1 Change log file for Exim from version 4.21
2 -------------------------------------------
8 TF/01 Correctly close the server side of TLS when forking for delivery.
10 When a message was received over SMTP with TLS, Exim failed to clear up
11 the incoming connection properly after forking off the child process to
12 deliver the message. In some situations the subsequent outgoing
13 delivery connection happened to have the same fd number as the incoming
14 connection previously had. Exim would try to use TLS and fail, logging
15 a "Bad file descriptor" error.
17 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
18 utilities have not been installed.
20 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
21 temporary space as the ACL may create new global variables.
23 TL/01 LDAP support uses per connection or global context settings, depending
24 upon the detected version of the libraries at build time.
26 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
27 to extract and use the src ip:port in logging and expansions as if it
28 were a direct connection from the outside internet.
30 JH/02 Add ${listextract {number}{list}{success}{fail}}.
32 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
33 Properly escape header and check for NULL return.
35 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
38 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
40 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
41 characters in header names, implemented as a verify condition.
42 Contributed by Michael Fischer v. Mollard.
44 TL/05 Rename SPF condition results err_perm and err_temp to standardized
45 results permerror and temperror. Is a backward incompatibility if
46 the ACL tests for either of these two results. Patch contributed by
47 user bes-internal on the mailing list.
49 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
51 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
52 selectors, in both main and reject logs.
54 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
57 JH/07 Add malware type "sock" for talking to simple daemon.
63 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
65 PP/02 Make -n do something, by making it not do something.
66 When combined with -bP, the name of an option is not output.
68 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
71 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
72 $sender_host_name and config options to manage this, and basic check
75 PP/05 DSCP support for outbound connections and control modifier for inbound.
77 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
78 (Only plugin which currently uses this is kerberos4, which nobody should
79 be using, but we should make it available and other future plugins might
80 conceivably use it, even though it would break NAT; stuff *should* be
81 using channel bindings instead).
83 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
84 name; added for Sendmail compatibility; requires admin caller.
85 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
86 just ignore it); requires trusted caller.
87 Also parse but ignore: -Ac -Am -X<logfile>
90 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
92 TL/02 Add +smtp_confirmation as a default logging option.
94 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
95 Patch by Magnus Holmgren from 2007-02-20.
97 TL/04 Bugzilla 1281 - Spec typo.
98 Bugzilla 1283 - Spec typo.
99 Bugzilla 1290 - Spec grammar fixes.
101 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
103 TL/06 Add Experimental DMARC support using libopendmarc libraries.
105 TL/07 Fix an out of order global option causing a segfault. Reported to dev
106 mailing list by by Dmitry Isaikin.
108 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
110 JH/02 Support "G" suffix to numbers in ${if comparisons.
112 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
114 NM/01 Bugzilla 1197 - Spec typo
115 Bugzilla 1196 - Spec examples corrections
117 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
119 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
120 gnutls_enable_pkcs11, but renamed to more accurately indicate its
123 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
124 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
126 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
127 "acl {{name}{arg}...}", and optional args on acl condition
130 JH/05 Permit multiple router/transport headers_add/remove lines.
132 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
134 JH/07 Avoid using a waiting database for a single-message-only transport.
135 Performance patch from Paul Fisher. Bugzilla 1262.
137 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
140 JH/09 Add $headers_added variable, with content from use of ACL modifier
141 add_header (but not yet added to the message). Bugzilla 199.
143 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
144 Pulled from Bugzilla 817 by Wolfgang Breyha.
146 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
148 (nb: this is the same fix as in Exim 4.80.1)
150 JH/11 Add A= logging on delivery lines, and a client_set_id option on
153 JH/12 Add optional authenticated_sender logging to A= and a log_selector
156 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
158 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
159 advertise SMTP AUTH mechanism to us, instead of a generic
160 protocol violation error. Also, make Exim more robust to bad
161 data from the Dovecot auth socket.
163 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
165 When a queue runner is handling a message, Exim first routes the
166 recipient addresses, during which it prunes them based on the retry
167 hints database. After that it attempts to deliver the message to
168 any remaining recipients. It then updates the hints database using
171 So if a recipient address works intermittently, it can get repeatedly
172 deferred at routing time. The retry hints record remains fresh so the
173 address never reaches the final cutoff time.
175 This is a fairly common occurrence when a user is bumping up against
176 their storage quota. Exim had some logic in its local delivery code
177 to deal with this. However it did not apply to per-recipient defers
178 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
180 This change adds a proper retry rule check during routing so that the
181 final cutoff time is checked against the message's age. We only do
182 this check if there is an address retry record and there is not a
183 domain retry record; this implies that previous attempts to handle
184 the address had the retry_use_local_parts option turned on. We use
185 this as an approximation for the destination being like a local
186 delivery, as in LMTP.
188 I suspect this new check makes the old local delivery cutoff check
189 redundant, but I have not verified this so I left the code in place.
191 TF/02 Correct gecos expansion when From: is a prefix of the username.
193 Test 0254 submits a message to Exim with the header
197 When I ran the test suite under the user fanf2, Exim expanded
198 the header to contain my full name, whereas it should have added
199 a Resent-Sender: header. It erroneously treats any prefix of the
200 username as equal to the username.
202 This change corrects that bug.
204 GF/01 DCC debug and logging tidyup
205 Error conditions log to paniclog rather than rejectlog.
206 Debug lines prefixed by "DCC: " to remove any ambiguity.
208 TF/03 Avoid unnecessary rebuilds of lookup-related code.
210 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
211 Bug spotted by Jeremy Harris; was flawed since initial commit.
212 Would have resulted in OCSP responses post-SNI triggering an Exim
213 NULL dereference and crash.
215 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
217 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
218 Bug detection, analysis and fix by Samuel Thibault.
219 Bugzilla 1331, Debian bug #698092.
221 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
223 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
224 Server implementation by Todd Lyons, client by JH.
225 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
226 config variable "prdr_enable" controls whether the server
227 advertises the facility. If the client requests PRDR a new
228 acl_data_smtp_prdr ACL is called once for each recipient, after
229 the body content is received and before the acl_smtp_data ACL.
230 The client is controlled by bolth of: a hosts_try_prdr option
231 on the smtp transport, and the server advertisement.
232 Default client logging of deliveries and rejections involving
233 PRDR are flagged with the string "PRDR".
235 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
236 fclose(). Diagnosis by Todd Lyons.
238 PP/17 Update configure.default to handle IPv6 localhost better.
239 Patch by Alain Williams (plus minor tweaks).
242 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
243 This is now consistent with GnuTLS, and is now documented: the
244 previous undocumented portable approach to treating the option as
245 unset was to force an expansion failure. That still works, and
246 an empty string is now equivalent.
248 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
249 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
250 not performing validation itself.
252 PP/20 Added force_command boolean option to pipe transport.
253 Patch from Nick Koston, of cPanel Inc.
255 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
258 TF/04 Added udpsend ACL modifer and hexquote expansion operator
260 PP/21 Fix eximon continuous updating with timestamped log-files.
261 Broken in a format-string cleanup in 4.80, missed when I repaired the
262 other false fix of the same issue.
263 Report and fix from Heiko Schlichting.
266 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
267 Report from Prashanth Katuri.
269 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
270 It's SecureTransport, so affects any MacOS clients which use the
271 system-integrated TLS libraries, including email clients.
273 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
274 using a MIME ACL for non-SMTP local injection.
275 Report and assistance in diagnosis by Warren Baker.
277 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
279 JH/16 Fix comparisons for 64b. Bugzilla 1385.
281 TL/09 Add expansion variable $authenticated_fail_id to keep track of
282 last id that failed so it may be referenced in subsequent ACL's.
284 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
287 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
288 ldap library initialization, allowing self-signed CA's to be
289 used. Also properly sets require_cert option later in code by
290 using NULL (global ldap config) instead of ldap handle (per
291 session). Bug diagnosis and testing by alxgomz.
293 TL/12 Enhanced documentation in the ratelimit.pl script provided in
294 the src/util/ subdirectory.
296 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
297 renamed to Transport Post Delivery Action by Jeremy Harris, as
300 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
301 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
302 redis_servers = needs to be configured which will be used by the redis
303 lookup. Patch from Warren Baker, of The Packet Hub.
305 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
307 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
308 hostname or reverse DNS when processing a host list. Used suggestions
309 from multiple comments on this bug.
311 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
313 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
314 Missed a few lines, added it to make the runtest require no keyboard
317 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
318 contains upper case chars. Make router use caseful_local_part.
320 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
321 support when GnuTLS has been built with p11-kit.
327 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
329 This, or similar/improved, will also be change PP/11 of 4.82.
335 PP/01 Handle short writes when writing local log-files.
336 In practice, only affects FreeBSD (8 onwards).
337 Bugzilla 1053, with thanks to Dmitry Isaikin.
339 NM/01 Bugzilla 949 - Documentation tweak
341 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
344 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
346 PP/02 Implemented gsasl authenticator.
348 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
350 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
351 `pkg-config foo` for cflags/libs.
353 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
354 with rest of GSASL and with heimdal_gssapi.
356 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
357 `pkg-config foo` for cflags/libs for the TLS implementation.
359 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
360 properties get this fed in as external SSF. A number of robustness
361 and debugging improvements to the cyrus_sasl authenticator.
363 PP/08 cyrus_sasl server now expands the server_realm option.
365 PP/09 Bugzilla 1214 - Log authentication information in reject log.
366 Patch by Jeremy Harris.
368 PP/10 Added dbmjz lookup type.
370 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
372 PP/12 MAIL args handles TAB as well as SP, for better interop with
373 non-compliant senders.
374 Analysis and variant patch by Todd Lyons.
376 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
377 Bug report from Lars Müller <lars@samba.org> (via SUSE),
378 Patch from Dirk Mueller <dmueller@suse.com>
380 PP/13 tls_peerdn now print-escaped for spool files.
381 Observed some $tls_peerdn in wild which contained \n, which resulted
382 in spool file corruption.
384 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
385 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
386 or write after TLS renegotiation, which otherwise led to messages
389 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
390 as a tracking header (ie: a signed header comes before the signature).
391 Patch from Wolfgang Breyha.
393 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
394 comma-sep list; embedded commas doubled.
396 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
398 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
400 Report and patch from Dmitry Banschikov.
402 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
403 Removed SSL_clear() after SSL_new() which led to protocol negotiation
404 failures. We appear to now support TLS1.1+ with Exim.
406 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
407 lets Exim select keys and certificates based upon TLS SNI from client.
408 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
409 before an outbound SMTP session. New log_selector, +tls_sni.
411 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
412 NULL dereference. Report and patch from Alun Jones.
414 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
415 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
416 Not seeing resolver debug output on NetBSD, but suspect this is a
417 resolver implementation change.
419 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
420 Left warnings. Added "eximon gdb" invocation mode.
422 PP/21 Defaulting "accept_8bitmime" to true, not false.
424 PP/22 Added -bw for inetd wait mode support.
426 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
427 locate the relevant includes and libraries. Made this the default.
429 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
430 Bugzilla 1246, report and most of solution from Tomasz Kusy.
432 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
433 This may cause build issues on older platforms.
435 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
436 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
437 gnutls_require_mac & gnutls_require_protocols (no longer supported).
438 Added SNI support via GnuTLS too.
439 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
441 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
443 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
444 Applied second patch from Janne, implementing suggestion to default
445 multiple-strings-in-record handling to match SPF spec.
447 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
449 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
450 read-only, out of scope).
451 Patch from Wolfgang Breyha, report from Stuart Northfield.
453 PP/29 Fix three issues highlighted by clang analyser static analysis.
454 Only crash-plausible issue would require the Cambridge-specific
455 iplookup router and a misconfiguration.
456 Report from Marcin Mirosław.
458 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
460 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
461 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
462 As part of this, removing so much warning spew let me fix some minor
463 real issues in debug logging.
465 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
466 assignment on my part. Fixed.
468 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
469 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
470 Janne Snabb (who went above and beyond: thank you).
472 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
473 string otherwise requires a connection and a bunch more work and it's
474 relatively easy to get wrong. Should also expose TLS library linkage
477 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
478 64-bit ${eval} (JH/03).
480 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
481 GNU libc to support some of the 64-bit stuff, should not lead to
482 conflicts. Defined before os.h is pulled in, so if a given platform
483 needs to override this, it can.
485 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
486 protection layer was required, which is not implemented.
487 Bugzilla 1254, patch from Wolfgang Breyha.
489 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
490 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
491 tls_dhparam take prime identifiers. Also unbreak combination of
492 OpenSSL+DH_params+TLSSNI.
494 PP/39 Disable SSLv2 by default in OpenSSL support.
500 PP/01 Solaris build fix for Oracle's LDAP libraries.
501 Bugzilla 1109, patch from Stephen Usher.
503 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
505 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
508 TF/02 Fix a couple more cases where we did not log the error message
509 when unlink() failed. See also change 4.74-TF/03.
511 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
512 lock up or crash if it happened to be inside a call to libc when it
513 got a SIGUSR1 from exiwhat.
515 The SIGUSR1 handler appends the current process status to the process
516 log which is later printed by exiwhat. It used to use the general
517 purpose logging code to do this, but several functions it calls are
518 not safe for signals.
520 The new output code in the SIGUSR1 handler is specific to the process
521 log, and simple enough that it's easy to inspect for signal safety.
522 Removing some special cases also simplifies the general logging code.
523 Removing the spurious timestamps from the process log simplifies
526 TF/04 Improved ratelimit ACL condition.
528 The /noupdate option has been deprecated in favour of /readonly which
529 has clearer semantics. The /leaky, /strict, and /readonly update modes
530 are mutually exclusive. The update mode is no longer included in the
531 database key; it just determines when the database is updated. (This
532 means that when you upgrde Exim will forget old rate measurements.)
534 Exim now checks that the per_* options are used with an update mode that
535 makes sense for the current ACL. For example, when Exim is processing a
536 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
537 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
538 must specify per_mail/readonly. If you omit the update mode it defaults to
539 /leaky where that makes sense (as before) or /readonly where required.
541 The /noupdate option is now undocumented but still supported for
542 backwards compatibility. It is equivalent to /readonly except that in
543 ACLs where /readonly is required you may specify /leaky/noupdate or
544 /strict/noupdate which are treated the same as /readonly.
546 A useful new feature is the /count= option. This is a generalization
547 of the per_byte option, so that you can measure the throughput of other
548 aggregate values. For example, the per_byte option is now equivalent
549 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
551 The per_rcpt option has been generalized using the /count= mechanism
552 (though it's more complicated than the per_byte equivalence). When it is
553 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
554 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
555 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
556 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
557 Note that using per_rcpt with a non-readonly update mode in more than
558 one ACL will cause the recipients to be double-counted. (The per_mail
559 and per_byte options don't have this problem.)
561 The handling of very low rates has changed slightly. If the computed rate
562 is less than the event's count (usually one) then this event is the first
563 after a long gap. In this case the rate is set to the same as this event's
564 count, so that the first message of a spam run is counted properly.
566 The major new feature is a mechanism for counting the rate of unique
567 events. The new per_addr option counts the number of different
568 recipients that someone has sent messages to in the last time period. It
569 behaves like per_rcpt if all the recipient addresses are different, but
570 duplicate recipient addresses do not increase the measured rate. Like
571 the /count= option this is a general mechanism, so the per_addr option
572 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
573 example, measure the rate that a client uses different sender addresses
574 with the options per_mail/unique=$sender_address. There are further
575 details in the main documentation.
577 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
579 TF/06 Removed a few PCRE remnants.
581 TF/07 Automatically extract Exim's version number from tags in the git
582 repository when doing development or release builds.
584 PP/02 Raise smtp_cmd_buffer_size to 16kB.
585 Bugzilla 879. Patch from Paul Fisher.
587 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
588 Heavily based on revision 40f9a89a from Simon Arlott's tree.
591 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
593 PP/05 Variable $av_failed, true if the AV scanner deferred.
594 Bugzilla 1078. Patch from John Horne.
596 PP/06 Stop make process more reliably on build failure.
597 Bugzilla 1087. Patch from Heiko Schlittermann.
599 PP/07 Make maildir_use_size_file an _expandable_ boolean.
600 Bugzilla 1089. Patch from Heiko Schlittermann.
602 PP/08 Handle ${run} returning more data than OS pipe buffer size.
603 Bugzilla 1131. Patch from Holger Weiß.
605 PP/09 Handle IPv6 addresses with SPF.
606 Bugzilla 860. Patch from Wolfgang Breyha.
608 PP/10 GnuTLS: support TLS 1.2 & 1.1.
610 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
613 PP/11 match_* no longer expand right-hand-side by default.
614 New compile-time build option, EXPAND_LISTMATCH_RHS.
615 New expansion conditions, "inlist", "inlisti".
617 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
619 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
621 PP/14 fix log_write() format string regression from TF/03.
622 Bugzilla 1152. Patch from Dmitry Isaikin.
628 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
630 PP/02 Harmonised TLS library version reporting; only show if debugging.
631 Layout now matches that introduced for other libraries in 4.74 PP/03.
633 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
635 PP/04 New "dns_use_edns0" global option.
637 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
640 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
641 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
643 TK/01 Updated PolarSSL code to 0.14.2.
644 Bugzilla 1097. Patch from Andreas Metzler.
646 PP/07 Catch divide-by-zero in ${eval:...}.
649 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
652 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
653 format-string attack -- SECURITY: remote arbitrary code execution.
655 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
656 time unintentionally subject to list matching rules, letting the header
657 cause arbitrary Exim lookups (of items which can occur in lists, *not*
658 arbitrary string expansion). This allowed for information disclosure.
660 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
661 INT_MIN/-1 -- value coerced to INT_MAX.
667 NM/01 Workround for PCRE version dependency in version reporting
670 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
671 This fixes portability to compilers other than gcc, notably
672 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
674 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
675 makefiles for portability to HP-UX and POSIX correctness.
677 PP/01 Permit LOOKUP_foo enabling on the make command-line.
678 Also via indented variable definition in the Makefile.
679 (Debugging by Oliver Heesakkers).
681 PP/02 Restore caching of spamd results with expanded spamd_address.
682 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
684 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
685 Improves build reliability. Fix from: Frank Elsner
687 NM/02 Fix wide character breakage in the rfc2047 coding
688 Fixes bug 1064. Patch from Andrey N. Oktyabrski
690 NM/03 Allow underscore in dnslist lookups
691 Fixes bug 1026. Patch from Graeme Fowler
693 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
694 Code patches from Adam Ciarcinski of NetBSD.
696 NM/04 Fixed exiqgrep to cope with mailq missing size issue
699 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
700 is logged, to avoid truncation. Patch from John Horne.
702 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
703 Patch from Jakob Hirsch.
705 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
706 SQL string expansion failure details.
707 Patch from Andrey Oktyabrski.
709 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
710 Patch from Simon Arlott.
712 PP/09 New lookups functionality failed to compile on old gcc which rejects
713 extern declarations in function scope.
714 Patch from Oliver Fleischmann
716 PP/10 Use sig_atomic_t for flags set from signal handlers.
717 Check getgroups() return and improve debugging.
718 Fixed developed for diagnosis in bug 927 (which turned out to be
721 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
722 Patch from Mark Zealey.
724 PP/12 Bugzilla 1056: Improved spamd server selection.
725 Patch from Mark Zealey.
727 PP/13 Bugzilla 1086: Deal with maildir quota file races.
728 Based on patch from Heiko Schlittermann.
730 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
731 Patch from Uwe Doering, sign-off by Michael Haardt.
733 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
734 variable declaration deep within a block. Bug and patch from
737 PP/15 lookups-Makefile IRIX compatibilty coercion.
739 PP/16 Make DISABLE_DKIM build knob functional.
741 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
742 Patch by Simon Arlott
744 TF/03 Fix valgrind.h portability to C89 compilers that do not support
745 variable argument macros. Our copy now differs from upstream.
751 TF/01 Failure to get a lock on a hints database can have serious
752 consequences so log it to the panic log.
754 TF/02 Log LMTP confirmation messages in the same way as SMTP,
755 controlled using the smtp_confirmation log selector.
757 TF/03 Include the error message when we fail to unlink a spool file.
759 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
760 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
761 for maintaining out-of-tree patches for some time.
763 PP/01 Bugzilla 139: Documentation and portability issues.
764 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
765 Handle per-OS dynamic-module compilation flags.
767 PP/02 Let /dev/null have normal permissions.
768 The 4.73 fixes were a little too stringent and complained about the
769 permissions on /dev/null. Exempt it from some checks.
770 Reported by Andreas M. Kirchwitz.
772 PP/03 Report version information for many libraries, including
773 Exim version information for dynamically loaded libraries. Created
774 version.h, now support a version extension string for distributors
775 who patch heavily. Dynamic module ABI change.
777 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
778 privilege escalation vulnerability whereby the Exim run-time user
779 can cause root to append content of the attacker's choosing to
782 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
785 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
786 If dropping privileges for untrusted macros, we disabled normal logging
787 on the basis that it would fail; for the Exim run-time user, this is not
788 the case, and it resulted in successful deliveries going unlogged.
789 Fixed. Reported by Andreas Metzler.
795 PP/01 Date: & Message-Id: revert to normally being appended to a message,
796 only prepend for the Resent-* case. Fixes regression introduced in
797 Exim 4.70 by NM/22 for Bugzilla 607.
799 PP/02 Include check_rfc2047_length in configure.default because we're seeing
800 increasing numbers of administrators be bitten by this.
802 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
804 PP/03 Bugzilla 994: added openssl_options main configuration option.
806 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
808 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
810 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
812 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
813 without a peer certificate, leading to a segfault because of an
814 assumption that peers always have certificates. Be a little more
815 paranoid. Problem reported by Martin Tscholak.
817 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
818 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
819 NB: ClamAV planning to remove STREAM in "middle of 2010".
820 CL also introduces -bmalware, various -d+acl logging additions and
821 more caution in buffer sizes.
823 PP/09 Implemented reverse_ip expansion operator.
825 PP/10 Bugzilla 937: provide a "debug" ACL control.
827 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
829 PP/12 Bugzilla 973: Implement --version.
831 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
833 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
835 PP/15 Bugzilla 816: support multiple condition rules on Routers.
837 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
838 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
839 ignore trailing whitespace.
841 JJ/02 prevent non-panic DKIM error from being sent to paniclog
843 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
846 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
847 Notification from Dr Andrew Aitchison.
849 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
850 ExtendedDetectionInfo response format.
851 Notification from John Horne.
853 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
856 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
857 XSL and documented dependency on system catalogs, with examples of how
860 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
863 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
864 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
865 configuration file which is writeable by the Exim user or group.
867 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
868 of configuration files to cover files specified with the -C option if
869 they are going to be used with root privileges, not just the default
872 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
873 option (effectively making it always true).
875 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
876 files to be used while preserving root privileges.
878 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
879 that rogue child processes cannot use them.
881 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
882 run-time user, instead of root.
884 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
885 Exim run-time user without dropping privileges.
887 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
888 result string, instead of calling string_vformat() twice with the same
891 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
892 for other users. Others should always drop root privileges if they use
893 -C on the command line, even for a whitelisted configure file.
895 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
897 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
903 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
904 $data_path, and $header_path variables; fixed documentation bugs and
907 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
908 exipick to access non-standard spools, including the "frozen" queue
911 NM/01 Bugzilla 965: Support mysql stored procedures.
912 Patch from Alain Williams
914 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
916 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
917 Patch from Andreas Metzler
919 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
920 Patch from Kirill Miazine
922 NM/05 Bugzilla 671: Added umask to procmail example.
924 JJ/03 installed exipick 20100323.0, fixing doc bug
926 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
927 directory. Notification and patch from Dan Rosenberg.
929 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
931 TK/02 Improve log output when DKIM signing operation fails.
933 MH/01 Treat the transport option dkim_domain as a colon separated
934 list, not as a single string, and sign the message with each element,
935 omitting multiple occurences of the same signer.
937 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
938 Bugzilla 985, 986. Patch by Simon Arlott
940 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
941 Patch by Simon Arlott
943 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
944 MBX locking. Notification from Dan Rosenberg.
950 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
952 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
954 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
956 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
958 NM/04 Bugzilla 915: Flag broken perl installation during build.
964 TK/01 Added patch by Johannes Berg that expands the main option
965 "spamd_address" if it starts with a dollar sign.
967 TK/02 Write list of recipients to X-Envelope-Sender header when building
968 the mbox-format spool file for content scanning (suggested by Jakob
971 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
972 (http://www.dcc-servers.net/) support via dccifd. Activated by
973 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
975 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
976 by Mark Daniel Reidel <mr@df.eu>.
978 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
979 When building exim an external PCRE library is now needed -
980 PCRE is a system library on the majority of modern systems.
981 See entry on PCRE_LIBS in EDITME file.
983 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
984 conversation. Added nologin parameter to request.
985 Patch contributed by Kirill Miazine.
987 TF/01 Do not log submission mode rewrites if they do not change the address.
989 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
991 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
992 log files in place. Contributed by Roberto Lima.
994 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
996 TF/03 Bugzilla 615: When checking the local_parts router precondition
997 after a local_part_suffix or local_part_prefix option, Exim now
998 does not use the address's named list lookup cache, since this
999 contains cached lookups for the whole local part.
1001 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1002 Robert Millan. Documentation is in experimental-spec.txt.
1004 TF/04 Bugzilla 668: Fix parallel build (make -j).
1006 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1008 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1009 Patch provided by Jan Srzednicki.
1011 TF/05 Leading white space used to be stripped from $spam_report which
1012 wrecked the formatting. Now it is preserved.
1014 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1015 that they are available at delivery time.
1017 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1019 TF/08 TLS error reporting now respects the incoming_interface and
1020 incoming_port log selectors.
1022 TF/09 Produce a more useful error message if an SMTP transport's hosts
1023 setting expands to an empty string.
1025 NM/06 Bugzilla 744: EXPN did not work under TLS.
1026 Patch provided by Phil Pennock.
1028 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1029 Patch provided by Richard Godbee.
1031 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1032 acl_smtp_notquit, added index entry.
1034 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1035 Patch provided by Eugene Bujak.
1037 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1038 accept(). Patch provided by Maxim Dounin.
1040 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1041 Patch provided by Phil Pennock.
1043 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1045 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1046 Patch provided by Brad "anomie" Jorsch.
1048 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1049 Patch provided by Dean Brooks.
1051 TK/05 Add native DKIM support (does not depend on external libraries).
1053 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1054 Patch provided by Graeme Fowler.
1056 NM/16 Bugzilla 851: Documentation example syntax fix.
1058 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1060 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1063 NM/19 Bugzilla 745: TLS version reporting.
1064 Patch provided by Phil Pennock.
1066 NM/20 Bugzilla 167: bool: condition support.
1067 Patch provided by Phil Pennock.
1069 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1070 clients. Patch provided by Phil Pennock.
1072 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1073 Patch provided by Brad "anomie" Jorsch.
1075 NM/23 Bugzilla 687: Fix misparses in eximstats.
1076 Patch provided by Heiko Schlittermann.
1078 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1079 Patch provided by Heiko Schlittermann.
1081 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1082 plus update to original patch.
1084 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1086 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1087 Patch provided by David Brownlee.
1089 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1091 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1093 NM/30 Bugzilla 888: TLS documentation bugfixes.
1095 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1097 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1098 Unlike the original bugzilla I have changed all shell scripts in src tree.
1100 NM/33 Bugzilla 898: Transport filter timeout fix.
1101 Patch by Todd Rinaldo.
1103 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1104 Patch by Serge Demonchaux.
1106 NM/35 Bugzilla 39: Base64 decode bug fixes.
1107 Patch by Jakob Hirsch.
1109 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1111 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1113 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1115 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1121 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1122 ALT-N's libdkim that I have put here:
1123 http://duncanthrax.net/exim-experimental/
1125 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1126 were called 'true' and it seems that C99 defines that as a reserved
1127 keyword to be used with 'bool' variable types. That means you could
1128 not include C99-style headers which use bools without triggering
1129 build errors in sieve.c.
1131 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1132 as mailq or other aliases. Changed the --help handling significantly
1133 to do whats expected. exim_usage() emits usage/help information.
1135 SC/01 Added the -bylocaldomain option to eximstats.
1137 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1139 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1141 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1147 PH/01 Another patch from the Sieve maintainer.
1149 PH/02 When an IPv6 address is converted to a string for single-key lookup
1150 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1151 dots are used instead of colons so that keys in lsearch files need not
1152 contain colons. This was done some time before quoting was made available
1153 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1154 (notated using the quote facility) so as to distinguish them from IPv4
1155 keys. This meant that lookups for IP addresses in host lists did not work
1156 for iplsearch lookups.
1158 This has been fixed by arranging for IPv6 addresses to be expressed with
1159 colons if the lookup type is iplsearch. This is not incompatible, because
1160 previously such lookups could never work.
1162 The situation is now rather anomolous, since one *can* have colons in
1163 ordinary lsearch keys. However, making the change in all cases is
1164 incompatible and would probably break a number of configurations.
1166 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1169 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1170 conversion specification without a maximum field width, thereby enabling
1171 a rogue spamd server to cause a buffer overflow. While nobody in their
1172 right mind would setup Exim to query an untrusted spamd server, an
1173 attacker that gains access to a server running spamd could potentially
1174 exploit this vulnerability to run arbitrary code as the Exim user.
1176 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1177 $primary_hostname instead of what libspf2 thinks the hosts name is.
1179 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1180 a directory entry by the name of the lookup key. Previously, if a
1181 symlink pointed to a non-existing file or a file in a directory that
1182 Exim lacked permissions to read, a lookup for a key matching that
1183 symlink would fail. Now it is enough that a matching directory entry
1184 exists, symlink or not. (Bugzilla 503.)
1186 PH/03 The body_linecount and body_zerocount variables are now exported in the
1189 PH/04 Added the $dnslist_matched variable.
1191 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1192 This means they are set thereafter only if the connection becomes
1195 PH/06 Added the client_condition to authenticators so that some can be skipped
1196 by clients under certain conditions.
1198 PH/07 The error message for a badly-placed control=no_multiline_responses left
1199 "_responses" off the end of the name.
1201 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1203 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1204 (without spaces) instead of just copying the configuration text.
1206 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1208 PH/11 Added $max_received_linelength.
1210 PH/12 Added +ignore_defer and +include_defer to host lists.
1212 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1213 way in which PCRE > 7.0 is built.
1215 PH/14 Implemented queue_only_load_latch.
1217 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1218 MAIL command. The effect was to mangle the value on 64-bit systems.
1220 PH/16 Another patch from the Sieve maintainer.
1222 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1224 PH/18 If a system quota error occurred while trying to create the file for
1225 a maildir delivery, the message "Mailbox is full" was not appended to the
1226 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1227 only to a quota excession during the actual writing of the file.
1229 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1230 characters?) which causes problems in log lines. The DN values are now
1231 passed through string_printing() before being added to log lines.
1233 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1234 and InterBase are left for another time.)
1236 PH/21 Added message_body_newlines option.
1238 PH/22 Guard against possible overflow in moan_check_errorcopy().
1240 PH/23 POSIX allows open() to be a macro; guard against that.
1242 PH/24 If the recipient of an error message contained an @ in the local part
1243 (suitably quoted, of course), incorrect values were put in $domain and
1244 $local_part during the evaluation of errors_copy.
1250 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1251 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1254 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1255 issue a MAIL command.
1257 PH/02 In an ACL statement such as
1259 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1261 if a client was not listed at all, or was listed with a value other than
1262 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1263 the condition was not true (as it should be), so access was not denied.
1264 The bug was that the ! inversion was incorrectly passed on to the second
1265 item. This has been fixed.
1267 PH/03 Added additional dnslists conditions == and =& which are different from
1268 = and & when the dns lookup returns more than one IP address.
1270 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1271 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1273 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1274 FSYNC, which compiles an option called disable_fsync that allows for
1275 bypassing fsync(). The documentation is heavily laced with warnings.
1277 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1279 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1280 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1281 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1282 including adding "make clean"; (3) Added -fPIC when compiling the test
1283 dynamically loaded module, to get rid of a warning.
1285 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1286 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1287 = 0s. The bug is otherwise harmless.
1289 PH/07 There was a bug in the dovecot authenticator such that the value of
1290 $auth1 could be overwritten, and so not correctly preserved, after a
1291 successful authentication. This usually meant that the value preserved by
1292 the server_setid option was incorrect.
1294 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1296 PH/09 Installed PCRE release 7.0.
1298 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1299 run for batched SMTP input. It is now run at the start of every message
1300 in the batch. While fixing this I discovered that the process information
1301 (output by running exiwhat) was not always getting set for -bs and -bS
1302 input. This is fixed, and it now also says "batched" for BSMTP.
1304 PH/11 Added control=no_pipelining.
1306 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1307 patch, slightly modified), and move the expansion of helo_data till after
1308 the connection is made in the smtp transport (so it can use these
1311 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1313 PH/14 Added log_selector = +pid.
1315 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1317 PH/16 Add ${if forany and ${if forall.
1319 PH/17 Added dsn_from option to vary the From: line in DSNs.
1321 PH/18 Flush SMTP output before performing a callout, unless control =
1322 no_callout_flush is set.
1324 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1325 was true (the default) a successful delivery failed to delete the retry
1326 item, thus causing premature timeout of the address. The bug is now
1329 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1331 PH/21 Long custom messages for fakedefer and fakereject are now split up
1332 into multiline reponses in the same way that messages for "deny" and
1333 other ACL rejections are.
1335 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1336 with slight modification.
1338 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1339 draft, changing the syntax and factoring some duplicate code".
1341 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1342 for deliveries of the second and subsequent messages over the same SMTP
1345 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1346 ${reduce, with only minor "tidies".
1348 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1350 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1351 expansion side effects.
1353 PH/27 When a message times out after an over-quota error from an Exim-imposed
1354 quota, the bounce message says "mailbox is full". This message was not
1355 being given when it was a system quota that was exceeded. It now should
1358 MH/03 Made $recipients available in local_scan(). local_scan() already has
1359 better access to the recipient list through recipients_list[], but
1360 $recipients can be useful in postmaster-provided expansion strings.
1362 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1363 in the case of a MAIL command with additional options following the
1364 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1365 were accidentally chopped off.
1367 PH/29 SMTP synchronization checks are implemented when a command is read -
1368 there is a check that no more input is waiting when there shouldn't be
1369 any. However, for some commands, a delay in an ACL can mean that it is
1370 some time before the response is written. In this time, more input might
1371 arrive, invalidly. So now there are extra checks after an ACL has run for
1372 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1373 pipelining has not been advertised.
1375 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1377 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1378 returned in the SMTP response when smtp_return_error_details was false.
1379 This has been fixed.
1381 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1382 instead of the C I/O that was originally supplied, because problems were
1383 reported on Solaris.
1385 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1386 Exim which did not show up earlier: it was assuming that a call to
1387 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1388 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1389 was a macro that became an assignment, so it seemed to work. This has
1390 changed to a proper function call with a void return, hence the compile
1391 error. Exim's code has been fixed.
1393 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1396 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1398 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1400 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1401 because it turns out to be too big or there is a timeout) there is no
1402 "Completed" line in the log. When some messages of this type were
1403 selected by exigrep, they were listed as "not completed". Others were
1404 picked up by some special patterns. I have improved the selection
1405 criteria to be more general.
1407 PH/38 The host_find_failed option in the manualroute router can now be set
1408 to "ignore", to completely ignore a host whose IP address cannot be
1409 found. If all hosts are ignored, the behaviour is controlled by the new
1410 host_all_ignored option.
1412 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1413 homing or because of multiple MX records with /mx) generated more than
1414 one IP address, and the following item turned out to be the local host,
1415 all the secondary addresses of the first item were incorrectly removed
1416 from the list, along with the local host and any following hosts (which
1417 is what is supposed to happen).
1419 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1420 whoever called Exim into the -H file. In the case of the daemon it was
1421 behaving confusingly. When first started, it used values for whoever
1422 started the daemon, but after a SIGHUP it used the Exim user (because it
1423 calls itself on a restart). I have changed the code so that it now always
1426 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1427 message are rejected with the same error (e.g. no authentication or bad
1428 sender address), and a DATA command is nevertheless sent (as can happen
1429 with PIPELINING or a stupid MUA), the error message that was given to the
1430 RCPT commands is included in the rejection of the DATA command. This is
1431 intended to be helpful for MUAs that show only the final error to their
1434 PH/42 Another patch from the Sieve maintainer.
1436 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1437 Eximstats - Fixed some broken HTML links and added missing column headers
1439 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1440 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1442 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1444 PH/43 Yet another patch from the Sieve maintainer.
1446 PH/44 I found a way to check for a TCP/IP connection going away before sending
1447 the response to the final '.' that terminates a message, but only in the
1448 case where the client has not sent further data following the '.'
1449 (unfortunately, this is allowed). However, in many cases there won't be
1450 any further data because there won't be any more messages to send. A call
1451 to select() can be used: if it shows that the input is "ready", there is
1452 either input waiting, or the socket has been closed. An attempt to read
1453 the next input character can distinguish the two cases. Previously, Exim
1454 would have sent an OK response which the client would never have see.
1455 This could lead to message repetition. This fix should cure that, at
1456 least in a lot of common cases.
1458 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1459 advertised in response to EHLO.
1465 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1466 fixed by 4.65/MH/01 (is this a record?) are fixed:
1468 (i) An empty string was always treated as zero by the numeric comparison
1469 operators. This behaviour has been restored.
1471 (ii) It is documented that the numeric comparison operators always treat
1472 their arguments as decimal numbers. This was broken in that numbers
1473 starting with 0 were being interpreted as octal.
1475 While fixing these problems I realized that there was another issue that
1476 hadn't been noticed. Values of message_size_limit (both the global option
1477 and the transport option) were treated as octal if they started with 0.
1478 The documentation was vague. These values are now always treated as
1479 decimal, and I will make that clear in the documentation.
1485 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1486 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1489 MH/01 Don't check that the operands of numeric comparison operators are
1490 integers when their expansion is in "skipping" mode (fixes bug
1491 introduced by 4.64-PH/07).
1493 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1494 child addresses, Exim now panics and dies. Previously, because the count
1495 is held in a short int, deliveries were likely to be lost. As such a
1496 large number of recipients for a single message is ridiculous
1497 (performance will be very, very poor), I have chosen to impose a limit
1498 rather than extend the field.
1504 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1505 leftover -K file (the existence of which was triggered by #402).
1506 While we were at it, introduced process PID as part of the -K
1507 filename. This should rule out race conditions when creating
1510 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1511 processing considerably. Previous code took too long for large mails,
1512 triggering a timeout which in turn triggers #401.
1514 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1515 in the DK code in transports.c. sendfile() is not really portable,
1516 hence the _LINUX specificness.
1518 TF/01 In the add_headers option to the mail command in an Exim filter,
1519 there was a bug that Exim would claim a syntax error in any
1520 header after the first one which had an odd number of characters
1523 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1524 callout verification, Exim cached a "reject" for the entire domain. This
1525 is correct for most verifications, but it is not correct for a recipient
1526 verification with use_sender or use_postmaster set, because in that case
1527 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1528 case of MAIL FROM:<> rejection from other early rejections (e.g.
1529 rejection of HELO). When verifying a recipient using a non-null MAIL
1530 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1531 Whatever the result of the callout, the value of the domain cache is
1532 left unchanged (for any other kind of callout, getting as far as trying
1533 RCPT means that the domain itself is ok).
1535 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1538 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1539 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1540 session, the daemon ignores ECONNECT errors and logs others; it now
1541 ignores EPIPE as well.
1543 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1544 (quoted-printable decoding).
1546 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1547 later the small subsequent patch to fix an introduced bug.
1549 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1551 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1553 PH/08 An error is now given if message_size_limit is specified negative.
1555 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1556 to be given (somewhat) arbitrary names.
1558 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1561 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1562 miscellaneous code fixes
1564 PH/10 Added the log_reject_target ACL modifier to specify where to log
1567 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1568 hostname. This is wrong, because it relates to the incoming message (and
1569 probably the interface on which it is arriving) and not to the outgoing
1570 callout (which could be using a different interface). This has been
1571 changed to use the value of the helo_data option from the smtp transport
1572 instead - this is what is used when a message is actually being sent. If
1573 there is no remote transport (possible with a router that sets up host
1574 addresses), $smtp_active_hostname is used.
1576 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1577 tweaks were necessary in order to get it to work (see also 21 below):
1578 (a) The code assumed that strncpy() returns a negative number on buffer
1579 overflow, which isn't the case. Replaced with Exim's string_format()
1581 (b) There were several signed/unsigned issues. I just did the minimum
1582 hacking in of casts. There is scope for a larger refactoring.
1583 (c) The code used strcasecmp() which is not a standard C function.
1584 Replaced with Exim's strcmpic() function.
1585 (d) The code set only $1; it now sets $auth1 as well.
1586 (e) A simple test gave the error "authentication client didn't specify
1587 service in request". It would seem that Dovecot has changed its
1588 interface. Fortunately there's a specification; I followed it and
1589 changed what the client sends and it appears to be working now.
1591 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1594 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1595 address A is aliased to B and C, where B exists and C does not. Without
1596 -v the output is "A verified" because verification stops after a
1597 successful redirection if more than one address is generated. However,
1598 with -v the child addresses are also verified. Exim was outputting "A
1599 failed to verify" and then showing the successful verification for C,
1600 with its parentage. It now outputs "B failed to verify", showing B's
1601 parentage before showing the successful verification of C.
1603 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1604 look up a TXT record in a specific list after matching in a combined
1607 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1608 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1609 they consult the DNS. I had assumed they would set it the way they
1610 wanted; and indeed my experiments on Linux seem to show that in some
1611 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1612 To be on the safe side, however, I have now made the interface to
1613 host_find_byname() similar to host_find_bydns(), with an argument
1614 containing the DNS resolver options. The host_find_byname() function now
1615 sets these options at its start, just as host_find_bydns() does. The smtp
1616 transport options dns_qualify_single and dns_search_parents are passed to
1617 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1618 of host_find_byname() use the default settings of RES_DEFNAMES
1619 (qualify_single) but not RES_DNSRCH (search_parents).
1621 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1622 spool_read_header() do less string testing, by means of a preliminary
1623 switch on the second character of optional "-foo" lines. (This is
1624 overdue, caused by the large number of possibilities that now exist.
1625 Originally there were few.) While I was there, I also converted the
1626 str(n)cmp tests so they don't re-test the leading "-" and the first
1627 character, in the hope this might squeeze out yet more improvement.
1629 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1630 flag allowing group syntax was set by the header_syntax check but not
1631 turned off, possible causing trouble later; (2) The flag was not being
1632 set at all for the header_verify test, causing "group"-style headers to
1633 be rejected. I have now set it in this case, and also caused header_
1634 verify to ignore an empty address taken from a group. While doing this, I
1635 came across some other cases where the code for allowing group syntax
1636 while scanning a header line wasn't quite right (mostly, not resetting
1637 the flag correctly in the right place). These bugs could have caused
1638 trouble for malformed header lines. I hope it is now all correct.
1640 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1641 with the "reply" argument non-NULL. The code, however (which originally
1642 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1643 but it didn't always do it. This confused somebody who was copying the
1644 code for some other use. I have removed all the tests.
1646 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1647 feature that was used to support insecure browsers during the U.S. crypto
1648 embargo. It requires special client support, and Exim is probably the
1649 only MTA that supported it -- and would never use it because real RSA is
1650 always available. This code has been removed, because it had the bad
1651 effect of slowing Exim down by computing (never used) parameters for the
1652 RSA_EXPORT functionality.
1654 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1655 authenticator to fail if there's a tab character in the incoming data
1656 (there should never be unless someone is messing about, as it's supposed
1657 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1658 if the connection is using TLS or if the remote IP is the same as the
1659 local IP, and the "valid-client-cert option" if a client certificate has
1662 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1663 authenticators. This can be used for authorization after authentication
1664 succeeds. (In the case of plaintext, it servers for both authentication
1667 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1668 if any retry times were supplied.
1670 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1671 connection, where there is no client IP address to check. In this
1672 situation, the verify now always succeeds.
1674 PH/25 Applied John Jetmore's -Mset patch.
1676 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1678 PH/27 In a string expansion for a processed (not raw) header when multiple
1679 headers of the same name were present, leading whitespace was being
1680 removed from all of them, but trailing whitespace was being removed only
1681 from the last one. Now trailing whitespace is removed from each header
1682 before concatenation. Completely empty headers in a concatenation (as
1683 before) are ignored.
1685 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1686 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1688 PH/29 [Removed. This was a change that I later backed out, and forgot to
1689 correct the ChangeLog entry (that I had efficiently created) before
1690 committing the later change.]
1692 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1693 address errors (4xx response to RCPT) over the same connection as other
1694 messages routed to the same hosts. Such deliveries are always "forced",
1695 so retry times are not inspected. This resulted in far too many retries
1696 for the affected addresses. The effect occurred only when there were more
1697 hosts than the hosts_max_try setting in the smtp transport when it had
1698 the 4xx errors. Those hosts that it had tried were not added to the list
1699 of hosts for which the message was waiting, so if all were tried, there
1700 was no problem. Two fixes have been applied:
1702 (i) If there are any address or message errors in an SMTP delivery, none
1703 of the hosts (tried or untried) are now added to the list of hosts
1704 for which the message is waiting, so the message should not be a
1705 candidate for sending over the same connection that was used for a
1706 successful delivery of some other message. This seems entirely
1707 reasonable: after all the message is NOT "waiting for some host".
1708 This is so "obvious" that I'm not sure why it wasn't done
1709 previously. Hope I haven't missed anything, but it can't do any
1710 harm, as the worst effect is to miss an optimization.
1712 (ii) If, despite (i), such a delivery is accidentally attempted, the
1713 routing retry time is respected, so at least it doesn't keep
1714 hammering the server.
1716 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1717 in ${readsocket because some servers need this prod.
1719 PH/32 Added some extra debug output when updating a wait-xxx database.
1721 PH/33 The hint "could be header name not terminated by colon", which has been
1722 given for certain expansion errors for a long time, was not being given
1723 for the ${if def:h_colon_omitted{... case.
1725 PH/34 The spec says: "With one important exception, whenever a domain list is
1726 being scanned, $domain contains the subject domain." There was at least
1727 one case where this was not true.
1729 PH/35 The error "getsockname() failed: connection reset by peer" was being
1730 written to the panic log as well as the main log, but it isn't really
1731 panic-worthy as it just means the connection died rather early on. I have
1732 removed the panic log writing for the ECONNRESET error when getsockname()
1735 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1736 runs only) independently of the message's sender address. This meant
1737 that, if the 4xx error was in fact related to the sender, a different
1738 message to the same recipient with a different sender could confuse
1739 things. In particualar, this can happen when sending to a greylisting
1740 server, but other circumstances could also provoke similar problems.
1741 I have changed the default so that the retry time for these errors is now
1742 based a combination of the sender and recipient addresses. This change
1743 can be overridden by setting address_retry_include_sender=false in the
1746 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1747 remote server are returned as part of bounce messages. This was not
1748 happening for LMTP over a pipe (the lmtp transport), but now it is the
1749 same for both kinds of LMTP.
1751 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1752 in header lines that were in fact CNAMEs. This is no longer the case.
1754 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1755 and queue runs started by the daemon processed all messages. This has
1756 been fixed so that -R and -S can now usefully be given with -q<time>.
1758 PH/40 Import PCRE release 6.7 (fixes some bugs).
1760 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1762 PH/42 Give an error if -q is specified more than once.
1764 PH/43 Renamed the variables $interface_address and $interface_port as
1765 $received_ip_address and $received_port, to make it clear that these
1766 values apply to message reception, and not to the outgoing interface when
1767 a message is delivered. (The old names remain recognized, of course.)
1769 PH/44 There was no timeout on the connect() call when using a Unix domain
1770 socket in the ${readsocket expansion. There now is.
1772 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1773 be meaningful with "accept".
1775 SC/01 Eximstats V1.43
1776 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1778 SC/02 Eximstats V1.44
1779 Use a glob alias rather than an array ref in the generated
1780 parser. This improves both readability and performance.
1782 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1783 Collect SpamAssassin and rejection statistics.
1784 Don't display local sender or destination tables unless
1785 there is data to show.
1786 Added average volumes into the top table text output.
1788 SC/04 Eximstats V1.46
1789 Collect data on the number of addresses (recipients)
1790 as well as the number of messages.
1792 SC/05 Eximstats V1.47
1793 Added 'Message too big' to the list of mail rejection
1794 reasons (thanks to Marco Gaiarin).
1796 SC/06 Eximstats V1.48
1797 Mainlog lines which have GMT offsets and are too short to
1798 have a flag are now skipped.
1800 SC/07 Eximstats V1.49 (Alain Williams)
1801 Added the -emptyok flag.
1803 SC/08 Eximstats V1.50
1804 Fixes for obtaining the IP address from reject messages.
1806 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1807 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1808 whitesspace changes from 4.64-PH/27
1810 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1813 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1814 are found, allow negative numbers in numeric criteria)
1816 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1818 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1819 to match changes made in 4.64-PH/43
1821 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1823 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1824 "domain" is actually an IP address, and force a failure. This locks out
1825 those revolvers/nameservers that support "A-for-A" lookups, in
1826 contravention of the specifications.
1828 PH/48 When a host name was looked up from an IP address, and the subsequent
1829 forward lookup of the name timed out, the host name was left in
1830 $sender_host_name, contrary to the specification.
1832 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1833 restricted to single-key lookups, Exim was not diagnosing an error if
1834 * or *@ was used with a query-style lookup.
1836 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1838 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1839 long ago, but noone interested enough thought of it. Let's just say that
1840 the "1.1" means that there are some new functions that weren't there at
1841 some point in the past.
1843 PH/51 Error processing for expansion failure of helo_data from an smtp
1844 transport during callout processing was broken.
1846 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1847 tested/used via the -bh/-bhc/-bs options.
1849 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1850 bug, fixed in subsequent PCRE releases).
1852 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1853 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1855 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1861 SC/01 Use a glob alias rather than an array ref in eximstats generated
1862 parser. This improves both readability and performance.
1864 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1865 Don't display local sender or destination tables in eximstats unless
1866 there is data to show.
1867 Added average volumes into the eximstats top table text output.
1869 SC/03 Collect data on the number of addresses (recipients) as well
1870 as the number of messages in eximstats.
1872 TF/01 Correct an error in the documentation for the redirect router. Exim
1873 does (usually) call initgroups() when daemonizing.
1875 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1876 with consistent privilege compared to when running as a daemon.
1878 TF/03 Note in the spec that $authenticated_id is not set for local
1879 submissions from trusted users.
1881 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1882 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1884 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1885 by adding some example configuration directives to the default
1886 configuration file. A little bit of work is required to uncomment the
1887 directives and define how usernames and passwords are checked, but
1888 there is now a framework to start from.
1890 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1891 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1892 without this. I don't know how relevant this is to other LDAP libraries.
1894 PH/02 Add the verb name to the "unknown ACL verb" error.
1896 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1898 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1900 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1901 directory not expanded when it should be if an expanded home directory
1902 was set for the address (which is overridden by the transport).
1904 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1907 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1908 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1909 because it is too late at that time, and has no effect.
1911 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1912 security issue with \' (bugzilla #107). I could not use the
1913 PQescapeStringConn() function, because it needs a PGconn value as one of
1916 PH/08 When testing addresses using -bt, indicate those final addresses that
1917 are duplicates that would not cause an additional delivery. At least one
1918 person was confused, thinking that -bt output corresponded to deliveries.
1919 (Suppressing duplicates isn't a good idea as you lose the information
1920 about possibly different redirections that led to the duplicates.)
1922 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1923 systems where poll() doesn't work, in particular OS X.
1925 PH/10 Added more information to debugging output for retry time not reached.
1927 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1928 operations in malware.c.
1930 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1933 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1934 syslog_duplication set false, log lines that would normally be written
1935 both the the main log and to the reject log were not written to syslog at
1938 PH/14 In the default configuration, change the use of "message" in ACL warn
1939 statements to "add_header".
1941 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1942 not followed by a command (e.g. "seen endif").
1944 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1945 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1948 PH/17 Added extra conditions to the default value of delay_warning_condition
1952 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1953 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1954 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1957 The Auto-Submitted: and various List- headers are standardised, whereas I
1958 don't think Precedence: ever was.
1960 PH/18 Refactored debugging code in route_finduser() to show more information,
1961 in particular, the error code if getpwnam() issues one.
1963 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1964 This is apparently needed in addition to the PH/07 change above to avoid
1965 any possible encoding problems.
1967 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1968 but not after initializing Perl.
1970 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1971 output them only if debugging. By default they are written stderr,
1972 apparently, which is not desirable.
1974 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1977 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1980 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1982 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1983 authenticated or an ident call has been made. Suppress the default
1984 values for $authenticated_id and $authenticated_sender (but permit -oMai
1985 and -oMas) when testing with -bh.
1987 PH/24 Re-jigged the order of the tests in the default configuration so that the
1988 tests for valid domains and recipients precede the DNS black list and CSA
1989 tests, on the grounds that those ones are more expensive.
1991 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1992 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1993 This bug exists in every version of Exim that I still have, right back to
1996 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1997 However, an attempt to turn on case-sensitivity in a regex key by
1998 including (?-i) didn't work because the subject string was already
1999 lowercased, and the effects were non-intuitive. It turns out that a
2000 one-line patch can be used to allow (?-i) to work as expected.
2006 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2007 other effects) broke the use of negated acl sub-conditions.
2009 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2012 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2013 "Deny" causes Exim to reject the incoming connection with a 554 error.
2014 Unfortunately, if there is a major crisis, such as a disk failure,
2015 tcp-wrappers gives "deny", whereas what one would like would be some
2016 kind of temporary error. A kludge has been added to help with this.
2017 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2018 554 error is used if errno is still zero or contains ENOENT (which occurs
2019 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2022 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2024 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2025 errors. Otherwise a message that provokes a temporary error (when other
2026 messages do not) can cause a whole host to time out.
2028 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2029 addresses were routed directly to files or pipes from a redirect router.
2030 File deliveries just didn't batch; pipe deliveries might have suffered
2033 PH/06 A failure to get a lock for a hints database would erroneously always say
2034 "Failed to get write lock", even when it was really a read lock.
2036 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2037 of 0600. This has been changed to use the value of the lockfile_mode
2038 option (which defaults to 0600).
2040 PH/08 Applied small patch from the Sieve maintainer.
2042 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2043 folder from quota calculations, a direct delivery into this folder messed
2044 up the contents of the maildirsize file. This was because the regex was
2045 used only to exclude .Trash (or whatever) when the size of the mailbox
2046 was calculated. There was no check that a delivery was happening into an
2047 excluded directory. This bug has been fixed by ignoring all quota
2048 processing for deliveries into excluded directories.
2050 PH/10 Added the maildirfolder_create_regex option to appendfile.
2056 PH/01 The code for finding all the local interface addresses on a FreeBSD
2057 system running IPv6 was broken. This may well have applied to all BSD
2058 systems, as well as to others that have similar system calls. The broken
2059 code found IPv4 interfaces correctly, but gave incorrect values for the
2060 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2061 that it would not match correctly against @[] and not recognize the IPv6
2064 PH/02 The ipliteral router was not recognizing addresses of the form user@
2065 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2067 PH/03 Added disable_ipv6.
2069 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2070 decoded form, because it is most often used to construct To: headers
2071 lines in autoreplies, and the decoded form may well be syntactically
2072 invalid. However, $reply_address has leading white space removed, and all
2073 newlines turned into spaces so that the autoreply transport does not
2076 PH/05 If group was specified without a user on a router, and no group or user
2077 was specified on a transport, the group from the router was ignored.
2079 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2080 for visible compile-time settings that can be used to change these
2081 numbers, for those that want even more. Backwards compatibility with old
2082 spool files has been maintained. However, going back to a previous Exim
2083 release will lost any variables that are in spool files.
2085 PH/07 Two small changes when running in the test harness: increase delay when
2086 passing a TCP/IP connection to a new process, in case the original
2087 process has to generate a bounce, and remove special handling of
2088 127.0.0.2 (sic), which is no longer necessary.
2090 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2091 be the same on different OS.
2093 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2096 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2097 whether --show-vars was specified or not
2099 JJ/02 exipick: Added support for new ACL variable spool format introduced
2102 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2103 syntactically invalid From: or Reply-to: line, and a filter used this to
2104 generate an autoreply, and therefore failed to obtain an address for the
2105 autoreply, Exim could try to deliver to a non-existent relative file
2106 name, causing unrelated and misleading errors. What now happens is that
2107 it logs this as a hard delivery error, but does not attempt to create a
2110 PH/11 The exinext utility has a -C option for testing purposes, but although
2111 the given file was scanned by exinext itself; it wasn't being passed on
2112 when Exim was called.
2114 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2115 an end-of-file indication when reading a command response.
2117 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2118 compiled. In many other places in Exim, IPv6 addresses are always
2119 recognized, so I have changed this. It also means that IPv4 domain
2120 literals of the form [IPV4:n.n.n.n] are now always recognized.
2122 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2123 used if the router is not running as root, for example, when verifying at
2124 ACL time, or when using -bh. The debugging output from this situation was
2125 non-existent - all you got was a failure to exec. I have made two
2128 (a) Failures to set uid/gid, the current directory, or a process leader
2129 in a subprocess such as that created by queryprogram now generate
2130 suitable debugging ouput when -d is set.
2132 (b) The queryprogram router detects when it is not running as root,
2133 outputs suitable debugging information if -d is set, and then runs
2134 the subprocess without attempting to change uid/gid.
2136 PH/15 Minor change to Makefile for building test_host (undocumented testing
2139 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2140 additional section of a DNS packet that returns MX or SRV records.
2141 Instead, it always explicitly searches for A/AAAA records. This avoids
2142 major problems that occur when a DNS server includes only records of one
2143 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2144 fixed another bug: if SRV records were looked up and the corresponding
2145 address records were *not* found in the additional section, the port
2146 values from the SRV records were lost.
2148 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2149 using the correct key (the original address) when searching the retry
2150 rules in order to find which one to use for generating the retry hint.
2152 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2153 adding the default one. Similarly, if it contains a Reply-To: header, the
2154 errors_reply_to option, if set, is not used.
2156 PH/19 When calculating a retry time, Exim used to measure the "time since
2157 failure" by looking at the "first failed" field in the retry record. Now
2158 it does not use this if it is later than than the arrival time of the
2159 message. Instead it uses the arrival time. This makes for better
2160 behaviour in cases where some deliveries succeed, thus re-setting the
2161 "first failed" field. An example is a quota failure for a huge message
2162 when small messages continue to be delivered. Without this change, the
2163 "time since failure" will always be short, possible causing more frequent
2164 delivery attempts for the huge message than are intended.
2165 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2167 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2168 $1, $2, $3) because the numerical variables can be reset during some
2169 expansion items (e.g. "match"), thereby losing the authentication data.
2171 PH/21 Make -bV show the size of off_t variables so that the test suite can
2172 decide whether to run tests for quotas > 2G.
2174 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2175 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2176 filecount value is greater than 2G or if a quota value is greater than 2G
2177 on a system where the size of off_t is not greater than 4, a panic error
2180 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2181 never match. The debug and -bh output now contains an explicit error
2182 message indicating a malformed IPv4 address or mask.
2184 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2185 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2186 PH/23 above applies.
2188 PH/25 Do not write to syslog when running in the test harness. The only
2189 occasion when this arises is a failure to open the main or panic logs
2190 (for which there is an explicit test).
2192 PH/26 Added the /no_tell option to "control=freeze".
2194 PH/27 If a host name lookup failed very early in a connection, for example, if
2195 the IP address matched host_lookup and the reverse lookup yielded a name
2196 that did not have a forward lookup, an error message of the form "no IP
2197 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2198 could be logged. Now it outputs the IP address instead of "NULL".
2200 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2201 allows the sender and the authenticated sender to be set when
2202 submitting a message from within Exim. Since child_open_exim() is
2203 documented for local_scan(), the new function should be too.
2205 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2206 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2207 results in an empty string is now treated as unset.
2209 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2211 PH/31 Added sender_verify_fail logging option.
2213 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2214 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2215 tidied the source and removed it altogether.
2217 PH/33 When a queue run was abandoned because the load average was too high, a
2218 log line was always written; now it is written only if the queue_run log
2219 selector is set. In addition, the log line for abandonment now contains
2220 information about the queue run such as the pid. This is always present
2221 in "start" and "stop" lines but was omitted from the "abandon" line.
2223 PH/34 Omit spaces between a header name and the colon in the error message that
2224 is given when verify = headers_syntax fails (if there are lots of them,
2225 the message gets confusing).
2227 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2228 names, as there are now some PTR records that contain slashes. This check
2229 is only to protect against broken name servers that fall over on strange
2230 characters, so the fact that it applies to all lookups doesn't matter.
2232 PH/36 Now that the new test suite is complete, we can remove some of the
2233 special code in Exim that was needed for the old test suite. For example,
2234 sorting DNS records because real resolvers return them in an arbitrary
2235 order. The new test suite's fake resolver always returns records in the
2238 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2239 bounces) except when queue_only is set, to avoid logging races between
2240 the different processes.
2242 PH/38 Panic-die if .include specifies a non-absolute path.
2244 PH/39 A tweak to the "H" retry rule from its user.
2246 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2247 a label. They prevented compilation on older perls.
2249 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2250 a warning to be raised on newish perls.
2252 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2253 on queue. Changes to match documented behaviour of showing count of
2254 messages matching specified criteria.
2256 PH/40 Changed the default ident timeout from 30s to 5s.
2258 PH/41 Added support for the use of login_cap features, on those BSD systems
2259 that have them, for controlling the resources used by pipe deliveries.
2261 PH/42 The content-scanning code uses fopen() to create files in which to put
2262 message data. Previously it was not paying any attention to the mode of
2263 the files. Exim runs with umask(0) because the rest of the code creates
2264 files with open(), and sets the required mode explicitly. Thus, these
2265 files were ending up world-writeable. This was not a big issue, because,
2266 being within the spool directory, they were not world-accessible. I have
2267 created a function called modefopen, which takes an additional mode
2268 argument. It sets umask(777), creates the file, chmods it to the required
2269 mode, then resets the umask. All the relevant calls to fopen() in the
2270 content scanning code have been changed to use this function.
2272 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2273 to 24 hours. This avoids potential overflow problems when processing G
2274 and H retry rules. I suspect nobody ever tinkers with this value.
2276 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2278 PH/45 When the plaintext authenticator is running as a client, the server's
2279 challenges are checked to ensure they are valid base64 strings. By
2280 default, the authentication attempt is cancelled if an invalid string is
2281 received. Setting client_ignore_invalid_base64 true ignores these errors.
2282 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2283 they are received. Thus, the responses can be made to depend on the
2284 challenges. If an invalid string is ignored, an empty string is placed in
2287 PH/46 Messages that are created by the autoreply transport now contains a
2288 References: header, in accordance with RFCs 2822 and 3834.
2290 PH/47 Added authenticated_sender_force to the smtp transport.
2292 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2294 PH/49 Installed latest patch from the Sieve maintainer.
2296 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2297 was also set, the appendfile transport was unnecessarily scanning a
2298 directory of message files (e.g. for maildir delivery) to find the count
2299 of files (along with the size), even though it did not need this
2300 information. It now does the scan only if it needs to find either the
2301 size of the count of files.
2303 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2305 PH/52 Two bugs concerned with error handling when the smtp transport is
2308 (i) Exim was not creating retry information for temporary errors given
2309 for individual recipients after the DATA command when the smtp transport
2310 was used in LMTP mode. This meant that they could be retried too
2311 frequently, and not timed out correctly.
2313 (ii) Exim was setting the flag that allows error details to be returned
2314 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2315 recipients that were returned after the DATA command.
2317 PH/53 This is related to PH/52, but is more general: for any failing address,
2318 when detailed error information was permitted to be returned to the
2319 sender, but the error was temporary, then after the final timeout, only
2320 "retry timeout exceeded" was returned. Now it returns the full error as
2321 well as "retry timeout exceeded".
2323 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2324 do this, and (what is worse) MTAs that accept it.
2326 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2327 will now be deprecated.
2329 PH/56 New os.c-cygwin from the Cygwin maintainer.
2331 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2332 formats (previously only available in exim formats via -bpr, -bpru,
2333 and -bpra. Now also available in native and exiqgrep formats)
2335 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2336 with very large, slow to parse queues
2338 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2340 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2342 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2343 responses to authentication challenges, though it was showing the
2344 challenges; (ii) I've removed the CR characters from the debug output for
2347 PH/58 Allow for the insertion of a newline as well as a space when a string
2348 is turned into more than one encoded-word during RFC 2047 encoding. The
2349 Sieve code now uses this.
2351 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2352 data_4xx, lost_connection, tls_required.
2354 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2355 message was being sent as an SMTP response.
2357 PH/61 Add -l and -k options to exicyclog.
2359 PH/62 When verifying, if an address was redirected to one new address, so that
2360 verification continued, and the new address failed or deferred after
2361 having set something in $address_data, the value of $address_data was not
2362 passed back to the ACL. This was different to the case when no
2363 redirection occurred. The value is now passed back in both cases.
2365 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2366 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2367 use login_cap.h, so on its own it isn't the distinguishing feature. The
2368 new name refers directly to the setclassresources() function.
2370 PH/65 Added configuration files for NetBSD3.
2372 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2374 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2375 is preferred over IPv4.
2377 PH/68 The bounce_return_message and bounce_return_body options were not being
2378 honoured for bounces generated during the reception of non-SMTP messages.
2379 In particular, this applied to messages rejected by the ACL. This bug has
2380 been fixed. However, if bounce_return_message is true and bounce_return_
2381 body is false, the headers that are returned for a non-SMTP message
2382 include only those that have been read before the error was detected.
2383 (In the case of an ACL rejection, they have all been read.)
2385 PH/69 The HTML version of the specification is now built in a directory called
2386 spec_html instead of spec.html, because the latter looks like a path with
2387 a MIME-type, and this confuses some software.
2389 PH/70 Catch two compiler warnings in sieve.c.
2391 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2392 function verify_get_ident() calls ip_connect() to connect a socket, but
2393 if the "connect()" function timed out, ip_connect() used to close the
2394 socket. However, verify_get_ident() also closes the socket later, and in
2395 between Exim writes to the log, which may get opened at this point. When
2396 the socket was closed in ip_connect(), the log could get the same file
2397 descriptor number as the socket. This naturally causes chaos. The fix is
2398 not to close the socket in ip_connect(); the socket should be closed by
2399 the function that creates it. There was only one place in the code where
2400 this was missing, in the iplookup router, which I don't think anybody now
2401 uses, but I've fixed it anyway.
2403 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2404 well as to direct DNS lookups. Otherwise the handling of names in host
2405 lists is inconsistent and therefore confusing.
2411 PH/01 Two changes to the default runtime configuration:
2413 (1) Move the checks for relay_from_hosts and authenticated clients from
2414 after to before the (commented out) DNS black list checks.
2416 (2) Add control=submission to the relay_from_hosts and authenticated
2417 clients checks, on the grounds that messages accepted by these
2418 statements are most likely to be submissions.
2420 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2422 (1) Generate an error if the third argument for the ${prvs expansion is
2425 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2428 (3) Reset the variables that are obtained from the first argument of
2429 ${prvscheck and used in the second argument before leaving the code,
2430 because their memory is reclaimed, so using them afterwards may do
2433 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2434 one (it's much easier than Tom thought :-).
2436 (5) Because of (4), we can now allow for the use of $prvscheck_result
2437 inside the third argument.
2439 PH/03 For some reason, the default setting of PATH when running a command from
2440 a pipe transport was just "/usr/bin". I have changed it to
2443 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2444 anything to be listed in the output from -bV.
2446 PH/05 When a filter generated an autoreply, the entire To: header line was
2447 quoted in the delivery log line, like this:
2449 => >A.N.Other <ano@some.domain> <original@ddress> ...
2451 This has been changed so that it extracts the operative address. There
2452 may be more than one such address. If so, they are comma-separated, like
2455 => >ano@some.domain,ona@other.domain <original@ddress> ...
2457 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2458 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2459 not being looked up in the rDNS to get a host name, Exim was showing the
2460 IP address twice in Received: lines, even though the IP addresses were
2461 identical. For example:
2463 Received: from [1.2.3.4] (helo=[1.2.3.4])
2465 However, if the real host name was known, it was omitting the HELO data
2466 if it matched the actual IP address. This has been tidied up so that it
2467 doesn't show the same IP address twice.
2469 PH/07 When both +timestamp and +memory debugging was on, the value given by
2470 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2471 called by the debug printing, thereby overwriting the timestamp buffer.
2472 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2474 PH/08 When the original message was included in an autoreply transport, it
2475 always said "this is a copy of the message, including all the headers",
2476 even if body_only or headers_only was set. It now gives an appropriate
2479 PH/09 Applied a patch from the Sieve maintainer which:
2481 o fixes some comments
2482 o adds the (disabled) notify extension core
2483 o adds some debug output for the result of if/elsif tests
2484 o points to the current vacation draft in the documentation
2485 and documents the missing references header update
2489 o fixes a bug in processing the envelope test (when testing
2490 multiple envelope elements, the last element determinted the
2493 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2494 Electronic Mail") by including:
2496 Auto-submitted: auto-generated
2498 in the messages that it generates (bounce messages and others, such as
2499 warnings). In the case of bounce messages for non-SMTP mesages, there was
2500 also a typo: it was using "Auto_submitted" (underscore instead of
2501 hyphen). Since every message generated by Exim is necessarily in response
2502 to another message, thes have all been changed to:
2504 Auto-Submitted: auto-replied
2506 in accordance with these statements in the RFC:
2508 The auto-replied keyword:
2510 - SHOULD be used on messages sent in direct response to another
2511 message by an automatic process,
2513 - MUST NOT be used on manually-generated messages,
2515 - MAY be used on Delivery Status Notifications (DSNs) and Message
2516 Disposition Notifications (MDNs),
2518 - MUST NOT be used on messages generated by automatic or periodic
2519 processes, except for messages which are automatic responses to
2522 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2523 to the default Received: header definition.
2525 PH/12 Added log selector acl_warn_skipped (default on).
2527 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2528 variables because (a) they are in the wrong storage pool and (b) even if
2529 they were copied, it wouldn't work properly because of the caching.
2531 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2532 checking when decoding. Apparently there are clients that generate
2533 overlong encoded strings. Why am I not surprised?
2535 PH/15 If the first argument of "${if match_address" was not empty, but did not
2536 contain an "@" character, Exim crashed. Now it writes a panic log message
2537 and treats the condition as false.
2539 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2541 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2542 extension "envelope-auth". The code is finished and in agreement with
2543 other implementations, but there is no documentation so far and in fact,
2544 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2545 not changing the active code.
2547 Print executed "if" and "elsif" statements when debugging is used. This
2548 helps a great deal to understand what a filter does.
2550 Document more things not specified clearly in RFC3028. I had all this
2551 sorted out, when out of a sudden new issues came to my mind. Oops."
2553 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2556 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2557 canonical form (as documented). However, after a host name lookup from
2558 the IP address, check_host() was doing a simple string comparison with
2559 addresses acquired from the DNS when checking that the found name did
2560 have the original IP as one of its addresses. Since any found IPv6
2561 addresses are likely to be in abbreviated form, the comparison could
2562 fail. Luckily, there already exists a function for doing the comparison
2563 by converting both addresses to binary, so now that is used instead of
2564 the text comparison.
2566 PH/20 There was another similar case to PH/19, when a complete host name was
2567 given in a host list; looking up its IP address could give an abbreviated
2568 form, whereas the current host's name might or might not be abbreviated.
2569 The same fix has been applied.
2575 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2576 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2579 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2580 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2582 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2584 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2585 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2586 Also, for "Auto-Submitted", treat anything other than "no" as
2587 non-personal, in accordance with RFC 3834. (Previously it treated
2588 anything starting "auto-" as non-personal.)
2590 TF/01 The control=submission/name=... option had a problem with syntax
2591 errors if the name included a slash character. The /name= option
2592 now slurps the rest of the string, so it can include any characters
2593 but it must come last in the list of options (after /sender_retain
2596 PH/05 Some modifications to the interface to the fake nameserver for the new
2604 TK/01 Added the "success_on_redirect" address verification option. See
2605 NewStuff for rationale and an example.
2607 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2609 PH/02 Patch to exigrep to allow it to work on syslog lines.
2611 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2612 fread() to skip over the body file's header line, because in Cygwin the
2613 header line is locked and is inaccessible.
2615 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2616 co-exist for some time) to make it clear that it is the Exim ID that is
2617 referenced, not the Message-ID: header line.
2619 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2620 string_format() function, because snprintf() does not exist on all
2623 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2624 ${stat: expansion item.
2626 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2627 protocol synchronization error", to keep the pedants happy.
2629 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2630 well as for IRIX systems, when gcc is being used. See the host.c source
2633 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2635 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2638 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2639 transport and to the smtp transport in LMTP mode.
2641 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2643 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2644 run a filter in a subprocess. This could lead to confusion in subsequent
2645 lookups in the parent process. There should also be a search_tidyup() at
2646 the end of the subprocess.
2648 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2649 only if the host matched helo_try_verify_hosts, which caused the
2650 verification to occur when the EHLO/HELO command was issued. The ACL just
2651 tested the remembered result. Now, if a previous verification attempt has
2652 not happened, "verify = helo" does it there and then.
2654 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2656 TK/03 Fix log output including CR from clamd.
2658 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2659 exist provoked a memory error which could cause a segfault.
2661 PH/15 Installed PCRE 6.2
2663 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2665 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2666 of the problem. Specifically, suggested +O2 rather than +O1 for the
2669 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2671 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2672 record, the port was not correctly logged when the outgoing_port log
2673 selector was set (it logged the transort's default port).
2675 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2676 fallback_hosts, and "hosts" in the smtp transport.
2678 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2679 host errors such as "Connection refused".
2681 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2682 authentication with radiusclient 0.4.9:
2684 - Error returned from rc_read_config was caught wrongly
2685 - Username/password not passed on to radius server due to wrong length.
2687 The presumption is that some radiusclient API changes for 4.51/PH/17
2688 were not taken care of correctly. The code is still untested by me (my
2689 Linux distribution still has 0.3.2 of radiusclient), but it was
2690 contributed by a Radius user.
2692 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2693 expanding the "port" option of the smtp transport.
2695 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2696 while reading a MIME header. Thanks to Tom Hughes for a patch.
2698 PH/24 Include config.h inside local_scan.h so that configuration settings are
2701 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2702 that in an ACL for RCPT (for example), you can examine exactly what was
2705 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2706 commands, but it was not correctly comparing the address with the actual
2707 client host address. Thus, it would show the EHLO address in Received:
2708 header lines when this was not necessary.
2710 PH/27 Added the % operator to ${eval:}.
2712 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2713 it should be ignoring failures (because with -C, for example, it has lost
2714 privilege). It wasn't ignoring creation failures other than "already
2717 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2720 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2721 because an input line was too long, either on its own, or by virtue of
2722 too many continuations, the temporary file was not being removed, and the
2723 return code was incorrect.
2725 PH/31 Missing "BOOL" in function definition in filtertest.c.
2727 PH/32 Applied Sieve patches from the maintainer.
2729 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2731 PH/33 Added "verify = not_blind".
2733 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2734 Local/Makefile (with some defaults set). These are used in built scripts
2735 such as exicyclog, but they have never been used in the exim_install
2736 script (though there are many overriding facilities there). I have
2737 arranged that the exim_install script now takes note of these two
2740 PH/35 Installed configuration files for Dragonfly.
2742 PH/36 When a locally submitted message by a trusted user did not contain a
2743 From: header, and the sender address was obtained from -f or from an SMTP
2744 MAIL command, and the trusted user did not use -F to supply a sender
2745 name, $originator_name was incorrectly used when constructing a From:
2746 header. Furthermore, $originator_name was used for submission mode
2747 messages from external hosts without From: headers in a similar way,
2748 which is clearly wrong.
2750 PH/37 Added control=suppress_local_fixups.
2752 PH/38 When log_selector = +received_sender was set, and the addition of the
2753 sender made the log line's construction buffer exactly full, or one byte
2754 less than full, an overflow happened when the terminating "\n" was
2757 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2758 when the result of a list match is failure because a DNS lookup failed.
2760 PH/40 RM_COMMAND is now used in the building process.
2762 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2763 the "build-* directories that it finds.
2765 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2766 address in a domain literal was a prefix of an interface address.
2768 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2769 when verifying a sender address, unless rewrite_headers is false.
2771 PH/44 Wrote a long comment about why errors_to addresses are verified as
2772 recipients, not senders.
2774 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2775 the ratelimit ACL was added.
2777 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2779 PH/46 Added extra information about PostgreSQL errors to the error string.
2781 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2782 avoiding the need to install special zones in a real server. This is
2783 backwards compatible; if it can't find the fake resolver, it drops back.
2784 Thus, both old and new test suites can be run.
2786 TF/02 Added util/ratelimit.pl
2788 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2789 clock is set back in time.
2791 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2792 Candler <B.Candler@pobox.com>.
2794 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2795 OK for addresses that are the result of redirections.
2797 PH/48 A number of further additions for the benefit of the new test suite,
2798 including a fake gethostbyname() that interfaces to the fake DNS resolver
2801 TF/06 The fix for widen_domains has also been applied to qualify_single and
2802 search_parents which are the other dnslookup options that can cause
2805 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2808 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2810 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2811 service" if it encountered an MX record with an empty target hostname.
2812 The message is now "an MX or SRV record indicated no SMTP service".
2814 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2815 if the DNS of the sending site is misconfigured. This is quite a
2816 common situation. This change restores the behaviour of treating a
2817 helo verification defer as a failure.
2819 PH/51 If self=fail was set on a router, the bounce message did not include the
2820 actual error message.
2826 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2828 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2829 command itself ended in error, the underlying message about the transport
2830 filter timeout was being overwritten with the pipe command error. Now the
2831 underlying error message should be appended to the second error message.
2833 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2835 PH/02 Reduce the amount of output that "make" produces by default. Full output
2836 can still be requested.
2838 PH/03 The warning log line about a condition test deferring for a "warn" verb
2839 was being output only once per connection, rather than after each
2840 occurrence (because it was using the same function as for successful
2841 "warn" verbs). This seems wrong, so I have changed it.
2843 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2844 it should not have, which might have caused a crash in the right
2845 circumstances, but probably never did.
2847 PH/04 Installed a modified version of Tony Finch's patch to make submission
2848 mode fix the return path as well as the Sender: header line, and to
2849 add a /name= option so that you can make the user's friendly name appear
2852 TF/03 Added the control = fakedefer ACL modifier.
2854 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2855 Mark Lowes for thorough testing.
2857 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2859 TK/03 Merged latest SRS patch from Miles Wilton.
2861 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2862 with the definition in sysexits.h (which is #included earlier).
2863 Fortunately, Exim does not actually use EX_OK. The code used to try to
2864 preserve the sysexits.h value, by assumimg that macro definitions were
2865 scanned for macro replacements. I have been disabused of this notion,
2866 so now the code just undefines EX_OK before #including unistd.h.
2868 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2869 in the smtp transport. When a block could not be written in a single
2870 write() function, the timeout was being re-applied to each part-write.
2871 This seems wrong - if the receiver was accepting one byte at a time it
2872 would take for ever. The timeout is now adjusted when this happens. It
2873 doesn't have to be particularly precise.
2875 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2876 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2878 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2879 without a domain if the check to <postmaster@domain> fails.
2881 SC/01 Eximstats: added -xls and the ability to specify output files
2882 (patch written by Frank Heydlauf).
2884 SC/02 Eximstats: use FileHandles for outputing results.
2886 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2888 SC/04 Eximstats: fixed display of large numbers with -nvr option
2890 SC/05 Eximstats: fixed merging of reports with empty tables.
2892 SC/06 Eximstats: added the -include_original_destination flag
2894 SC/07 Eximstats: removed tabs and trailing whitespace.
2896 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2898 TK/06 MBOX spool code: Add real "From " MBOX separator line
2899 so the .eml file is really in mbox format (even though
2900 most programs do not really care). Patch from Alex Miller.
2902 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2903 The latter is generated from $received_to and is only set if the
2904 message has one envelope recipient. SA can use these headers,
2905 obviously out-of-the-box. Patch from Alex Miller.
2907 PH/08 The ${def test on a variable was returning false if the variable's
2908 value was "0", contrary to what the specification has always said!
2909 The result should be true unless the variable is empty.
2911 PH/09 The syntax error of a character other than { following "${if
2912 def:variable_name" (after optional whitespace) was not being diagnosed.
2913 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2914 accidental colon was present, for example, could give incorrect results.
2916 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2917 result is used (not including appendfile, where other changes are about
2920 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2921 This involved changing a lot of size variables from int to off_t. It
2922 should work with maildirs and everything.
2924 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2925 spamd dying while we are connected to it.
2927 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2930 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2931 function for BDB 4.3.
2933 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2935 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2936 the sources for PCRE have been re-arranged and more files are now
2939 PH/15 The code I had for printing potentially long long variables in PH/11
2940 above was not the best (it lost precision). The length of off_t variables
2941 is now inspected at build time, and an appropriate printing format (%ld
2942 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2943 to be "long long int" or "long int". This is needed for the internal
2944 formatting function string_vformat().
2946 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2947 the configuration file to be ":syslog", then the script "guesses" where
2948 the logs files are, rather than using the compiled in default. In our
2949 case the guess is not the same as the compiled default, so the script
2950 suddenly stopped working when I started to use syslog. The patch checks
2951 to see if log_file_path is "". If so, it attempts to read it from exim
2952 with no configuration file to get the compiled in version, before it
2953 falls back to the previous guessing code."
2955 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2956 implementing BATV in an Exim configuration. See NewStuff for the gory
2959 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2960 Makefile that are specific to HP-UX.
2962 PH/18 If the "use_postmaster" option was set for a recipient callout together
2963 with the "random" option, the postmaster address was used as the MAIL
2964 FROM address for the random test, but not for the subsequent recipient
2965 test. It is now used for both.
2967 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2968 patch removes a few documentation additions to RFC 3028, because the
2969 latest draft now contains them. It adds the new en;ascii-case comparator
2970 and a new error check for 8bit text in MIME parts. Comparator and
2971 require names are now matched exactly. I enabled the subaddress
2972 extension, but it is not well tested yet (read: it works for me)."
2974 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2975 rework some of the code of TK/09 above to avoid the hardwired use of
2976 "%lld" and "long long". Replaced the call to snprintf() with a call to
2979 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2980 records point to non-existent hosts", "retry timeout exceeded", and
2981 "retry time not reached for any host after a long failure period".
2983 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2984 experimental DomainKeys support:
2986 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2987 (2) On an error such as an illegally used "control", the wrong name for
2988 the control was given.
2990 These problems did NOT occur unless DomainKeys support was compiled.
2992 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2994 PH/24 Added ${if match_ip condition.
2996 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2997 databases so that it will be absolutely obvious if a crash occurs in the
2998 DB library. This is a regular occurrence (often caused by mis-matched
3001 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3002 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3003 picked up on a user's system that detects such things. There doesn't seem
3004 to be a gcc warning option for this - only an attribute that has to be
3005 put on the function's prototype. It seems that in Fedora Core 4 they have
3006 set this on a number of new functions. No doubt there will be more in due
3009 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3010 specify a transport. However, if an address that was verified by such a
3011 router was the subject of a callout, Exim crashed because it tried to
3012 read the rcpt_include_affixes from the non-existent transport. Now it
3013 just assumes that the setting of that option is false. This bug was
3014 introduced by 4.51/PH/31.
3016 PH/28 Changed -d+all to exclude +memory, because that information is very
3017 rarely of interest, but it makes the output a lot bigger. People tend to
3018 do -d+all out of habit.
3020 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3021 code in os-type was giving problems when libc.so lives in lib64, like on
3024 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3025 aren't the modern standard, and it seems that some systems' include files
3026 don't always have them. Exim was already checking for some of the newer
3027 ones like T_AAAA, and defining it itself. I've added checks for all the
3028 record types that Exim uses.
3030 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3031 not automatically generating a new one, as it is supposed to. This
3032 prevented TLS from working. If the file did exist, but contained invalid
3033 data, a new version was generated, as expected. It was only the case of a
3034 non-existent file that was broken.
3036 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3037 with a change in libdomainkeys > 0.64.
3039 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3040 from DNS. If the selector record carries the flag, it now has
3041 precedence over the domain-wide flag.
3043 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3045 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3046 the use of a port name that isn't defined in /etc/services) occurred, the
3047 message was deferred as in a normal delivery, and thus remained on the
3048 spool, instead of being failed because of the mua_wrapper setting. This
3049 is now fixed, and I tidied up some of the mua_wrapper messages at the
3052 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3053 the messages in a hash of arrays rather than using individual hashes.
3054 This is a bit cleaner and results in dramatic memory savings, albeit
3055 at a slight CPU cost.
3057 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3058 as requested by Marc Sherman.
3060 SC/10 Eximstats: added histograms for user specified patterns as requested
3063 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3065 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3066 fopen() in the content-scanning modules that did not already have it.
3072 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3073 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3075 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3077 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3079 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3080 to test Sieve filters that use "vacation".
3082 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3083 that changes the way the GnuTLS parameters are stored in the cache file.
3084 The new format can be generated externally. For backward compatibility,
3085 if the data in the cache doesn't make sense, Exim assumes it has read an
3086 old-format file, and it generates new data and writes a new file. This
3087 means that you can't go back to an older release without removing the
3090 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3091 work if there are any delivery delays because "one_time" forces the
3092 parent to be marked "delivered", so its unseen clone is never tried
3093 again. For this reason, Exim now forbids the simultaneous setting of
3096 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3097 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3098 entry if you want to know the details. The fix, however, neglected to
3099 consider the case when local delivery batching is involved. The test for
3100 "previously delivered" was not happening when checking to see if an
3101 address could be batched with a previous (undelivered) one; under
3102 certain circumstances this could lead to multiple deliveries to the same
3105 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3106 in its include files, and this causes problems building Exim.
3108 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3109 header_syntax) but Exim was just ignoring anything given after a slash.
3110 In particular, this caused confusion with an attempt to use "verify =
3111 reverse_host_lookup/defer_ok". An error is now given when options are
3112 supplied for verify items that do not have them. (Maybe reverse_host_
3113 lookup should have a defer_ok option, but that's a different point.)
3115 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3116 defined by RFC 821) to 2048, because there were problems with some AUTH
3117 commands, and RFC 1869 says the size should be increased for extended
3118 SMTP commands that take arguments.
3120 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3123 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3124 "unknown" error; now it says that the functionality isn't in the binary.
3126 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3127 an address' error message when a string expansion fails (syntax or
3128 whatever). Otherwise the password may appear in the log. Following change
3129 PH/42 below, there is no longer a chance of it appearing in a bounce
3132 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3134 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3135 of its addresses was ever tried. (Bugzilla bug #2.)
3137 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3138 the result incorrectly in the debug output. (It correctly added a newline
3139 to what was transported.)
3141 TF/01 Added $received_time.
3143 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3144 commented out examples of how to interface to a virus scanner and to
3145 SpamAssassin. Also added commented examples of av_scanner and
3146 spamd_address settings.
3148 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3149 and controls are allowed in which ACLs. There were a couple of minor
3150 errors. Some of the entries in the conditions table (which is a table of
3151 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3152 negation of where the condition IS allowed.
3154 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3156 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3157 header file does not have a version number, so I've had to invent a new
3158 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3159 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3160 radiusclient), but it was contributed by a Radius user.
3162 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3163 files or directories, for OpenSSL.
3165 PH/19 When an Exim process that is running as root has to create an Exim log
3166 file, it does so in a subprocess that runs as exim:exim so as to get the
3167 ownership right at creation (otherwise, other Exim processes might see
3168 the file with the wrong ownership). There was no test for failure of this
3169 fork() call, which would lead to the process getting stuck as it waited
3170 for a non-existent subprocess. Forks do occasionally fail when resources
3171 run out. I reviewed all the other calls to fork(); they all seem to check
3174 PH/20 When checking for unexpected SMTP input at connect time (before writing
3175 the banner), Exim was not dealing correctly with a non-positive return
3176 from the read() function. If the client had disconnected by this time,
3177 the result was a log entry for a synchronization error with an empty
3178 string after "input=" when read() returned zero. If read() returned -1
3179 (an event I could not check), uninitialized data bytes were printed.
3180 There were reports of junk text (parts of files, etc) appearing after
3183 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3185 PH/22 Added support for macro redefinition, and (re)definition in between
3186 driver and ACL definitions.
3188 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3189 forgetting to use the resulting value; it was using the unexpanded value.
3191 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3192 hadn't been configured. The fix is from Juergen Kreileder, who
3193 understands it better than I do:
3195 "Here's what I see happening with three configured cyrus_sasl
3196 authenticators configured (plain, login, cram-md5):
3198 On startup auth_cyrus_sasl_init() gets called for each of these.
3199 This means three calls to sasl_listmech() without a specified mech_list.
3200 => SASL tests which mechs of all available mechs actually work
3201 => three warnings about OTP not working
3202 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3204 With the patch, sasl_listmech() also gets called three times. But now
3205 SASL's mech_list option is set to the server_mech specified in the the
3206 authenticator. Or in other words, the answer from sasl_listmech()
3207 gets limited to just the mech you're testing for (which is different
3209 => the return list contains just 'plain' or 'login', 'cram-md5' or
3210 nothing depending on the value of ob->server_mech.
3212 I've just tested the patch: Authentication still works fine,
3213 unavailable mechs specified in the exim configuration are still
3214 caught, and the auth.log warnings about OTP are gone."
3216 PH/25 When debugging is enabled, the contents of the command line are added
3217 to the debugging output, even when log_selector=+arguments is not
3220 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3221 answer is "GNU", and only if the return is "GNU/something" is the answer
3224 PH/27 $acl_verify_message is now set immediately after the failure of a
3225 verification in an ACL, and so is available in subsequent modifiers. In
3226 particular, the message can be preserved by coding like this:
3228 warn !verify = sender
3229 set acl_m0 = $acl_verify_message
3231 Previously, $acl_verify_message was set only while expanding "message"
3232 and "log_message" when a very denied access.
3234 PH/28 Modified OS/os.c-Linux with
3236 -#ifndef OS_LOAD_AVERAGE
3237 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3239 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3240 nomenclature these days.)
3242 PH/29 Installed patch from the Sieve maintainer that adds the options
3243 sieve_useraddress and sieve_subaddress to the redirect router.
3245 PH/30 In these circumstances:
3246 . Two addresses routed to the same list of hosts;
3247 . First host does not offer TLS;
3248 . First host accepts first address;
3249 . First host gives temporary error to second address;
3250 . Second host offers TLS and a TLS session is established;
3251 . Second host accepts second address.
3252 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3253 and peerdn, if requested) that were in fact used only for the second
3256 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3257 attention to any local part prefix or suffix that was matched by the
3258 router that accepted the address. It now behaves in the same way as it
3259 does for delivery: the affixes are removed from the local part unless
3260 rcpt_include_affixes is set on the transport.
3262 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3263 timeout during the DATA phase of an incoming message.
3265 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3266 applied a patch sanctioned by the Sieve maintainer.
3268 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3269 the uid or gid is negative. A case of a negative gid caused this to be
3270 noticed. The fix allows for either to be negative.
3272 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3273 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3276 PH/36 The change PH/12 above was broken. Fixed it.
3278 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3279 the grounds that routing the same address twice would always produce the
3280 same answer. This might have been true once, but it is certainly no
3281 longer true now. Routing a child address may depend on the previous
3282 routing that produced that child. Some complicated redirection strategies
3283 went wrong when messages had multiple recipients, and made Exim's
3284 behaviour dependent on the order in which the addresses were given.
3286 I have moved the duplicate checking until after the routing is complete.
3287 Exim scans the addresses that are assigned to local and remote
3288 transports, and removes any duplicates. This means that more work will be
3289 done, as duplicates will always all be routed, but duplicates are
3290 presumably rare, so I don't expect this is of any significance.
3292 For deliveries to pipes, files, and autoreplies, the duplicate checking
3293 still happens during the routing process, since they are not going to be
3296 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3297 It corrects a timeout issue with spamd. This is Ian's comment: "The
3298 background is that sometimes spamd either never reads data from a
3299 connection it has accepted, or it never writes response data. The exiscan
3300 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3301 blindly assumes that writes won't block so it may never time out."
3303 PH/39 Allow G after quota size as well as K and M.
3305 PH/40 The value set for $authenticated_id in an authenticator may not contain
3306 binary zeroes or newlines because the value is written to log lines and
3307 to spool files. There was no check on this. Now the value is run through
3308 the string_printing() function so that such characters are converted to
3309 printable escape sequences.
3311 PH/41 $message_linecount is a new variable that contains the total number of
3312 lines in the message. Compare $body_linecount, which is the count for the
3315 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3316 bounce and delay warning messages, except in certain special cases, which
3319 (a) An SMTP error message from a remote host;
3320 (b) A message specified in a :fail: redirection;
3321 (c) A message specified in a "fail" command in a system filter;
3322 (d) A message specified in a FAIL return from the queryprogram router;
3323 (e) A message specified by the cannot_route_message router option.
3325 In these cases only, Exim does include the error details in bounce and
3326 warning messages. There are also a few cases where bland messages such
3327 as "unrouteable address" or "local delivery error" are given.
3329 PH/43 $value is now also set for the "else" part of a ${run expansion.
3331 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3332 being worked on, but at least Exim now implements the latest version to
3335 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3336 process to complete was treated as a delivery failure, a timeout while
3337 writing the message to the pipe was logged, but erroneously treated as a
3338 successful delivery. Such timeouts include transport filter timeouts. For
3339 consistency with the overall process timeout, these timeouts are now
3340 treated as errors, giving rise to delivery failures by default. However,
3341 there is now a new Boolean option for the pipe transport called
3342 timeout_defer, which, if set TRUE, converts the failures into defers for
3343 both kinds of timeout. A transport filter timeout is now identified in
3346 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3347 systems where "make" and "gmake" are different, calling "gmake" at top
3348 level broke things. I've arranged for the value of $(MAKE) to be passed
3349 from the Makefile to this script so that it can call the same version of
3353 A note about Exim versions 4.44 and 4.50
3354 ----------------------------------------
3356 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3357 changes of various kinds. As a consequence, a big documentation update was
3358 needed. This delayed the release for rather longer than seemed good, especially
3359 in the light of a couple of (minor) security issues. Therefore, the changes
3360 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3361 release. So 4.44 and 4.50 are in effect two different branches that both start
3364 I have left the 4.50 change log unchanged; it contains all the changes since
3365 4.43. The change log for 4.44 is below; many of its items are identical to
3366 those for 4.50. This seems to be the most sensible way to preserve the
3367 historical information.
3373 1. Minor wording change to the doc/README.SIEVE file.
3375 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3376 computation of the current number of files was incorrect.
3378 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3379 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3382 4. Give more explanation in the error message when the command for a transport
3383 filter fails to execute.
3385 5. There are several places where Exim runs a non-Exim command in a
3386 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3387 was being done only for the command run by the queryprogram router. It is
3388 now done for all such subprocesses. The other cases are: ${run, transport
3389 filters, and the commands run by the lmtp and pipe transports.
3391 6. Added CONFIGURE_GROUP build-time option.
3393 7. Some older OS have a limit of 256 on the maximum number of file
3394 descriptors. Exim was using setrlimit() to set 1000 as a large value
3395 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3396 systems. I've change it so that if it can't get 1000, it tries for 256.
3398 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3399 was an oversight, and furthermore, ever since the addition of extra
3400 controls (e.g. 4.43/32), the checks on when to allow different forms of
3401 "control" were broken. There should now be diagnostics for all cases when a
3402 control that does not make sense is encountered.
3404 9. Added the /retain_sender option to "control=submission".
3406 10. $recipients is now available in the predata ACL (oversight).
3408 11. Tidy the search cache before the fork to do a delivery from a message
3409 received from the command line. Otherwise the child will trigger a lookup
3410 failure and thereby defer the delivery if it tries to use (for example) a
3411 cached ldap connection that the parent has called unbind on.
3413 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3414 of $address_data from the recipient verification was clobbered by the
3415 sender verification.
3417 13. The value of address_data from a sender verification is now available in
3418 $sender_address_data in subsequent conditions in the ACL statement.
3420 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3422 15. Added a new option "connect=<time>" to callout options, to set a different
3425 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3426 was its contents. (It was OK if the option was not defined at all.)
3428 17. A "Completed" log line is now written for messages that are removed from
3429 the spool by the -Mrm option.
3431 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3432 information about exactly what failed.
3434 19. Added -dd to debug only the daemon process.
3436 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3437 handles timeouts, both on the server side and network timeouts. Renamed the
3438 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3440 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3441 It is now set to "smtps".
3443 22. $host_address is now set to the target address during the checking of
3444 ignore_target_hosts.
3446 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3447 being passed; this would have caused $sender_host_name to have been used if
3448 matching the list had actually called for a host name (not very likely,
3449 since this list is usually IP addresses). A host name is now passed as
3452 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3453 code that specifies a non-restarting handler (typically sigaction() in
3454 modern systems) in an attempt to fix a rare and obscure crash bug.
3456 25. Narrowed the window for a race in the daemon that could cause it to ignore
3457 SIGCHLD signals. This is not a major problem, because they are used only to
3458 wake it up if nothing else does.
3460 26. A malformed maildirsize file could cause Exim to calculate negative values
3461 for the mailbox size or file count. Odd effects could occur as a result.
3462 The maildirsize information is now recalculated if the size or filecount
3465 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3466 support for a long time. Removed HAVE_SYS_VFS_H.
3468 28. Installed the latest version of exipick from John Jetmore.
3470 29. In an address list, if the pattern was not a regular expression, an empty
3471 subject address (from a bounce message) matched only if the pattern was an
3472 empty string. Non-empty patterns were not even tested. This was the wrong
3473 because it is perfectly reasonable to use an empty address as part of a
3474 database query. An empty address is now tested by patterns that are
3475 lookups. However, all the other forms of pattern expect the subject to
3476 contain a local part and a domain, and therefore, for them, an empty
3477 address still always fails if the pattern is not itself empty.
3479 30. Exim went into a mad DNS loop when attempting to do a callout where the
3480 host was specified on an smtp transport, and looking it up yielded more
3481 than one IP address.
3483 31. Re-factored the code for checking spool and log partition space into a
3484 function that finds that data and another that does the check. The former
3485 is then used to implement four new variables: $spool_space, $log_space,
3486 $spool_inodes, and $log_inodes.
3488 32. The RFC2047 encoding function was originally intended for short strings
3489 such as real names; it was not keeping to the 75-character limit for
3490 encoded words that the RFC imposes. It now respects the limit, and
3491 generates multiple encoded words if necessary. To be on the safe side, I
3492 have increased the buffer size for the ${rfc2047: expansion operator from
3495 33. It is now permitted to omit both strings after an "if" condition; if the
3496 condition is true, the result is "true". As before, when the second string
3497 is omitted, a false condition yields an empty string. This makes it less
3498 cumbersome to write custom ACL and router conditions.
3500 34. Failure to deliver a bounce message always caused it to be frozen, even if
3501 there was an errors_to setting on the router. The errors_to setting is now
3504 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3505 canonical form (fully expanded) before being placed in
3506 $sender_host_address.
3508 36. The table in the code that translates DNS record types into text (T_A to
3509 "A" for instance) was missing entries for NS and CNAME. It is just possible
3510 that this could have caused confusion if both these types were looked up
3511 for the same domain, because the text type is used as part of Exim's
3512 per-process caching. But the chance of anyone hitting this buglet seems
3515 37. The dnsdb lookup has been extended in a number of ways.
3517 (1) There is a new type, "zns", which walks up the domain tree until it
3518 finds some nameserver records. It should be used with care.
3520 (2) There is a new type, "mxh", which is like "mx" except that it returns
3521 just the host names, not the priorities.
3523 (3) It is now possible to give a list of domains (or IP addresses) to be
3524 looked up. The behaviour when one of the lookups defers can be
3525 controlled by a keyword.
3527 (4) It is now possible to specify the separator character for use when
3528 multiple records are returned.
3530 38. The dnslists ACL condition has been extended: it is now possible to supply
3531 a list of IP addresses and/or domains to be looked up in a particular DNS
3534 39. Added log_selector=+queue_time_overall.
3536 40. When running the queue in the test harness, wait just a tad after forking a
3537 delivery process, to get repeatability of debugging output.
3539 41. Include certificate and key file names in error message when GnuTLS fails
3540 to set them up, because the GnuTLS error message doesn't include the name
3541 of the failing file when there is a problem reading it.
3543 42. Allow both -bf and -bF in the same test run.
3545 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3547 44. The "Exiscan patch" is now merged into the mainline Exim source.
3549 45. Sometimes the final signoff response after QUIT could fail to get
3550 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3551 < 0 before doing a fflush(). This bug looks as though it goes back to the
3552 introduction of TLS in release 3.20, but "sometimes" must have been rare
3553 because the tests only now provoked it.
3555 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3556 (this can affect the format of dates).
3558 47. exim_tidydb, when checking for the continued existence of a message for
3559 which it has found a message-specific retry record, was not finding
3560 messages that were in split spool directories. Consequently, it was
3561 deleting retry records that should have stayed in existence.
3563 48. Steve fixed some bugs in eximstats.
3565 49. The SPA authentication driver was not abandoning authentication and moving
3566 on to the next authenticator when an expansion was forced to fail,
3567 contradicting the general specification for all authenticators. Instead it
3568 was generating a temporary error. It now behaves as specified.
3570 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3571 (the order specifies the preference for clients). The order is now AES256,
3572 AES128, 3DES, ARCFOUR128.
3574 51. Small patch to Sieve code - explicitly set From: when generating an
3577 52. Exim crashed if a remote delivery caused a very long error message to be
3578 recorded - for instance if somebody sent an entire SpamAssassin report back
3579 as a large number of 550 error lines. This bug was coincidentally fixed by
3580 increasing the size of one of Exim's internal buffers (big_buffer) that
3581 happened as part of the Exiscan merge. However, to be on the safe side, I
3582 have made the code more robust (and fixed the comments that describe what
3585 53. Now that there can be additional text after "Completed" in log lines (if
3586 the queue_time_overall log selector is set), a one-byte patch to exigrep
3587 was needed to allow it to recognize "Completed" as not the last thing in
3590 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3591 patch that reportedly fixes this has been added. I am not expert enough to
3592 create a test for it. This is what the patch creator wrote:
3594 "I found a little strange behaviour of ldap code when working with
3595 Windows 2003 AD Domain, where users was placed in more than one
3596 Organization Units. When I tried to give exim partial DN, the exit code
3597 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3598 But simultaneously result of request was absolutely normal ldap result,
3599 so I produce this patch..."
3601 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3602 REFERENCE, so I have modified the code to exclude the patch when that macro
3605 55. Some experimental protocols are using DNS PTR records for new purposes. The
3606 keys for these records are domain names, not reversed IP addresses. The
3607 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3608 leaves it alone. Component reversal etc. now happens only for IP addresses.
3611 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3613 57. Double the size of the debug message buffer (to 2048) so that more of very
3614 long debug lines gets shown.
3616 58. The exicyclog utility now does better if the number of log files to keep
3617 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3619 59. Two changes related to the smtp_active_hostname option:
3621 (1) $smtp_active_hostname is now available as a variable.
3622 (2) The default for smtp_banner uses $smtp_active_hostname instead
3623 of $primary_hostname.
3625 60. The host_aton() function is supposed to be passed a string that is known
3626 to be a valid IP address. However, in the case of IPv6 addresses, it was
3627 not checking this. This is a hostage to fortune. Exim now panics and dies
3628 if the condition is not met. A case was found where this could be provoked
3629 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3630 components; fortuitously, this particular loophole had already been fixed
3631 by change 4.50/55 above.
3633 If there are any other similar loopholes, the new check in host_aton()
3634 itself should stop them being exploited. The report I received stated that
3635 data on the command line could provoke the exploit when Exim was running as
3636 exim, but did not say which command line option was involved. All I could
3637 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3638 running as the user.
3641 61. There was a buffer overflow vulnerability in the SPA authentication code
3642 (which came originally from the Samba project). I have added a test to the
3643 spa_base64_to_bits() function which I hope fixes it.
3646 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3647 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3649 63. The daemon start-up calls getloadavg() while still root for those OS that
3650 need the first call to be done as root, but it missed one case: when
3651 deliver_queue_load_max is set with deliver_drop_privilege. This is
3652 necessary for the benefit of the queue runner, because there is no re-exec
3653 when deliver_drop_privilege is set.
3655 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3656 This has been fixed.
3658 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3659 was in use, was not putting the data itself into the right store pool;
3660 consequently, it could be overwritten for a subsequent message in the same
3661 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3664 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3666 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3667 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3668 the function were treating the return as a boolean value, which happened to
3669 work because 0=false and not-0=true, but is not correct code.
3671 68. The host_aton() function was not handling scoped IPv6 addresses (those
3672 with, for example, "%eth0" on the end) correctly.
3674 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3675 negated items (that is, ~something) in unsigned ints. Some compilers
3676 apparently mutter when there is no cast.
3678 70. If an address verification called from an ACL failed, and did not produce a
3679 user-specific message (i.e. there was only a "system" message), nothing was
3680 put in $acl_verify_message. In this situation, it now puts the system
3683 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3684 session; change 4.31/43 added the unwanted input to the log line - except
3685 that it did not do this in the start of session case. It now does.
3687 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3688 This is wrong and can cause the other end to generate a synchronization
3689 error if it is another Exim or anything else that does the synchronization
3690 check. A QUIT command is no longer sent after a timeout.
3692 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3693 during host lookups.
3695 74. The defer_ok option of callout verification was not working if it was used
3696 when verifying addresses in header lines, that is, for this case:
3698 verify = header_sender/callout=defer_ok
3700 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3701 those file descriptors could be used for SMTP connections. If anything
3702 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3703 could be sent to the SMTP client, causing chaos. The daemon now opens
3704 stdin, stdout, and stderr to /dev/null when it puts itself into the
3707 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3708 log by default. The user can override this with suitable Perl magic.
3710 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3711 the log message when discard triggers, was not working for the DATA ACL or
3712 for the non-SMTP ACL.
3714 78. Error message wording change in sieve.c.
3716 79. If smtp_accept_max_per_host was set, the number of connections could be
3717 restricted to fewer than expected, because the daemon was trying to set up
3718 a new connection before checking whether the processes handling previous
3719 connections had finished. The check for completed processes is now done
3720 earlier. On busy systems, this bug wouldn't be noticed because something
3721 else would have woken the daemon, and it would have reaped the completed
3724 80. If a message was submitted locally by a user whose login name contained one
3725 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3726 It caused a spool format error. I have fixed the spool reading code. A
3727 related problem was that the "from" clause in the Received: line became
3728 illegal because of the space(s). It is now covered by ${quote_local_part.
3730 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3733 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3735 83. Patch from Sieve maintainer to fix unterminated string problem in
3736 "vacation" handling.
3738 84. Some minor changes to the Linux configuration files to help with other
3739 OS variants using glibc.
3741 85. One more patch for Sieve to update vacation handling to latest spec.
3744 ----------------------------------------------------
3745 See the note above about the 4.44 and 4.50 releases.
3746 ----------------------------------------------------
3752 1. Change 4.43/35 introduced a bug that caused file counts to be
3753 incorrectly computed when quota_filecount was set in an appendfile
3756 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3757 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3760 3. Give more explanation in the error message when the command for a transport
3761 filter fails to execute.
3763 4. There are several places where Exim runs a non-Exim command in a
3764 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3765 was being done only for the command run by the queryprogram router. It is
3766 now done for all such subprocesses. The other cases are: ${run, transport
3767 filters, and the commands run by the lmtp and pipe transports.
3769 5. Some older OS have a limit of 256 on the maximum number of file
3770 descriptors. Exim was using setrlimit() to set 1000 as a large value
3771 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3772 systems. I've change it so that if it can't get 1000, it tries for 256.
3774 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3775 was an oversight, and furthermore, ever since the addition of extra
3776 controls (e.g. 4.43/32), the checks on when to allow different forms of
3777 "control" were broken. There should now be diagnostics for all cases when a
3778 control that does not make sense is encountered.
3780 7. $recipients is now available in the predata ACL (oversight).
3782 8. Tidy the search cache before the fork to do a delivery from a message
3783 received from the command line. Otherwise the child will trigger a lookup
3784 failure and thereby defer the delivery if it tries to use (for example) a
3785 cached ldap connection that the parent has called unbind on.
3787 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3788 of $address_data from the recipient verification was clobbered by the
3789 sender verification.
3791 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3792 was its contents. (It was OK if the option was not defined at all.)
3794 11. A "Completed" log line is now written for messages that are removed from
3795 the spool by the -Mrm option.
3797 12. $host_address is now set to the target address during the checking of
3798 ignore_target_hosts.
3800 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3801 being passed; this would have caused $sender_host_name to have been used if
3802 matching the list had actually called for a host name (not very likely,
3803 since this list is usually IP addresses). A host name is now passed as
3806 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3807 code that specifies a non-restarting handler (typically sigaction() in
3808 modern systems) in an attempt to fix a rare and obscure crash bug.
3810 15. Narrowed the window for a race in the daemon that could cause it to ignore
3811 SIGCHLD signals. This is not a major problem, because they are used only to
3812 wake it up if nothing else does.
3814 16. A malformed maildirsize file could cause Exim to calculate negative values
3815 for the mailbox size or file count. Odd effects could occur as a result.
3816 The maildirsize information is now recalculated if the size or filecount
3819 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3820 support for a long time. Removed HAVE_SYS_VFS_H.
3822 18. Updated exipick to current release from John Jetmore.
3824 19. Allow an empty sender to be matched against a lookup in an address list.
3825 Previously the only cases considered were a regular expression, or an
3828 20. Exim went into a mad DNS lookup loop when doing a callout where the
3829 host was specified on the transport, if the DNS lookup yielded more than
3832 21. The RFC2047 encoding function was originally intended for short strings
3833 such as real names; it was not keeping to the 75-character limit for
3834 encoded words that the RFC imposes. It now respects the limit, and
3835 generates multiple encoded words if necessary. To be on the safe side, I
3836 have increased the buffer size for the ${rfc2047: expansion operator from
3839 22. Failure to deliver a bounce message always caused it to be frozen, even if
3840 there was an errors_to setting on the router. The errors_to setting is now
3843 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3844 canonical form (fully expanded) before being placed in
3845 $sender_host_address.
3847 24. Updated eximstats to version 1.33
3849 25. Include certificate and key file names in error message when GnuTLS fails
3850 to set them up, because the GnuTLS error message doesn't include the name
3851 of the failing file when there is a problem reading it.
3853 26. Expand error message when OpenSSL has problems setting up cert/key files.
3856 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3857 (this can affect the format of dates).
3859 28. exim_tidydb, when checking for the continued existence of a message for
3860 which it has found a message-specific retry record, was not finding
3861 messages that were in split spool directories. Consequently, it was
3862 deleting retry records that should have stayed in existence.
3864 29. eximstats updated to version 1.35
3865 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3866 1.35 - bugfix such that pie charts by volume are generated correctly
3868 30. The SPA authentication driver was not abandoning authentication and moving
3869 on to the next authenticator when an expansion was forced to fail,
3870 contradicting the general specification for all authenticators. Instead it
3871 was generating a temporary error. It now behaves as specified.
3873 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3874 (the order specifies the preference for clients). The order is now AES256,
3875 AES128, 3DES, ARCFOUR128.
3877 31. Small patch to Sieve code - explicitly set From: when generating an
3880 32. Exim crashed if a remote delivery caused a very long error message to be
3881 recorded - for instance if somebody sent an entire SpamAssassin report back
3882 as a large number of 550 error lines. This bug was coincidentally fixed by
3883 increasing the size of one of Exim's internal buffers (big_buffer) that
3884 happened as part of the Exiscan merge. However, to be on the safe side, I
3885 have made the code more robust (and fixed the comments that describe what
3888 33. Some experimental protocols are using DNS PTR records for new purposes. The
3889 keys for these records are domain names, not reversed IP addresses. The
3890 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3891 leaves it alone. Component reversal etc. now happens only for IP addresses.
3894 34. The host_aton() function is supposed to be passed a string that is known
3895 to be a valid IP address. However, in the case of IPv6 addresses, it was
3896 not checking this. This is a hostage to fortune. Exim now panics and dies
3897 if the condition is not met. A case was found where this could be provoked
3898 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3899 components; fortuitously, this particular loophole had already been fixed
3900 by change 4.50/55 or 4.44/33 above.
3902 If there are any other similar loopholes, the new check in host_aton()
3903 itself should stop them being exploited. The report I received stated that
3904 data on the command line could provoke the exploit when Exim was running as
3905 exim, but did not say which command line option was involved. All I could
3906 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3907 running as the user.
3910 35. There was a buffer overflow vulnerability in the SPA authentication code
3911 (which came originally from the Samba project). I have added a test to the
3912 spa_base64_to_bits() function which I hope fixes it.
3915 36. The daemon start-up calls getloadavg() while still root for those OS that
3916 need the first call to be done as root, but it missed one case: when
3917 deliver_queue_load_max is set with deliver_drop_privilege. This is
3918 necessary for the benefit of the queue runner, because there is no re-exec
3919 when deliver_drop_privilege is set.
3921 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3922 was in use, was not putting the data itself into the right store pool;
3923 consequently, it could be overwritten for a subsequent message in the same
3924 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3927 38. Sometimes the final signoff response after QUIT could fail to get
3928 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3929 < 0 before doing a fflush(). This bug looks as though it goes back to the
3930 introduction of TLS in release 3.20, but "sometimes" must have been rare
3931 because the tests only now provoked it.
3937 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3938 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3939 Now it returns an integer. A number of calls were still expecting a T/F
3940 return. Fortuitously, in all cases, the tests worked in OK situations,
3941 which is the norm. However, things would have gone wrong on any write
3942 failures on the smtp file descriptor. This function is used when sending
3943 messages over SMTP and also when doing verify callouts.
3945 2. When Exim is called to do synchronous delivery of a locally submitted
3946 message (the -odf or -odi options), it no longer closes stderr before doing
3949 3. Implemented the mua_wrapper option.
3951 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3953 5. Implemented the functions header_remove(), header_testname(),
3954 header_add_at_position(), and receive_remove_recipient(), and exported them
3957 6. If an ACL "warn" statement specified the addition of headers, Exim already
3958 inserted X-ACL-Warn: at the start if there was no header name. However, it
3959 was not making this test for the second and subsequent header lines if
3960 there were newlines in the string. This meant that an invalid header could
3961 be inserted if Exim was badly configured.
3963 7. Allow an ACL "warn" statement to add header lines at the start or after all
3964 the Received: headers, as well as at the end.
3966 8. Added the rcpt_4xx retry error code.
3968 9. Added postmaster_mailfrom=xxx to callout verification option.
3970 10. Added mailfrom=xxxx to the callout verification option, for verify=
3973 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3974 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3976 12. Inserted some casts to stop certain compilers warning when using pointer
3977 differences as field lengths or precisions in printf-type calls (mostly
3978 affecting debugging statements).
3980 13. Added optional readline() support for -be (dynamically loaded).
3982 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3983 same clock tick as a message's arrival, so that its received time was the
3984 same as the "first fail" time on the retry record, and that message
3985 remained on the queue past the ultimate address timeout, every queue runner
3986 would try a delivery (because it was past the ultimate address timeout) but
3987 after another failure, the ultimate address timeout, which should have then
3988 bounced the address, did not kick in. This was a "< instead of <=" error;
3989 in most cases the first failure would have been in the next clock tick
3990 after the received time, and all would be well.
3992 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3993 being recognized when the domain list was tested by the match_domain
3994 condition in an expansion string.
3996 16. Added the ${str2b64: operator.
3998 17. Exim was always calling setrlimit() to set a large limit for the number of
3999 processes, without checking whether the existing limit was already
4000 adequate. (It did check for the limit on file descriptors.) Furthermore,
4001 errors from getrlimit() and setrlimit() were being ignored. Now they are
4002 logged to the main and panic logs, but Exim does carry on, to try to do its
4003 job under whatever limits there are.
4005 18. Imported PCRE 5.0.
4007 19. Trivial typo in log message " temporarily refused connection" (the leading
4010 20. If the log selector return_path_on_delivery was set and an address was
4011 redirected to /dev/null, the delivery process crashed because it assumed
4012 that a return path would always be set for a "successful" delivery. In this
4013 case, the whole delivery is bypassed as an optimization, and therefore no
4016 21. Internal re-arrangement: the function for sending a challenge and reading
4017 a response while authentication was assuming a zero-terminated challenge
4018 string. It's now changed to take a pointer and a length, to allow for
4019 binary data in such strings.
4021 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4023 23. Exim was not respecting finduser_retries when seeking the login of the
4024 uid under which it was called; it was always trying 10 times. (The default
4025 setting of finduser_retries is zero.) Also, it was sleeping after the final
4026 failure, which is pointless.
4028 24. Implemented tls_on_connect_ports.
4030 25. Implemented acl_smtp_predata.
4032 26. If the domain in control=submission is set empty, Exim assumes that the
4033 authenticated id is a complete email address when it generates From: or
4034 Sender: header lines.
4036 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4037 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4038 chown and chgrp in /bin and hostname in /usr/bin.
4040 28. Exim was keeping the "process log" file open after each use, just as it
4041 does for the main log. This opens the possibility of it remaining open for
4042 long periods when the USR1 signal hits a daemon. Occasional processlog
4043 errors were reported, that could have been caused by this. Anyway, it seems
4044 much more sensible not to leave this file open at all, so that is what now
4047 29. The long-running daemon process does not normally write to the log once it
4048 has entered its main loop, and it closes the log before doing so. This is
4049 so that log files can straightforwardly be renamed and moved. However,
4050 there are a couple of unusual error situations where the daemon does write
4051 log entries, and I had neglected to close the log afterwards.
4053 30. The text of an SMTP error response that was received during a remote
4054 delivery was being truncated at 512 bytes. This is too short for some of
4055 the long messages that one sometimes sees. I've increased the limit to
4058 31. It is now possible to make retry rules that apply only when a message has a
4059 specific sender, in particular, an empty sender.
4061 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4062 it possible to be selective about when SMTP synchronization is enforced.
4064 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4066 32. Implemented hosts_connection_nolog.
4068 33. Added an ACL for QUIT.
4070 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4073 35. Added mailbox_size and mailbox_filecount to appendfile.
4075 36. Added control = no_multiline_responses to ACLs.
4077 37. There was a bug in the logic of the code that waits for the clock to tick
4078 in the case where the clock went backwards by a substantial amount such
4079 that the microsecond fraction of "now" was more than the microsecond
4080 fraction of "then" (but the whole seconds number was less).
4082 38. Added support for the libradius Radius client library this is found on
4083 FreeBSD (previously only the radiusclient library was supported).
4089 1. When certain lookups returned multiple values in the form name=value, the
4090 quoting of the values was not always being done properly. Specifically:
4091 (a) If the value started with a double quote, but contained no whitespace,
4093 (b) If the value contained whitespace other than a space character (i.e.
4094 tabs or newlines or carriage returns) it was not quoted.
4095 This fix has been applied to the mysql and pgsql lookups by writing a
4096 separate quoting function and calling it from the lookup code. The fix
4097 should probably also be applied to nisplus, ibase and oracle lookups, but
4098 since I cannot test any of those, I have not disturbed their existing code.
4100 2. A hit in the callout cache for a specific address caused a log line with no
4101 reason for rejecting RCPT. Now it says "Previous (cached) callout
4102 verification failure".
4104 3. There was an off-by-one bug in the queryprogram router. An over-long
4105 return line was truncated at 256 instead of 255 characters, thereby
4106 overflowing its buffer with the terminating zero. As well as fixing this, I
4107 have increased the buffer size to 1024 (and made a note to document this).
4109 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4110 when Exim is waiting for an SMTP response from a remote server, Exim
4111 restarts its select() call on the socket, thereby resetting its timeout.
4112 This is not a problem when such interrupts are rare. Somebody set up a cron
4113 job to run exiwhat every 2 minutes, which is less than the normal select()
4114 timeout (5 or 10 minutes). This meant that the select() timeout never
4115 kicked in because it was always reset. I have fixed this by comparing the
4116 time when an interrupt arrives with the time at the start of the first call
4117 to select(). If more time than the timeout has elapsed, the interrupt is
4118 treated as a timeout.
4120 5. Some internal re-factoring in preparation for the addition of Sieve
4121 extensions (by MH). In particular, the "personal" test is moved to a
4122 separate function, and given an option for scanning Cc: and Bcc: (which is
4123 not set for Exim filters).
4125 6. When Exim created an email address using the login of the caller as the
4126 local part (e.g. when creating a From: or Sender: header line), it was not
4127 quoting the local part when it contained special characters such as @.
4129 7. Installed new OpenBSD configuration files.
4131 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4132 try to make them clearer.
4134 9. Callout options, other than the timeout value, were being ignored when
4135 verifying sender addresses in header lines. For example, when using
4137 verify = header_sender/callout=no_cache
4139 the cache was (incorrectly) being used.
4141 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4142 only the Cygwin environment.
4144 11. When return_path_on_delivery was set as a log selector, if different remote
4145 addresses in the same message used different return paths and parallel
4146 remote delivery occurred, the wrong values would sometimes be logged.
4147 (Whenever a remote delivery process finished, the return path value from
4148 the most recently started remote delivery process was logged.)
4150 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4151 lines when AUTH and/or TLS are in use. This is the "received protocol"
4152 field. Exim used to use "asmtp" for authenticated SMTP, without any
4153 indication (in the protocol name) for TLS use. Now it follows the RFC and
4154 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4155 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4156 names appear in log lines as well as in Received: header lines.
4158 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4159 extensions, and comparison tests, and to fix some bugs.
4161 14. Changes to the "personal" filter test:
4163 (1) The test was buggy in that it was just doing the equivalent of
4164 "contains" tests on header lines. For example, if a user's address was
4165 anne@some.where, the "personal" test would incorrectly be true for
4167 To: susanne@some.where
4169 This test is now done by extracting each address from the header in turn,
4170 and checking the entire address. Other tests that are part of "personal"
4171 are now done using regular expressions (for example, to check local parts
4172 of addresses in From: header lines).
4174 (2) The list of non-personal local parts in From: addresses has been
4175 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4176 taken from the Sieve specification recommendations.
4178 (3) If the message contains any header line starting with "List-" it is
4179 treated as non-personal.
4181 (4) The test for "circular" in the Subject: header line has been removed
4182 because it now seems ill-conceived.
4184 15. Minor typos in src/EDITME comments corrected.
4186 16. Installed latest exipick from John Jetmore.
4188 17. If headers_add on a router specified a text string that was too long for
4189 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4190 of string_sprintf() is now avoided.
4192 18. $message_body_size was not set (it was always zero) when running the DATA
4193 ACL and the local_scan() function.
4195 19. For the "mail" command in an Exim filter, no default was being set for
4196 the once_repeat time, causing a random time value to be used if "once" was
4197 specified. (If the value happened to be <= 0, no repeat happened.) The
4198 default is now 0s, meaning "never repeat". The "vacation" command was OK
4199 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4200 (I found it when inspecting the code).
4202 20. There is now an overall timeout for performing a callout verification. It
4203 defaults to 4 times the callout timeout, which applies to individual SMTP
4204 commands during the callout. The overall timeout applies when there is more
4205 than one host that can be tried. The timeout is checked before trying the
4206 next host. This prevents very long delays if there are a large number of
4207 hosts and all are timing out (e.g. when the network connections are timing
4208 out). The value of the overall timeout can be changed by specifying an
4209 additional sub-option for "callout", called "maxwait". For example:
4211 verify = sender/callout=5s,maxwait=20s
4213 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4214 to the end before writing, but this should make it even safer).
4216 22. Exim was forgetting that it had advertised PIPELINING for the second and
4217 subsequent messages on an SMTP connection. It was also not resetting its
4218 memory on STARTTLS and an internal HELO.
4220 23. When Exim logs an SMTP synchronization error within a session, it now
4221 records whether PIPELINING has been advertised or not.
4223 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4224 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4225 rather than long int.
4227 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4229 26. Added the never_mail option to autoreply.
4235 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4236 crash if the getsockname() call failed; this can happen if a connection is
4237 closed very soon after it is established. The problem was simply in the
4238 order in which certain operations were done, causing Exim to try to write
4239 to the SMTP stream before it had set up the file descriptor. The bug has
4240 been fixed by making things happen in the correct order.
4246 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4247 before the connection was closed, thus losing the rejection response.
4249 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4250 some early Solaris releases, but causes trouble in current releases where
4251 socklen_t is defined.
4253 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4256 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4259 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4260 that releasing the top of it at the end releases what was used for sub-
4261 expansions (unless the block got too big). However, discard this block if
4262 the first thing is a variable or header, so that we can use its block when
4263 it is dynamic (useful for very large $message_headers, for example).
4265 6. Lookups now cache *every* query, not just the most recent. A new, separate
4266 store pool is used for this. It can be recovered when all lookup caches are
4267 flushed. Lookups now release memory at the end of their result strings.
4268 This has involved some general refactoring of the lookup sources.
4270 7. Some code has been added to the store_xxx() functions to reduce the amount
4271 of flapping under certain conditions.
4273 8. log_incoming_interface used to affect only the <= reception log lines. Now
4274 it causes the local interface and port to be added to several more SMTP log
4275 lines, for example "SMTP connection from", and rejection lines.
4277 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4279 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4281 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4282 could be overwritten at the end of the current message (or the start of a
4283 new message if it was set in a HELO ACL). The value is now preserved for
4284 the duration of the SMTP connection.
4286 12. If a transport had a headers_rewrite setting, and a matching header line
4287 contained an unqualified address, that address was qualified, even if it
4288 did not match any rewriting rules. The underlying bug was that the values
4289 of the flags that permit the existence of unqualified sender and recipient
4290 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4291 non-local messages, and by -bnq for local messages) were not being
4292 preserved with the message after it was received.
4294 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4295 "next input=" as part of the text comprising the host identity instead of
4296 the correct text. The code was using the same buffer for two different
4297 strings. However, depending on which order the printing function evaluated
4298 its arguments, the bug did not always show up. Under Linux, for example, my
4299 test suite worked just fine.
4301 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4302 doesn't work with some older versions of Perl. It has been changed to "my",
4303 which in any case is probably the better facility to use.
4305 15. A really picky compiler found some instances of statements for creating
4306 error messages that either had too many or two few arguments for the format
4309 16. The size of the buffer for calls to the DNS resolver has been increased
4310 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4311 for addresses that have a lot of PTR records. This alleviates a problem; it
4312 does not fully solve it.
4314 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4315 buffer now truncates the list and logs the incident, which is the same
4316 action as happens when Exim is looking up a host name and its aliases.
4317 Previously in this situation something unpredictable would happen;
4318 sometimes it was "internal error: store_reset failed".
4320 18. If a server dropped the connection unexpectedly when an Exim client was
4321 using GnuTLS and trying to read a response, the client delivery process
4322 crashed while trying to generate an error log message.
4324 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4325 string, for example:
4327 warn message = H1: something\nH2: something
4329 the text was added as a single header line from Exim's point of view
4330 though it ended up OK in the delivered message. However, searching for the
4331 second and subsequent header lines using $h_h2: did not work. This has been
4332 fixed. Similarly, if a system filter added multiple headers in this way,
4333 the routers could not see them.
4335 20. Expanded the error message when iplsearch is called with an invalid key to
4336 suggest using net-iplsearch in a host list.
4338 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4339 ACLs are no longer actually imposed (and a message to that effect is
4342 22. If a "gecos" field in a passwd entry contained escaped characters, in
4343 particular, if it contained a \" sequence, Exim got it wrong when building
4344 a From: or a Sender: header from that name. A second bug also caused
4345 incorrect handling when an unquoted " was present following a character
4346 that needed quoting.
4348 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4349 was not being matched caselessly.
4351 24. Arranged for all hyphens in the exim.8 source to be escaped with
4354 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4355 properly. Recipient callout cache records were still being keyed to include
4356 the sender, even when use_sender was set false. This led to far more
4357 callouts that were necessary. The sender is no longer included in the key
4358 when use_sender is false.
4360 26. Added "control = submission" modifier to ACLs.
4362 27. Added the ${base62d: operator to decode base 62 numbers.
4364 28. dnsdb lookups can now access SRV records.
4366 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4367 the configuration file.
4369 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4370 (-v) mode. This makes the output for a verbose queue run more intelligible.
4372 31. Added a use_postmaster feature to recipient callouts.
4374 32. Added the $body_zerocount variable, containing the number of binary zero
4375 bytes in the message body.
4377 33. The time of last modification of the "new" subdirectory is now used as the
4378 "mailbox time last read" when there is a quota error for a maildir
4381 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4383 35. Added +ignore_unknown as a special item in host lists.
4385 36. Code for decoding IPv6 addresses in host lists is now included, even if
4386 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4387 address was recognized as an IP address, but was then not correctly decoded
4388 into binary, causing unexpected and incorrect effects when compared with
4395 1. Very minor rewording of debugging text in manualroute to say "list of
4396 hosts" instead of "hostlist".
4398 2. If verify=header_syntax was set, and a header line with an unqualified
4399 address (no domain) and a large number of spaces between the end of the
4400 name and the colon was received, the reception process suffered a buffer
4401 overflow, and (when I tested it) crashed. This was caused by some obsolete
4402 code that should have been removed. The fix is to remove it!
4404 3. When running in the test harness, delay a bit after writing a bounce
4405 message to get a bit more predictability in the log output.
4407 4. Added a call to search_tidyup() just before forking a reception process. In
4408 theory, someone could use a lookup in the expansion of smtp_accept_max_
4409 per_host which, without the tidyup, could leave open a database connection.
4411 5. Added the variables $recipient_data and $sender_data which get set from a
4412 lookup success in an ACL "recipients" or "senders" condition, or a router
4413 "senders" option, similar to $domain_data and $local_part_data.
4415 6. Moved the writing of debug_print from before to after the "senders" test
4418 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4419 problems for message scanning, either using a data ACL, or using
4420 local_scan() because the Received: header was not generated till after they
4421 were called (in order to set the time as the time of reception completion).
4422 I have revised the way this works. The header is now generated after the
4423 body is received, but before the ACL or local_scan() are called. After they
4424 are run, the timestamp in the header is updated.
4430 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4431 before starting a queue runner without re-exec. This happened only when
4432 deliver_drop_privilege was set or when the Exim user was set to root. The
4433 effect of the bug was that timeouts during subsequent deliveries caused
4434 crashes instead of being properly handled. The handler is now left at its
4435 default (and expected) setting.
4437 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4438 message, again when deliver_drop_privilege is set or Exim is run as root.
4439 The bug described in (1) was not present in this case, but the tidying up
4440 of the other signals was missing. I have made the two cases consistent.
4442 3. The ignore_target_hosts setting on a manualroute router was being ignored
4443 for hosts that were looked up using the /MX notation.
4445 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4448 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4449 operated on the sender address. After changing the $sender_address to <>
4450 for the sender address verify, Exim was re-instated it as the original
4451 (before rewriting) address, but remembering that it had rewritten it, so it
4452 wasn't rewriting it again. This bug also had the effect of breaking the
4453 sender address verification caching when the sender address was rewritten.
4455 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4456 This has been changed so that if the ip literal address matches
4457 ignore_target_hosts, the router declines.
4459 7. Added expansion conditions match_domain, match_address, and match_local_
4460 part (NOT match_host).
4462 8. The placeholder for the Received: header didn't have a length field set.
4464 9. Added code to Exim itself and to exim_lock to test for a specific race
4465 condition that could lead to file corruption when using MBX delivery. The
4466 issue is with the lockfile that is created in /tmp. If this file is removed
4467 after a process has opened it but before that process has acquired a lock,
4468 there is the potential for a second process to recreate the file and also
4469 acquire a lock. This could lead to two Exim processes writing to the file
4470 at the same time. The added code performs the same test as UW imapd; it
4471 checks after acquiring the lock that its file descriptor still refers to
4472 the same named file.
4474 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4475 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4478 11. Added the smtp_active_hostname option. If used, this will typically be made
4479 to depend on the incoming interface address. Because $interface_address is
4480 not set up until the daemon has forked a reception process, error responses
4481 that can happen earlier (such as "too many connections") no longer contain
4484 12. If an expansion in a condition on a "warn" statement fails because a lookup
4485 defers, the "warn" statement is abandoned, and the next ACL statement is
4486 processed. Previously this caused the whole ACL to be aborted.
4488 13. Added the iplsearch lookup type.
4490 14. Added ident_timeout as a log selector.
4492 15. Added tls_certificate_verified as a log selector.
4494 16. Added a global option tls_require_ciphers (compare the smtp transport
4495 option of the same name). This controls incoming TLS connections.
4497 17. I finally figured out how to make tls_require_ciphers do a similar thing
4498 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4499 before starting the TLS session.
4501 18. Tabs are now shown as \t in -bP output.
4503 19. If the log selector return_path_on_delivery was set, Exim crashed when
4504 bouncing a message because it had too many Received: header lines.
4506 20. If two routers both had headers_remove settings, and the first one included
4507 a superfluous trailing colon, the final name in the first list and the
4508 first name in the second list were incorrectly joined into one item (with a
4509 colon in the middle).
4515 1. Added -C and -D options to the exinext utility, mainly to make it easier
4516 to include in the automated testing, but these could be helpful when
4517 multiple configurations are in use.
4519 2. The exinext utility was not formatting the output nicely when there was
4520 an alternate port involved in the retry record key, nor when there was a
4521 message id as well (for retries that were specific to a specific message
4522 and a specific host). It was also confused by IPv6 addresses, because of
4523 the additional colons they contain. I have fixed the IPv4 problem, and
4524 patched it up to do a reasonable job for IPv6.
4526 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4527 delivery, the log line now contains "pipelined" if PIPELINING was used.
4529 4. An SMTP transport process used to panic and die if the bind() call to set
4530 an explicit outgoing interface failed. This has been changed; it is now
4531 treated in the same way as a connect() failure.
4533 5. A reference to $sender_host_name in the part of a conditional expansion
4534 that was being skipped was still causing a DNS lookup. This no longer
4537 6. The def: expansion condition was not recognizing references to header lines
4538 that used bh_ and bheader_.
4540 7. Added the _cache feature to named lists.
4542 8. The code for checking quota_filecount in the appendfile transport was
4543 allowing one more file than it should have been.
4545 9. For compatibility with Sendmail, the command line option
4553 and sets the incoming protocol and host name (for trusted callers). The
4554 host name and its colon can be omitted when only the protocol is to be set.
4555 Note the Exim already has two private options, -pd and -ps, that refer to
4556 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4557 "s", but I don't think that's a major issue.
4559 10. A number of refactoring changes to the code, none of which should affect
4562 (a) The number of logging options was getting close to filling up the
4563 32-bit word that was used as a bit map. I have split them into two classes:
4564 those that are passed in the argument to log_write(), and those that are
4565 only ever tested independently outside of that function. These are now in
4566 separate 32-bit words, so there is plenty of room for expansion again.
4567 There is no change in the user interface or the logging behaviour.
4569 (b) When building, for example, log lines, the code previously used a
4570 macro that called string_cat() twice, in order to add two strings. This is
4571 not really sufficiently general. Furthermore, there was one instance where
4572 it was actually wrong because one of the argument was used twice, and in
4573 one call a function was used. (As it happened, calling the function twice
4574 did not affect the overall behaviour.) The macro has been replaced by a
4575 function that can join an arbitrary number of extra strings onto a growing
4578 (c) The code for expansion conditions now uses a table and a binary chop
4579 instead of a serial search (which was left over from when there were very
4580 few conditions). Also, it now recognizes conditions like "pam" even when
4581 the relevant support is not compiled in: a suitably worded error message is
4582 given if an attempt is made to use such a condition.
4584 11. Added ${time_interval:xxxxx}.
4586 12. A bug was causing one of the ddress fields not to be passed back correctly
4587 from remote delivery subprocesses. The field in question was not being
4588 subsequently used, so this caused to problems in practice.
4590 13. Added new log selectors queue_time and deliver_time.
4592 14. Might have fixed a bug in maildirsizefile handling that threw up
4593 "unexpected character" debug warnings, and recalculated the data
4594 unnecessarily. In any case, I expanded the warning message to give more
4597 15. Added the message "Restricted characters in address" to the statements in
4598 the default ACL that block characters like @ and % in local parts.
4600 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4601 Three changes have been made:
4603 (a) There was a serious bug; a negative response to MAIL caused the whole
4604 recipient domain to be cached as invalid, thereby blocking all messages
4605 to all local parts at the same domain, from all senders. This bug has
4606 been fixed. The domain is no longer cached after a negative response to
4607 MAIL if the sender used is not empty.
4609 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4612 (c) A new callout option, "use_sender" has been added for people who want
4613 the modified behaviour.
4619 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4622 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4623 indeed breaks things for older releases.
4625 3. Added additional logging to the case where there is a problem reading data
4626 from a filter that is running in a subprocess using a pipe, in order to
4627 try to track down a specific problem.
4629 4. Testing facility fudge: when running in the test harness and attempting
4630 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4631 getting "No route to host". Convert this to a timeout.
4633 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4636 6. Some OS don't have socklen_t but use size_t instead. This affects the
4637 fifth argument of getsockopt() amongst other things. This is now
4638 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4639 can be set for individual OS. I have set it for SunOS5, OSF1, and
4640 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4641 some earlier ones do not.
4643 7. Change 4.30/15 was not doing the test caselessly.
4645 8. The standard form for an IPv6 address literal was being rejected by address
4646 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4647 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4648 this, as well as the form without the "IPv6" on the front (but only when
4649 address literals are enabled, of course).
4651 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4653 10. Exim crashed if a message with an empty sender address specified by -f
4654 encountered a router with an errors_to setting. This could be provoked only
4655 by a command such as
4659 where an empty string was supplied; "<>" did not hit this bug.
4661 11. Installed PCRE release 4.5.
4663 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4664 remained set. It is now erased.
4666 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4667 times from message ids (which are base 36 rather than the normal 62).
4669 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4670 were being counted as actual protocol errors, and logged if the log
4671 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4672 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4673 and DATA following a set of rejected RCPTs do not count as protocol errors.
4674 In other words, Exim assumes they were pipelined, though this may not
4675 actually be the case. Of course, in all cases the client gets an
4676 appropriate error code.
4678 15. If a lookup fails in an ACL condition, a message about the failure may
4679 be available; it is used if testing the ACL cannot continue, because most
4680 such messages specify what the cause of the deferral is. However, some
4681 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4682 that caused an old message to be retained and used if a later statement
4683 caused a defer, replacing the real cause of the deferral.
4685 16. If an IP address had so many PTR records that the DNS lookup buffer
4686 was not large enough to hold them, Exim could crash while trying to process
4687 the truncated data. It now detects and logs this case.
4689 17. Further to 4.21/58, another change has been made: if (and only if) the
4690 first line of a message (the first header line) ends with CRLF, a bare LF
4691 in a subsequent header line has a space inserted after it, so as not to
4692 terminate the header.
4694 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4695 unnecessarily, used atoi() instead of strtol(), and didn't check the
4696 termination when getting file sizes from file names by regex.
4698 19. Completely re-implemented the support for maildirsize files, in the light
4699 of a number of problems with the previous contributed implementation
4700 (4.30/29). In particular:
4702 . If the quota is zero, the maildirsize file is maintained, but no quota is
4705 . If the maildir directory does not exist, it is created before any attempt
4706 to write a maildirsize file.
4708 . The quota value in the file is just a cache; if the quota is changed in
4709 the transport, the new value overrides.
4711 . A regular expression is available for excluding directories from the
4714 20. The autoreply transport checks the characters in options that define the
4715 message's headers; it allows continued headers, but it was checking with
4716 isspace() after an embedded newline instead of explicitly looking for a
4719 21. If all the "regular" hosts to which an address was routed had passed their
4720 expiry times, and had not reached their retry times, the address was
4721 bounced, even if fallback hosts were defined. Now Exim should go on to try
4724 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4725 equivalent code in the SMTP transport. Some hosts send humungous responses
4726 to HELO/EHLO, more than 1024 it seems.
4728 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4729 gives compiler warnings in some environments. I've now done it "properly",
4732 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4733 (because of problems with the built-in one) was declared to return uschar *
4734 instead of char *, causing compiler failure.
4736 25. Fixed a file descriptor leak when processing alias/forward files.
4738 26. Fixed a minor format string issue in dbfn.c.
4740 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4742 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4743 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4746 29. When a sender address is verified, it is cached, to save repeating the test
4747 when there is more than one recipient in a message. However, when the
4748 verification involves a callout, it is possible for different callout
4749 options to be set for different recipients. It is too complicated to keep
4750 track of this in the cache, so now Exim always runs a verification when a
4751 callout is required, relying on the callout cache for the optimization.
4752 The overhead is duplication of the address routing, but this should not be
4755 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4756 to be verified with callout=postmaster, and the main callout worked but the
4757 postmaster check failed, the verification correctly failed. However, if a
4758 subsequent RCPT command asked for sender verification *without* the
4759 postmaster check, incorrect caching caused this verification also to fail,
4762 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4763 it was not caching the DNS options (qualify_single, search_parents) that
4764 were used when the lookup failed. A subsequent lookup with different
4765 options therefore always gave the same answer, though there were cases
4766 where it should not have. (Example: a "domains = !$mx_any" option on a
4767 dnslookup router: the "domains" option is always processed without any
4768 widening, but the router might have qualify_single set.) Now Exim uses the
4769 cached value only when the same options are set.
4771 32. Added John Jetmore's "exipick" utility to the distribution.
4773 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4774 than a timeout (e.g. a certificate is required, and is not provided), an
4775 Exim server now closes the connection immediately. Previously it waited for
4776 the client to close - but if the client is SSL, it seems that they each
4777 wait for each other, leading to a delay before one of them times out.
4779 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4780 maintained 0.8.x compatibility because I don't think many are using it, and
4781 it is clearly obsolete.
4783 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4786 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4787 client certificate was expired. A simple patch fixes this, though I don't
4788 understand the full logic of why the verify callback is called multiple
4791 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4792 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4793 which causes problems with some clients (such as the Certicom SSL Plus
4794 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4795 disables the coutermeasure allowing Eudora to connect."
4797 38. Exim was not checking that a write() to a log file succeeded. This could
4798 lead to Bad Things if a log got too big, in particular if it hit a file
4799 size limit. Exim now panics and dies if it cannot write to a log file, just
4800 as it does if it cannot open a log file.
4802 39. Modified OS/Makefile-Linux so that it now contains
4804 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4806 The two -D definitions ensure that Exim is compiled with large file
4807 support, which makes it possible to handle log files that are bigger than
4810 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4811 instance) a domain was checked against a named list that involved a lookup,
4812 causing $domain_data to be set, then another domain was checked against the
4813 same list, then the first domain was re-checked, the value of $domain_data
4814 after the final check could be wrong. In particular, if the second check
4815 failed, it could be set empty. This bug probably also applied to
4818 41. The strip_trailing_dot option was not being applied to the address given
4819 with the -f command-line option.
4821 42. The code for reading a message's header from the spool was incrementing
4822 $received_count, but never initializing it. This meant that the value was
4823 incorrect (doubled) while delivering a message in the same process in which
4824 it was received. In the most common configuration of Exim, this never
4825 happens - a fresh exec is done - but it can happen when
4826 deliver_drop_privilege is set.
4828 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4829 it now includes up to 150 characters of the unexpected data in the log
4832 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4833 and building data strings. The size of both of these buffers was 10 000
4834 bytes - far larger than anybody would *ever* want, thought I. Needless to
4835 say, somebody hit the limit. I have increased the maximum line length to
4836 20 000 and the maximum data length of concatenated lines to 100 000. I have
4837 also fixed two bugs, because there was no checking on these buffers. Tsk,
4838 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4839 buffer is too small.
4841 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4842 lsearch lookups. Now it does.
4844 46. When parsing a route_list item in a manualroute router, a fixed-length
4845 buffer was used for the list of hosts. I made this 1024 bytes long,
4846 thinking that nobody would ever have a list of hosts that long. Wrong.
4847 Somebody had a whole pile of complicated expansion conditions, and the
4848 string was silently truncated, leading to an expansion error. It turns out
4849 that it is easier to change to an unlimited length (owing to other changes
4850 that have happened since this code was originally written) than to build
4851 structure for giving a limitation error. The length of the item that
4852 expands into the list of hosts is now unlimited.
4854 47. The lsearch lookup could not handle data where the length of text line was
4855 more than 4095 characters. Such lines were truncated, leading to shortened
4856 data being returned. It should now handle lines of any length.
4858 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4859 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4862 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4863 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4864 no longer happens in comment lines. A list of replacements is now placed
4865 at the head of all of the source files, except those whose only change is
4866 to replace PERL_COMMAND in the very first #! line.
4868 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4869 messages on the queue, with a bottom-up merge sort, using code contributed
4870 by Michael Haardt. This should make operations like -bp somewhat faster on
4871 large queues. It won't affect queue runners, except when queue_run_in_order
4874 51. Installed eximstats 1.31 in the distribution.
4876 52. Added support for SRV lookups to the dnslookup router.
4878 53. If an ACL referred to $message_body or $message_body_end, the value was not
4879 reset for any messages that followed in the same SMTP session.
4881 54. The store-handling optimization for building very long strings was not
4882 differentiating between the different store pools. I don't think this
4883 actually made any difference in practice, but I've tidied it.
4885 55. While running the routers to verify a sender address, $sender_address
4886 was still set to the sender address. This is wrong, because when routing to
4887 send a bounce to the sender, it would be empty. Therefore, I have changed
4888 it so that, while verifying a sender address, $sender_address is set to <>.
4889 (There is no change to what happens when verifying a recipient address.)
4891 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4892 target A or AAAA records (if not already returned) without resetting the
4893 qualify_single or search_parents options of the DNS resolver. These are
4894 inappropriate in this case because the targets of MX and SRV records must
4895 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4896 target that, when qualified, matched something in the local domain. These
4897 two options are now turned off when doing these lookups.
4899 57. It seems that at least some releases of Reiserfs (which does not have the
4900 concept of a fixed number of inodes) returns zero and not -1 for the
4901 number of available inodes. This interacted badly with check_spool_inodes,
4902 which assumed that -1 was the "no such thing" setting. What I have done is
4903 to check that the total number of inodes is greater than zero before doing
4904 the test of how many are available.
4906 58. When a "warn" ACL statement has a log_message modifier, the message is
4907 remembered, and not repeated. This is to avoid a lot of repetition when a
4908 message has many recipients that cause the same warning to be written.
4909 Howewer, Exim was preserving the list of already written lines for an
4910 entire SMTP session, which doesn't seem right. The memory is now reset if a
4911 new message is started.
4913 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4914 debugging output unless log_rewrite was also set.
4916 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4917 of (int)(handle) when we know that handle contains (void *)(-1).
4919 61. The Exim daemon panic-logs an error return when it closes the incoming
4920 connection. However "connection reset by peer" seems to be common, and
4921 isn't really an error worthy of noting specially, so that particular error
4924 62. When Exim is trying to find all the local interfaces, it used to panic and
4925 die if the ioctl to get the interface flags failed. However, it seems that
4926 on at least one OS (Solaris 9) it is possible to have an interface that is
4927 included in the list of interfaces, but for which you get a failure error
4928 for this call. This happens when the interface is not "plumbed" into a
4929 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4930 failure of the "get flags" call assumes that the interface is down.
4932 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4933 makes life easier for people who are doing arithmetic on fields extracted
4934 from dates, where you often get leading zeros that should not be
4935 interpreted as octal.
4937 64. Added qualify_domain to the redirect router, to override the global
4940 65. If a pathologically long header line contained very many addresses (the
4941 report of this problem mentioned 10 000) and each of them was rewritten,
4942 Exim could use up a very large amount of memory. (It kept on making new
4943 copies of the header line as it rewrote, and never released the old ones.)
4944 At the expense of a bit more processing, the header rewriting function has
4945 been changed so that it no longer eats memory in this way.
4947 66. The generation of the Received: header has been moved from the time that a
4948 message starts to be received, to the time that it finishes. The timestamp
4949 in the Received: header should now be very close to that of the <= log
4950 line. There are two side-effects of this change:
4952 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4953 logged header lines no longer include the local Received: line, because
4954 it has not yet been created. The same applies to a copy of the message
4955 that is returned to a non-SMTP sender when a message is rejected.
4957 (b) When a filter file is tested using -bf, no additional Received: header
4958 is added to the test message. After some thought, I decided that this
4961 This change does not affect the value of $received_for. It is still set
4962 after address rewriting, but before local_scan() is called.
4964 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4966 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4967 gave an unhelpful panic error message, and a defer error. I have managed to
4968 change this behaviour so that it now rejects any supplied certificate,
4969 which seems right, as the list of acceptable certificates is empty.
4971 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4972 gave an unhelpful defer error. I have not managed to make this reject any
4973 supplied certificates, but the error message it gives is "no certificate
4974 supplied", which is not helpful.
4976 70. exigrep's output now also includes lines that are not associated with any
4977 message, but which match the given pattern. Implemented by a patch from
4978 Martin Sluka, which also tidied up the Perl a bit.
4980 71. Recipient callout verification, like sender verification, was using <> in
4981 the MAIL FROM command. This isn't really the right thing, since the actual
4982 sender may affect whether the remote host accepts the recipient or not. I
4983 have changed it to use the actual sender in the callout; this means that
4984 the cache record is now keyed on a recipient/sender pair, not just the
4985 recipient address. There doesn't seem to be a real danger of callout loops,
4986 since a callout by the remote host to check the sender would use <>.
4987 [SEE ABOVE: changed after hitting problems.]
4989 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4990 temporary errors. However, in the case of such a code being given after
4991 the end of a data transmission (i.e. after ".") Exim was failing to write
4992 a retry record for the message. (Yes, there was some broken host that was
4993 actually sending 8xx at this point.)
4995 73. An unknown lookup type in a host list could cause Exim to panic-die when
4996 the list was checked. (An example that provoked this was putting <; in the
4997 middle of a list instead of at the start.) If this happened during a DATA
4998 ACL check, a -D file could be left lying around. This kind of configuration
4999 error no longer causes Exim to die; instead it causes a defer errror. The
5000 incident is still logged to the main and panic logs.
5002 74. Buglet left over from Exim 3 conversion. The message "too many messages
5003 in one connection" was written to the rejectlog but not the mainlog, except
5004 when address rewriting (yes!) was being logged.
5006 75. Added write_rejectlog option.
5008 76. When a system filter was run not as root (that is, when system_filter_user
5009 was set), the values of the $n variables were not being returned to the
5010 main process; thus, they were not subsequently available in the $sn
5013 77. Added +return_path_on_delivery log selector.
5015 78. A connection timeout was being treated differently from recipients deferred
5016 when testing hosts_max_try with a message that was older than the host's
5017 retry timeout. (The host should not be counted, thus allowing all hosts to
5018 be tried at least once before bouncing.) This may have been the cause of an
5019 occasionally reported bug whereby a message would remain on the queue
5020 longer than the retry timeout, but would be bounced if a delivery was
5021 forced. I say "may" because I never totally pinned down the problem;
5022 setting up timeout/retry tests is difficult. See also the next item.
5024 79. The ultimate address timeout was not being applied to errors that involved
5025 a combination of host plus message (for example, a timeout on a MAIL
5026 command). When an address resolved to a number of possible hosts, and they
5027 were not all tried for each delivery (e.g. because of hosts_max_try), a
5028 message could remain on the queue longer than the retry timeout.
5030 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5033 81. Fixed an obscure SMTP outgoing bug which required at least the following
5034 conditions: (a) there was another message waiting for the same server;
5035 (b) the server returned 5xx to all RCPT commands in the first message so
5036 that the message was not completed; (c) the server dropped the connection
5037 or gave a negative response to the RSET that Exim sends to abort the
5038 transaction. The observed case was a dropped connection after DATA that had
5039 been sent in pipelining mode. That is, the server had advertised PIPELINING
5040 but was not implementing it correctly. The effect of the bug was incorrect
5041 behaviour, such as trying another host, and this could lead to a crash.
5047 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5048 and daemon.c were passed as pointers to ints; they should have been
5049 pointers to socklen_t variables (which are typically unsigned ints).
5051 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5054 3. Fixed a really odd bug that affected only the testing scheme; patching a
5055 certain fixed string in the binary changed the value of another string that
5056 happened to be identical to the end of the original first string.
5058 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5059 name", it returns that address as the IP address. On some operating
5060 systems (e.g. Solaris), it also passes back the IP address string as the
5061 "host name". However, on others (e.g. Linux), it passes back an empty
5062 string. Exim wasn't checking for this, and was changing the host name to an
5063 empty string, assuming it had been canonicized.
5065 5. Although rare, it is permitted to have more than one PTR record for a given
5066 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5067 all the names associated with an address, because they do in Solaris.
5068 However, it seems that they do not in Linux for data that comes from the
5069 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5070 I found this out when I moved to a new Linux workstation and tried to run
5071 the Exim test suite.
5073 To get round this problem I have changed the code so that it now does its
5074 own call to the DNS to look up PTR records when searching for a host name.
5075 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5076 addresses that are only in /etc/hosts are still found.
5078 This behaviour is, however, controlled by an option called host_lookup_
5079 order, which defaults to "bydns:byaddr". If people want to use the other
5080 order, or indeed, just use one or the other means of lookup, they can
5081 specify it in this variable.
5083 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5084 some operating systems, this comes back from gethostbyaddr() as an empty
5085 string, and this is what Exim used to test for. However, it seems that in
5086 other systems, "." is yielded. Exim now tests for this case too.
5088 7. The values of check_spool_space and check_log_space are now held internally
5089 as a number of kilobytes instead of an absolute number of bytes. If a
5090 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5091 kilobyte. This means that much larger values can be stored.
5093 8. Exim monitor: an attempt to get the action menu when not actually pointing
5094 at a message produces an empty menu entitled "No message selected". This
5095 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5096 no entries in it ("Shell widget menu has zero width and/or height"). So I
5097 have added a single, blank menu entry in this case.
5099 9. Added ${quote_local_part.
5101 10. MIME decoding is now applied to the contents of Subject: header lines when
5104 11. Now that a reference to $sender_host_address automatically causes a reverse
5105 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5106 host lookup before query-style lookups in lists that might use this
5107 variable. This has therefore been abolished, and the "net-" prefix is no
5108 longer necessary for query-style lookups.
5110 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5111 have been a typo for LFLAGS, so it has been changed.
5113 13. The install script calls Exim with "-C /dev/null" in order to find the
5114 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5115 to be output. Howeve, since Exim outputs its version number before the
5116 error, it didn't break the script. It just looked ugly. I fixed this by
5117 always allowing "-C /dev/null" if the caller is root.
5119 14. Ignore overlarge ACL variable number when reading spool file - insurance
5120 against a later release with more variables having written the file.
5122 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5123 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5124 this, as well as the form without the "IPv6" on the front.
5126 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5127 OS/Makefile-Darwin file.
5129 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5130 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5132 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5133 variable. However, some OS use ints for inodes. I've added cast to long int
5134 to get rid of the compiler warning.
5136 19. I had forgotten to lock out "/../" in configuration file names when
5137 ALT_CONFIG_PREFIX was set.
5139 20. Routers used for verification do not need to specify transports. However,
5140 if such a router generated a host list, and callout was configured, Exim
5141 crashed, because it could not find a port number from the (non-existent)
5142 transport. It now assumes port 25 in this circumstance.
5144 21. Added the -t option to exigrep.
5146 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5147 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5148 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5149 them is set, LOOKUP_LSEARCH is forced.
5151 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5153 24. Added sender and host information to the "rejected by local_scan()" log
5154 line; previously there was no indication of these.
5156 25. Added .include_if_exists.
5158 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5159 Linux. It turns out that not all file systems support this. Apparently some
5160 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5161 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5162 not supported on the file descriptor, is now ignored when Exim is trying to
5163 sync a directory. This applies only to Linux.
5165 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5167 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5168 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5171 29. Added support for maildirsize files from supplied patch (modified a bit).
5173 30. The use of :fail: followed by an empty string could lead Exim to respond to
5174 sender verification failures with (e.g.):
5176 550 Verification failed for <xxx>
5177 550 Sender verify failed
5179 where the first response line was missing the '-' that indicates it is not
5180 the final line of the response.
5182 31. The loop for finding the name of the user that called Exim had a hardwired
5183 limit of 10; it now uses the value of finduser_retries, which is used for
5184 all other user lookups.
5186 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5189 33. Exim was neglecting to zero errno before one call of strtol() when
5190 expanding a string and expecting an integer value. On some systems this
5191 resulted in spurious "integer overflow" errors. Also, it was casting the
5192 result into an int without checking.
5194 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5195 did not work. The code looks as if it has *never* worked, though it appears
5196 to have been documented since at least releast 1.62. I have made it work.
5198 35. The "timeout_DNS" error in retry rules, also documented since at least
5199 1.62, also never worked. As it isn't clear exactly what this means, and
5200 clearly it isn't a major issue, I have abolished the feature by treating it
5201 as "timeout", and writing a warning to the main and panic logs.
5203 36. The display of retry rules for -brt wasn't always showing the error code
5206 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5207 timeout_connect_A, timeout_connect_MX.
5209 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5210 to the empty sender.
5212 39. The daemon was not analysing the content of -oX till after it had closed
5213 stderr and disconnected from the controlling terminal. This meant that any
5214 syntax errors were only noted on the panic log, and the return code from
5215 the command was 0. By re-arranging the code a little, I've made the
5216 decoding happen first, so such errors now appear on stderr, and the return
5217 code is 1. However, the actual setting up of the sockets still happens in
5218 the disconnected process, so errors there are still only recorded on the
5221 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5222 connections (as happens on some IP stacks) was logged at start up time as
5223 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5224 it from "IPv6 and IPv4", which means that two separate sockets are being
5227 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5228 says whether AF_INET or AF_INET6 was passed as an argument.
5230 42. Exiwhat output was messed up when time zones were included in log
5233 43. Exiwhat now gives more information about the daemon's listening ports,
5234 and whether -tls-on-connect was used.
5236 44. The "port" option of the smtp transport is now expanded.
5238 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5239 silently ignored. Now an error message is written to the main and panic
5242 46. There's a new ACL modifier called "logwrite" which writes to a log file
5243 as soon as it is encountered.
5245 47. Added $local_user_uid and $local_user_gid at routing time.
5247 48. Exim crashed when trying to verify a sender address that was being
5250 49. Exim was recognizing only a space character after ".include". It now also
5251 recognizes a tab character.
5253 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5254 extracting the relevant information from the specification. The man page no
5255 longer contains scrambled data for the -d option, and I've added a section
5256 at the front about calling Exim under different names.
5258 51. Added "extra_headers" argument to the "mail" command in filter files.
5260 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5263 53. Installed eximstats 1.29.
5265 54. Added transport_filter_timeout as a generic transport option.
5267 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5268 Cc: header lines. This was required by RFC 822, but it not required by RFC
5271 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5272 incoming messages that did not have them. Now it does so only if the
5273 message originates locally, that is, if there is no associated remote host
5274 address. When Resent- header lines are present, this applies to the Resent-
5275 lines rather than the non-Resent- lines.
5277 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5278 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5280 58. Messages for configuration errors now include the name of the main
5281 configuration file - useful now that there may be more than one file in a
5282 list (.included file names were always shown).
5284 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5285 for those rare installations that do not start the daemon as root or run it
5286 setuid root. I've cut out the call to initgroups() if the daemon is not
5289 60. The Exim user and group can now be bound into the binary as text strings
5290 that are looked up at the start of Exim's processing.
5292 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5294 62. Added $mailstore_basename variable.
5296 63. Installed patch to sieve.c from Michael Haardt.
5298 64. When Exim failed to open the panic log after failing to open the main log,
5299 the original message it was trying to log was written to stderr and debug
5300 output, but if they were not available (the usual case in production), it
5301 was lost. Now it is written to syslog before the two lines that record the
5302 failures to open the logs.
5304 65. Users' Exim filters run in subprocesses under the user's uid. It is
5305 possible for a "deliver" command or an alias in a "personal" command to
5306 provoke an address rewrite. If logging of address rewriting is configured,
5307 this fails because the process is not running as root or exim. There may be
5308 a better way of dealing with this, but for the moment (because 4.30 needs
5309 to be released), I have disabled address rewrite logging when running a
5310 filter in a non-root, non-exim process.
5316 1. The buildconfig auxiliary program wasn't quoting the value set for
5317 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5318 not defined. This bug was masked in 4.22 by the effect that was fixed in
5321 2. Some messages that were rejected after a message id was allocated were
5322 shown as "incomplete" by exigrep. It no longer does this for messages that
5323 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5325 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5326 have allow_domain_literals set, the ID did not get logged in the <= line.
5327 Domain literals are now always recognized in Message-ID: header lines.
5329 4. The first argument for a ${extract expansion item is the key name or field
5330 number. Leading and trailing spaces in this item were not being ignored,
5331 causing some misleading effects.
5333 5. When deliver_drop_privilege was set, single queue runner processes started
5334 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5335 same command in the process it spins off) were not dropping privilege.
5337 6. When the daemon running as "exim" started a queue runner, it always
5338 re-executed Exim in the spun-off process. This is a waste of effort when
5339 deliver_drop_privilege is set. The new process now just calls the
5340 queue-runner function directly.
5346 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5349 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5350 ignored. Though the use of -oP was forcing the writing of a pid file, it
5351 was always written to the default place.
5353 3. If the message "no IP address found for host xxxx" is generated during
5354 incoming verification, it is now followed by identification of the incoming
5355 connection (so you can more easily find what provoked it).
5357 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5359 5. Added some features to "harden" Exim a bit more against certain attacks:
5361 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5362 be put in Local/Makefile. This is like the never_users runtime option,
5363 but it cannot be overridden. The default setting is "root".
5365 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5366 prefix string with which any file named in a -C command line option
5369 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5370 is retained for -C and -D only if the caller of Exim is root. Without
5371 it, the exim user may also use -C and -D and retain privilege.
5373 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5374 command line option is disabled.
5376 6. Macro names set by the -D option must start with an upper case letter, just
5377 like macro names defined in the configuration file.
5379 7. Added "dereference=" facility to LDAP.
5381 8. Two instances of the typo "uknown" in the source files are fixed.
5383 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5384 the Configure-Makefile script screwed up while processing it.
5386 10. Incorporated PCRE 4.4.
5388 11. The SMTP synchronization check was not operating right at the start of an
5389 SMTP session. For example, it could not catch a HELO sent before the client
5390 waited for the greeting. There is now a check for outstanding input at the
5391 point when the greeting is written. Because of the duplex, asynchronous
5392 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5393 way, but not yet received, when the check is performed.
5395 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5396 on TCP/IP sockets, because this apparently causes some broken clients to
5399 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5400 unchanged) from the Cygwin maintainer.
5402 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5403 was supported instead of testing for mbx. Effectively a typo.
5405 15. The spa authenticator server code was not checking that the input it
5406 received was valid base64.
5408 16. The debug output line for the "set" modifier in ACLs was not showing the
5409 name of the variable that was being set.
5411 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5413 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5414 lookup on its own. Something else was needed to trigger the lookup. For
5415 example, a match in host_lookup or the need for a host name in a host list.
5416 Now, if $sender_host_name is referenced and the host name has not yet been
5417 looked up, a lookup is performed. If the lookup fails, the variable remains
5418 empty, and $host_lookup_failed is set to "1".
5420 19. Added "eqi" as a case-independent comparison operator.
5422 20. The saslauthd authentication condition could segfault if neither service
5423 nor realm was specified.
5425 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5426 error message that was logged was misleading, and incoming SMTP
5427 connections were dropped. The message is now more accurate, and temporary
5428 errors are given to SMTP connections.
5430 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5431 (except RSET) with a 421 error, until QUIT is received. However, it was
5432 failing to send a response to QUIT.
5434 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5435 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5436 ACL is now run after the helo_try_verify_hosts code.
5438 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5439 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5440 case-independent, so other case variants are also recognized). Apparently
5441 some systems use these upper case variants.
5443 25. If more than two messages were waiting for the same host, and a transport
5444 filter was specified for the transport, Exim sent two messages over the
5445 same TCP/IP connection, and then failed with "socket operation on non-
5446 socket" when it tried to send the third.
5448 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5450 27. The extern definition of crypt16() in expand.c was not being excluded when
5451 the OS had its own crypt16() function.
5453 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5454 as a preferred synonym for return_size_limit, both as an option and as an
5457 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5459 30. Changed the default configuration ACL to relax the local part checking rule
5460 for addresses that are not in any local domains. For these addresses,
5461 slashes and pipe symbols are allowed within local parts, but the sequence
5462 /../ is explicitly forbidden.
5464 31. SPA server authentication was not clearing the challenge buffer before
5467 32. log_message in a "warn" ACL statement was writing to the reject log as
5468 well as to the main log, which contradicts the documentation and doesn't
5469 seem right (because no rejection is happening). So I have stopped it.
5471 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5472 However, I am unable to do any testing of this.
5474 34. Fixed an infelicity in the appendfile transport. When checking directories
5475 for a mailbox, to see if any needed to be created, it was accidentally
5476 using path names with one or more superfluous leading slashes; tracing
5477 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5479 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5480 contents are added to the log line that is written for every discarded
5481 recipient. (Previously a log_message setting was ignored.)
5483 36. The ${quote: operator now quotes the string if it is empty.
5485 37. The install script runs exim in order to find its version number. If for
5486 some reason other than non-existence or emptiness, which it checks, it
5487 could not run './exim', it was installing it with an empty version number,
5488 i.e. as "exim-". This error state is now caught, and the installation is
5491 38. An argument was missing from the function that creates an error message
5492 when Exim fails to connect to the socket for saslauthd authentication.
5493 This could cause Exim to crash, or give a corrupted message.
5495 39. Added isip, isip4, and isip6 to ${if conditions.
5497 40. The ACL variables $acl_xx are now saved with the message, and can be
5498 accessed later in routers, transports, and filters.
5500 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5501 strings in the file are not string-expanded.
5503 42. If a MAIL command specified a SIZE value that was too large to fit into an
5504 int variable, the check against message_size_limit failed. Such values are
5505 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5506 day this will have to be increased, but I don't think I want to be around
5507 when emails are that large.
5514 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5515 iconv() is not standard in FreeBSD.
5517 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5518 IPv6 enabled. The observed symptom was a segmentation fault on return from
5519 the function os_common_find_running_interfaces() in src/os.c.
5521 3. In the check_special_case() function in daemon.c I had used "errno" as an
5522 argument name, which causes warnings on some systems. This was basically a
5523 typo, since it was named "eno" in the comments!
5525 4. The code that waits for the clock to tick (at a resolution of some fraction
5526 of a second) so as to ensure message-id uniqueness was always waiting for
5527 at least one whole tick, when it could have waited for less. [This is
5528 almost certainly not relevant at current processor speeds, where it is
5529 unlikely to ever wait at all. But we try to future-proof.]
5531 5. The function that sleeps for a time interval that includes fractions of a
5532 second contained a race. It did not block SIGALRM between setting the
5533 timer, and suspending (a couple of lines later). If the interval was short
5534 and the sigsuspend() was delayed until after it had expired, the suspension
5535 never ended. On busy systems this could lead to processes getting stuck for
5538 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5539 process, before it forks any delivery processes. The open lookup caching
5540 mechanism meant that the open file or database connection was passed into
5541 the delivery process. The problem was that delivery processes always tidy
5542 up cached lookup data. This could cause a problem for the next delivery
5543 process started by the queue runner, because the external queue runner
5544 process does not know about the closure. So the next delivery process
5545 still has data in the lookup cache. In the case of a file lookup, there was
5546 no problem because closing a file descriptor in a subprocess doesn't affect
5547 the parent. However, if the lookup was caching a connection to a database,
5548 the connection was closed, and the second delivery process was likely to
5549 see errors such as "PGSQL: query failed: server closed the connection
5550 unexpectedly". The problem has been fixed by closing all cached lookups
5551 in a queue runner before running a delivery process.
5553 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5554 seem to have the "const" qualifier which it has on other OS. I've
5557 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5558 *of the same type* (client or server) with the same public name that an
5559 error should be diagnosed.
5561 9. When Exim looked up a host name for an IP address, but failed to find the
5562 original IP address when looking up the host name (a safety check), it
5563 output the message "<ip address> does not match any IP for NULL", which was
5564 confusing, to say the least. The bug was that the host name should have
5565 appeared instead of "NULL".
5567 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5568 user that is built into the binary, and the -C or -D options is used, root
5569 privilege is dropped before the configuration file is read. In addition,
5570 logging is switched to stderr instead of the normal log files. If the
5571 configuration then re-defines the Exim user, the unprivileged environment
5572 is probably not what is expected, so Exim logs a panic warning message (but
5575 However, if deliver_drop_privilege is set, the unprivileged state may well
5576 be exactly what is intended, so the warning has been cut out in that case,
5577 and Exim is allowed to try to write to its normal log files.
5583 1. smtp_return_error_details was not giving details for temporary sender
5584 or receiver verification errors.
5586 2. Diagnose a configuration error if two authenticators have the same public
5589 3. Exim used not to create the message log file for a message until the first
5590 delivery attempt. This could be confusing when incoming messages were held
5591 for policy or load reasons. The message log file is now created at the time
5592 the message is received, and an initial "Received" line is written to it.
5594 4. The automatically generated man page for command line options had a minor
5595 bug that caused no ill effects; however, a more serious problem was that
5596 the procedure for building the man page automatically didn't always
5597 operate. Consequently, release 4.20 contains an out-of-date version. This
5598 shouldn't happen again.
5600 5. When building Exim with embedded Perl support, the script that builds the
5601 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5602 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5604 6. The freeze_tell option was not being used for messages that were frozen on
5605 arrival, either by an ACL or by local_scan().
5607 7. Added the smtp_incomplete_transaction log selector.
5609 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5610 was accepting AUTH without a new EHLO.
5612 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5613 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5616 10. Logging of TCP/IP connections (when configured) now happens in the main
5617 daemon process instead of the child process, so that the TCP/IP connection
5618 count is more accurate (but it can never be perfect).
5620 11. The use of "drop" in a nested ACL was not being handled correctly in the
5621 outer ACL. Now, if condition failure induced by the nested "drop" causes
5622 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5623 or "require"), the connection is dropped.
5625 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5626 that yield "discard" can now be used with an "accept" or a "discard" verb,
5627 but an error is generated for any others (because I can't see a useful way
5628 to define what should happen).
5630 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5631 are now processed in the same way as lines in the Exim configuration file.
5632 In particular, continuation lines are supported.
5634 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5636 15. Added -ti meaning -t -i.
5638 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5639 domains, and warn by logging if others are found.
5641 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5642 returned from GIFCONF when Exim is trying to find the list of interfaces on
5643 a host. The code in os.c has been modified to copy each ifreq to an aligned
5644 structure in all cases.
5646 Also, in some cases, the returned ifreq's were being copied to a 'struct
5647 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5648 means the last couple of bytes of an IPv6 address could be chopped if the
5649 ifreq contained only a normal sockaddr (14 bytes storage).
5651 18. Named domain lists were not supported in the hosts_treat_as_local option.
5652 An entry such as +xxxx was not recognized, and was treated as a literal
5655 19. Ensure that header lines added by a DATA ACL are included in the reject log
5656 if the ACL subsequently rejects the message.
5658 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5659 MD5 (which is deprecated).
5661 21. When testing a filter file using -bf, Exim was writing a message when it
5662 took the sender from a "From " line in the message, but it was not doing so
5663 when it took $return_path from a Return-Path: header line. It now does.
5665 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5666 with a valid header line field name (a series of printing characters
5667 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5669 23. Changed "disc" in the source to "disk" to conform to the documentation and
5670 the book and for uniformity.
5672 24. Ignore Sendmail's -Ooption=value command line item.
5674 25. When execve() failed while trying to run a command in a pipe transport,
5675 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5676 could be confused with a return value of 69 from the command itself. This
5677 has been changed to 127, the value the shell returns if it is asked to run
5678 a non-existent command. The wording for the related log line suggests a
5679 non-existent command as the problem.
5681 26. If received_header_text expands to an empty string, do not add a Received:
5682 header line to the message. (Well, it adds a token one on the spool, but
5683 marks it "old" so that it doesn't get used or transmitted.)
5685 27. Installed eximstats 1.28 (addition of -nt option).
5687 28. There was no check for failure on the call to getsockname() in the daemon
5688 code. This can fail if there is a shortage of resources on the system, with
5689 ENOMEM, for example. A temporary error is now given on failure.
5691 29. Contrary to the C standard, it seems that in some environments, the
5692 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5693 program. Exim now does this explicitly; it affects the formatting of
5694 timestamps using strftime().
5696 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5697 complaints. I've now initialized all the variables, to avoid this.
5699 32. Header lines added by a system filter were not being "seen" during
5700 transport-time rewrites.
5702 33. The info_callback() function passed to OpenSSL is set up with type void
5703 (*)(SSL *, int, int), as described somewhere. However, when calling the
5704 function (actually a macro) that sets it up, the type void(*)() is
5705 expected. I've put in a cast to prevent warnings from picky compilers.
5707 34. If a DNS black list lookup found a CNAME record, but there were no A
5708 records associated with the domain it pointed at, Exim crashed.
5710 35. If a DNS black list lookup returned more than one A record, Exim ignored
5711 all but the first. It now scans all returned addresses if a particular IP
5712 value is being sought. In this situation, the contents of the
5713 $dnslist_value variable are a list of all the addresses, separated by a
5716 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5717 to accept a host name and all its aliases if the forward lookup for any of
5718 them yielded the IP address of the incoming connection. Now it accepts only
5719 those names whose forward lookup yields the correct IP address. Any other
5720 names are discarded. This closes a loophole whereby a rogue DNS
5721 administrator could create reverse DNS records to break through a
5722 wildcarded host restriction in an ACL.
5724 37. If a user filter or a system filter that ran in a subprocess used any of
5725 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5726 the wrong values were passed to the pipe command ($thisaddress had the
5727 value of $0, $0 had the value of $1, etc). This bug was introduced by
5728 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5730 38. Improved the line breaking for long SMTP error messages from ACLs.
5731 Previously, if there was no break point between 40 and 75 characters, Exim
5732 left the rest of the message alone. Two changes have been made: (a) I've
5733 reduced the minimum length to 35 characters; (b) if it can't find a break
5734 point between 35 and 75 characters, it looks ahead and uses the first one
5735 that it finds. This may give the occasional overlong line, but at least the
5736 remaining text gets split now.
5738 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5739 file descriptors might be low, and that setting 1000 would always raise it.
5740 It turns out that in some environments, the limit is already over 1000 and
5741 that lowering it causes trouble. So now Exim takes care not to decrease it.
5743 40. When delivering a message, the value of $return_path is set to $sender_
5744 address at the start of routing (routers may change the value). By an
5745 oversight, this default was not being set up when an address was tested by
5746 -bt or -bv, which affected the outcome if any router or filter referred to
5749 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5750 log only when "message" or "log_message" are set. However, if one of the
5751 conditions was an address verification, or a call to a nested ACL, the
5752 messages generated by the underlying test were being passed through. This
5753 no longer happens. The underlying message is available in $acl_verify_
5754 message for both "message" and "log_message" expansions, so it can be
5755 passed through if needed.
5757 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5758 new expansion $bh_ to give the encoded byte string without charset
5759 translation. Translation happens only if iconv() is available; HAVE_ICONV
5760 indicates this at build time. HEADERS_CHARSET gives the charset to
5761 translate to; headers_charset can change it in the configuration, and
5762 "headers charset" can change it in an individual filter file.
5764 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5765 that creates RFC 2047 encoded "words" labels them as that charset instead
5766 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5767 expansion operator; (ii) when Exim creates a From: line for a local
5768 message; (iii) when a header line is rewritten to include a "phrase" part.
5770 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5771 buggy, causing it to skip the first lines of messages whose message ID
5772 ended in 'D'. This would not have bitten before Exim release 4.14, because
5773 message IDs were unlikely to end in 'D' before then. The effect was to have
5774 incorrect size information for certain domains.
5776 45. #include "config.h" was missing at the start of the crypt16.c module. This
5777 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5780 46. If there was a timeout during a "random" callout check, Exim treated it as
5781 a failure of the random address, and carried on sending RSET and the real
5782 address. If the delay was just some slowness somewhere, the response to the
5783 original RCPT would be taken as a response to RSET and so on, causing
5784 mayhem of various kinds.
5786 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5787 when I implemented it. It didn't allow for the fact that some option values
5788 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5789 the right test for positive values.
5791 48. Domain names in DNS records are case-independent. Exim always looks them up
5792 in lower case. Some resolvers return domain names in exactly the case they
5793 appear in the zone file, that is, they may contain uppercase letters. Not
5794 all resolvers do this - some return always lower case. Exim was treating a
5795 change of case by a resolver as a change of domain, similar to a widening
5796 of a domain abbreviation. This triggered its re-routing code and so it was
5797 trying to route what was effectively the same domain again. This normally
5798 caused routing to fail (because the router wouldn't handle the domain
5799 twice). Now Exim checks for this case specially, and just changes the
5800 casing of the domain that it ultimately uses when it transmits the message
5803 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5806 50. If a filter generated a file delivery with a non-absolute name (possible if
5807 no home directory exists for the router), the forbid_file option was not
5810 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5811 the existing equality matching.
5813 52. Exim was using ints instead of ino_t variables in some places where it was
5814 dealing with inode numbers.
5816 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5817 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5818 called TMPDIR, and if it finds it is different, it changes its value.
5820 54. The smtp_printf() function is now made available to local_scan() so
5821 additional output lines can be written before returning. There is also an
5822 smtp_fflush() function to enable the detection of a dropped connection.
5823 The variables smtp_input and smtp_batched_input are exported to
5826 55. Changed the default runtime configuration: the message "Unknown user"
5827 has been removed from the ACL, and instead placed on the localuser router,
5828 using the cannot_route_message feature. This means that any verification
5829 failures that generate their own messages won't get overridden. Similarly,
5830 the "Unrouteable address" message that was in the ACL for unverifiable
5831 relay addresses has also been removed.
5833 56. Added hosts_avoid_esmtp to the smtp transport.
5835 57. The exicyclog script was not checking for the esoteric option
5836 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5837 will work only if exicyclog is run under the appropriate euid.
5839 58. Following a discussion on the list, the rules by which Exim recognises line
5840 endings on incoming messages have been changed. The -dropcr and drop_cr
5841 options are now no-ops, retained only for backwards compatibility. The
5842 following line terminators are recognized: LF CRLF CR. However, special
5843 processing applies to CR:
5845 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5846 nor a local message in the state where . is a terminator.
5848 (ii) If a bare CR is encountered in a header line, an extra space is added
5849 after the line terminator so as not to end the header. The reasoning
5850 behind this is that bare CRs in header lines are most likely either
5851 to be mistakes, or people trying to play silly games.
5853 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5854 was being incorrectly given as 18 bytes larger than it should have been.
5855 This is a VOB (very old bug).
5857 60. This may never have affected anything current, but just in case it has:
5858 When the local host is found other than at the start of a list of hosts,
5859 the local host, those with the same MX, and any that follow, are discarded.
5860 When the list in question was part of a longer list of hosts, the following
5861 hosts (not currently being processed) were also being discarded. This no
5862 longer happens. I'm not sure if this situation could ever has previously
5865 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5868 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5869 header. This is something that is recommended in a new Internet Draft, and
5870 is something that is documented as being done by Sendmail. There are two
5871 possible values. For messages generated by the autoreply transport, Exim
5874 Auto-Submitted: auto-replied
5876 whereas for all other generated messages (e.g. bounces) it adds
5878 Auto-Submitted: auto-generated
5880 63. The "personal" condition in filters now includes a test for the
5881 Auto-Submitted: header. If it contains the string "auto-" the message it
5882 not considered personal.
5884 64. Added rcpt_include_affixes as a generic transport option.
5886 65. Added queue_only_override (default true).
5888 66. Added the syslog_duplication option.
5890 67. If what should have been the first header line of a message consisted of
5891 a space followed by a colon, Exim was mis-interpreting it as a header line.
5892 It isn't of course - it is syntactically invalid and should therefore be
5893 treated as the start of the message body. The misbehaviour could have
5894 caused a number of strange effects, including loss of data in subsequent
5895 header lines, and spool format errors.
5897 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5898 client host had authenticated. This control can now be exercised by an ACL
5899 for more flexibility.
5901 69. By default, callouts do not happen when testing with -bh. There is now a
5902 variant, -bhc, which does actually run the callout code, including
5903 consulting and updating the callout cache.
5905 70. Added support for saslauthd authentication, courtesy of Alexander
5908 71. If statvfs() failed on the spool or log directories while checking their
5909 size for availability, Exim confusingly gave the error "space shortage".
5910 Furthermore, in debugging mode it crashed with a floating point exception.
5911 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5912 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5913 serious problem, Exim now writes to the main and panic logs when this
5914 happens, with details of the failure. It then refuses to accept the
5915 incoming message, giving the message "spool directory problem" or "log
5916 directory problem" with a 421 code for SMTP messages.
5918 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5919 0, 1, and 2 exist, because some OS complain for execs without them (see
5920 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5921 descriptors. However, the code omitted to check that the open succeeded,
5922 causing mysterious errors if for some reason the permissions on /dev/null
5923 got screwed. Now Exim writes a message to the main and panic logs, and
5924 bombs out if it can't open /dev/null.
5926 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5927 interact so that it is all more flexible. It is supposed to remain
5928 backwards compatible. Also added extra_local_interfaces.
5930 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5931 to bomb out with an assertion failure - to the client this appears as a
5932 connection drop. This problem occurs in the part of the code that was taken
5933 from the Samba project. Fortunately, the assertion is in a very simple
5934 function, so I have fixed this by reproducing the function inline in the
5935 one place where it is called, and arranging for authentication to fail
5936 instead of killing the process with assert().
5938 75. The SPA client code was not working when the server requested OEM rather
5939 than Unicode encoding.
5941 76. Added code to make require_files with a specific uid setting more usable in
5942 the case where statting the file as root fails - usually a non-root-mounted
5943 NFS file system. When this happens and the failure is EACCES, Exim now
5944 forks a subprocess and does the per-uid checking as the relevant uid.
5946 77. Added process_log_path.
5948 78. If log_file_path was not explicitly set, a setting of check_log_space or
5949 check_log_inodes was ignored.
5951 79. If a space check for the spool or log partitions fails, the incident is now
5952 logged. Of course, in the latter case the data may get lost...
5954 80. Added the %p formatting code to string_format() so that it can be used to
5955 print addresses in debug_print(). Adjusted all the address printing in the
5956 debugging in store.c to use %p rather than %d.
5958 81. There was a concern that a line of code in smtp_in.c could overflow a
5959 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5960 initially expressed, the concern was not well-founded, because trailing
5961 spaces are removed early. However, if the trailing spaces were followed by
5962 a NULL, they did not get removed, so the overflow was possible. Two fixes
5965 (a) I re-wrote the offending code in a cleaner fashion.
5966 (b) If an incoming SMTP command contains a NULL character, it is rejected
5969 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5970 runs initgroups(), so that if the Exim user is in any additional groups,
5971 they will be used during message reception.
5977 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob