
4009 that a return path would always be set for a "successful" delivery. In this
4010 case, the whole delivery is bypassed as an optimization, and therefore no
4013 21. Internal re-arrangement: the function for sending a challenge and reading
4014 a response while authentication was assuming a zero-terminated challenge
4015 string. It's now changed to take a pointer and a length, to allow for
4016 binary data in such strings.
4018 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4020 23. Exim was not respecting finduser_retries when seeking the login of the
4021 uid under which it was called; it was always trying 10 times. (The default
4022 setting of finduser_retries is zero.) Also, it was sleeping after the final
4023 failure, which is pointless.
4025 24. Implemented tls_on_connect_ports.
4027 25. Implemented acl_smtp_predata.
4029 26. If the domain in control=submission is set empty, Exim assumes that the
4030 authenticated id is a complete email address when it generates From: or
4031 Sender: header lines.
4033 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4034 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4035 chown and chgrp in /bin and hostname in /usr/bin.
4037 28. Exim was keeping the "process log" file open after each use, just as it
4038 does for the main log. This opens the possibility of it remaining open for
4039 long periods when the USR1 signal hits a daemon. Occasional processlog
4040 errors were reported, that could have been caused by this. Anyway, it seems
4041 much more sensible not to leave this file open at all, so that is what now
4044 29. The long-running daemon process does not normally write to the log once it
4045 has entered its main loop, and it closes the log before doing so. This is
4046 so that log files can straightforwardly be renamed and moved. However,
4047 there are a couple of unusual error situations where the daemon does write
4048 log entries, and I had neglected to close the log afterwards.
4050 30. The text of an SMTP error response that was received during a remote
4051 delivery was being truncated at 512 bytes. This is too short for some of
4052 the long messages that one sometimes sees. I've increased the limit to
4055 31. It is now possible to make retry rules that apply only when a message has a
4056 specific sender, in particular, an empty sender.
4058 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4059 it possible to be selective about when SMTP synchronization is enforced.
4061 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4063 32. Implemented hosts_connection_nolog.
4065 33. Added an ACL for QUIT.
4067 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4070 35. Added mailbox_size and mailbox_filecount to appendfile.
4072 36. Added control = no_multiline_responses to ACLs.
4074 37. There was a bug in the logic of the code that waits for the clock to tick
4075 in the case where the clock went backwards by a substantial amount such
4076 that the microsecond fraction of "now" was more than the microsecond
4077 fraction of "then" (but the whole seconds number was less).
4079 38. Added support for the libradius Radius client library this is found on
4080 FreeBSD (previously only the radiusclient library was supported).
4086 1. When certain lookups returned multiple values in the form name=value, the
4087 quoting of the values was not always being done properly. Specifically:
4088 (a) If the value started with a double quote, but contained no whitespace,
4090 (b) If the value contained whitespace other than a space character (i.e.
4091 tabs or newlines or carriage returns) it was not quoted.
4092 This fix has been applied to the mysql and pgsql lookups by writing a
4093 separate quoting function and calling it from the lookup code. The fix
4094 should probably also be applied to nisplus, ibase and oracle lookups, but
4095 since I cannot test any of those, I have not disturbed their existing code.
4097 2. A hit in the callout cache for a specific address caused a log line with no
4098 reason for rejecting RCPT. Now it says "Previous (cached) callout
4099 verification failure".
4101 3. There was an off-by-one bug in the queryprogram router. An over-long
4102 return line was truncated at 256 instead of 255 characters, thereby
4103 overflowing its buffer with the terminating zero. As well as fixing this, I
4104 have increased the buffer size to 1024 (and made a note to document this).
4106 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4107 when Exim is waiting for an SMTP response from a remote server, Exim
4108 restarts its select() call on the socket, thereby resetting its timeout.
4109 This is not a problem when such interrupts are rare. Somebody set up a cron
4110 job to run exiwhat every 2 minutes, which is less than the normal select()
4111 timeout (5 or 10 minutes). This meant that the select() timeout never
4112 kicked in because it was always reset. I have fixed this by comparing the
4113 time when an interrupt arrives with the time at the start of the first call
4114 to select(). If more time than the timeout has elapsed, the interrupt is
4115 treated as a timeout.
4117 5. Some internal re-factoring in preparation for the addition of Sieve
4118 extensions (by MH). In particular, the "personal" test is moved to a
4119 separate function, and given an option for scanning Cc: and Bcc: (which is
4120 not set for Exim filters).
4122 6. When Exim created an email address using the login of the caller as the
4123 local part (e.g. when creating a From: or Sender: header line), it was not
4124 quoting the local part when it contained special characters such as @.
4126 7. Installed new OpenBSD configuration files.
4128 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4129 try to make them clearer.
4131 9. Callout options, other than the timeout value, were being ignored when
4132 verifying sender addresses in header lines. For example, when using
4134 verify = header_sender/callout=no_cache
4136 the cache was (incorrectly) being used.
4138 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4139 only the Cygwin environment.
4141 11. When return_path_on_delivery was set as a log selector, if different remote
4142 addresses in the same message used different return paths and parallel
4143 remote delivery occurred, the wrong values would sometimes be logged.
4144 (Whenever a remote delivery process finished, the return path value from
4145 the most recently started remote delivery process was logged.)
4147 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4148 lines when AUTH and/or TLS are in use. This is the "received protocol"
4149 field. Exim used to use "asmtp" for authenticated SMTP, without any
4150 indication (in the protocol name) for TLS use. Now it follows the RFC and
4151 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4152 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4153 names appear in log lines as well as in Received: header lines.
4155 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4156 extensions, and comparison tests, and to fix some bugs.
4158 14. Changes to the "personal" filter test:
4160 (1) The test was buggy in that it was just doing the equivalent of
4161 "contains" tests on header lines. For example, if a user's address was
4162 anne@some.where, the "personal" test would incorrectly be true for
4164 To: susanne@some.where
4166 This test is now done by extracting each address from the header in turn,
4167 and checking the entire address. Other tests that are part of "personal"
4168 are now done using regular expressions (for example, to check local parts
4169 of addresses in From: header lines).
4171 (2) The list of non-personal local parts in From: addresses has been
4172 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4173 taken from the Sieve specification recommendations.
4175 (3) If the message contains any header line starting with "List-" it is
4176 treated as non-personal.
4178 (4) The test for "circular" in the Subject: header line has been removed
4179 because it now seems ill-conceived.
4181 15. Minor typos in src/EDITME comments corrected.
4183 16. Installed latest exipick from John Jetmore.
4185 17. If headers_add on a router specified a text string that was too long for
4186 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4187 of string_sprintf() is now avoided.
4189 18. $message_body_size was not set (it was always zero) when running the DATA
4190 ACL and the local_scan() function.
4192 19. For the "mail" command in an Exim filter, no default was being set for
4193 the once_repeat time, causing a random time value to be used if "once" was
4194 specified. (If the value happened to be <= 0, no repeat happened.) The
4195 default is now 0s, meaning "never repeat". The "vacation" command was OK
4196 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4197 (I found it when inspecting the code).
4199 20. There is now an overall timeout for performing a callout verification. It
4200 defaults to 4 times the callout timeout, which applies to individual SMTP
4201 commands during the callout. The overall timeout applies when there is more
4202 than one host that can be tried. The timeout is checked before trying the
4203 next host. This prevents very long delays if there are a large number of
4204 hosts and all are timing out (e.g. when the network connections are timing
4205 out). The value of the overall timeout can be changed by specifying an
4206 additional sub-option for "callout", called "maxwait". For example:
4208 verify = sender/callout=5s,maxwait=20s
4210 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4211 to the end before writing, but this should make it even safer).
4213 22. Exim was forgetting that it had advertised PIPELINING for the second and
4214 subsequent messages on an SMTP connection. It was also not resetting its
4215 memory on STARTTLS and an internal HELO.
4217 23. When Exim logs an SMTP synchronization error within a session, it now
4218 records whether PIPELINING has been advertised or not.
4220 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4221 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4222 rather than long int.
4224 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4226 26. Added the never_mail option to autoreply.
4232 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4233 crash if the getsockname() call failed; this can happen if a connection is
4234 closed very soon after it is established. The problem was simply in the
4235 order in which certain operations were done, causing Exim to try to write
4236 to the SMTP stream before it had set up the file descriptor. The bug has
4237 been fixed by making things happen in the correct order.
4243 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4244 before the connection was closed, thus losing the rejection response.
4246 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4247 some early Solaris releases, but causes trouble in current releases where
4248 socklen_t is defined.
4250 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4253 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4256 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4257 that releasing the top of it at the end releases what was used for sub-
4258 expansions (unless the block got too big). However, discard this block if
4259 the first thing is a variable or header, so that we can use its block when
4260 it is dynamic (useful for very large $message_headers, for example).
4262 6. Lookups now cache *every* query, not just the most recent. A new, separate
4263 store pool is used for this. It can be recovered when all lookup caches are
4264 flushed. Lookups now release memory at the end of their result strings.
4265 This has involved some general refactoring of the lookup sources.
4267 7. Some code has been added to the store_xxx() functions to reduce the amount
4268 of flapping under certain conditions.
4270 8. log_incoming_interface used to affect only the <= reception log lines. Now
4271 it causes the local interface and port to be added to several more SMTP log
4272 lines, for example "SMTP connection from", and rejection lines.
4274 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4276 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4278 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4279 could be overwritten at the end of the current message (or the start of a
4280 new message if it was set in a HELO ACL). The value is now preserved for
4281 the duration of the SMTP connection.
4283 12. If a transport had a headers_rewrite setting, and a matching header line
4284 contained an unqualified address, that address was qualified, even if it
4285 did not match any rewriting rules. The underlying bug was that the values
4286 of the flags that permit the existence of unqualified sender and recipient
4287 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4288 non-local messages, and by -bnq for local messages) were not being
4289 preserved with the message after it was received.
4291 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4292 "next input=" as part of the text comprising the host identity instead of
4293 the correct text. The code was using the same buffer for two different
4294 strings. However, depending on which order the printing function evaluated
4295 its arguments, the bug did not always show up. Under Linux, for example, my
4296 test suite worked just fine.
4298 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4299 doesn't work with some older versions of Perl. It has been changed to "my",
4300 which in any case is probably the better facility to use.
4302 15. A really picky compiler found some instances of statements for creating
4303 error messages that either had too many or two few arguments for the format
4306 16. The size of the buffer for calls to the DNS resolver has been increased
4307 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4308 for addresses that have a lot of PTR records. This alleviates a problem; it
4309 does not fully solve it.
4311 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4312 buffer now truncates the list and logs the incident, which is the same
4313 action as happens when Exim is looking up a host name and its aliases.
4314 Previously in this situation something unpredictable would happen;
4315 sometimes it was "internal error: store_reset failed".
4317 18. If a server dropped the connection unexpectedly when an Exim client was
4318 using GnuTLS and trying to read a response, the client delivery process
4319 crashed while trying to generate an error log message.
4321 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4322 string, for example:
4324 warn message = H1: something\nH2: something
4326 the text was added as a single header line from Exim's point of view
4327 though it ended up OK in the delivered message. However, searching for the
4328 second and subsequent header lines using $h_h2: did not work. This has been
4329 fixed. Similarly, if a system filter added multiple headers in this way,
4330 the routers could not see them.
4332 20. Expanded the error message when iplsearch is called with an invalid key to
4333 suggest using net-iplsearch in a host list.
4335 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4336 ACLs are no longer actually imposed (and a message to that effect is
4339 22. If a "gecos" field in a passwd entry contained escaped characters, in
4340 particular, if it contained a \" sequence, Exim got it wrong when building
4341 a From: or a Sender: header from that name. A second bug also caused
4342 incorrect handling when an unquoted " was present following a character
4343 that needed quoting.
4345 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4346 was not being matched caselessly.
4348 24. Arranged for all hyphens in the exim.8 source to be escaped with
4351 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4352 properly. Recipient callout cache records were still being keyed to include
4353 the sender, even when use_sender was set false. This led to far more
4354 callouts that were necessary. The sender is no longer included in the key
4355 when use_sender is false.
4357 26. Added "control = submission" modifier to ACLs.
4359 27. Added the ${base62d: operator to decode base 62 numbers.
4361 28. dnsdb lookups can now access SRV records.
4363 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4364 the configuration file.
4366 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4367 (-v) mode. This makes the output for a verbose queue run more intelligible.
4369 31. Added a use_postmaster feature to recipient callouts.
4371 32. Added the $body_zerocount variable, containing the number of binary zero
4372 bytes in the message body.
4374 33. The time of last modification of the "new" subdirectory is now used as the
4375 "mailbox time last read" when there is a quota error for a maildir
4378 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4380 35. Added +ignore_unknown as a special item in host lists.
4382 36. Code for decoding IPv6 addresses in host lists is now included, even if
4383 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4384 address was recognized as an IP address, but was then not correctly decoded
4385 into binary, causing unexpected and incorrect effects when compared with
4392 1. Very minor rewording of debugging text in manualroute to say "list of
4393 hosts" instead of "hostlist".
4395 2. If verify=header_syntax was set, and a header line with an unqualified
4396 address (no domain) and a large number of spaces between the end of the
4397 name and the colon was received, the reception process suffered a buffer
4398 overflow, and (when I tested it) crashed. This was caused by some obsolete
4399 code that should have been removed. The fix is to remove it!
4401 3. When running in the test harness, delay a bit after writing a bounce
4402 message to get a bit more predictability in the log output.
4404 4. Added a call to search_tidyup() just before forking a reception process. In
4405 theory, someone could use a lookup in the expansion of smtp_accept_max_
4406 per_host which, without the tidyup, could leave open a database connection.
4408 5. Added the variables $recipient_data and $sender_data which get set from a
4409 lookup success in an ACL "recipients" or "senders" condition, or a router
4410 "senders" option, similar to $domain_data and $local_part_data.
4412 6. Moved the writing of debug_print from before to after the "senders" test
4415 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4416 problems for message scanning, either using a data ACL, or using
4417 local_scan() because the Received: header was not generated till after they
4418 were called (in order to set the time as the time of reception completion).
4419 I have revised the way this works. The header is now generated after the
4420 body is received, but before the ACL or local_scan() are called. After they
4421 are run, the timestamp in the header is updated.
4427 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4428 before starting a queue runner without re-exec. This happened only when
4429 deliver_drop_privilege was set or when the Exim user was set to root. The
4430 effect of the bug was that timeouts during subsequent deliveries caused
4431 crashes instead of being properly handled. The handler is now left at its
4432 default (and expected) setting.
4434 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4435 message, again when deliver_drop_privilege is set or Exim is run as root.
4436 The bug described in (1) was not present in this case, but the tidying up
4437 of the other signals was missing. I have made the two cases consistent.
4439 3. The ignore_target_hosts setting on a manualroute router was being ignored
4440 for hosts that were looked up using the /MX notation.
4442 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4445 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4446 operated on the sender address. After changing the $sender_address to <>
4447 for the sender address verify, Exim was re-instated it as the original
4448 (before rewriting) address, but remembering that it had rewritten it, so it
4449 wasn't rewriting it again. This bug also had the effect of breaking the
4450 sender address verification caching when the sender address was rewritten.
4452 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4453 This has been changed so that if the ip literal address matches
4454 ignore_target_hosts, the router declines.
4456 7. Added expansion conditions match_domain, match_address, and match_local_
4457 part (NOT match_host).
4459 8. The placeholder for the Received: header didn't have a length field set.
4461 9. Added code to Exim itself and to exim_lock to test for a specific race
4462 condition that could lead to file corruption when using MBX delivery. The
4463 issue is with the lockfile that is created in /tmp. If this file is removed
4464 after a process has opened it but before that process has acquired a lock,
4465 there is the potential for a second process to recreate the file and also
4466 acquire a lock. This could lead to two Exim processes writing to the file
4467 at the same time. The added code performs the same test as UW imapd; it
4468 checks after acquiring the lock that its file descriptor still refers to
4469 the same named file.
4471 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4472 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4475 11. Added the smtp_active_hostname option. If used, this will typically be made
4476 to depend on the incoming interface address. Because $interface_address is
4477 not set up until the daemon has forked a reception process, error responses
4478 that can happen earlier (such as "too many connections") no longer contain
4481 12. If an expansion in a condition on a "warn" statement fails because a lookup
4482 defers, the "warn" statement is abandoned, and the next ACL statement is
4483 processed. Previously this caused the whole ACL to be aborted.
4485 13. Added the iplsearch lookup type.
4487 14. Added ident_timeout as a log selector.
4489 15. Added tls_certificate_verified as a log selector.
4491 16. Added a global option tls_require_ciphers (compare the smtp transport
4492 option of the same name). This controls incoming TLS connections.
4494 17. I finally figured out how to make tls_require_ciphers do a similar thing
4495 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4496 before starting the TLS session.
4498 18. Tabs are now shown as \t in -bP output.
4500 19. If the log selector return_path_on_delivery was set, Exim crashed when
4501 bouncing a message because it had too many Received: header lines.
4503 20. If two routers both had headers_remove settings, and the first one included
4504 a superfluous trailing colon, the final name in the first list and the
4505 first name in the second list were incorrectly joined into one item (with a
4506 colon in the middle).
4512 1. Added -C and -D options to the exinext utility, mainly to make it easier
4513 to include in the automated testing, but these could be helpful when
4514 multiple configurations are in use.
4516 2. The exinext utility was not formatting the output nicely when there was
4517 an alternate port involved in the retry record key, nor when there was a
4518 message id as well (for retries that were specific to a specific message
4519 and a specific host). It was also confused by IPv6 addresses, because of
4520 the additional colons they contain. I have fixed the IPv4 problem, and
4521 patched it up to do a reasonable job for IPv6.
4523 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4524 delivery, the log line now contains "pipelined" if PIPELINING was used.
4526 4. An SMTP transport process used to panic and die if the bind() call to set
4527 an explicit outgoing interface failed. This has been changed; it is now
4528 treated in the same way as a connect() failure.
4530 5. A reference to $sender_host_name in the part of a conditional expansion
4531 that was being skipped was still causing a DNS lookup. This no longer
4534 6. The def: expansion condition was not recognizing references to header lines
4535 that used bh_ and bheader_.
4537 7. Added the _cache feature to named lists.
4539 8. The code for checking quota_filecount in the appendfile transport was
4540 allowing one more file than it should have been.
4542 9. For compatibility with Sendmail, the command line option
4550 and sets the incoming protocol and host name (for trusted callers). The
4551 host name and its colon can be omitted when only the protocol is to be set.
4552 Note the Exim already has two private options, -pd and -ps, that refer to
4553 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4554 "s", but I don't think that's a major issue.
4556 10. A number of refactoring changes to the code, none of which should affect
4559 (a) The number of logging options was getting close to filling up the
4560 32-bit word that was used as a bit map. I have split them into two classes:
4561 those that are passed in the argument to log_write(), and those that are
4562 only ever tested independently outside of that function. These are now in
4563 separate 32-bit words, so there is plenty of room for expansion again.
4564 There is no change in the user interface or the logging behaviour.
4566 (b) When building, for example, log lines, the code previously used a
4567 macro that called string_cat() twice, in order to add two strings. This is
4568 not really sufficiently general. Furthermore, there was one instance where
4569 it was actually wrong because one of the argument was used twice, and in
4570 one call a function was used. (As it happened, calling the function twice
4571 did not affect the overall behaviour.) The macro has been replaced by a
4572 function that can join an arbitrary number of extra strings onto a growing
4575 (c) The code for expansion conditions now uses a table and a binary chop
4576 instead of a serial search (which was left over from when there were very
4577 few conditions). Also, it now recognizes conditions like "pam" even when
4578 the relevant support is not compiled in: a suitably worded error message is
4579 given if an attempt is made to use such a condition.
4581 11. Added ${time_interval:xxxxx}.
4583 12. A bug was causing one of the ddress fields not to be passed back correctly
4584 from remote delivery subprocesses. The field in question was not being
4585 subsequently used, so this caused to problems in practice.
4587 13. Added new log selectors queue_time and deliver_time.
4589 14. Might have fixed a bug in maildirsizefile handling that threw up
4590 "unexpected character" debug warnings, and recalculated the data
4591 unnecessarily. In any case, I expanded the warning message to give more
4594 15. Added the message "Restricted characters in address" to the statements in
4595 the default ACL that block characters like @ and % in local parts.
4597 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4598 Three changes have been made:
4600 (a) There was a serious bug; a negative response to MAIL caused the whole
4601 recipient domain to be cached as invalid, thereby blocking all messages
4602 to all local parts at the same domain, from all senders. This bug has
4603 been fixed. The domain is no longer cached after a negative response to
4604 MAIL if the sender used is not empty.
4606 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4609 (c) A new callout option, "use_sender" has been added for people who want
4610 the modified behaviour.
4616 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4619 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4620 indeed breaks things for older releases.
4622 3. Added additional logging to the case where there is a problem reading data
4623 from a filter that is running in a subprocess using a pipe, in order to
4624 try to track down a specific problem.
4626 4. Testing facility fudge: when running in the test harness and attempting
4627 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4628 getting "No route to host". Convert this to a timeout.
4630 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4633 6. Some OS don't have socklen_t but use size_t instead. This affects the
4634 fifth argument of getsockopt() amongst other things. This is now
4635 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4636 can be set for individual OS. I have set it for SunOS5, OSF1, and
4637 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4638 some earlier ones do not.
4640 7. Change 4.30/15 was not doing the test caselessly.
4642 8. The standard form for an IPv6 address literal was being rejected by address
4643 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4644 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4645 this, as well as the form without the "IPv6" on the front (but only when
4646 address literals are enabled, of course).
4648 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4650 10. Exim crashed if a message with an empty sender address specified by -f
4651 encountered a router with an errors_to setting. This could be provoked only
4652 by a command such as
4656 where an empty string was supplied; "<>" did not hit this bug.
4658 11. Installed PCRE release 4.5.
4660 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4661 remained set. It is now erased.
4663 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4664 times from message ids (which are base 36 rather than the normal 62).
4666 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4667 were being counted as actual protocol errors, and logged if the log
4668 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4669 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4670 and DATA following a set of rejected RCPTs do not count as protocol errors.
4671 In other words, Exim assumes they were pipelined, though this may not
4672 actually be the case. Of course, in all cases the client gets an
4673 appropriate error code.
4675 15. If a lookup fails in an ACL condition, a message about the failure may
4676 be available; it is used if testing the ACL cannot continue, because most
4677 such messages specify what the cause of the deferral is. However, some
4678 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4679 that caused an old message to be retained and used if a later statement
4680 caused a defer, replacing the real cause of the deferral.
4682 16. If an IP address had so many PTR records that the DNS lookup buffer
4683 was not large enough to hold them, Exim could crash while trying to process
4684 the truncated data. It now detects and logs this case.
4686 17. Further to 4.21/58, another change has been made: if (and only if) the
4687 first line of a message (the first header line) ends with CRLF, a bare LF
4688 in a subsequent header line has a space inserted after it, so as not to
4689 terminate the header.
4691 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4692 unnecessarily, used atoi() instead of strtol(), and didn't check the
4693 termination when getting file sizes from file names by regex.
4695 19. Completely re-implemented the support for maildirsize files, in the light
4696 of a number of problems with the previous contributed implementation
4697 (4.30/29). In particular:
4699 . If the quota is zero, the maildirsize file is maintained, but no quota is
4702 . If the maildir directory does not exist, it is created before any attempt
4703 to write a maildirsize file.
4705 . The quota value in the file is just a cache; if the quota is changed in
4706 the transport, the new value overrides.
4708 . A regular expression is available for excluding directories from the
4711 20. The autoreply transport checks the characters in options that define the
4712 message's headers; it allows continued headers, but it was checking with
4713 isspace() after an embedded newline instead of explicitly looking for a
4716 21. If all the "regular" hosts to which an address was routed had passed their
4717 expiry times, and had not reached their retry times, the address was
4718 bounced, even if fallback hosts were defined. Now Exim should go on to try
4721 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4722 equivalent code in the SMTP transport. Some hosts send humungous responses
4723 to HELO/EHLO, more than 1024 it seems.
4725 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4726 gives compiler warnings in some environments. I've now done it "properly",
4729 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4730 (because of problems with the built-in one) was declared to return uschar *
4731 instead of char *, causing compiler failure.
4733 25. Fixed a file descriptor leak when processing alias/forward files.
4735 26. Fixed a minor format string issue in dbfn.c.
4737 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4739 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4740 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4743 29. When a sender address is verified, it is cached, to save repeating the test
4744 when there is more than one recipient in a message. However, when the
4745 verification involves a callout, it is possible for different callout
4746 options to be set for different recipients. It is too complicated to keep
4747 track of this in the cache, so now Exim always runs a verification when a
4748 callout is required, relying on the callout cache for the optimization.
4749 The overhead is duplication of the address routing, but this should not be
4752 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4753 to be verified with callout=postmaster, and the main callout worked but the
4754 postmaster check failed, the verification correctly failed. However, if a
4755 subsequent RCPT command asked for sender verification *without* the
4756 postmaster check, incorrect caching caused this verification also to fail,
4759 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4760 it was not caching the DNS options (qualify_single, search_parents) that
4761 were used when the lookup failed. A subsequent lookup with different
4762 options therefore always gave the same answer, though there were cases
4763 where it should not have. (Example: a "domains = !$mx_any" option on a
4764 dnslookup router: the "domains" option is always processed without any
4765 widening, but the router might have qualify_single set.) Now Exim uses the
4766 cached value only when the same options are set.
4768 32. Added John Jetmore's "exipick" utility to the distribution.
4770 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4771 than a timeout (e.g. a certificate is required, and is not provided), an
4772 Exim server now closes the connection immediately. Previously it waited for
4773 the client to close - but if the client is SSL, it seems that they each
4774 wait for each other, leading to a delay before one of them times out.
4776 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4777 maintained 0.8.x compatibility because I don't think many are using it, and
4778 it is clearly obsolete.
4780 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4783 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4784 client certificate was expired. A simple patch fixes this, though I don't
4785 understand the full logic of why the verify callback is called multiple
4788 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4789 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4790 which causes problems with some clients (such as the Certicom SSL Plus
4791 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4792 disables the coutermeasure allowing Eudora to connect."
4794 38. Exim was not checking that a write() to a log file succeeded. This could
4795 lead to Bad Things if a log got too big, in particular if it hit a file
4796 size limit. Exim now panics and dies if it cannot write to a log file, just
4797 as it does if it cannot open a log file.
4799 39. Modified OS/Makefile-Linux so that it now contains
4801 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4803 The two -D definitions ensure that Exim is compiled with large file
4804 support, which makes it possible to handle log files that are bigger than
4807 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4808 instance) a domain was checked against a named list that involved a lookup,
4809 causing $domain_data to be set, then another domain was checked against the
4810 same list, then the first domain was re-checked, the value of $domain_data
4811 after the final check could be wrong. In particular, if the second check
4812 failed, it could be set empty. This bug probably also applied to
4815 41. The strip_trailing_dot option was not being applied to the address given
4816 with the -f command-line option.
4818 42. The code for reading a message's header from the spool was incrementing
4819 $received_count, but never initializing it. This meant that the value was
4820 incorrect (doubled) while delivering a message in the same process in which
4821 it was received. In the most common configuration of Exim, this never
4822 happens - a fresh exec is done - but it can happen when
4823 deliver_drop_privilege is set.
4825 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4826 it now includes up to 150 characters of the unexpected data in the log
4829 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4830 and building data strings. The size of both of these buffers was 10 000
4831 bytes - far larger than anybody would *ever* want, thought I. Needless to
4832 say, somebody hit the limit. I have increased the maximum line length to
4833 20 000 and the maximum data length of concatenated lines to 100 000. I have
4834 also fixed two bugs, because there was no checking on these buffers. Tsk,
4835 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4836 buffer is too small.
4838 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4839 lsearch lookups. Now it does.
4841 46. When parsing a route_list item in a manualroute router, a fixed-length
4842 buffer was used for the list of hosts. I made this 1024 bytes long,
4843 thinking that nobody would ever have a list of hosts that long. Wrong.
4844 Somebody had a whole pile of complicated expansion conditions, and the
4845 string was silently truncated, leading to an expansion error. It turns out
4846 that it is easier to change to an unlimited length (owing to other changes
4847 that have happened since this code was originally written) than to build
4848 structure for giving a limitation error. The length of the item that
4849 expands into the list of hosts is now unlimited.
4851 47. The lsearch lookup could not handle data where the length of text line was
4852 more than 4095 characters. Such lines were truncated, leading to shortened
4853 data being returned. It should now handle lines of any length.
4855 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4856 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4859 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4860 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4861 no longer happens in comment lines. A list of replacements is now placed
4862 at the head of all of the source files, except those whose only change is
4863 to replace PERL_COMMAND in the very first #! line.
4865 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4866 messages on the queue, with a bottom-up merge sort, using code contributed
4867 by Michael Haardt. This should make operations like -bp somewhat faster on
4868 large queues. It won't affect queue runners, except when queue_run_in_order
4871 51. Installed eximstats 1.31 in the distribution.
4873 52. Added support for SRV lookups to the dnslookup router.
4875 53. If an ACL referred to $message_body or $message_body_end, the value was not
4876 reset for any messages that followed in the same SMTP session.
4878 54. The store-handling optimization for building very long strings was not
4879 differentiating between the different store pools. I don't think this
4880 actually made any difference in practice, but I've tidied it.
4882 55. While running the routers to verify a sender address, $sender_address
4883 was still set to the sender address. This is wrong, because when routing to
4884 send a bounce to the sender, it would be empty. Therefore, I have changed
4885 it so that, while verifying a sender address, $sender_address is set to <>.
4886 (There is no change to what happens when verifying a recipient address.)
4888 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4889 target A or AAAA records (if not already returned) without resetting the
4890 qualify_single or search_parents options of the DNS resolver. These are
4891 inappropriate in this case because the targets of MX and SRV records must
4892 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4893 target that, when qualified, matched something in the local domain. These
4894 two options are now turned off when doing these lookups.
4896 57. It seems that at least some releases of Reiserfs (which does not have the
4897 concept of a fixed number of inodes) returns zero and not -1 for the
4898 number of available inodes. This interacted badly with check_spool_inodes,
4899 which assumed that -1 was the "no such thing" setting. What I have done is
4900 to check that the total number of inodes is greater than zero before doing
4901 the test of how many are available.
4903 58. When a "warn" ACL statement has a log_message modifier, the message is
4904 remembered, and not repeated. This is to avoid a lot of repetition when a
4905 message has many recipients that cause the same warning to be written.
4906 Howewer, Exim was preserving the list of already written lines for an
4907 entire SMTP session, which doesn't seem right. The memory is now reset if a
4908 new message is started.
4910 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4911 debugging output unless log_rewrite was also set.
4913 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4914 of (int)(handle) when we know that handle contains (void *)(-1).
4916 61. The Exim daemon panic-logs an error return when it closes the incoming
4917 connection. However "connection reset by peer" seems to be common, and
4918 isn't really an error worthy of noting specially, so that particular error
4921 62. When Exim is trying to find all the local interfaces, it used to panic and
4922 die if the ioctl to get the interface flags failed. However, it seems that
4923 on at least one OS (Solaris 9) it is possible to have an interface that is
4924 included in the list of interfaces, but for which you get a failure error
4925 for this call. This happens when the interface is not "plumbed" into a
4926 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4927 failure of the "get flags" call assumes that the interface is down.
4929 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4930 makes life easier for people who are doing arithmetic on fields extracted
4931 from dates, where you often get leading zeros that should not be
4932 interpreted as octal.
4934 64. Added qualify_domain to the redirect router, to override the global
4937 65. If a pathologically long header line contained very many addresses (the
4938 report of this problem mentioned 10 000) and each of them was rewritten,
4939 Exim could use up a very large amount of memory. (It kept on making new
4940 copies of the header line as it rewrote, and never released the old ones.)
4941 At the expense of a bit more processing, the header rewriting function has
4942 been changed so that it no longer eats memory in this way.
4944 66. The generation of the Received: header has been moved from the time that a
4945 message starts to be received, to the time that it finishes. The timestamp
4946 in the Received: header should now be very close to that of the <= log
4947 line. There are two side-effects of this change:
4949 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4950 logged header lines no longer include the local Received: line, because
4951 it has not yet been created. The same applies to a copy of the message
4952 that is returned to a non-SMTP sender when a message is rejected.
4954 (b) When a filter file is tested using -bf, no additional Received: header
4955 is added to the test message. After some thought, I decided that this
4958 This change does not affect the value of $received_for. It is still set
4959 after address rewriting, but before local_scan() is called.
4961 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4963 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4964 gave an unhelpful panic error message, and a defer error. I have managed to
4965 change this behaviour so that it now rejects any supplied certificate,
4966 which seems right, as the list of acceptable certificates is empty.
4968 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4969 gave an unhelpful defer error. I have not managed to make this reject any
4970 supplied certificates, but the error message it gives is "no certificate
4971 supplied", which is not helpful.
4973 70. exigrep's output now also includes lines that are not associated with any
4974 message, but which match the given pattern. Implemented by a patch from
4975 Martin Sluka, which also tidied up the Perl a bit.
4977 71. Recipient callout verification, like sender verification, was using <> in
4978 the MAIL FROM command. This isn't really the right thing, since the actual
4979 sender may affect whether the remote host accepts the recipient or not. I
4980 have changed it to use the actual sender in the callout; this means that
4981 the cache record is now keyed on a recipient/sender pair, not just the
4982 recipient address. There doesn't seem to be a real danger of callout loops,
4983 since a callout by the remote host to check the sender would use <>.
4984 [SEE ABOVE: changed after hitting problems.]
4986 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4987 temporary errors. However, in the case of such a code being given after
4988 the end of a data transmission (i.e. after ".") Exim was failing to write
4989 a retry record for the message. (Yes, there was some broken host that was
4990 actually sending 8xx at this point.)
4992 73. An unknown lookup type in a host list could cause Exim to panic-die when
4993 the list was checked. (An example that provoked this was putting <; in the
4994 middle of a list instead of at the start.) If this happened during a DATA
4995 ACL check, a -D file could be left lying around. This kind of configuration
4996 error no longer causes Exim to die; instead it causes a defer errror. The
4997 incident is still logged to the main and panic logs.
4999 74. Buglet left over from Exim 3 conversion. The message "too many messages
5000 in one connection" was written to the rejectlog but not the mainlog, except
5001 when address rewriting (yes!) was being logged.
5003 75. Added write_rejectlog option.
5005 76. When a system filter was run not as root (that is, when system_filter_user
5006 was set), the values of the $n variables were not being returned to the
5007 main process; thus, they were not subsequently available in the $sn
5010 77. Added +return_path_on_delivery log selector.
5012 78. A connection timeout was being treated differently from recipients deferred
5013 when testing hosts_max_try with a message that was older than the host's
5014 retry timeout. (The host should not be counted, thus allowing all hosts to
5015 be tried at least once before bouncing.) This may have been the cause of an
5016 occasionally reported bug whereby a message would remain on the queue
5017 longer than the retry timeout, but would be bounced if a delivery was
5018 forced. I say "may" because I never totally pinned down the problem;
5019 setting up timeout/retry tests is difficult. See also the next item.
5021 79. The ultimate address timeout was not being applied to errors that involved
5022 a combination of host plus message (for example, a timeout on a MAIL
5023 command). When an address resolved to a number of possible hosts, and they
5024 were not all tried for each delivery (e.g. because of hosts_max_try), a
5025 message could remain on the queue longer than the retry timeout.
5027 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5030 81. Fixed an obscure SMTP outgoing bug which required at least the following
5031 conditions: (a) there was another message waiting for the same server;
5032 (b) the server returned 5xx to all RCPT commands in the first message so
5033 that the message was not completed; (c) the server dropped the connection
5034 or gave a negative response to the RSET that Exim sends to abort the
5035 transaction. The observed case was a dropped connection after DATA that had
5036 been sent in pipelining mode. That is, the server had advertised PIPELINING
5037 but was not implementing it correctly. The effect of the bug was incorrect
5038 behaviour, such as trying another host, and this could lead to a crash.
5044 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5045 and daemon.c were passed as pointers to ints; they should have been
5046 pointers to socklen_t variables (which are typically unsigned ints).
5048 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5051 3. Fixed a really odd bug that affected only the testing scheme; patching a
5052 certain fixed string in the binary changed the value of another string that
5053 happened to be identical to the end of the original first string.
5055 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5056 name", it returns that address as the IP address. On some operating
5057 systems (e.g. Solaris), it also passes back the IP address string as the
5058 "host name". However, on others (e.g. Linux), it passes back an empty
5059 string. Exim wasn't checking for this, and was changing the host name to an
5060 empty string, assuming it had been canonicized.
5062 5. Although rare, it is permitted to have more than one PTR record for a given
5063 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5064 all the names associated with an address, because they do in Solaris.
5065 However, it seems that they do not in Linux for data that comes from the
5066 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5067 I found this out when I moved to a new Linux workstation and tried to run
5068 the Exim test suite.
5070 To get round this problem I have changed the code so that it now does its
5071 own call to the DNS to look up PTR records when searching for a host name.
5072 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5073 addresses that are only in /etc/hosts are still found.
5075 This behaviour is, however, controlled by an option called host_lookup_
5076 order, which defaults to "bydns:byaddr". If people want to use the other
5077 order, or indeed, just use one or the other means of lookup, they can
5078 specify it in this variable.
5080 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5081 some operating systems, this comes back from gethostbyaddr() as an empty
5082 string, and this is what Exim used to test for. However, it seems that in
5083 other systems, "." is yielded. Exim now tests for this case too.
5085 7. The values of check_spool_space and check_log_space are now held internally
5086 as a number of kilobytes instead of an absolute number of bytes. If a
5087 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5088 kilobyte. This means that much larger values can be stored.
5090 8. Exim monitor: an attempt to get the action menu when not actually pointing
5091 at a message produces an empty menu entitled "No message selected". This
5092 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5093 no entries in it ("Shell widget menu has zero width and/or height"). So I
5094 have added a single, blank menu entry in this case.
5096 9. Added ${quote_local_part.
5098 10. MIME decoding is now applied to the contents of Subject: header lines when
5101 11. Now that a reference to $sender_host_address automatically causes a reverse
5102 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5103 host lookup before query-style lookups in lists that might use this
5104 variable. This has therefore been abolished, and the "net-" prefix is no
5105 longer necessary for query-style lookups.
5107 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5108 have been a typo for LFLAGS, so it has been changed.
5110 13. The install script calls Exim with "-C /dev/null" in order to find the
5111 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5112 to be output. Howeve, since Exim outputs its version number before the
5113 error, it didn't break the script. It just looked ugly. I fixed this by
5114 always allowing "-C /dev/null" if the caller is root.
5116 14. Ignore overlarge ACL variable number when reading spool file - insurance
5117 against a later release with more variables having written the file.
5119 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5120 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5121 this, as well as the form without the "IPv6" on the front.
5123 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5124 OS/Makefile-Darwin file.
5126 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5127 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5129 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5130 variable. However, some OS use ints for inodes. I've added cast to long int
5131 to get rid of the compiler warning.
5133 19. I had forgotten to lock out "/../" in configuration file names when
5134 ALT_CONFIG_PREFIX was set.
5136 20. Routers used for verification do not need to specify transports. However,
5137 if such a router generated a host list, and callout was configured, Exim
5138 crashed, because it could not find a port number from the (non-existent)
5139 transport. It now assumes port 25 in this circumstance.
5141 21. Added the -t option to exigrep.
5143 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5144 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5145 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5146 them is set, LOOKUP_LSEARCH is forced.
5148 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5150 24. Added sender and host information to the "rejected by local_scan()" log
5151 line; previously there was no indication of these.
5153 25. Added .include_if_exists.
5155 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5156 Linux. It turns out that not all file systems support this. Apparently some
5157 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5158 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5159 not supported on the file descriptor, is now ignored when Exim is trying to
5160 sync a directory. This applies only to Linux.
5162 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5164 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5165 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5168 29. Added support for maildirsize files from supplied patch (modified a bit).
5170 30. The use of :fail: followed by an empty string could lead Exim to respond to
5171 sender verification failures with (e.g.):
5173 550 Verification failed for <xxx>
5174 550 Sender verify failed
5176 where the first response line was missing the '-' that indicates it is not
5177 the final line of the response.
5179 31. The loop for finding the name of the user that called Exim had a hardwired
5180 limit of 10; it now uses the value of finduser_retries, which is used for
5181 all other user lookups.
5183 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5186 33. Exim was neglecting to zero errno before one call of strtol() when
5187 expanding a string and expecting an integer value. On some systems this
5188 resulted in spurious "integer overflow" errors. Also, it was casting the
5189 result into an int without checking.
5191 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5192 did not work. The code looks as if it has *never* worked, though it appears
5193 to have been documented since at least releast 1.62. I have made it work.
5195 35. The "timeout_DNS" error in retry rules, also documented since at least
5196 1.62, also never worked. As it isn't clear exactly what this means, and
5197 clearly it isn't a major issue, I have abolished the feature by treating it
5198 as "timeout", and writing a warning to the main and panic logs.
5200 36. The display of retry rules for -brt wasn't always showing the error code
5203 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5204 timeout_connect_A, timeout_connect_MX.
5206 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5207 to the empty sender.
5209 39. The daemon was not analysing the content of -oX till after it had closed
5210 stderr and disconnected from the controlling terminal. This meant that any
5211 syntax errors were only noted on the panic log, and the return code from
5212 the command was 0. By re-arranging the code a little, I've made the
5213 decoding happen first, so such errors now appear on stderr, and the return
5214 code is 1. However, the actual setting up of the sockets still happens in
5215 the disconnected process, so errors there are still only recorded on the
5218 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5219 connections (as happens on some IP stacks) was logged at start up time as
5220 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5221 it from "IPv6 and IPv4", which means that two separate sockets are being
5224 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5225 says whether AF_INET or AF_INET6 was passed as an argument.
5227 42. Exiwhat output was messed up when time zones were included in log
5230 43. Exiwhat now gives more information about the daemon's listening ports,
5231 and whether -tls-on-connect was used.
5233 44. The "port" option of the smtp transport is now expanded.
5235 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5236 silently ignored. Now an error message is written to the main and panic
5239 46. There's a new ACL modifier called "logwrite" which writes to a log file
5240 as soon as it is encountered.
5242 47. Added $local_user_uid and $local_user_gid at routing time.
5244 48. Exim crashed when trying to verify a sender address that was being
5247 49. Exim was recognizing only a space character after ".include". It now also
5248 recognizes a tab character.
5250 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5251 extracting the relevant information from the specification. The man page no
5252 longer contains scrambled data for the -d option, and I've added a section
5253 at the front about calling Exim under different names.
5255 51. Added "extra_headers" argument to the "mail" command in filter files.
5257 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5260 53. Installed eximstats 1.29.
5262 54. Added transport_filter_timeout as a generic transport option.
5264 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5265 Cc: header lines. This was required by RFC 822, but it not required by RFC
5268 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5269 incoming messages that did not have them. Now it does so only if the
5270 message originates locally, that is, if there is no associated remote host
5271 address. When Resent- header lines are present, this applies to the Resent-
5272 lines rather than the non-Resent- lines.
5274 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5275 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5277 58. Messages for configuration errors now include the name of the main
5278 configuration file - useful now that there may be more than one file in a
5279 list (.included file names were always shown).
5281 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5282 for those rare installations that do not start the daemon as root or run it
5283 setuid root. I've cut out the call to initgroups() if the daemon is not
5286 60. The Exim user and group can now be bound into the binary as text strings
5287 that are looked up at the start of Exim's processing.
5289 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5291 62. Added $mailstore_basename variable.
5293 63. Installed patch to sieve.c from Michael Haardt.
5295 64. When Exim failed to open the panic log after failing to open the main log,
5296 the original message it was trying to log was written to stderr and debug
5297 output, but if they were not available (the usual case in production), it
5298 was lost. Now it is written to syslog before the two lines that record the
5299 failures to open the logs.
5301 65. Users' Exim filters run in subprocesses under the user's uid. It is
5302 possible for a "deliver" command or an alias in a "personal" command to
5303 provoke an address rewrite. If logging of address rewriting is configured,
5304 this fails because the process is not running as root or exim. There may be
5305 a better way of dealing with this, but for the moment (because 4.30 needs
5306 to be released), I have disabled address rewrite logging when running a
5307 filter in a non-root, non-exim process.
5313 1. The buildconfig auxiliary program wasn't quoting the value set for
5314 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5315 not defined. This bug was masked in 4.22 by the effect that was fixed in
5318 2. Some messages that were rejected after a message id was allocated were
5319 shown as "incomplete" by exigrep. It no longer does this for messages that
5320 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5322 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5323 have allow_domain_literals set, the ID did not get logged in the <= line.
5324 Domain literals are now always recognized in Message-ID: header lines.
5326 4. The first argument for a ${extract expansion item is the key name or field
5327 number. Leading and trailing spaces in this item were not being ignored,
5328 causing some misleading effects.
5330 5. When deliver_drop_privilege was set, single queue runner processes started
5331 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5332 same command in the process it spins off) were not dropping privilege.
5334 6. When the daemon running as "exim" started a queue runner, it always
5335 re-executed Exim in the spun-off process. This is a waste of effort when
5336 deliver_drop_privilege is set. The new process now just calls the
5337 queue-runner function directly.
5343 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5346 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5347 ignored. Though the use of -oP was forcing the writing of a pid file, it
5348 was always written to the default place.
5350 3. If the message "no IP address found for host xxxx" is generated during
5351 incoming verification, it is now followed by identification of the incoming
5352 connection (so you can more easily find what provoked it).
5354 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5356 5. Added some features to "harden" Exim a bit more against certain attacks:
5358 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5359 be put in Local/Makefile. This is like the never_users runtime option,
5360 but it cannot be overridden. The default setting is "root".
5362 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5363 prefix string with which any file named in a -C command line option
5366 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5367 is retained for -C and -D only if the caller of Exim is root. Without
5368 it, the exim user may also use -C and -D and retain privilege.
5370 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5371 command line option is disabled.
5373 6. Macro names set by the -D option must start with an upper case letter, just
5374 like macro names defined in the configuration file.
5376 7. Added "dereference=" facility to LDAP.
5378 8. Two instances of the typo "uknown" in the source files are fixed.
5380 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5381 the Configure-Makefile script screwed up while processing it.
5383 10. Incorporated PCRE 4.4.
5385 11. The SMTP synchronization check was not operating right at the start of an
5386 SMTP session. For example, it could not catch a HELO sent before the client
5387 waited for the greeting. There is now a check for outstanding input at the
5388 point when the greeting is written. Because of the duplex, asynchronous
5389 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5390 way, but not yet received, when the check is performed.
5392 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5393 on TCP/IP sockets, because this apparently causes some broken clients to
5396 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5397 unchanged) from the Cygwin maintainer.
5399 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5400 was supported instead of testing for mbx. Effectively a typo.
5402 15. The spa authenticator server code was not checking that the input it
5403 received was valid base64.
5405 16. The debug output line for the "set" modifier in ACLs was not showing the
5406 name of the variable that was being set.
5408 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5410 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5411 lookup on its own. Something else was needed to trigger the lookup. For
5412 example, a match in host_lookup or the need for a host name in a host list.
5413 Now, if $sender_host_name is referenced and the host name has not yet been
5414 looked up, a lookup is performed. If the lookup fails, the variable remains
5415 empty, and $host_lookup_failed is set to "1".
5417 19. Added "eqi" as a case-independent comparison operator.
5419 20. The saslauthd authentication condition could segfault if neither service
5420 nor realm was specified.
5422 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5423 error message that was logged was misleading, and incoming SMTP
5424 connections were dropped. The message is now more accurate, and temporary
5425 errors are given to SMTP connections.
5427 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5428 (except RSET) with a 421 error, until QUIT is received. However, it was
5429 failing to send a response to QUIT.
5431 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5432 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5433 ACL is now run after the helo_try_verify_hosts code.
5435 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5436 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5437 case-independent, so other case variants are also recognized). Apparently
5438 some systems use these upper case variants.
5440 25. If more than two messages were waiting for the same host, and a transport
5441 filter was specified for the transport, Exim sent two messages over the
5442 same TCP/IP connection, and then failed with "socket operation on non-
5443 socket" when it tried to send the third.
5445 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5447 27. The extern definition of crypt16() in expand.c was not being excluded when
5448 the OS had its own crypt16() function.
5450 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5451 as a preferred synonym for return_size_limit, both as an option and as an
5454 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5456 30. Changed the default configuration ACL to relax the local part checking rule
5457 for addresses that are not in any local domains. For these addresses,
5458 slashes and pipe symbols are allowed within local parts, but the sequence
5459 /../ is explicitly forbidden.
5461 31. SPA server authentication was not clearing the challenge buffer before
5464 32. log_message in a "warn" ACL statement was writing to the reject log as
5465 well as to the main log, which contradicts the documentation and doesn't
5466 seem right (because no rejection is happening). So I have stopped it.
5468 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5469 However, I am unable to do any testing of this.
5471 34. Fixed an infelicity in the appendfile transport. When checking directories
5472 for a mailbox, to see if any needed to be created, it was accidentally
5473 using path names with one or more superfluous leading slashes; tracing
5474 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5476 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5477 contents are added to the log line that is written for every discarded
5478 recipient. (Previously a log_message setting was ignored.)
5480 36. The ${quote: operator now quotes the string if it is empty.
5482 37. The install script runs exim in order to find its version number. If for
5483 some reason other than non-existence or emptiness, which it checks, it
5484 could not run './exim', it was installing it with an empty version number,
5485 i.e. as "exim-". This error state is now caught, and the installation is
5488 38. An argument was missing from the function that creates an error message
5489 when Exim fails to connect to the socket for saslauthd authentication.
5490 This could cause Exim to crash, or give a corrupted message.
5492 39. Added isip, isip4, and isip6 to ${if conditions.
5494 40. The ACL variables $acl_xx are now saved with the message, and can be
5495 accessed later in routers, transports, and filters.
5497 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5498 strings in the file are not string-expanded.
5500 42. If a MAIL command specified a SIZE value that was too large to fit into an
5501 int variable, the check against message_size_limit failed. Such values are
5502 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5503 day this will have to be increased, but I don't think I want to be around
5504 when emails are that large.
5511 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5512 iconv() is not standard in FreeBSD.
5514 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5515 IPv6 enabled. The observed symptom was a segmentation fault on return from
5516 the function os_common_find_running_interfaces() in src/os.c.
5518 3. In the check_special_case() function in daemon.c I had used "errno" as an
5519 argument name, which causes warnings on some systems. This was basically a
5520 typo, since it was named "eno" in the comments!
5522 4. The code that waits for the clock to tick (at a resolution of some fraction
5523 of a second) so as to ensure message-id uniqueness was always waiting for
5524 at least one whole tick, when it could have waited for less. [This is
5525 almost certainly not relevant at current processor speeds, where it is
5526 unlikely to ever wait at all. But we try to future-proof.]
5528 5. The function that sleeps for a time interval that includes fractions of a
5529 second contained a race. It did not block SIGALRM between setting the
5530 timer, and suspending (a couple of lines later). If the interval was short
5531 and the sigsuspend() was delayed until after it had expired, the suspension
5532 never ended. On busy systems this could lead to processes getting stuck for
5535 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5536 process, before it forks any delivery processes. The open lookup caching
5537 mechanism meant that the open file or database connection was passed into
5538 the delivery process. The problem was that delivery processes always tidy
5539 up cached lookup data. This could cause a problem for the next delivery
5540 process started by the queue runner, because the external queue runner
5541 process does not know about the closure. So the next delivery process
5542 still has data in the lookup cache. In the case of a file lookup, there was
5543 no problem because closing a file descriptor in a subprocess doesn't affect
5544 the parent. However, if the lookup was caching a connection to a database,
5545 the connection was closed, and the second delivery process was likely to
5546 see errors such as "PGSQL: query failed: server closed the connection
5547 unexpectedly". The problem has been fixed by closing all cached lookups
5548 in a queue runner before running a delivery process.
5550 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5551 seem to have the "const" qualifier which it has on other OS. I've
5554 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5555 *of the same type* (client or server) with the same public name that an
5556 error should be diagnosed.
5558 9. When Exim looked up a host name for an IP address, but failed to find the
5559 original IP address when looking up the host name (a safety check), it
5560 output the message "<ip address> does not match any IP for NULL", which was
5561 confusing, to say the least. The bug was that the host name should have
5562 appeared instead of "NULL".
5564 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5565 user that is built into the binary, and the -C or -D options is used, root
5566 privilege is dropped before the configuration file is read. In addition,
5567 logging is switched to stderr instead of the normal log files. If the
5568 configuration then re-defines the Exim user, the unprivileged environment
5569 is probably not what is expected, so Exim logs a panic warning message (but
5572 However, if deliver_drop_privilege is set, the unprivileged state may well
5573 be exactly what is intended, so the warning has been cut out in that case,
5574 and Exim is allowed to try to write to its normal log files.
5580 1. smtp_return_error_details was not giving details for temporary sender
5581 or receiver verification errors.
5583 2. Diagnose a configuration error if two authenticators have the same public
5586 3. Exim used not to create the message log file for a message until the first
5587 delivery attempt. This could be confusing when incoming messages were held
5588 for policy or load reasons. The message log file is now created at the time
5589 the message is received, and an initial "Received" line is written to it.
5591 4. The automatically generated man page for command line options had a minor
5592 bug that caused no ill effects; however, a more serious problem was that
5593 the procedure for building the man page automatically didn't always
5594 operate. Consequently, release 4.20 contains an out-of-date version. This
5595 shouldn't happen again.
5597 5. When building Exim with embedded Perl support, the script that builds the
5598 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5599 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5601 6. The freeze_tell option was not being used for messages that were frozen on
5602 arrival, either by an ACL or by local_scan().
5604 7. Added the smtp_incomplete_transaction log selector.
5606 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5607 was accepting AUTH without a new EHLO.
5609 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5610 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5613 10. Logging of TCP/IP connections (when configured) now happens in the main
5614 daemon process instead of the child process, so that the TCP/IP connection
5615 count is more accurate (but it can never be perfect).
5617 11. The use of "drop" in a nested ACL was not being handled correctly in the
5618 outer ACL. Now, if condition failure induced by the nested "drop" causes
5619 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5620 or "require"), the connection is dropped.
5622 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5623 that yield "discard" can now be used with an "accept" or a "discard" verb,
5624 but an error is generated for any others (because I can't see a useful way
5625 to define what should happen).
5627 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5628 are now processed in the same way as lines in the Exim configuration file.
5629 In particular, continuation lines are supported.
5631 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5633 15. Added -ti meaning -t -i.
5635 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5636 domains, and warn by logging if others are found.
5638 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5639 returned from GIFCONF when Exim is trying to find the list of interfaces on
5640 a host. The code in os.c has been modified to copy each ifreq to an aligned
5641 structure in all cases.
5643 Also, in some cases, the returned ifreq's were being copied to a 'struct
5644 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5645 means the last couple of bytes of an IPv6 address could be chopped if the
5646 ifreq contained only a normal sockaddr (14 bytes storage).
5648 18. Named domain lists were not supported in the hosts_treat_as_local option.
5649 An entry such as +xxxx was not recognized, and was treated as a literal
5652 19. Ensure that header lines added by a DATA ACL are included in the reject log
5653 if the ACL subsequently rejects the message.
5655 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5656 MD5 (which is deprecated).
5658 21. When testing a filter file using -bf, Exim was writing a message when it
5659 took the sender from a "From " line in the message, but it was not doing so
5660 when it took $return_path from a Return-Path: header line. It now does.
5662 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5663 with a valid header line field name (a series of printing characters
5664 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5666 23. Changed "disc" in the source to "disk" to conform to the documentation and
5667 the book and for uniformity.
5669 24. Ignore Sendmail's -Ooption=value command line item.
5671 25. When execve() failed while trying to run a command in a pipe transport,
5672 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5673 could be confused with a return value of 69 from the command itself. This
5674 has been changed to 127, the value the shell returns if it is asked to run
5675 a non-existent command. The wording for the related log line suggests a
5676 non-existent command as the problem.
5678 26. If received_header_text expands to an empty string, do not add a Received:
5679 header line to the message. (Well, it adds a token one on the spool, but
5680 marks it "old" so that it doesn't get used or transmitted.)
5682 27. Installed eximstats 1.28 (addition of -nt option).
5684 28. There was no check for failure on the call to getsockname() in the daemon
5685 code. This can fail if there is a shortage of resources on the system, with
5686 ENOMEM, for example. A temporary error is now given on failure.
5688 29. Contrary to the C standard, it seems that in some environments, the
5689 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5690 program. Exim now does this explicitly; it affects the formatting of
5691 timestamps using strftime().
5693 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5694 complaints. I've now initialized all the variables, to avoid this.
5696 32. Header lines added by a system filter were not being "seen" during
5697 transport-time rewrites.
5699 33. The info_callback() function passed to OpenSSL is set up with type void
5700 (*)(SSL *, int, int), as described somewhere. However, when calling the
5701 function (actually a macro) that sets it up, the type void(*)() is
5702 expected. I've put in a cast to prevent warnings from picky compilers.
5704 34. If a DNS black list lookup found a CNAME record, but there were no A
5705 records associated with the domain it pointed at, Exim crashed.
5707 35. If a DNS black list lookup returned more than one A record, Exim ignored
5708 all but the first. It now scans all returned addresses if a particular IP
5709 value is being sought. In this situation, the contents of the
5710 $dnslist_value variable are a list of all the addresses, separated by a
5713 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5714 to accept a host name and all its aliases if the forward lookup for any of
5715 them yielded the IP address of the incoming connection. Now it accepts only
5716 those names whose forward lookup yields the correct IP address. Any other
5717 names are discarded. This closes a loophole whereby a rogue DNS
5718 administrator could create reverse DNS records to break through a
5719 wildcarded host restriction in an ACL.
5721 37. If a user filter or a system filter that ran in a subprocess used any of
5722 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5723 the wrong values were passed to the pipe command ($thisaddress had the
5724 value of $0, $0 had the value of $1, etc). This bug was introduced by
5725 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5727 38. Improved the line breaking for long SMTP error messages from ACLs.
5728 Previously, if there was no break point between 40 and 75 characters, Exim
5729 left the rest of the message alone. Two changes have been made: (a) I've
5730 reduced the minimum length to 35 characters; (b) if it can't find a break
5731 point between 35 and 75 characters, it looks ahead and uses the first one
5732 that it finds. This may give the occasional overlong line, but at least the
5733 remaining text gets split now.
5735 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5736 file descriptors might be low, and that setting 1000 would always raise it.
5737 It turns out that in some environments, the limit is already over 1000 and
5738 that lowering it causes trouble. So now Exim takes care not to decrease it.
5740 40. When delivering a message, the value of $return_path is set to $sender_
5741 address at the start of routing (routers may change the value). By an
5742 oversight, this default was not being set up when an address was tested by
5743 -bt or -bv, which affected the outcome if any router or filter referred to
5746 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5747 log only when "message" or "log_message" are set. However, if one of the
5748 conditions was an address verification, or a call to a nested ACL, the
5749 messages generated by the underlying test were being passed through. This
5750 no longer happens. The underlying message is available in $acl_verify_
5751 message for both "message" and "log_message" expansions, so it can be
5752 passed through if needed.
5754 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5755 new expansion $bh_ to give the encoded byte string without charset
5756 translation. Translation happens only if iconv() is available; HAVE_ICONV
5757 indicates this at build time. HEADERS_CHARSET gives the charset to
5758 translate to; headers_charset can change it in the configuration, and
5759 "headers charset" can change it in an individual filter file.
5761 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5762 that creates RFC 2047 encoded "words" labels them as that charset instead
5763 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5764 expansion operator; (ii) when Exim creates a From: line for a local
5765 message; (iii) when a header line is rewritten to include a "phrase" part.
5767 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5768 buggy, causing it to skip the first lines of messages whose message ID
5769 ended in 'D'. This would not have bitten before Exim release 4.14, because
5770 message IDs were unlikely to end in 'D' before then. The effect was to have
5771 incorrect size information for certain domains.
5773 45. #include "config.h" was missing at the start of the crypt16.c module. This
5774 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5777 46. If there was a timeout during a "random" callout check, Exim treated it as
5778 a failure of the random address, and carried on sending RSET and the real
5779 address. If the delay was just some slowness somewhere, the response to the
5780 original RCPT would be taken as a response to RSET and so on, causing
5781 mayhem of various kinds.
5783 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5784 when I implemented it. It didn't allow for the fact that some option values
5785 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5786 the right test for positive values.
5788 48. Domain names in DNS records are case-independent. Exim always looks them up
5789 in lower case. Some resolvers return domain names in exactly the case they
5790 appear in the zone file, that is, they may contain uppercase letters. Not
5791 all resolvers do this - some return always lower case. Exim was treating a
5792 change of case by a resolver as a change of domain, similar to a widening
5793 of a domain abbreviation. This triggered its re-routing code and so it was
5794 trying to route what was effectively the same domain again. This normally
5795 caused routing to fail (because the router wouldn't handle the domain
5796 twice). Now Exim checks for this case specially, and just changes the
5797 casing of the domain that it ultimately uses when it transmits the message
5800 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5803 50. If a filter generated a file delivery with a non-absolute name (possible if
5804 no home directory exists for the router), the forbid_file option was not
5807 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5808 the existing equality matching.
5810 52. Exim was using ints instead of ino_t variables in some places where it was
5811 dealing with inode numbers.
5813 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5814 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5815 called TMPDIR, and if it finds it is different, it changes its value.
5817 54. The smtp_printf() function is now made available to local_scan() so
5818 additional output lines can be written before returning. There is also an
5819 smtp_fflush() function to enable the detection of a dropped connection.
5820 The variables smtp_input and smtp_batched_input are exported to
5823 55. Changed the default runtime configuration: the message "Unknown user"
5824 has been removed from the ACL, and instead placed on the localuser router,
5825 using the cannot_route_message feature. This means that any verification
5826 failures that generate their own messages won't get overridden. Similarly,
5827 the "Unrouteable address" message that was in the ACL for unverifiable
5828 relay addresses has also been removed.
5830 56. Added hosts_avoid_esmtp to the smtp transport.
5832 57. The exicyclog script was not checking for the esoteric option
5833 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5834 will work only if exicyclog is run under the appropriate euid.
5836 58. Following a discussion on the list, the rules by which Exim recognises line
5837 endings on incoming messages have been changed. The -dropcr and drop_cr
5838 options are now no-ops, retained only for backwards compatibility. The
5839 following line terminators are recognized: LF CRLF CR. However, special
5840 processing applies to CR:
5842 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5843 nor a local message in the state where . is a terminator.
5845 (ii) If a bare CR is encountered in a header line, an extra space is added
5846 after the line terminator so as not to end the header. The reasoning
5847 behind this is that bare CRs in header lines are most likely either
5848 to be mistakes, or people trying to play silly games.
5850 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5851 was being incorrectly given as 18 bytes larger than it should have been.
5852 This is a VOB (very old bug).
5854 60. This may never have affected anything current, but just in case it has:
5855 When the local host is found other than at the start of a list of hosts,
5856 the local host, those with the same MX, and any that follow, are discarded.
5857 When the list in question was part of a longer list of hosts, the following
5858 hosts (not currently being processed) were also being discarded. This no
5859 longer happens. I'm not sure if this situation could ever has previously
5862 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5865 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5866 header. This is something that is recommended in a new Internet Draft, and
5867 is something that is documented as being done by Sendmail. There are two
5868 possible values. For messages generated by the autoreply transport, Exim
5871 Auto-Submitted: auto-replied
5873 whereas for all other generated messages (e.g. bounces) it adds
5875 Auto-Submitted: auto-generated
5877 63. The "personal" condition in filters now includes a test for the
5878 Auto-Submitted: header. If it contains the string "auto-" the message it
5879 not considered personal.
5881 64. Added rcpt_include_affixes as a generic transport option.
5883 65. Added queue_only_override (default true).
5885 66. Added the syslog_duplication option.
5887 67. If what should have been the first header line of a message consisted of
5888 a space followed by a colon, Exim was mis-interpreting it as a header line.
5889 It isn't of course - it is syntactically invalid and should therefore be
5890 treated as the start of the message body. The misbehaviour could have
5891 caused a number of strange effects, including loss of data in subsequent
5892 header lines, and spool format errors.
5894 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5895 client host had authenticated. This control can now be exercised by an ACL
5896 for more flexibility.
5898 69. By default, callouts do not happen when testing with -bh. There is now a
5899 variant, -bhc, which does actually run the callout code, including
5900 consulting and updating the callout cache.
5902 70. Added support for saslauthd authentication, courtesy of Alexander
5905 71. If statvfs() failed on the spool or log directories while checking their
5906 size for availability, Exim confusingly gave the error "space shortage".
5907 Furthermore, in debugging mode it crashed with a floating point exception.
5908 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5909 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5910 serious problem, Exim now writes to the main and panic logs when this
5911 happens, with details of the failure. It then refuses to accept the
5912 incoming message, giving the message "spool directory problem" or "log
5913 directory problem" with a 421 code for SMTP messages.
5915 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5916 0, 1, and 2 exist, because some OS complain for execs without them (see
5917 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5918 descriptors. However, the code omitted to check that the open succeeded,
5919 causing mysterious errors if for some reason the permissions on /dev/null
5920 got screwed. Now Exim writes a message to the main and panic logs, and
5921 bombs out if it can't open /dev/null.
5923 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5924 interact so that it is all more flexible. It is supposed to remain
5925 backwards compatible. Also added extra_local_interfaces.
5927 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5928 to bomb out with an assertion failure - to the client this appears as a
5929 connection drop. This problem occurs in the part of the code that was taken
5930 from the Samba project. Fortunately, the assertion is in a very simple
5931 function, so I have fixed this by reproducing the function inline in the
5932 one place where it is called, and arranging for authentication to fail
5933 instead of killing the process with assert().
5935 75. The SPA client code was not working when the server requested OEM rather
5936 than Unicode encoding.
5938 76. Added code to make require_files with a specific uid setting more usable in
5939 the case where statting the file as root fails - usually a non-root-mounted
5940 NFS file system. When this happens and the failure is EACCES, Exim now
5941 forks a subprocess and does the per-uid checking as the relevant uid.
5943 77. Added process_log_path.
5945 78. If log_file_path was not explicitly set, a setting of check_log_space or
5946 check_log_inodes was ignored.
5948 79. If a space check for the spool or log partitions fails, the incident is now
5949 logged. Of course, in the latter case the data may get lost...
5951 80. Added the %p formatting code to string_format() so that it can be used to
5952 print addresses in debug_print(). Adjusted all the address printing in the
5953 debugging in store.c to use %p rather than %d.
5955 81. There was a concern that a line of code in smtp_in.c could overflow a
5956 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5957 initially expressed, the concern was not well-founded, because trailing
5958 spaces are removed early. However, if the trailing spaces were followed by
5959 a NULL, they did not get removed, so the overflow was possible. Two fixes
5962 (a) I re-wrote the offending code in a cleaner fashion.
5963 (b) If an incoming SMTP command contains a NULL character, it is rejected
5966 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5967 runs initgroups(), so that if the Exim user is in any additional groups,
5968 they will be used during message reception.
5974 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob