1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
9 PP/02 Make -n do something, by making it not do something.
10 When combined with -bP, the name of an option is not output.
12 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
15 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
16 $sender_host_name and config options to manage this, and basic check
19 PP/05 DSCP support for outbound connections and control modifier for inbound.
21 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
22 (Only plugin which currently uses this is kerberos4, which nobody should
23 be using, but we should make it available and other future plugins might
24 conceivably use it, even though it would break NAT; stuff *should* be
25 using channel bindings instead).
27 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
28 name; added for Sendmail compatibility; requires admin caller.
29 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
30 just ignore it); requires trusted caller.
31 Also parse but ignore: -Ac -Am -X<logfile>
34 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
36 TL/02 Add +smtp_confirmation as a default logging option.
38 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
39 Patch by Magnus Holmgren from 2007-02-20.
41 TL/04 Bugzilla 1281 - Spec typo.
42 Bugzilla 1283 - Spec typo.
43 Bugzilla 1290 - Spec grammar fixes.
45 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
47 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
49 JH/02 Support "G" suffix to numbers in ${if comparisons.
51 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
53 NM/01 Bugzilla 1197 - Spec typo
54 Bugzilla 1196 - Spec examples corrections
56 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
58 PP/09 Add gnutls_enable_pkcs11 option.
60 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
61 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
63 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
64 "acl {{name}{arg}...}", and optional args on acl condition
67 JH/05 Permit multiple router/transport headers_add/remove lines.
69 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
71 JH/07 Avoid using a waiting database for a single-message-only transport.
72 Performance patch from Paul Fisher. Bugzilla 1262.
74 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
77 JH/09 Add $headers_added variable, with content from use of ACL modifier
78 add_header (but not yet added to the message). Bugzilla 199.
80 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
81 Pulled from Bugzilla 817 by Wolfgang Breyha.
83 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
86 JH/11 Add A= logging on delivery lines, and a client_set_id option on
89 JH/12 Add optional authenticated_sender logging to A= and a log_selector
92 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
94 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
95 advertise SMTP AUTH mechanism to us, instead of a generic
96 protocol violation error. Also, make Exim more robust to bad
97 data from the Dovecot auth socket.
99 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
101 When a queue runner is handling a message, Exim first routes the
102 recipient addresses, during which it prunes them based on the retry
103 hints database. After that it attempts to deliver the message to
104 any remaining recipients. It then updates the hints database using
107 So if a recipient address works intermittently, it can get repeatedly
108 deferred at routing time. The retry hints record remains fresh so the
109 address never reaches the final cutoff time.
111 This is a fairly common occurrence when a user is bumping up against
112 their storage quota. Exim had some logic in its local delivery code
113 to deal with this. However it did not apply to per-recipient defers
114 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
116 This change adds a proper retry rule check during routing so that the
117 final cutoff time is checked against the message's age. We only do
118 this check if there is an address retry record and there is not a
119 domain retry record; this implies that previous attempts to handle
120 the address had the retry_use_local_parts option turned on. We use
121 this as an approximation for the destination being like a local
122 delivery, as in LMTP.
124 I suspect this new check makes the old local delivery cutoff check
125 redundant, but I have not verified this so I left the code in place.
127 TF/02 Correct gecos expansion when From: is a prefix of the username.
129 Test 0254 submits a message to Exim with the header
133 When I ran the test suite under the user fanf2, Exim expanded
134 the header to contain my full name, whereas it should have added
135 a Resent-Sender: header. It erroneously treats any prefix of the
136 username as equal to the username.
138 This change corrects that bug.
144 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
146 This, or similar/improved, will also be change PP/11 of 4.82.
152 PP/01 Handle short writes when writing local log-files.
153 In practice, only affects FreeBSD (8 onwards).
154 Bugzilla 1053, with thanks to Dmitry Isaikin.
156 NM/01 Bugzilla 949 - Documentation tweak
158 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
161 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
163 PP/02 Implemented gsasl authenticator.
165 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
167 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
168 `pkg-config foo` for cflags/libs.
170 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
171 with rest of GSASL and with heimdal_gssapi.
173 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
174 `pkg-config foo` for cflags/libs for the TLS implementation.
176 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
177 properties get this fed in as external SSF. A number of robustness
178 and debugging improvements to the cyrus_sasl authenticator.
180 PP/08 cyrus_sasl server now expands the server_realm option.
182 PP/09 Bugzilla 1214 - Log authentication information in reject log.
183 Patch by Jeremy Harris.
185 PP/10 Added dbmjz lookup type.
187 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
189 PP/12 MAIL args handles TAB as well as SP, for better interop with
190 non-compliant senders.
191 Analysis and variant patch by Todd Lyons.
193 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
194 Bug report from Lars Müller <lars@samba.org> (via SUSE),
195 Patch from Dirk Mueller <dmueller@suse.com>
197 PP/13 tls_peerdn now print-escaped for spool files.
198 Observed some $tls_peerdn in wild which contained \n, which resulted
199 in spool file corruption.
201 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
202 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
203 or write after TLS renegotiation, which otherwise led to messages
206 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
207 as a tracking header (ie: a signed header comes before the signature).
208 Patch from Wolfgang Breyha.
210 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
211 comma-sep list; embedded commas doubled.
213 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
215 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
217 Report and patch from Dmitry Banschikov.
219 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
220 Removed SSL_clear() after SSL_new() which led to protocol negotiation
221 failures. We appear to now support TLS1.1+ with Exim.
223 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
224 lets Exim select keys and certificates based upon TLS SNI from client.
225 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
226 before an outbound SMTP session. New log_selector, +tls_sni.
228 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
229 NULL dereference. Report and patch from Alun Jones.
231 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
232 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
233 Not seeing resolver debug output on NetBSD, but suspect this is a
234 resolver implementation change.
236 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
237 Left warnings. Added "eximon gdb" invocation mode.
239 PP/21 Defaulting "accept_8bitmime" to true, not false.
241 PP/22 Added -bw for inetd wait mode support.
243 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
244 locate the relevant includes and libraries. Made this the default.
246 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
247 Bugzilla 1246, report and most of solution from Tomasz Kusy.
249 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
250 This may cause build issues on older platforms.
252 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
253 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
254 gnutls_require_mac & gnutls_require_protocols (no longer supported).
255 Added SNI support via GnuTLS too.
256 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
258 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
260 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
261 Applied second patch from Janne, implementing suggestion to default
262 multiple-strings-in-record handling to match SPF spec.
264 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
266 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
267 read-only, out of scope).
268 Patch from Wolfgang Breyha, report from Stuart Northfield.
270 PP/29 Fix three issues highlighted by clang analyser static analysis.
271 Only crash-plausible issue would require the Cambridge-specific
272 iplookup router and a misconfiguration.
273 Report from Marcin Mirosław.
275 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
277 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
278 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
279 As part of this, removing so much warning spew let me fix some minor
280 real issues in debug logging.
282 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
283 assignment on my part. Fixed.
285 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
286 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
287 Janne Snabb (who went above and beyond: thank you).
289 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
290 string otherwise requires a connection and a bunch more work and it's
291 relatively easy to get wrong. Should also expose TLS library linkage
294 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
295 64-bit ${eval} (JH/03).
297 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
298 GNU libc to support some of the 64-bit stuff, should not lead to
299 conflicts. Defined before os.h is pulled in, so if a given platform
300 needs to override this, it can.
302 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
303 protection layer was required, which is not implemented.
304 Bugzilla 1254, patch from Wolfgang Breyha.
306 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
307 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
308 tls_dhparam take prime identifiers. Also unbreak combination of
309 OpenSSL+DH_params+TLSSNI.
311 PP/39 Disable SSLv2 by default in OpenSSL support.
317 PP/01 Solaris build fix for Oracle's LDAP libraries.
318 Bugzilla 1109, patch from Stephen Usher.
320 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
322 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
325 TF/02 Fix a couple more cases where we did not log the error message
326 when unlink() failed. See also change 4.74-TF/03.
328 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
329 lock up or crash if it happened to be inside a call to libc when it
330 got a SIGUSR1 from exiwhat.
332 The SIGUSR1 handler appends the current process status to the process
333 log which is later printed by exiwhat. It used to use the general
334 purpose logging code to do this, but several functions it calls are
335 not safe for signals.
337 The new output code in the SIGUSR1 handler is specific to the process
338 log, and simple enough that it's easy to inspect for signal safety.
339 Removing some special cases also simplifies the general logging code.
340 Removing the spurious timestamps from the process log simplifies
343 TF/04 Improved ratelimit ACL condition.
345 The /noupdate option has been deprecated in favour of /readonly which
346 has clearer semantics. The /leaky, /strict, and /readonly update modes
347 are mutually exclusive. The update mode is no longer included in the
348 database key; it just determines when the database is updated. (This
349 means that when you upgrde Exim will forget old rate measurements.)
351 Exim now checks that the per_* options are used with an update mode that
352 makes sense for the current ACL. For example, when Exim is processing a
353 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
354 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
355 must specify per_mail/readonly. If you omit the update mode it defaults to
356 /leaky where that makes sense (as before) or /readonly where required.
358 The /noupdate option is now undocumented but still supported for
359 backwards compatibility. It is equivalent to /readonly except that in
360 ACLs where /readonly is required you may specify /leaky/noupdate or
361 /strict/noupdate which are treated the same as /readonly.
363 A useful new feature is the /count= option. This is a generalization
364 of the per_byte option, so that you can measure the throughput of other
365 aggregate values. For example, the per_byte option is now equivalent
366 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
368 The per_rcpt option has been generalized using the /count= mechanism
369 (though it's more complicated than the per_byte equivalence). When it is
370 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
371 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
372 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
373 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
374 Note that using per_rcpt with a non-readonly update mode in more than
375 one ACL will cause the recipients to be double-counted. (The per_mail
376 and per_byte options don't have this problem.)
378 The handling of very low rates has changed slightly. If the computed rate
379 is less than the event's count (usually one) then this event is the first
380 after a long gap. In this case the rate is set to the same as this event's
381 count, so that the first message of a spam run is counted properly.
383 The major new feature is a mechanism for counting the rate of unique
384 events. The new per_addr option counts the number of different
385 recipients that someone has sent messages to in the last time period. It
386 behaves like per_rcpt if all the recipient addresses are different, but
387 duplicate recipient addresses do not increase the measured rate. Like
388 the /count= option this is a general mechanism, so the per_addr option
389 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
390 example, measure the rate that a client uses different sender addresses
391 with the options per_mail/unique=$sender_address. There are further
392 details in the main documentation.
394 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
396 TF/06 Removed a few PCRE remnants.
398 TF/07 Automatically extract Exim's version number from tags in the git
399 repository when doing development or release builds.
401 PP/02 Raise smtp_cmd_buffer_size to 16kB.
402 Bugzilla 879. Patch from Paul Fisher.
404 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
405 Heavily based on revision 40f9a89a from Simon Arlott's tree.
408 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
410 PP/05 Variable $av_failed, true if the AV scanner deferred.
411 Bugzilla 1078. Patch from John Horne.
413 PP/06 Stop make process more reliably on build failure.
414 Bugzilla 1087. Patch from Heiko Schlittermann.
416 PP/07 Make maildir_use_size_file an _expandable_ boolean.
417 Bugzilla 1089. Patch from Heiko Schlittermann.
419 PP/08 Handle ${run} returning more data than OS pipe buffer size.
420 Bugzilla 1131. Patch from Holger Weiß.
422 PP/09 Handle IPv6 addresses with SPF.
423 Bugzilla 860. Patch from Wolfgang Breyha.
425 PP/10 GnuTLS: support TLS 1.2 & 1.1.
427 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
430 PP/11 match_* no longer expand right-hand-side by default.
431 New compile-time build option, EXPAND_LISTMATCH_RHS.
432 New expansion conditions, "inlist", "inlisti".
434 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
436 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
438 PP/14 fix log_write() format string regression from TF/03.
439 Bugzilla 1152. Patch from Dmitry Isaikin.
445 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
447 PP/02 Harmonised TLS library version reporting; only show if debugging.
448 Layout now matches that introduced for other libraries in 4.74 PP/03.
450 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
452 PP/04 New "dns_use_edns0" global option.
454 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
457 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
458 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
460 TK/01 Updated PolarSSL code to 0.14.2.
461 Bugzilla 1097. Patch from Andreas Metzler.
463 PP/07 Catch divide-by-zero in ${eval:...}.
466 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
469 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
470 format-string attack -- SECURITY: remote arbitrary code execution.
472 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
473 time unintentionally subject to list matching rules, letting the header
474 cause arbitrary Exim lookups (of items which can occur in lists, *not*
475 arbitrary string expansion). This allowed for information disclosure.
477 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
478 INT_MIN/-1 -- value coerced to INT_MAX.
484 NM/01 Workround for PCRE version dependency in version reporting
487 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
488 This fixes portability to compilers other than gcc, notably
489 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
491 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
492 makefiles for portability to HP-UX and POSIX correctness.
494 PP/01 Permit LOOKUP_foo enabling on the make command-line.
495 Also via indented variable definition in the Makefile.
496 (Debugging by Oliver Heesakkers).
498 PP/02 Restore caching of spamd results with expanded spamd_address.
499 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
501 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
502 Improves build reliability. Fix from: Frank Elsner
504 NM/02 Fix wide character breakage in the rfc2047 coding
505 Fixes bug 1064. Patch from Andrey N. Oktyabrski
507 NM/03 Allow underscore in dnslist lookups
508 Fixes bug 1026. Patch from Graeme Fowler
510 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
511 Code patches from Adam Ciarcinski of NetBSD.
513 NM/04 Fixed exiqgrep to cope with mailq missing size issue
516 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
517 is logged, to avoid truncation. Patch from John Horne.
519 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
520 Patch from Jakob Hirsch.
522 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
523 SQL string expansion failure details.
524 Patch from Andrey Oktyabrski.
526 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
527 Patch from Simon Arlott.
529 PP/09 New lookups functionality failed to compile on old gcc which rejects
530 extern declarations in function scope.
531 Patch from Oliver Fleischmann
533 PP/10 Use sig_atomic_t for flags set from signal handlers.
534 Check getgroups() return and improve debugging.
535 Fixed developed for diagnosis in bug 927 (which turned out to be
538 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
539 Patch from Mark Zealey.
541 PP/12 Bugzilla 1056: Improved spamd server selection.
542 Patch from Mark Zealey.
544 PP/13 Bugzilla 1086: Deal with maildir quota file races.
545 Based on patch from Heiko Schlittermann.
547 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
548 Patch from Uwe Doering, sign-off by Michael Haardt.
550 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
551 variable declaration deep within a block. Bug and patch from
554 PP/15 lookups-Makefile IRIX compatibilty coercion.
556 PP/16 Make DISABLE_DKIM build knob functional.
558 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
559 Patch by Simon Arlott
561 TF/03 Fix valgrind.h portability to C89 compilers that do not support
562 variable argument macros. Our copy now differs from upstream.
568 TF/01 Failure to get a lock on a hints database can have serious
569 consequences so log it to the panic log.
571 TF/02 Log LMTP confirmation messages in the same way as SMTP,
572 controlled using the smtp_confirmation log selector.
574 TF/03 Include the error message when we fail to unlink a spool file.
576 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
577 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
578 for maintaining out-of-tree patches for some time.
580 PP/01 Bugzilla 139: Documentation and portability issues.
581 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
582 Handle per-OS dynamic-module compilation flags.
584 PP/02 Let /dev/null have normal permissions.
585 The 4.73 fixes were a little too stringent and complained about the
586 permissions on /dev/null. Exempt it from some checks.
587 Reported by Andreas M. Kirchwitz.
589 PP/03 Report version information for many libraries, including
590 Exim version information for dynamically loaded libraries. Created
591 version.h, now support a version extension string for distributors
592 who patch heavily. Dynamic module ABI change.
594 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
595 privilege escalation vulnerability whereby the Exim run-time user
596 can cause root to append content of the attacker's choosing to
599 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
602 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
603 If dropping privileges for untrusted macros, we disabled normal logging
604 on the basis that it would fail; for the Exim run-time user, this is not
605 the case, and it resulted in successful deliveries going unlogged.
606 Fixed. Reported by Andreas Metzler.
612 PP/01 Date: & Message-Id: revert to normally being appended to a message,
613 only prepend for the Resent-* case. Fixes regression introduced in
614 Exim 4.70 by NM/22 for Bugzilla 607.
616 PP/02 Include check_rfc2047_length in configure.default because we're seeing
617 increasing numbers of administrators be bitten by this.
619 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
621 PP/03 Bugzilla 994: added openssl_options main configuration option.
623 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
625 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
627 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
629 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
630 without a peer certificate, leading to a segfault because of an
631 assumption that peers always have certificates. Be a little more
632 paranoid. Problem reported by Martin Tscholak.
634 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
635 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
636 NB: ClamAV planning to remove STREAM in "middle of 2010".
637 CL also introduces -bmalware, various -d+acl logging additions and
638 more caution in buffer sizes.
640 PP/09 Implemented reverse_ip expansion operator.
642 PP/10 Bugzilla 937: provide a "debug" ACL control.
644 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
646 PP/12 Bugzilla 973: Implement --version.
648 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
650 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
652 PP/15 Bugzilla 816: support multiple condition rules on Routers.
654 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
655 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
656 ignore trailing whitespace.
658 JJ/02 prevent non-panic DKIM error from being sent to paniclog
660 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
663 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
664 Notification from Dr Andrew Aitchison.
666 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
667 ExtendedDetectionInfo response format.
668 Notification from John Horne.
670 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
673 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
674 XSL and documented dependency on system catalogs, with examples of how
677 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
680 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
681 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
682 configuration file which is writeable by the Exim user or group.
684 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
685 of configuration files to cover files specified with the -C option if
686 they are going to be used with root privileges, not just the default
689 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
690 option (effectively making it always true).
692 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
693 files to be used while preserving root privileges.
695 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
696 that rogue child processes cannot use them.
698 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
699 run-time user, instead of root.
701 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
702 Exim run-time user without dropping privileges.
704 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
705 result string, instead of calling string_vformat() twice with the same
708 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
709 for other users. Others should always drop root privileges if they use
710 -C on the command line, even for a whitelisted configure file.
712 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
714 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
720 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
721 $data_path, and $header_path variables; fixed documentation bugs and
724 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
725 exipick to access non-standard spools, including the "frozen" queue
728 NM/01 Bugzilla 965: Support mysql stored procedures.
729 Patch from Alain Williams
731 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
733 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
734 Patch from Andreas Metzler
736 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
737 Patch from Kirill Miazine
739 NM/05 Bugzilla 671: Added umask to procmail example.
741 JJ/03 installed exipick 20100323.0, fixing doc bug
743 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
744 directory. Notification and patch from Dan Rosenberg.
746 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
748 TK/02 Improve log output when DKIM signing operation fails.
750 MH/01 Treat the transport option dkim_domain as a colon separated
751 list, not as a single string, and sign the message with each element,
752 omitting multiple occurences of the same signer.
754 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
755 Bugzilla 985, 986. Patch by Simon Arlott
757 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
758 Patch by Simon Arlott
760 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
761 MBX locking. Notification from Dan Rosenberg.
767 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
769 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
771 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
773 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
775 NM/04 Bugzilla 915: Flag broken perl installation during build.
781 TK/01 Added patch by Johannes Berg that expands the main option
782 "spamd_address" if it starts with a dollar sign.
784 TK/02 Write list of recipients to X-Envelope-Sender header when building
785 the mbox-format spool file for content scanning (suggested by Jakob
788 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
789 (http://www.dcc-servers.net/) support via dccifd. Activated by
790 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
792 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
793 by Mark Daniel Reidel <mr@df.eu>.
795 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
796 When building exim an external PCRE library is now needed -
797 PCRE is a system library on the majority of modern systems.
798 See entry on PCRE_LIBS in EDITME file.
800 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
801 conversation. Added nologin parameter to request.
802 Patch contributed by Kirill Miazine.
804 TF/01 Do not log submission mode rewrites if they do not change the address.
806 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
808 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
809 log files in place. Contributed by Roberto Lima.
811 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
813 TF/03 Bugzilla 615: When checking the local_parts router precondition
814 after a local_part_suffix or local_part_prefix option, Exim now
815 does not use the address's named list lookup cache, since this
816 contains cached lookups for the whole local part.
818 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
819 Robert Millan. Documentation is in experimental-spec.txt.
821 TF/04 Bugzilla 668: Fix parallel build (make -j).
823 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
825 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
826 Patch provided by Jan Srzednicki.
828 TF/05 Leading white space used to be stripped from $spam_report which
829 wrecked the formatting. Now it is preserved.
831 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
832 that they are available at delivery time.
834 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
836 TF/08 TLS error reporting now respects the incoming_interface and
837 incoming_port log selectors.
839 TF/09 Produce a more useful error message if an SMTP transport's hosts
840 setting expands to an empty string.
842 NM/06 Bugzilla 744: EXPN did not work under TLS.
843 Patch provided by Phil Pennock.
845 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
846 Patch provided by Richard Godbee.
848 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
849 acl_smtp_notquit, added index entry.
851 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
852 Patch provided by Eugene Bujak.
854 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
855 accept(). Patch provided by Maxim Dounin.
857 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
858 Patch provided by Phil Pennock.
860 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
862 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
863 Patch provided by Brad "anomie" Jorsch.
865 NM/14 Bugzilla 622: Added timeout setting to transport filter.
866 Patch provided by Dean Brooks.
868 TK/05 Add native DKIM support (does not depend on external libraries).
870 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
871 Patch provided by Graeme Fowler.
873 NM/16 Bugzilla 851: Documentation example syntax fix.
875 NM/17 Changed NOTICE file to remove references to embedded PCRE.
877 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
880 NM/19 Bugzilla 745: TLS version reporting.
881 Patch provided by Phil Pennock.
883 NM/20 Bugzilla 167: bool: condition support.
884 Patch provided by Phil Pennock.
886 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
887 clients. Patch provided by Phil Pennock.
889 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
890 Patch provided by Brad "anomie" Jorsch.
892 NM/23 Bugzilla 687: Fix misparses in eximstats.
893 Patch provided by Heiko Schlittermann.
895 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
896 Patch provided by Heiko Schlittermann.
898 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
899 plus update to original patch.
901 NM/26 Bugzilla 799: Documentation correction for ratelimit.
903 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
904 Patch provided by David Brownlee.
906 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
908 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
910 NM/30 Bugzilla 888: TLS documentation bugfixes.
912 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
914 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
915 Unlike the original bugzilla I have changed all shell scripts in src tree.
917 NM/33 Bugzilla 898: Transport filter timeout fix.
918 Patch by Todd Rinaldo.
920 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
921 Patch by Serge Demonchaux.
923 NM/35 Bugzilla 39: Base64 decode bug fixes.
924 Patch by Jakob Hirsch.
926 NM/36 Bugzilla 909: Correct connect() call in dcc code.
928 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
930 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
932 NM/39 Bugzilla 911: Fixed MakeLinks build script.
938 TK/01 Add preliminary DKIM support. Currently requires a forked version of
939 ALT-N's libdkim that I have put here:
940 http://duncanthrax.net/exim-experimental/
942 Note to Michael Haardt: I had to rename some vars in sieve.c. They
943 were called 'true' and it seems that C99 defines that as a reserved
944 keyword to be used with 'bool' variable types. That means you could
945 not include C99-style headers which use bools without triggering
946 build errors in sieve.c.
948 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
949 as mailq or other aliases. Changed the --help handling significantly
950 to do whats expected. exim_usage() emits usage/help information.
952 SC/01 Added the -bylocaldomain option to eximstats.
954 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
956 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
958 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
964 PH/01 Another patch from the Sieve maintainer.
966 PH/02 When an IPv6 address is converted to a string for single-key lookup
967 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
968 dots are used instead of colons so that keys in lsearch files need not
969 contain colons. This was done some time before quoting was made available
970 in lsearch files. However, iplsearch files do require colons in IPv6 keys
971 (notated using the quote facility) so as to distinguish them from IPv4
972 keys. This meant that lookups for IP addresses in host lists did not work
973 for iplsearch lookups.
975 This has been fixed by arranging for IPv6 addresses to be expressed with
976 colons if the lookup type is iplsearch. This is not incompatible, because
977 previously such lookups could never work.
979 The situation is now rather anomolous, since one *can* have colons in
980 ordinary lsearch keys. However, making the change in all cases is
981 incompatible and would probably break a number of configurations.
983 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
986 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
987 conversion specification without a maximum field width, thereby enabling
988 a rogue spamd server to cause a buffer overflow. While nobody in their
989 right mind would setup Exim to query an untrusted spamd server, an
990 attacker that gains access to a server running spamd could potentially
991 exploit this vulnerability to run arbitrary code as the Exim user.
993 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
994 $primary_hostname instead of what libspf2 thinks the hosts name is.
996 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
997 a directory entry by the name of the lookup key. Previously, if a
998 symlink pointed to a non-existing file or a file in a directory that
999 Exim lacked permissions to read, a lookup for a key matching that
1000 symlink would fail. Now it is enough that a matching directory entry
1001 exists, symlink or not. (Bugzilla 503.)
1003 PH/03 The body_linecount and body_zerocount variables are now exported in the
1006 PH/04 Added the $dnslist_matched variable.
1008 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1009 This means they are set thereafter only if the connection becomes
1012 PH/06 Added the client_condition to authenticators so that some can be skipped
1013 by clients under certain conditions.
1015 PH/07 The error message for a badly-placed control=no_multiline_responses left
1016 "_responses" off the end of the name.
1018 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1020 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1021 (without spaces) instead of just copying the configuration text.
1023 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1025 PH/11 Added $max_received_linelength.
1027 PH/12 Added +ignore_defer and +include_defer to host lists.
1029 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1030 way in which PCRE > 7.0 is built.
1032 PH/14 Implemented queue_only_load_latch.
1034 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1035 MAIL command. The effect was to mangle the value on 64-bit systems.
1037 PH/16 Another patch from the Sieve maintainer.
1039 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1041 PH/18 If a system quota error occurred while trying to create the file for
1042 a maildir delivery, the message "Mailbox is full" was not appended to the
1043 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1044 only to a quota excession during the actual writing of the file.
1046 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1047 characters?) which causes problems in log lines. The DN values are now
1048 passed through string_printing() before being added to log lines.
1050 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1051 and InterBase are left for another time.)
1053 PH/21 Added message_body_newlines option.
1055 PH/22 Guard against possible overflow in moan_check_errorcopy().
1057 PH/23 POSIX allows open() to be a macro; guard against that.
1059 PH/24 If the recipient of an error message contained an @ in the local part
1060 (suitably quoted, of course), incorrect values were put in $domain and
1061 $local_part during the evaluation of errors_copy.
1067 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1068 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1071 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1072 issue a MAIL command.
1074 PH/02 In an ACL statement such as
1076 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1078 if a client was not listed at all, or was listed with a value other than
1079 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1080 the condition was not true (as it should be), so access was not denied.
1081 The bug was that the ! inversion was incorrectly passed on to the second
1082 item. This has been fixed.
1084 PH/03 Added additional dnslists conditions == and =& which are different from
1085 = and & when the dns lookup returns more than one IP address.
1087 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1088 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1090 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1091 FSYNC, which compiles an option called disable_fsync that allows for
1092 bypassing fsync(). The documentation is heavily laced with warnings.
1094 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1096 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1097 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1098 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1099 including adding "make clean"; (3) Added -fPIC when compiling the test
1100 dynamically loaded module, to get rid of a warning.
1102 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1103 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1104 = 0s. The bug is otherwise harmless.
1106 PH/07 There was a bug in the dovecot authenticator such that the value of
1107 $auth1 could be overwritten, and so not correctly preserved, after a
1108 successful authentication. This usually meant that the value preserved by
1109 the server_setid option was incorrect.
1111 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1113 PH/09 Installed PCRE release 7.0.
1115 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1116 run for batched SMTP input. It is now run at the start of every message
1117 in the batch. While fixing this I discovered that the process information
1118 (output by running exiwhat) was not always getting set for -bs and -bS
1119 input. This is fixed, and it now also says "batched" for BSMTP.
1121 PH/11 Added control=no_pipelining.
1123 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1124 patch, slightly modified), and move the expansion of helo_data till after
1125 the connection is made in the smtp transport (so it can use these
1128 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1130 PH/14 Added log_selector = +pid.
1132 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1134 PH/16 Add ${if forany and ${if forall.
1136 PH/17 Added dsn_from option to vary the From: line in DSNs.
1138 PH/18 Flush SMTP output before performing a callout, unless control =
1139 no_callout_flush is set.
1141 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1142 was true (the default) a successful delivery failed to delete the retry
1143 item, thus causing premature timeout of the address. The bug is now
1146 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1148 PH/21 Long custom messages for fakedefer and fakereject are now split up
1149 into multiline reponses in the same way that messages for "deny" and
1150 other ACL rejections are.
1152 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1153 with slight modification.
1155 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1156 draft, changing the syntax and factoring some duplicate code".
1158 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1159 for deliveries of the second and subsequent messages over the same SMTP
1162 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1163 ${reduce, with only minor "tidies".
1165 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1167 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1168 expansion side effects.
1170 PH/27 When a message times out after an over-quota error from an Exim-imposed
1171 quota, the bounce message says "mailbox is full". This message was not
1172 being given when it was a system quota that was exceeded. It now should
1175 MH/03 Made $recipients available in local_scan(). local_scan() already has
1176 better access to the recipient list through recipients_list[], but
1177 $recipients can be useful in postmaster-provided expansion strings.
1179 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1180 in the case of a MAIL command with additional options following the
1181 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1182 were accidentally chopped off.
1184 PH/29 SMTP synchronization checks are implemented when a command is read -
1185 there is a check that no more input is waiting when there shouldn't be
1186 any. However, for some commands, a delay in an ACL can mean that it is
1187 some time before the response is written. In this time, more input might
1188 arrive, invalidly. So now there are extra checks after an ACL has run for
1189 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1190 pipelining has not been advertised.
1192 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1194 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1195 returned in the SMTP response when smtp_return_error_details was false.
1196 This has been fixed.
1198 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1199 instead of the C I/O that was originally supplied, because problems were
1200 reported on Solaris.
1202 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1203 Exim which did not show up earlier: it was assuming that a call to
1204 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1205 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1206 was a macro that became an assignment, so it seemed to work. This has
1207 changed to a proper function call with a void return, hence the compile
1208 error. Exim's code has been fixed.
1210 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1213 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1215 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1217 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1218 because it turns out to be too big or there is a timeout) there is no
1219 "Completed" line in the log. When some messages of this type were
1220 selected by exigrep, they were listed as "not completed". Others were
1221 picked up by some special patterns. I have improved the selection
1222 criteria to be more general.
1224 PH/38 The host_find_failed option in the manualroute router can now be set
1225 to "ignore", to completely ignore a host whose IP address cannot be
1226 found. If all hosts are ignored, the behaviour is controlled by the new
1227 host_all_ignored option.
1229 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1230 homing or because of multiple MX records with /mx) generated more than
1231 one IP address, and the following item turned out to be the local host,
1232 all the secondary addresses of the first item were incorrectly removed
1233 from the list, along with the local host and any following hosts (which
1234 is what is supposed to happen).
1236 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1237 whoever called Exim into the -H file. In the case of the daemon it was
1238 behaving confusingly. When first started, it used values for whoever
1239 started the daemon, but after a SIGHUP it used the Exim user (because it
1240 calls itself on a restart). I have changed the code so that it now always
1243 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1244 message are rejected with the same error (e.g. no authentication or bad
1245 sender address), and a DATA command is nevertheless sent (as can happen
1246 with PIPELINING or a stupid MUA), the error message that was given to the
1247 RCPT commands is included in the rejection of the DATA command. This is
1248 intended to be helpful for MUAs that show only the final error to their
1251 PH/42 Another patch from the Sieve maintainer.
1253 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1254 Eximstats - Fixed some broken HTML links and added missing column headers
1256 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1257 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1259 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1261 PH/43 Yet another patch from the Sieve maintainer.
1263 PH/44 I found a way to check for a TCP/IP connection going away before sending
1264 the response to the final '.' that terminates a message, but only in the
1265 case where the client has not sent further data following the '.'
1266 (unfortunately, this is allowed). However, in many cases there won't be
1267 any further data because there won't be any more messages to send. A call
1268 to select() can be used: if it shows that the input is "ready", there is
1269 either input waiting, or the socket has been closed. An attempt to read
1270 the next input character can distinguish the two cases. Previously, Exim
1271 would have sent an OK response which the client would never have see.
1272 This could lead to message repetition. This fix should cure that, at
1273 least in a lot of common cases.
1275 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1276 advertised in response to EHLO.
1282 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1283 fixed by 4.65/MH/01 (is this a record?) are fixed:
1285 (i) An empty string was always treated as zero by the numeric comparison
1286 operators. This behaviour has been restored.
1288 (ii) It is documented that the numeric comparison operators always treat
1289 their arguments as decimal numbers. This was broken in that numbers
1290 starting with 0 were being interpreted as octal.
1292 While fixing these problems I realized that there was another issue that
1293 hadn't been noticed. Values of message_size_limit (both the global option
1294 and the transport option) were treated as octal if they started with 0.
1295 The documentation was vague. These values are now always treated as
1296 decimal, and I will make that clear in the documentation.
1302 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1303 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1306 MH/01 Don't check that the operands of numeric comparison operators are
1307 integers when their expansion is in "skipping" mode (fixes bug
1308 introduced by 4.64-PH/07).
1310 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1311 child addresses, Exim now panics and dies. Previously, because the count
1312 is held in a short int, deliveries were likely to be lost. As such a
1313 large number of recipients for a single message is ridiculous
1314 (performance will be very, very poor), I have chosen to impose a limit
1315 rather than extend the field.
1321 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1322 leftover -K file (the existence of which was triggered by #402).
1323 While we were at it, introduced process PID as part of the -K
1324 filename. This should rule out race conditions when creating
1327 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1328 processing considerably. Previous code took too long for large mails,
1329 triggering a timeout which in turn triggers #401.
1331 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1332 in the DK code in transports.c. sendfile() is not really portable,
1333 hence the _LINUX specificness.
1335 TF/01 In the add_headers option to the mail command in an Exim filter,
1336 there was a bug that Exim would claim a syntax error in any
1337 header after the first one which had an odd number of characters
1340 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1341 callout verification, Exim cached a "reject" for the entire domain. This
1342 is correct for most verifications, but it is not correct for a recipient
1343 verification with use_sender or use_postmaster set, because in that case
1344 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1345 case of MAIL FROM:<> rejection from other early rejections (e.g.
1346 rejection of HELO). When verifying a recipient using a non-null MAIL
1347 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1348 Whatever the result of the callout, the value of the domain cache is
1349 left unchanged (for any other kind of callout, getting as far as trying
1350 RCPT means that the domain itself is ok).
1352 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1355 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1356 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1357 session, the daemon ignores ECONNECT errors and logs others; it now
1358 ignores EPIPE as well.
1360 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1361 (quoted-printable decoding).
1363 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1364 later the small subsequent patch to fix an introduced bug.
1366 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1368 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1370 PH/08 An error is now given if message_size_limit is specified negative.
1372 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1373 to be given (somewhat) arbitrary names.
1375 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1378 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1379 miscellaneous code fixes
1381 PH/10 Added the log_reject_target ACL modifier to specify where to log
1384 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1385 hostname. This is wrong, because it relates to the incoming message (and
1386 probably the interface on which it is arriving) and not to the outgoing
1387 callout (which could be using a different interface). This has been
1388 changed to use the value of the helo_data option from the smtp transport
1389 instead - this is what is used when a message is actually being sent. If
1390 there is no remote transport (possible with a router that sets up host
1391 addresses), $smtp_active_hostname is used.
1393 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1394 tweaks were necessary in order to get it to work (see also 21 below):
1395 (a) The code assumed that strncpy() returns a negative number on buffer
1396 overflow, which isn't the case. Replaced with Exim's string_format()
1398 (b) There were several signed/unsigned issues. I just did the minimum
1399 hacking in of casts. There is scope for a larger refactoring.
1400 (c) The code used strcasecmp() which is not a standard C function.
1401 Replaced with Exim's strcmpic() function.
1402 (d) The code set only $1; it now sets $auth1 as well.
1403 (e) A simple test gave the error "authentication client didn't specify
1404 service in request". It would seem that Dovecot has changed its
1405 interface. Fortunately there's a specification; I followed it and
1406 changed what the client sends and it appears to be working now.
1408 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1411 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1412 address A is aliased to B and C, where B exists and C does not. Without
1413 -v the output is "A verified" because verification stops after a
1414 successful redirection if more than one address is generated. However,
1415 with -v the child addresses are also verified. Exim was outputting "A
1416 failed to verify" and then showing the successful verification for C,
1417 with its parentage. It now outputs "B failed to verify", showing B's
1418 parentage before showing the successful verification of C.
1420 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1421 look up a TXT record in a specific list after matching in a combined
1424 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1425 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1426 they consult the DNS. I had assumed they would set it the way they
1427 wanted; and indeed my experiments on Linux seem to show that in some
1428 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1429 To be on the safe side, however, I have now made the interface to
1430 host_find_byname() similar to host_find_bydns(), with an argument
1431 containing the DNS resolver options. The host_find_byname() function now
1432 sets these options at its start, just as host_find_bydns() does. The smtp
1433 transport options dns_qualify_single and dns_search_parents are passed to
1434 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1435 of host_find_byname() use the default settings of RES_DEFNAMES
1436 (qualify_single) but not RES_DNSRCH (search_parents).
1438 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1439 spool_read_header() do less string testing, by means of a preliminary
1440 switch on the second character of optional "-foo" lines. (This is
1441 overdue, caused by the large number of possibilities that now exist.
1442 Originally there were few.) While I was there, I also converted the
1443 str(n)cmp tests so they don't re-test the leading "-" and the first
1444 character, in the hope this might squeeze out yet more improvement.
1446 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1447 flag allowing group syntax was set by the header_syntax check but not
1448 turned off, possible causing trouble later; (2) The flag was not being
1449 set at all for the header_verify test, causing "group"-style headers to
1450 be rejected. I have now set it in this case, and also caused header_
1451 verify to ignore an empty address taken from a group. While doing this, I
1452 came across some other cases where the code for allowing group syntax
1453 while scanning a header line wasn't quite right (mostly, not resetting
1454 the flag correctly in the right place). These bugs could have caused
1455 trouble for malformed header lines. I hope it is now all correct.
1457 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1458 with the "reply" argument non-NULL. The code, however (which originally
1459 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1460 but it didn't always do it. This confused somebody who was copying the
1461 code for some other use. I have removed all the tests.
1463 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1464 feature that was used to support insecure browsers during the U.S. crypto
1465 embargo. It requires special client support, and Exim is probably the
1466 only MTA that supported it -- and would never use it because real RSA is
1467 always available. This code has been removed, because it had the bad
1468 effect of slowing Exim down by computing (never used) parameters for the
1469 RSA_EXPORT functionality.
1471 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1472 authenticator to fail if there's a tab character in the incoming data
1473 (there should never be unless someone is messing about, as it's supposed
1474 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1475 if the connection is using TLS or if the remote IP is the same as the
1476 local IP, and the "valid-client-cert option" if a client certificate has
1479 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1480 authenticators. This can be used for authorization after authentication
1481 succeeds. (In the case of plaintext, it servers for both authentication
1484 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1485 if any retry times were supplied.
1487 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1488 connection, where there is no client IP address to check. In this
1489 situation, the verify now always succeeds.
1491 PH/25 Applied John Jetmore's -Mset patch.
1493 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1495 PH/27 In a string expansion for a processed (not raw) header when multiple
1496 headers of the same name were present, leading whitespace was being
1497 removed from all of them, but trailing whitespace was being removed only
1498 from the last one. Now trailing whitespace is removed from each header
1499 before concatenation. Completely empty headers in a concatenation (as
1500 before) are ignored.
1502 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1503 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1505 PH/29 [Removed. This was a change that I later backed out, and forgot to
1506 correct the ChangeLog entry (that I had efficiently created) before
1507 committing the later change.]
1509 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1510 address errors (4xx response to RCPT) over the same connection as other
1511 messages routed to the same hosts. Such deliveries are always "forced",
1512 so retry times are not inspected. This resulted in far too many retries
1513 for the affected addresses. The effect occurred only when there were more
1514 hosts than the hosts_max_try setting in the smtp transport when it had
1515 the 4xx errors. Those hosts that it had tried were not added to the list
1516 of hosts for which the message was waiting, so if all were tried, there
1517 was no problem. Two fixes have been applied:
1519 (i) If there are any address or message errors in an SMTP delivery, none
1520 of the hosts (tried or untried) are now added to the list of hosts
1521 for which the message is waiting, so the message should not be a
1522 candidate for sending over the same connection that was used for a
1523 successful delivery of some other message. This seems entirely
1524 reasonable: after all the message is NOT "waiting for some host".
1525 This is so "obvious" that I'm not sure why it wasn't done
1526 previously. Hope I haven't missed anything, but it can't do any
1527 harm, as the worst effect is to miss an optimization.
1529 (ii) If, despite (i), such a delivery is accidentally attempted, the
1530 routing retry time is respected, so at least it doesn't keep
1531 hammering the server.
1533 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1534 in ${readsocket because some servers need this prod.
1536 PH/32 Added some extra debug output when updating a wait-xxx database.
1538 PH/33 The hint "could be header name not terminated by colon", which has been
1539 given for certain expansion errors for a long time, was not being given
1540 for the ${if def:h_colon_omitted{... case.
1542 PH/34 The spec says: "With one important exception, whenever a domain list is
1543 being scanned, $domain contains the subject domain." There was at least
1544 one case where this was not true.
1546 PH/35 The error "getsockname() failed: connection reset by peer" was being
1547 written to the panic log as well as the main log, but it isn't really
1548 panic-worthy as it just means the connection died rather early on. I have
1549 removed the panic log writing for the ECONNRESET error when getsockname()
1552 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1553 runs only) independently of the message's sender address. This meant
1554 that, if the 4xx error was in fact related to the sender, a different
1555 message to the same recipient with a different sender could confuse
1556 things. In particualar, this can happen when sending to a greylisting
1557 server, but other circumstances could also provoke similar problems.
1558 I have changed the default so that the retry time for these errors is now
1559 based a combination of the sender and recipient addresses. This change
1560 can be overridden by setting address_retry_include_sender=false in the
1563 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1564 remote server are returned as part of bounce messages. This was not
1565 happening for LMTP over a pipe (the lmtp transport), but now it is the
1566 same for both kinds of LMTP.
1568 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1569 in header lines that were in fact CNAMEs. This is no longer the case.
1571 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1572 and queue runs started by the daemon processed all messages. This has
1573 been fixed so that -R and -S can now usefully be given with -q<time>.
1575 PH/40 Import PCRE release 6.7 (fixes some bugs).
1577 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1579 PH/42 Give an error if -q is specified more than once.
1581 PH/43 Renamed the variables $interface_address and $interface_port as
1582 $received_ip_address and $received_port, to make it clear that these
1583 values apply to message reception, and not to the outgoing interface when
1584 a message is delivered. (The old names remain recognized, of course.)
1586 PH/44 There was no timeout on the connect() call when using a Unix domain
1587 socket in the ${readsocket expansion. There now is.
1589 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1590 be meaningful with "accept".
1592 SC/01 Eximstats V1.43
1593 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1595 SC/02 Eximstats V1.44
1596 Use a glob alias rather than an array ref in the generated
1597 parser. This improves both readability and performance.
1599 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1600 Collect SpamAssassin and rejection statistics.
1601 Don't display local sender or destination tables unless
1602 there is data to show.
1603 Added average volumes into the top table text output.
1605 SC/04 Eximstats V1.46
1606 Collect data on the number of addresses (recipients)
1607 as well as the number of messages.
1609 SC/05 Eximstats V1.47
1610 Added 'Message too big' to the list of mail rejection
1611 reasons (thanks to Marco Gaiarin).
1613 SC/06 Eximstats V1.48
1614 Mainlog lines which have GMT offsets and are too short to
1615 have a flag are now skipped.
1617 SC/07 Eximstats V1.49 (Alain Williams)
1618 Added the -emptyok flag.
1620 SC/08 Eximstats V1.50
1621 Fixes for obtaining the IP address from reject messages.
1623 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1624 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1625 whitesspace changes from 4.64-PH/27
1627 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1630 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1631 are found, allow negative numbers in numeric criteria)
1633 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1635 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1636 to match changes made in 4.64-PH/43
1638 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1640 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1641 "domain" is actually an IP address, and force a failure. This locks out
1642 those revolvers/nameservers that support "A-for-A" lookups, in
1643 contravention of the specifications.
1645 PH/48 When a host name was looked up from an IP address, and the subsequent
1646 forward lookup of the name timed out, the host name was left in
1647 $sender_host_name, contrary to the specification.
1649 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1650 restricted to single-key lookups, Exim was not diagnosing an error if
1651 * or *@ was used with a query-style lookup.
1653 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1655 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1656 long ago, but noone interested enough thought of it. Let's just say that
1657 the "1.1" means that there are some new functions that weren't there at
1658 some point in the past.
1660 PH/51 Error processing for expansion failure of helo_data from an smtp
1661 transport during callout processing was broken.
1663 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1664 tested/used via the -bh/-bhc/-bs options.
1666 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1667 bug, fixed in subsequent PCRE releases).
1669 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1670 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1672 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1678 SC/01 Use a glob alias rather than an array ref in eximstats generated
1679 parser. This improves both readability and performance.
1681 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1682 Don't display local sender or destination tables in eximstats unless
1683 there is data to show.
1684 Added average volumes into the eximstats top table text output.
1686 SC/03 Collect data on the number of addresses (recipients) as well
1687 as the number of messages in eximstats.
1689 TF/01 Correct an error in the documentation for the redirect router. Exim
1690 does (usually) call initgroups() when daemonizing.
1692 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1693 with consistent privilege compared to when running as a daemon.
1695 TF/03 Note in the spec that $authenticated_id is not set for local
1696 submissions from trusted users.
1698 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1699 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1701 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1702 by adding some example configuration directives to the default
1703 configuration file. A little bit of work is required to uncomment the
1704 directives and define how usernames and passwords are checked, but
1705 there is now a framework to start from.
1707 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1708 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1709 without this. I don't know how relevant this is to other LDAP libraries.
1711 PH/02 Add the verb name to the "unknown ACL verb" error.
1713 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1715 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1717 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1718 directory not expanded when it should be if an expanded home directory
1719 was set for the address (which is overridden by the transport).
1721 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1724 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1725 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1726 because it is too late at that time, and has no effect.
1728 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1729 security issue with \' (bugzilla #107). I could not use the
1730 PQescapeStringConn() function, because it needs a PGconn value as one of
1733 PH/08 When testing addresses using -bt, indicate those final addresses that
1734 are duplicates that would not cause an additional delivery. At least one
1735 person was confused, thinking that -bt output corresponded to deliveries.
1736 (Suppressing duplicates isn't a good idea as you lose the information
1737 about possibly different redirections that led to the duplicates.)
1739 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1740 systems where poll() doesn't work, in particular OS X.
1742 PH/10 Added more information to debugging output for retry time not reached.
1744 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1745 operations in malware.c.
1747 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1750 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1751 syslog_duplication set false, log lines that would normally be written
1752 both the the main log and to the reject log were not written to syslog at
1755 PH/14 In the default configuration, change the use of "message" in ACL warn
1756 statements to "add_header".
1758 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1759 not followed by a command (e.g. "seen endif").
1761 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1762 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1765 PH/17 Added extra conditions to the default value of delay_warning_condition
1769 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1770 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1771 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1774 The Auto-Submitted: and various List- headers are standardised, whereas I
1775 don't think Precedence: ever was.
1777 PH/18 Refactored debugging code in route_finduser() to show more information,
1778 in particular, the error code if getpwnam() issues one.
1780 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1781 This is apparently needed in addition to the PH/07 change above to avoid
1782 any possible encoding problems.
1784 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1785 but not after initializing Perl.
1787 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1788 output them only if debugging. By default they are written stderr,
1789 apparently, which is not desirable.
1791 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1794 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1797 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1799 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1800 authenticated or an ident call has been made. Suppress the default
1801 values for $authenticated_id and $authenticated_sender (but permit -oMai
1802 and -oMas) when testing with -bh.
1804 PH/24 Re-jigged the order of the tests in the default configuration so that the
1805 tests for valid domains and recipients precede the DNS black list and CSA
1806 tests, on the grounds that those ones are more expensive.
1808 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1809 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1810 This bug exists in every version of Exim that I still have, right back to
1813 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1814 However, an attempt to turn on case-sensitivity in a regex key by
1815 including (?-i) didn't work because the subject string was already
1816 lowercased, and the effects were non-intuitive. It turns out that a
1817 one-line patch can be used to allow (?-i) to work as expected.
1823 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1824 other effects) broke the use of negated acl sub-conditions.
1826 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1829 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1830 "Deny" causes Exim to reject the incoming connection with a 554 error.
1831 Unfortunately, if there is a major crisis, such as a disk failure,
1832 tcp-wrappers gives "deny", whereas what one would like would be some
1833 kind of temporary error. A kludge has been added to help with this.
1834 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1835 554 error is used if errno is still zero or contains ENOENT (which occurs
1836 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1839 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1841 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1842 errors. Otherwise a message that provokes a temporary error (when other
1843 messages do not) can cause a whole host to time out.
1845 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1846 addresses were routed directly to files or pipes from a redirect router.
1847 File deliveries just didn't batch; pipe deliveries might have suffered
1850 PH/06 A failure to get a lock for a hints database would erroneously always say
1851 "Failed to get write lock", even when it was really a read lock.
1853 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1854 of 0600. This has been changed to use the value of the lockfile_mode
1855 option (which defaults to 0600).
1857 PH/08 Applied small patch from the Sieve maintainer.
1859 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1860 folder from quota calculations, a direct delivery into this folder messed
1861 up the contents of the maildirsize file. This was because the regex was
1862 used only to exclude .Trash (or whatever) when the size of the mailbox
1863 was calculated. There was no check that a delivery was happening into an
1864 excluded directory. This bug has been fixed by ignoring all quota
1865 processing for deliveries into excluded directories.
1867 PH/10 Added the maildirfolder_create_regex option to appendfile.
1873 PH/01 The code for finding all the local interface addresses on a FreeBSD
1874 system running IPv6 was broken. This may well have applied to all BSD
1875 systems, as well as to others that have similar system calls. The broken
1876 code found IPv4 interfaces correctly, but gave incorrect values for the
1877 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1878 that it would not match correctly against @[] and not recognize the IPv6
1881 PH/02 The ipliteral router was not recognizing addresses of the form user@
1882 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1884 PH/03 Added disable_ipv6.
1886 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1887 decoded form, because it is most often used to construct To: headers
1888 lines in autoreplies, and the decoded form may well be syntactically
1889 invalid. However, $reply_address has leading white space removed, and all
1890 newlines turned into spaces so that the autoreply transport does not
1893 PH/05 If group was specified without a user on a router, and no group or user
1894 was specified on a transport, the group from the router was ignored.
1896 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1897 for visible compile-time settings that can be used to change these
1898 numbers, for those that want even more. Backwards compatibility with old
1899 spool files has been maintained. However, going back to a previous Exim
1900 release will lost any variables that are in spool files.
1902 PH/07 Two small changes when running in the test harness: increase delay when
1903 passing a TCP/IP connection to a new process, in case the original
1904 process has to generate a bounce, and remove special handling of
1905 127.0.0.2 (sic), which is no longer necessary.
1907 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1908 be the same on different OS.
1910 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1913 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1914 whether --show-vars was specified or not
1916 JJ/02 exipick: Added support for new ACL variable spool format introduced
1919 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1920 syntactically invalid From: or Reply-to: line, and a filter used this to
1921 generate an autoreply, and therefore failed to obtain an address for the
1922 autoreply, Exim could try to deliver to a non-existent relative file
1923 name, causing unrelated and misleading errors. What now happens is that
1924 it logs this as a hard delivery error, but does not attempt to create a
1927 PH/11 The exinext utility has a -C option for testing purposes, but although
1928 the given file was scanned by exinext itself; it wasn't being passed on
1929 when Exim was called.
1931 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1932 an end-of-file indication when reading a command response.
1934 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1935 compiled. In many other places in Exim, IPv6 addresses are always
1936 recognized, so I have changed this. It also means that IPv4 domain
1937 literals of the form [IPV4:n.n.n.n] are now always recognized.
1939 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1940 used if the router is not running as root, for example, when verifying at
1941 ACL time, or when using -bh. The debugging output from this situation was
1942 non-existent - all you got was a failure to exec. I have made two
1945 (a) Failures to set uid/gid, the current directory, or a process leader
1946 in a subprocess such as that created by queryprogram now generate
1947 suitable debugging ouput when -d is set.
1949 (b) The queryprogram router detects when it is not running as root,
1950 outputs suitable debugging information if -d is set, and then runs
1951 the subprocess without attempting to change uid/gid.
1953 PH/15 Minor change to Makefile for building test_host (undocumented testing
1956 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1957 additional section of a DNS packet that returns MX or SRV records.
1958 Instead, it always explicitly searches for A/AAAA records. This avoids
1959 major problems that occur when a DNS server includes only records of one
1960 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1961 fixed another bug: if SRV records were looked up and the corresponding
1962 address records were *not* found in the additional section, the port
1963 values from the SRV records were lost.
1965 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1966 using the correct key (the original address) when searching the retry
1967 rules in order to find which one to use for generating the retry hint.
1969 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1970 adding the default one. Similarly, if it contains a Reply-To: header, the
1971 errors_reply_to option, if set, is not used.
1973 PH/19 When calculating a retry time, Exim used to measure the "time since
1974 failure" by looking at the "first failed" field in the retry record. Now
1975 it does not use this if it is later than than the arrival time of the
1976 message. Instead it uses the arrival time. This makes for better
1977 behaviour in cases where some deliveries succeed, thus re-setting the
1978 "first failed" field. An example is a quota failure for a huge message
1979 when small messages continue to be delivered. Without this change, the
1980 "time since failure" will always be short, possible causing more frequent
1981 delivery attempts for the huge message than are intended.
1982 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1984 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1985 $1, $2, $3) because the numerical variables can be reset during some
1986 expansion items (e.g. "match"), thereby losing the authentication data.
1988 PH/21 Make -bV show the size of off_t variables so that the test suite can
1989 decide whether to run tests for quotas > 2G.
1991 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1992 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1993 filecount value is greater than 2G or if a quota value is greater than 2G
1994 on a system where the size of off_t is not greater than 4, a panic error
1997 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1998 never match. The debug and -bh output now contains an explicit error
1999 message indicating a malformed IPv4 address or mask.
2001 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2002 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2003 PH/23 above applies.
2005 PH/25 Do not write to syslog when running in the test harness. The only
2006 occasion when this arises is a failure to open the main or panic logs
2007 (for which there is an explicit test).
2009 PH/26 Added the /no_tell option to "control=freeze".
2011 PH/27 If a host name lookup failed very early in a connection, for example, if
2012 the IP address matched host_lookup and the reverse lookup yielded a name
2013 that did not have a forward lookup, an error message of the form "no IP
2014 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2015 could be logged. Now it outputs the IP address instead of "NULL".
2017 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2018 allows the sender and the authenticated sender to be set when
2019 submitting a message from within Exim. Since child_open_exim() is
2020 documented for local_scan(), the new function should be too.
2022 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2023 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2024 results in an empty string is now treated as unset.
2026 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2028 PH/31 Added sender_verify_fail logging option.
2030 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2031 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2032 tidied the source and removed it altogether.
2034 PH/33 When a queue run was abandoned because the load average was too high, a
2035 log line was always written; now it is written only if the queue_run log
2036 selector is set. In addition, the log line for abandonment now contains
2037 information about the queue run such as the pid. This is always present
2038 in "start" and "stop" lines but was omitted from the "abandon" line.
2040 PH/34 Omit spaces between a header name and the colon in the error message that
2041 is given when verify = headers_syntax fails (if there are lots of them,
2042 the message gets confusing).
2044 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2045 names, as there are now some PTR records that contain slashes. This check
2046 is only to protect against broken name servers that fall over on strange
2047 characters, so the fact that it applies to all lookups doesn't matter.
2049 PH/36 Now that the new test suite is complete, we can remove some of the
2050 special code in Exim that was needed for the old test suite. For example,
2051 sorting DNS records because real resolvers return them in an arbitrary
2052 order. The new test suite's fake resolver always returns records in the
2055 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2056 bounces) except when queue_only is set, to avoid logging races between
2057 the different processes.
2059 PH/38 Panic-die if .include specifies a non-absolute path.
2061 PH/39 A tweak to the "H" retry rule from its user.
2063 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2064 a label. They prevented compilation on older perls.
2066 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2067 a warning to be raised on newish perls.
2069 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2070 on queue. Changes to match documented behaviour of showing count of
2071 messages matching specified criteria.
2073 PH/40 Changed the default ident timeout from 30s to 5s.
2075 PH/41 Added support for the use of login_cap features, on those BSD systems
2076 that have them, for controlling the resources used by pipe deliveries.
2078 PH/42 The content-scanning code uses fopen() to create files in which to put
2079 message data. Previously it was not paying any attention to the mode of
2080 the files. Exim runs with umask(0) because the rest of the code creates
2081 files with open(), and sets the required mode explicitly. Thus, these
2082 files were ending up world-writeable. This was not a big issue, because,
2083 being within the spool directory, they were not world-accessible. I have
2084 created a function called modefopen, which takes an additional mode
2085 argument. It sets umask(777), creates the file, chmods it to the required
2086 mode, then resets the umask. All the relevant calls to fopen() in the
2087 content scanning code have been changed to use this function.
2089 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2090 to 24 hours. This avoids potential overflow problems when processing G
2091 and H retry rules. I suspect nobody ever tinkers with this value.
2093 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2095 PH/45 When the plaintext authenticator is running as a client, the server's
2096 challenges are checked to ensure they are valid base64 strings. By
2097 default, the authentication attempt is cancelled if an invalid string is
2098 received. Setting client_ignore_invalid_base64 true ignores these errors.
2099 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2100 they are received. Thus, the responses can be made to depend on the
2101 challenges. If an invalid string is ignored, an empty string is placed in
2104 PH/46 Messages that are created by the autoreply transport now contains a
2105 References: header, in accordance with RFCs 2822 and 3834.
2107 PH/47 Added authenticated_sender_force to the smtp transport.
2109 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2111 PH/49 Installed latest patch from the Sieve maintainer.
2113 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2114 was also set, the appendfile transport was unnecessarily scanning a
2115 directory of message files (e.g. for maildir delivery) to find the count
2116 of files (along with the size), even though it did not need this
2117 information. It now does the scan only if it needs to find either the
2118 size of the count of files.
2120 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2122 PH/52 Two bugs concerned with error handling when the smtp transport is
2125 (i) Exim was not creating retry information for temporary errors given
2126 for individual recipients after the DATA command when the smtp transport
2127 was used in LMTP mode. This meant that they could be retried too
2128 frequently, and not timed out correctly.
2130 (ii) Exim was setting the flag that allows error details to be returned
2131 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2132 recipients that were returned after the DATA command.
2134 PH/53 This is related to PH/52, but is more general: for any failing address,
2135 when detailed error information was permitted to be returned to the
2136 sender, but the error was temporary, then after the final timeout, only
2137 "retry timeout exceeded" was returned. Now it returns the full error as
2138 well as "retry timeout exceeded".
2140 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2141 do this, and (what is worse) MTAs that accept it.
2143 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2144 will now be deprecated.
2146 PH/56 New os.c-cygwin from the Cygwin maintainer.
2148 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2149 formats (previously only available in exim formats via -bpr, -bpru,
2150 and -bpra. Now also available in native and exiqgrep formats)
2152 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2153 with very large, slow to parse queues
2155 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2157 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2159 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2160 responses to authentication challenges, though it was showing the
2161 challenges; (ii) I've removed the CR characters from the debug output for
2164 PH/58 Allow for the insertion of a newline as well as a space when a string
2165 is turned into more than one encoded-word during RFC 2047 encoding. The
2166 Sieve code now uses this.
2168 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2169 data_4xx, lost_connection, tls_required.
2171 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2172 message was being sent as an SMTP response.
2174 PH/61 Add -l and -k options to exicyclog.
2176 PH/62 When verifying, if an address was redirected to one new address, so that
2177 verification continued, and the new address failed or deferred after
2178 having set something in $address_data, the value of $address_data was not
2179 passed back to the ACL. This was different to the case when no
2180 redirection occurred. The value is now passed back in both cases.
2182 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2183 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2184 use login_cap.h, so on its own it isn't the distinguishing feature. The
2185 new name refers directly to the setclassresources() function.
2187 PH/65 Added configuration files for NetBSD3.
2189 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2191 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2192 is preferred over IPv4.
2194 PH/68 The bounce_return_message and bounce_return_body options were not being
2195 honoured for bounces generated during the reception of non-SMTP messages.
2196 In particular, this applied to messages rejected by the ACL. This bug has
2197 been fixed. However, if bounce_return_message is true and bounce_return_
2198 body is false, the headers that are returned for a non-SMTP message
2199 include only those that have been read before the error was detected.
2200 (In the case of an ACL rejection, they have all been read.)
2202 PH/69 The HTML version of the specification is now built in a directory called
2203 spec_html instead of spec.html, because the latter looks like a path with
2204 a MIME-type, and this confuses some software.
2206 PH/70 Catch two compiler warnings in sieve.c.
2208 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2209 function verify_get_ident() calls ip_connect() to connect a socket, but
2210 if the "connect()" function timed out, ip_connect() used to close the
2211 socket. However, verify_get_ident() also closes the socket later, and in
2212 between Exim writes to the log, which may get opened at this point. When
2213 the socket was closed in ip_connect(), the log could get the same file
2214 descriptor number as the socket. This naturally causes chaos. The fix is
2215 not to close the socket in ip_connect(); the socket should be closed by
2216 the function that creates it. There was only one place in the code where
2217 this was missing, in the iplookup router, which I don't think anybody now
2218 uses, but I've fixed it anyway.
2220 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2221 well as to direct DNS lookups. Otherwise the handling of names in host
2222 lists is inconsistent and therefore confusing.
2228 PH/01 Two changes to the default runtime configuration:
2230 (1) Move the checks for relay_from_hosts and authenticated clients from
2231 after to before the (commented out) DNS black list checks.
2233 (2) Add control=submission to the relay_from_hosts and authenticated
2234 clients checks, on the grounds that messages accepted by these
2235 statements are most likely to be submissions.
2237 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2239 (1) Generate an error if the third argument for the ${prvs expansion is
2242 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2245 (3) Reset the variables that are obtained from the first argument of
2246 ${prvscheck and used in the second argument before leaving the code,
2247 because their memory is reclaimed, so using them afterwards may do
2250 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2251 one (it's much easier than Tom thought :-).
2253 (5) Because of (4), we can now allow for the use of $prvscheck_result
2254 inside the third argument.
2256 PH/03 For some reason, the default setting of PATH when running a command from
2257 a pipe transport was just "/usr/bin". I have changed it to
2260 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2261 anything to be listed in the output from -bV.
2263 PH/05 When a filter generated an autoreply, the entire To: header line was
2264 quoted in the delivery log line, like this:
2266 => >A.N.Other <ano@some.domain> <original@ddress> ...
2268 This has been changed so that it extracts the operative address. There
2269 may be more than one such address. If so, they are comma-separated, like
2272 => >ano@some.domain,ona@other.domain <original@ddress> ...
2274 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2275 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2276 not being looked up in the rDNS to get a host name, Exim was showing the
2277 IP address twice in Received: lines, even though the IP addresses were
2278 identical. For example:
2280 Received: from [1.2.3.4] (helo=[1.2.3.4])
2282 However, if the real host name was known, it was omitting the HELO data
2283 if it matched the actual IP address. This has been tidied up so that it
2284 doesn't show the same IP address twice.
2286 PH/07 When both +timestamp and +memory debugging was on, the value given by
2287 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2288 called by the debug printing, thereby overwriting the timestamp buffer.
2289 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2291 PH/08 When the original message was included in an autoreply transport, it
2292 always said "this is a copy of the message, including all the headers",
2293 even if body_only or headers_only was set. It now gives an appropriate
2296 PH/09 Applied a patch from the Sieve maintainer which:
2298 o fixes some comments
2299 o adds the (disabled) notify extension core
2300 o adds some debug output for the result of if/elsif tests
2301 o points to the current vacation draft in the documentation
2302 and documents the missing references header update
2306 o fixes a bug in processing the envelope test (when testing
2307 multiple envelope elements, the last element determinted the
2310 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2311 Electronic Mail") by including:
2313 Auto-submitted: auto-generated
2315 in the messages that it generates (bounce messages and others, such as
2316 warnings). In the case of bounce messages for non-SMTP mesages, there was
2317 also a typo: it was using "Auto_submitted" (underscore instead of
2318 hyphen). Since every message generated by Exim is necessarily in response
2319 to another message, thes have all been changed to:
2321 Auto-Submitted: auto-replied
2323 in accordance with these statements in the RFC:
2325 The auto-replied keyword:
2327 - SHOULD be used on messages sent in direct response to another
2328 message by an automatic process,
2330 - MUST NOT be used on manually-generated messages,
2332 - MAY be used on Delivery Status Notifications (DSNs) and Message
2333 Disposition Notifications (MDNs),
2335 - MUST NOT be used on messages generated by automatic or periodic
2336 processes, except for messages which are automatic responses to
2339 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2340 to the default Received: header definition.
2342 PH/12 Added log selector acl_warn_skipped (default on).
2344 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2345 variables because (a) they are in the wrong storage pool and (b) even if
2346 they were copied, it wouldn't work properly because of the caching.
2348 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2349 checking when decoding. Apparently there are clients that generate
2350 overlong encoded strings. Why am I not surprised?
2352 PH/15 If the first argument of "${if match_address" was not empty, but did not
2353 contain an "@" character, Exim crashed. Now it writes a panic log message
2354 and treats the condition as false.
2356 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2358 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2359 extension "envelope-auth". The code is finished and in agreement with
2360 other implementations, but there is no documentation so far and in fact,
2361 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2362 not changing the active code.
2364 Print executed "if" and "elsif" statements when debugging is used. This
2365 helps a great deal to understand what a filter does.
2367 Document more things not specified clearly in RFC3028. I had all this
2368 sorted out, when out of a sudden new issues came to my mind. Oops."
2370 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2373 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2374 canonical form (as documented). However, after a host name lookup from
2375 the IP address, check_host() was doing a simple string comparison with
2376 addresses acquired from the DNS when checking that the found name did
2377 have the original IP as one of its addresses. Since any found IPv6
2378 addresses are likely to be in abbreviated form, the comparison could
2379 fail. Luckily, there already exists a function for doing the comparison
2380 by converting both addresses to binary, so now that is used instead of
2381 the text comparison.
2383 PH/20 There was another similar case to PH/19, when a complete host name was
2384 given in a host list; looking up its IP address could give an abbreviated
2385 form, whereas the current host's name might or might not be abbreviated.
2386 The same fix has been applied.
2392 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2393 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2396 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2397 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2399 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2401 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2402 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2403 Also, for "Auto-Submitted", treat anything other than "no" as
2404 non-personal, in accordance with RFC 3834. (Previously it treated
2405 anything starting "auto-" as non-personal.)
2407 TF/01 The control=submission/name=... option had a problem with syntax
2408 errors if the name included a slash character. The /name= option
2409 now slurps the rest of the string, so it can include any characters
2410 but it must come last in the list of options (after /sender_retain
2413 PH/05 Some modifications to the interface to the fake nameserver for the new
2421 TK/01 Added the "success_on_redirect" address verification option. See
2422 NewStuff for rationale and an example.
2424 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2426 PH/02 Patch to exigrep to allow it to work on syslog lines.
2428 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2429 fread() to skip over the body file's header line, because in Cygwin the
2430 header line is locked and is inaccessible.
2432 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2433 co-exist for some time) to make it clear that it is the Exim ID that is
2434 referenced, not the Message-ID: header line.
2436 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2437 string_format() function, because snprintf() does not exist on all
2440 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2441 ${stat: expansion item.
2443 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2444 protocol synchronization error", to keep the pedants happy.
2446 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2447 well as for IRIX systems, when gcc is being used. See the host.c source
2450 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2452 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2455 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2456 transport and to the smtp transport in LMTP mode.
2458 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2460 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2461 run a filter in a subprocess. This could lead to confusion in subsequent
2462 lookups in the parent process. There should also be a search_tidyup() at
2463 the end of the subprocess.
2465 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2466 only if the host matched helo_try_verify_hosts, which caused the
2467 verification to occur when the EHLO/HELO command was issued. The ACL just
2468 tested the remembered result. Now, if a previous verification attempt has
2469 not happened, "verify = helo" does it there and then.
2471 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2473 TK/03 Fix log output including CR from clamd.
2475 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2476 exist provoked a memory error which could cause a segfault.
2478 PH/15 Installed PCRE 6.2
2480 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2482 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2483 of the problem. Specifically, suggested +O2 rather than +O1 for the
2486 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2488 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2489 record, the port was not correctly logged when the outgoing_port log
2490 selector was set (it logged the transort's default port).
2492 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2493 fallback_hosts, and "hosts" in the smtp transport.
2495 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2496 host errors such as "Connection refused".
2498 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2499 authentication with radiusclient 0.4.9:
2501 - Error returned from rc_read_config was caught wrongly
2502 - Username/password not passed on to radius server due to wrong length.
2504 The presumption is that some radiusclient API changes for 4.51/PH/17
2505 were not taken care of correctly. The code is still untested by me (my
2506 Linux distribution still has 0.3.2 of radiusclient), but it was
2507 contributed by a Radius user.
2509 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2510 expanding the "port" option of the smtp transport.
2512 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2513 while reading a MIME header. Thanks to Tom Hughes for a patch.
2515 PH/24 Include config.h inside local_scan.h so that configuration settings are
2518 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2519 that in an ACL for RCPT (for example), you can examine exactly what was
2522 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2523 commands, but it was not correctly comparing the address with the actual
2524 client host address. Thus, it would show the EHLO address in Received:
2525 header lines when this was not necessary.
2527 PH/27 Added the % operator to ${eval:}.
2529 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2530 it should be ignoring failures (because with -C, for example, it has lost
2531 privilege). It wasn't ignoring creation failures other than "already
2534 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2537 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2538 because an input line was too long, either on its own, or by virtue of
2539 too many continuations, the temporary file was not being removed, and the
2540 return code was incorrect.
2542 PH/31 Missing "BOOL" in function definition in filtertest.c.
2544 PH/32 Applied Sieve patches from the maintainer.
2546 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2548 PH/33 Added "verify = not_blind".
2550 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2551 Local/Makefile (with some defaults set). These are used in built scripts
2552 such as exicyclog, but they have never been used in the exim_install
2553 script (though there are many overriding facilities there). I have
2554 arranged that the exim_install script now takes note of these two
2557 PH/35 Installed configuration files for Dragonfly.
2559 PH/36 When a locally submitted message by a trusted user did not contain a
2560 From: header, and the sender address was obtained from -f or from an SMTP
2561 MAIL command, and the trusted user did not use -F to supply a sender
2562 name, $originator_name was incorrectly used when constructing a From:
2563 header. Furthermore, $originator_name was used for submission mode
2564 messages from external hosts without From: headers in a similar way,
2565 which is clearly wrong.
2567 PH/37 Added control=suppress_local_fixups.
2569 PH/38 When log_selector = +received_sender was set, and the addition of the
2570 sender made the log line's construction buffer exactly full, or one byte
2571 less than full, an overflow happened when the terminating "\n" was
2574 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2575 when the result of a list match is failure because a DNS lookup failed.
2577 PH/40 RM_COMMAND is now used in the building process.
2579 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2580 the "build-* directories that it finds.
2582 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2583 address in a domain literal was a prefix of an interface address.
2585 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2586 when verifying a sender address, unless rewrite_headers is false.
2588 PH/44 Wrote a long comment about why errors_to addresses are verified as
2589 recipients, not senders.
2591 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2592 the ratelimit ACL was added.
2594 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2596 PH/46 Added extra information about PostgreSQL errors to the error string.
2598 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2599 avoiding the need to install special zones in a real server. This is
2600 backwards compatible; if it can't find the fake resolver, it drops back.
2601 Thus, both old and new test suites can be run.
2603 TF/02 Added util/ratelimit.pl
2605 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2606 clock is set back in time.
2608 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2609 Candler <B.Candler@pobox.com>.
2611 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2612 OK for addresses that are the result of redirections.
2614 PH/48 A number of further additions for the benefit of the new test suite,
2615 including a fake gethostbyname() that interfaces to the fake DNS resolver
2618 TF/06 The fix for widen_domains has also been applied to qualify_single and
2619 search_parents which are the other dnslookup options that can cause
2622 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2625 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2627 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2628 service" if it encountered an MX record with an empty target hostname.
2629 The message is now "an MX or SRV record indicated no SMTP service".
2631 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2632 if the DNS of the sending site is misconfigured. This is quite a
2633 common situation. This change restores the behaviour of treating a
2634 helo verification defer as a failure.
2636 PH/51 If self=fail was set on a router, the bounce message did not include the
2637 actual error message.
2643 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2645 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2646 command itself ended in error, the underlying message about the transport
2647 filter timeout was being overwritten with the pipe command error. Now the
2648 underlying error message should be appended to the second error message.
2650 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2652 PH/02 Reduce the amount of output that "make" produces by default. Full output
2653 can still be requested.
2655 PH/03 The warning log line about a condition test deferring for a "warn" verb
2656 was being output only once per connection, rather than after each
2657 occurrence (because it was using the same function as for successful
2658 "warn" verbs). This seems wrong, so I have changed it.
2660 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2661 it should not have, which might have caused a crash in the right
2662 circumstances, but probably never did.
2664 PH/04 Installed a modified version of Tony Finch's patch to make submission
2665 mode fix the return path as well as the Sender: header line, and to
2666 add a /name= option so that you can make the user's friendly name appear
2669 TF/03 Added the control = fakedefer ACL modifier.
2671 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2672 Mark Lowes for thorough testing.
2674 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2676 TK/03 Merged latest SRS patch from Miles Wilton.
2678 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2679 with the definition in sysexits.h (which is #included earlier).
2680 Fortunately, Exim does not actually use EX_OK. The code used to try to
2681 preserve the sysexits.h value, by assumimg that macro definitions were
2682 scanned for macro replacements. I have been disabused of this notion,
2683 so now the code just undefines EX_OK before #including unistd.h.
2685 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2686 in the smtp transport. When a block could not be written in a single
2687 write() function, the timeout was being re-applied to each part-write.
2688 This seems wrong - if the receiver was accepting one byte at a time it
2689 would take for ever. The timeout is now adjusted when this happens. It
2690 doesn't have to be particularly precise.
2692 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2693 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2695 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2696 without a domain if the check to <postmaster@domain> fails.
2698 SC/01 Eximstats: added -xls and the ability to specify output files
2699 (patch written by Frank Heydlauf).
2701 SC/02 Eximstats: use FileHandles for outputing results.
2703 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2705 SC/04 Eximstats: fixed display of large numbers with -nvr option
2707 SC/05 Eximstats: fixed merging of reports with empty tables.
2709 SC/06 Eximstats: added the -include_original_destination flag
2711 SC/07 Eximstats: removed tabs and trailing whitespace.
2713 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2715 TK/06 MBOX spool code: Add real "From " MBOX separator line
2716 so the .eml file is really in mbox format (even though
2717 most programs do not really care). Patch from Alex Miller.
2719 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2720 The latter is generated from $received_to and is only set if the
2721 message has one envelope recipient. SA can use these headers,
2722 obviously out-of-the-box. Patch from Alex Miller.
2724 PH/08 The ${def test on a variable was returning false if the variable's
2725 value was "0", contrary to what the specification has always said!
2726 The result should be true unless the variable is empty.
2728 PH/09 The syntax error of a character other than { following "${if
2729 def:variable_name" (after optional whitespace) was not being diagnosed.
2730 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2731 accidental colon was present, for example, could give incorrect results.
2733 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2734 result is used (not including appendfile, where other changes are about
2737 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2738 This involved changing a lot of size variables from int to off_t. It
2739 should work with maildirs and everything.
2741 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2742 spamd dying while we are connected to it.
2744 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2747 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2748 function for BDB 4.3.
2750 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2752 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2753 the sources for PCRE have been re-arranged and more files are now
2756 PH/15 The code I had for printing potentially long long variables in PH/11
2757 above was not the best (it lost precision). The length of off_t variables
2758 is now inspected at build time, and an appropriate printing format (%ld
2759 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2760 to be "long long int" or "long int". This is needed for the internal
2761 formatting function string_vformat().
2763 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2764 the configuration file to be ":syslog", then the script "guesses" where
2765 the logs files are, rather than using the compiled in default. In our
2766 case the guess is not the same as the compiled default, so the script
2767 suddenly stopped working when I started to use syslog. The patch checks
2768 to see if log_file_path is "". If so, it attempts to read it from exim
2769 with no configuration file to get the compiled in version, before it
2770 falls back to the previous guessing code."
2772 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2773 implementing BATV in an Exim configuration. See NewStuff for the gory
2776 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2777 Makefile that are specific to HP-UX.
2779 PH/18 If the "use_postmaster" option was set for a recipient callout together
2780 with the "random" option, the postmaster address was used as the MAIL
2781 FROM address for the random test, but not for the subsequent recipient
2782 test. It is now used for both.
2784 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2785 patch removes a few documentation additions to RFC 3028, because the
2786 latest draft now contains them. It adds the new en;ascii-case comparator
2787 and a new error check for 8bit text in MIME parts. Comparator and
2788 require names are now matched exactly. I enabled the subaddress
2789 extension, but it is not well tested yet (read: it works for me)."
2791 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2792 rework some of the code of TK/09 above to avoid the hardwired use of
2793 "%lld" and "long long". Replaced the call to snprintf() with a call to
2796 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2797 records point to non-existent hosts", "retry timeout exceeded", and
2798 "retry time not reached for any host after a long failure period".
2800 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2801 experimental DomainKeys support:
2803 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2804 (2) On an error such as an illegally used "control", the wrong name for
2805 the control was given.
2807 These problems did NOT occur unless DomainKeys support was compiled.
2809 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2811 PH/24 Added ${if match_ip condition.
2813 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2814 databases so that it will be absolutely obvious if a crash occurs in the
2815 DB library. This is a regular occurrence (often caused by mis-matched
2818 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2819 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2820 picked up on a user's system that detects such things. There doesn't seem
2821 to be a gcc warning option for this - only an attribute that has to be
2822 put on the function's prototype. It seems that in Fedora Core 4 they have
2823 set this on a number of new functions. No doubt there will be more in due
2826 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2827 specify a transport. However, if an address that was verified by such a
2828 router was the subject of a callout, Exim crashed because it tried to
2829 read the rcpt_include_affixes from the non-existent transport. Now it
2830 just assumes that the setting of that option is false. This bug was
2831 introduced by 4.51/PH/31.
2833 PH/28 Changed -d+all to exclude +memory, because that information is very
2834 rarely of interest, but it makes the output a lot bigger. People tend to
2835 do -d+all out of habit.
2837 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2838 code in os-type was giving problems when libc.so lives in lib64, like on
2841 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2842 aren't the modern standard, and it seems that some systems' include files
2843 don't always have them. Exim was already checking for some of the newer
2844 ones like T_AAAA, and defining it itself. I've added checks for all the
2845 record types that Exim uses.
2847 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2848 not automatically generating a new one, as it is supposed to. This
2849 prevented TLS from working. If the file did exist, but contained invalid
2850 data, a new version was generated, as expected. It was only the case of a
2851 non-existent file that was broken.
2853 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2854 with a change in libdomainkeys > 0.64.
2856 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2857 from DNS. If the selector record carries the flag, it now has
2858 precedence over the domain-wide flag.
2860 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2862 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2863 the use of a port name that isn't defined in /etc/services) occurred, the
2864 message was deferred as in a normal delivery, and thus remained on the
2865 spool, instead of being failed because of the mua_wrapper setting. This
2866 is now fixed, and I tidied up some of the mua_wrapper messages at the
2869 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2870 the messages in a hash of arrays rather than using individual hashes.
2871 This is a bit cleaner and results in dramatic memory savings, albeit
2872 at a slight CPU cost.
2874 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2875 as requested by Marc Sherman.
2877 SC/10 Eximstats: added histograms for user specified patterns as requested
2880 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2882 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2883 fopen() in the content-scanning modules that did not already have it.
2889 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2890 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2892 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2894 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2896 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2897 to test Sieve filters that use "vacation".
2899 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2900 that changes the way the GnuTLS parameters are stored in the cache file.
2901 The new format can be generated externally. For backward compatibility,
2902 if the data in the cache doesn't make sense, Exim assumes it has read an
2903 old-format file, and it generates new data and writes a new file. This
2904 means that you can't go back to an older release without removing the
2907 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2908 work if there are any delivery delays because "one_time" forces the
2909 parent to be marked "delivered", so its unseen clone is never tried
2910 again. For this reason, Exim now forbids the simultaneous setting of
2913 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2914 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2915 entry if you want to know the details. The fix, however, neglected to
2916 consider the case when local delivery batching is involved. The test for
2917 "previously delivered" was not happening when checking to see if an
2918 address could be batched with a previous (undelivered) one; under
2919 certain circumstances this could lead to multiple deliveries to the same
2922 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2923 in its include files, and this causes problems building Exim.
2925 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2926 header_syntax) but Exim was just ignoring anything given after a slash.
2927 In particular, this caused confusion with an attempt to use "verify =
2928 reverse_host_lookup/defer_ok". An error is now given when options are
2929 supplied for verify items that do not have them. (Maybe reverse_host_
2930 lookup should have a defer_ok option, but that's a different point.)
2932 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2933 defined by RFC 821) to 2048, because there were problems with some AUTH
2934 commands, and RFC 1869 says the size should be increased for extended
2935 SMTP commands that take arguments.
2937 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2940 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2941 "unknown" error; now it says that the functionality isn't in the binary.
2943 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2944 an address' error message when a string expansion fails (syntax or
2945 whatever). Otherwise the password may appear in the log. Following change
2946 PH/42 below, there is no longer a chance of it appearing in a bounce
2949 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2951 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2952 of its addresses was ever tried. (Bugzilla bug #2.)
2954 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2955 the result incorrectly in the debug output. (It correctly added a newline
2956 to what was transported.)
2958 TF/01 Added $received_time.
2960 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2961 commented out examples of how to interface to a virus scanner and to
2962 SpamAssassin. Also added commented examples of av_scanner and
2963 spamd_address settings.
2965 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2966 and controls are allowed in which ACLs. There were a couple of minor
2967 errors. Some of the entries in the conditions table (which is a table of
2968 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2969 negation of where the condition IS allowed.
2971 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2973 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2974 header file does not have a version number, so I've had to invent a new
2975 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2976 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2977 radiusclient), but it was contributed by a Radius user.
2979 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2980 files or directories, for OpenSSL.
2982 PH/19 When an Exim process that is running as root has to create an Exim log
2983 file, it does so in a subprocess that runs as exim:exim so as to get the
2984 ownership right at creation (otherwise, other Exim processes might see
2985 the file with the wrong ownership). There was no test for failure of this
2986 fork() call, which would lead to the process getting stuck as it waited
2987 for a non-existent subprocess. Forks do occasionally fail when resources
2988 run out. I reviewed all the other calls to fork(); they all seem to check
2991 PH/20 When checking for unexpected SMTP input at connect time (before writing
2992 the banner), Exim was not dealing correctly with a non-positive return
2993 from the read() function. If the client had disconnected by this time,
2994 the result was a log entry for a synchronization error with an empty
2995 string after "input=" when read() returned zero. If read() returned -1
2996 (an event I could not check), uninitialized data bytes were printed.
2997 There were reports of junk text (parts of files, etc) appearing after
3000 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3002 PH/22 Added support for macro redefinition, and (re)definition in between
3003 driver and ACL definitions.
3005 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3006 forgetting to use the resulting value; it was using the unexpanded value.
3008 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3009 hadn't been configured. The fix is from Juergen Kreileder, who
3010 understands it better than I do:
3012 "Here's what I see happening with three configured cyrus_sasl
3013 authenticators configured (plain, login, cram-md5):
3015 On startup auth_cyrus_sasl_init() gets called for each of these.
3016 This means three calls to sasl_listmech() without a specified mech_list.
3017 => SASL tests which mechs of all available mechs actually work
3018 => three warnings about OTP not working
3019 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3021 With the patch, sasl_listmech() also gets called three times. But now
3022 SASL's mech_list option is set to the server_mech specified in the the
3023 authenticator. Or in other words, the answer from sasl_listmech()
3024 gets limited to just the mech you're testing for (which is different
3026 => the return list contains just 'plain' or 'login', 'cram-md5' or
3027 nothing depending on the value of ob->server_mech.
3029 I've just tested the patch: Authentication still works fine,
3030 unavailable mechs specified in the exim configuration are still
3031 caught, and the auth.log warnings about OTP are gone."
3033 PH/25 When debugging is enabled, the contents of the command line are added
3034 to the debugging output, even when log_selector=+arguments is not
3037 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3038 answer is "GNU", and only if the return is "GNU/something" is the answer
3041 PH/27 $acl_verify_message is now set immediately after the failure of a
3042 verification in an ACL, and so is available in subsequent modifiers. In
3043 particular, the message can be preserved by coding like this:
3045 warn !verify = sender
3046 set acl_m0 = $acl_verify_message
3048 Previously, $acl_verify_message was set only while expanding "message"
3049 and "log_message" when a very denied access.
3051 PH/28 Modified OS/os.c-Linux with
3053 -#ifndef OS_LOAD_AVERAGE
3054 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3056 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3057 nomenclature these days.)
3059 PH/29 Installed patch from the Sieve maintainer that adds the options
3060 sieve_useraddress and sieve_subaddress to the redirect router.
3062 PH/30 In these circumstances:
3063 . Two addresses routed to the same list of hosts;
3064 . First host does not offer TLS;
3065 . First host accepts first address;
3066 . First host gives temporary error to second address;
3067 . Second host offers TLS and a TLS session is established;
3068 . Second host accepts second address.
3069 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3070 and peerdn, if requested) that were in fact used only for the second
3073 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3074 attention to any local part prefix or suffix that was matched by the
3075 router that accepted the address. It now behaves in the same way as it
3076 does for delivery: the affixes are removed from the local part unless
3077 rcpt_include_affixes is set on the transport.
3079 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3080 timeout during the DATA phase of an incoming message.
3082 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3083 applied a patch sanctioned by the Sieve maintainer.
3085 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3086 the uid or gid is negative. A case of a negative gid caused this to be
3087 noticed. The fix allows for either to be negative.
3089 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3090 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3093 PH/36 The change PH/12 above was broken. Fixed it.
3095 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3096 the grounds that routing the same address twice would always produce the
3097 same answer. This might have been true once, but it is certainly no
3098 longer true now. Routing a child address may depend on the previous
3099 routing that produced that child. Some complicated redirection strategies
3100 went wrong when messages had multiple recipients, and made Exim's
3101 behaviour dependent on the order in which the addresses were given.
3103 I have moved the duplicate checking until after the routing is complete.
3104 Exim scans the addresses that are assigned to local and remote
3105 transports, and removes any duplicates. This means that more work will be
3106 done, as duplicates will always all be routed, but duplicates are
3107 presumably rare, so I don't expect this is of any significance.
3109 For deliveries to pipes, files, and autoreplies, the duplicate checking
3110 still happens during the routing process, since they are not going to be
3113 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3114 It corrects a timeout issue with spamd. This is Ian's comment: "The
3115 background is that sometimes spamd either never reads data from a
3116 connection it has accepted, or it never writes response data. The exiscan
3117 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3118 blindly assumes that writes won't block so it may never time out."
3120 PH/39 Allow G after quota size as well as K and M.
3122 PH/40 The value set for $authenticated_id in an authenticator may not contain
3123 binary zeroes or newlines because the value is written to log lines and
3124 to spool files. There was no check on this. Now the value is run through
3125 the string_printing() function so that such characters are converted to
3126 printable escape sequences.
3128 PH/41 $message_linecount is a new variable that contains the total number of
3129 lines in the message. Compare $body_linecount, which is the count for the
3132 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3133 bounce and delay warning messages, except in certain special cases, which
3136 (a) An SMTP error message from a remote host;
3137 (b) A message specified in a :fail: redirection;
3138 (c) A message specified in a "fail" command in a system filter;
3139 (d) A message specified in a FAIL return from the queryprogram router;
3140 (e) A message specified by the cannot_route_message router option.
3142 In these cases only, Exim does include the error details in bounce and
3143 warning messages. There are also a few cases where bland messages such
3144 as "unrouteable address" or "local delivery error" are given.
3146 PH/43 $value is now also set for the "else" part of a ${run expansion.
3148 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3149 being worked on, but at least Exim now implements the latest version to
3152 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3153 process to complete was treated as a delivery failure, a timeout while
3154 writing the message to the pipe was logged, but erroneously treated as a
3155 successful delivery. Such timeouts include transport filter timeouts. For
3156 consistency with the overall process timeout, these timeouts are now
3157 treated as errors, giving rise to delivery failures by default. However,
3158 there is now a new Boolean option for the pipe transport called
3159 timeout_defer, which, if set TRUE, converts the failures into defers for
3160 both kinds of timeout. A transport filter timeout is now identified in
3163 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3164 systems where "make" and "gmake" are different, calling "gmake" at top
3165 level broke things. I've arranged for the value of $(MAKE) to be passed
3166 from the Makefile to this script so that it can call the same version of
3170 A note about Exim versions 4.44 and 4.50
3171 ----------------------------------------
3173 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3174 changes of various kinds. As a consequence, a big documentation update was
3175 needed. This delayed the release for rather longer than seemed good, especially
3176 in the light of a couple of (minor) security issues. Therefore, the changes
3177 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3178 release. So 4.44 and 4.50 are in effect two different branches that both start
3181 I have left the 4.50 change log unchanged; it contains all the changes since
3182 4.43. The change log for 4.44 is below; many of its items are identical to
3183 those for 4.50. This seems to be the most sensible way to preserve the
3184 historical information.
3190 1. Minor wording change to the doc/README.SIEVE file.
3192 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3193 computation of the current number of files was incorrect.
3195 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3196 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3199 4. Give more explanation in the error message when the command for a transport
3200 filter fails to execute.
3202 5. There are several places where Exim runs a non-Exim command in a
3203 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3204 was being done only for the command run by the queryprogram router. It is
3205 now done for all such subprocesses. The other cases are: ${run, transport
3206 filters, and the commands run by the lmtp and pipe transports.
3208 6. Added CONFIGURE_GROUP build-time option.
3210 7. Some older OS have a limit of 256 on the maximum number of file
3211 descriptors. Exim was using setrlimit() to set 1000 as a large value
3212 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3213 systems. I've change it so that if it can't get 1000, it tries for 256.
3215 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3216 was an oversight, and furthermore, ever since the addition of extra
3217 controls (e.g. 4.43/32), the checks on when to allow different forms of
3218 "control" were broken. There should now be diagnostics for all cases when a
3219 control that does not make sense is encountered.
3221 9. Added the /retain_sender option to "control=submission".
3223 10. $recipients is now available in the predata ACL (oversight).
3225 11. Tidy the search cache before the fork to do a delivery from a message
3226 received from the command line. Otherwise the child will trigger a lookup
3227 failure and thereby defer the delivery if it tries to use (for example) a
3228 cached ldap connection that the parent has called unbind on.
3230 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3231 of $address_data from the recipient verification was clobbered by the
3232 sender verification.
3234 13. The value of address_data from a sender verification is now available in
3235 $sender_address_data in subsequent conditions in the ACL statement.
3237 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3239 15. Added a new option "connect=<time>" to callout options, to set a different
3242 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3243 was its contents. (It was OK if the option was not defined at all.)
3245 17. A "Completed" log line is now written for messages that are removed from
3246 the spool by the -Mrm option.
3248 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3249 information about exactly what failed.
3251 19. Added -dd to debug only the daemon process.
3253 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3254 handles timeouts, both on the server side and network timeouts. Renamed the
3255 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3257 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3258 It is now set to "smtps".
3260 22. $host_address is now set to the target address during the checking of
3261 ignore_target_hosts.
3263 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3264 being passed; this would have caused $sender_host_name to have been used if
3265 matching the list had actually called for a host name (not very likely,
3266 since this list is usually IP addresses). A host name is now passed as
3269 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3270 code that specifies a non-restarting handler (typically sigaction() in
3271 modern systems) in an attempt to fix a rare and obscure crash bug.
3273 25. Narrowed the window for a race in the daemon that could cause it to ignore
3274 SIGCHLD signals. This is not a major problem, because they are used only to
3275 wake it up if nothing else does.
3277 26. A malformed maildirsize file could cause Exim to calculate negative values
3278 for the mailbox size or file count. Odd effects could occur as a result.
3279 The maildirsize information is now recalculated if the size or filecount
3282 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3283 support for a long time. Removed HAVE_SYS_VFS_H.
3285 28. Installed the latest version of exipick from John Jetmore.
3287 29. In an address list, if the pattern was not a regular expression, an empty
3288 subject address (from a bounce message) matched only if the pattern was an
3289 empty string. Non-empty patterns were not even tested. This was the wrong
3290 because it is perfectly reasonable to use an empty address as part of a
3291 database query. An empty address is now tested by patterns that are
3292 lookups. However, all the other forms of pattern expect the subject to
3293 contain a local part and a domain, and therefore, for them, an empty
3294 address still always fails if the pattern is not itself empty.
3296 30. Exim went into a mad DNS loop when attempting to do a callout where the
3297 host was specified on an smtp transport, and looking it up yielded more
3298 than one IP address.
3300 31. Re-factored the code for checking spool and log partition space into a
3301 function that finds that data and another that does the check. The former
3302 is then used to implement four new variables: $spool_space, $log_space,
3303 $spool_inodes, and $log_inodes.
3305 32. The RFC2047 encoding function was originally intended for short strings
3306 such as real names; it was not keeping to the 75-character limit for
3307 encoded words that the RFC imposes. It now respects the limit, and
3308 generates multiple encoded words if necessary. To be on the safe side, I
3309 have increased the buffer size for the ${rfc2047: expansion operator from
3312 33. It is now permitted to omit both strings after an "if" condition; if the
3313 condition is true, the result is "true". As before, when the second string
3314 is omitted, a false condition yields an empty string. This makes it less
3315 cumbersome to write custom ACL and router conditions.
3317 34. Failure to deliver a bounce message always caused it to be frozen, even if
3318 there was an errors_to setting on the router. The errors_to setting is now
3321 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3322 canonical form (fully expanded) before being placed in
3323 $sender_host_address.
3325 36. The table in the code that translates DNS record types into text (T_A to
3326 "A" for instance) was missing entries for NS and CNAME. It is just possible
3327 that this could have caused confusion if both these types were looked up
3328 for the same domain, because the text type is used as part of Exim's
3329 per-process caching. But the chance of anyone hitting this buglet seems
3332 37. The dnsdb lookup has been extended in a number of ways.
3334 (1) There is a new type, "zns", which walks up the domain tree until it
3335 finds some nameserver records. It should be used with care.
3337 (2) There is a new type, "mxh", which is like "mx" except that it returns
3338 just the host names, not the priorities.
3340 (3) It is now possible to give a list of domains (or IP addresses) to be
3341 looked up. The behaviour when one of the lookups defers can be
3342 controlled by a keyword.
3344 (4) It is now possible to specify the separator character for use when
3345 multiple records are returned.
3347 38. The dnslists ACL condition has been extended: it is now possible to supply
3348 a list of IP addresses and/or domains to be looked up in a particular DNS
3351 39. Added log_selector=+queue_time_overall.
3353 40. When running the queue in the test harness, wait just a tad after forking a
3354 delivery process, to get repeatability of debugging output.
3356 41. Include certificate and key file names in error message when GnuTLS fails
3357 to set them up, because the GnuTLS error message doesn't include the name
3358 of the failing file when there is a problem reading it.
3360 42. Allow both -bf and -bF in the same test run.
3362 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3364 44. The "Exiscan patch" is now merged into the mainline Exim source.
3366 45. Sometimes the final signoff response after QUIT could fail to get
3367 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3368 < 0 before doing a fflush(). This bug looks as though it goes back to the
3369 introduction of TLS in release 3.20, but "sometimes" must have been rare
3370 because the tests only now provoked it.
3372 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3373 (this can affect the format of dates).
3375 47. exim_tidydb, when checking for the continued existence of a message for
3376 which it has found a message-specific retry record, was not finding
3377 messages that were in split spool directories. Consequently, it was
3378 deleting retry records that should have stayed in existence.
3380 48. Steve fixed some bugs in eximstats.
3382 49. The SPA authentication driver was not abandoning authentication and moving
3383 on to the next authenticator when an expansion was forced to fail,
3384 contradicting the general specification for all authenticators. Instead it
3385 was generating a temporary error. It now behaves as specified.
3387 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3388 (the order specifies the preference for clients). The order is now AES256,
3389 AES128, 3DES, ARCFOUR128.
3391 51. Small patch to Sieve code - explicitly set From: when generating an
3394 52. Exim crashed if a remote delivery caused a very long error message to be
3395 recorded - for instance if somebody sent an entire SpamAssassin report back
3396 as a large number of 550 error lines. This bug was coincidentally fixed by
3397 increasing the size of one of Exim's internal buffers (big_buffer) that
3398 happened as part of the Exiscan merge. However, to be on the safe side, I
3399 have made the code more robust (and fixed the comments that describe what
3402 53. Now that there can be additional text after "Completed" in log lines (if
3403 the queue_time_overall log selector is set), a one-byte patch to exigrep
3404 was needed to allow it to recognize "Completed" as not the last thing in
3407 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3408 patch that reportedly fixes this has been added. I am not expert enough to
3409 create a test for it. This is what the patch creator wrote:
3411 "I found a little strange behaviour of ldap code when working with
3412 Windows 2003 AD Domain, where users was placed in more than one
3413 Organization Units. When I tried to give exim partial DN, the exit code
3414 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3415 But simultaneously result of request was absolutely normal ldap result,
3416 so I produce this patch..."
3418 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3419 REFERENCE, so I have modified the code to exclude the patch when that macro
3422 55. Some experimental protocols are using DNS PTR records for new purposes. The
3423 keys for these records are domain names, not reversed IP addresses. The
3424 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3425 leaves it alone. Component reversal etc. now happens only for IP addresses.
3428 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3430 57. Double the size of the debug message buffer (to 2048) so that more of very
3431 long debug lines gets shown.
3433 58. The exicyclog utility now does better if the number of log files to keep
3434 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3436 59. Two changes related to the smtp_active_hostname option:
3438 (1) $smtp_active_hostname is now available as a variable.
3439 (2) The default for smtp_banner uses $smtp_active_hostname instead
3440 of $primary_hostname.
3442 60. The host_aton() function is supposed to be passed a string that is known
3443 to be a valid IP address. However, in the case of IPv6 addresses, it was
3444 not checking this. This is a hostage to fortune. Exim now panics and dies
3445 if the condition is not met. A case was found where this could be provoked
3446 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3447 components; fortuitously, this particular loophole had already been fixed
3448 by change 4.50/55 above.
3450 If there are any other similar loopholes, the new check in host_aton()
3451 itself should stop them being exploited. The report I received stated that
3452 data on the command line could provoke the exploit when Exim was running as
3453 exim, but did not say which command line option was involved. All I could
3454 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3455 running as the user.
3458 61. There was a buffer overflow vulnerability in the SPA authentication code
3459 (which came originally from the Samba project). I have added a test to the
3460 spa_base64_to_bits() function which I hope fixes it.
3463 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3464 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3466 63. The daemon start-up calls getloadavg() while still root for those OS that
3467 need the first call to be done as root, but it missed one case: when
3468 deliver_queue_load_max is set with deliver_drop_privilege. This is
3469 necessary for the benefit of the queue runner, because there is no re-exec
3470 when deliver_drop_privilege is set.
3472 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3473 This has been fixed.
3475 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3476 was in use, was not putting the data itself into the right store pool;
3477 consequently, it could be overwritten for a subsequent message in the same
3478 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3481 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3483 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3484 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3485 the function were treating the return as a boolean value, which happened to
3486 work because 0=false and not-0=true, but is not correct code.
3488 68. The host_aton() function was not handling scoped IPv6 addresses (those
3489 with, for example, "%eth0" on the end) correctly.
3491 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3492 negated items (that is, ~something) in unsigned ints. Some compilers
3493 apparently mutter when there is no cast.
3495 70. If an address verification called from an ACL failed, and did not produce a
3496 user-specific message (i.e. there was only a "system" message), nothing was
3497 put in $acl_verify_message. In this situation, it now puts the system
3500 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3501 session; change 4.31/43 added the unwanted input to the log line - except
3502 that it did not do this in the start of session case. It now does.
3504 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3505 This is wrong and can cause the other end to generate a synchronization
3506 error if it is another Exim or anything else that does the synchronization
3507 check. A QUIT command is no longer sent after a timeout.
3509 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3510 during host lookups.
3512 74. The defer_ok option of callout verification was not working if it was used
3513 when verifying addresses in header lines, that is, for this case:
3515 verify = header_sender/callout=defer_ok
3517 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3518 those file descriptors could be used for SMTP connections. If anything
3519 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3520 could be sent to the SMTP client, causing chaos. The daemon now opens
3521 stdin, stdout, and stderr to /dev/null when it puts itself into the
3524 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3525 log by default. The user can override this with suitable Perl magic.
3527 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3528 the log message when discard triggers, was not working for the DATA ACL or
3529 for the non-SMTP ACL.
3531 78. Error message wording change in sieve.c.
3533 79. If smtp_accept_max_per_host was set, the number of connections could be
3534 restricted to fewer than expected, because the daemon was trying to set up
3535 a new connection before checking whether the processes handling previous
3536 connections had finished. The check for completed processes is now done
3537 earlier. On busy systems, this bug wouldn't be noticed because something
3538 else would have woken the daemon, and it would have reaped the completed
3541 80. If a message was submitted locally by a user whose login name contained one
3542 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3543 It caused a spool format error. I have fixed the spool reading code. A
3544 related problem was that the "from" clause in the Received: line became
3545 illegal because of the space(s). It is now covered by ${quote_local_part.
3547 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3550 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3552 83. Patch from Sieve maintainer to fix unterminated string problem in
3553 "vacation" handling.
3555 84. Some minor changes to the Linux configuration files to help with other
3556 OS variants using glibc.
3558 85. One more patch for Sieve to update vacation handling to latest spec.
3561 ----------------------------------------------------
3562 See the note above about the 4.44 and 4.50 releases.
3563 ----------------------------------------------------
3569 1. Change 4.43/35 introduced a bug that caused file counts to be
3570 incorrectly computed when quota_filecount was set in an appendfile
3573 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3574 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3577 3. Give more explanation in the error message when the command for a transport
3578 filter fails to execute.
3580 4. There are several places where Exim runs a non-Exim command in a
3581 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3582 was being done only for the command run by the queryprogram router. It is
3583 now done for all such subprocesses. The other cases are: ${run, transport
3584 filters, and the commands run by the lmtp and pipe transports.
3586 5. Some older OS have a limit of 256 on the maximum number of file
3587 descriptors. Exim was using setrlimit() to set 1000 as a large value
3588 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3589 systems. I've change it so that if it can't get 1000, it tries for 256.
3591 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3592 was an oversight, and furthermore, ever since the addition of extra
3593 controls (e.g. 4.43/32), the checks on when to allow different forms of
3594 "control" were broken. There should now be diagnostics for all cases when a
3595 control that does not make sense is encountered.
3597 7. $recipients is now available in the predata ACL (oversight).
3599 8. Tidy the search cache before the fork to do a delivery from a message
3600 received from the command line. Otherwise the child will trigger a lookup
3601 failure and thereby defer the delivery if it tries to use (for example) a
3602 cached ldap connection that the parent has called unbind on.
3604 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3605 of $address_data from the recipient verification was clobbered by the
3606 sender verification.
3608 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3609 was its contents. (It was OK if the option was not defined at all.)
3611 11. A "Completed" log line is now written for messages that are removed from
3612 the spool by the -Mrm option.
3614 12. $host_address is now set to the target address during the checking of
3615 ignore_target_hosts.
3617 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3618 being passed; this would have caused $sender_host_name to have been used if
3619 matching the list had actually called for a host name (not very likely,
3620 since this list is usually IP addresses). A host name is now passed as
3623 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3624 code that specifies a non-restarting handler (typically sigaction() in
3625 modern systems) in an attempt to fix a rare and obscure crash bug.
3627 15. Narrowed the window for a race in the daemon that could cause it to ignore
3628 SIGCHLD signals. This is not a major problem, because they are used only to
3629 wake it up if nothing else does.
3631 16. A malformed maildirsize file could cause Exim to calculate negative values
3632 for the mailbox size or file count. Odd effects could occur as a result.
3633 The maildirsize information is now recalculated if the size or filecount
3636 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3637 support for a long time. Removed HAVE_SYS_VFS_H.
3639 18. Updated exipick to current release from John Jetmore.
3641 19. Allow an empty sender to be matched against a lookup in an address list.
3642 Previously the only cases considered were a regular expression, or an
3645 20. Exim went into a mad DNS lookup loop when doing a callout where the
3646 host was specified on the transport, if the DNS lookup yielded more than
3649 21. The RFC2047 encoding function was originally intended for short strings
3650 such as real names; it was not keeping to the 75-character limit for
3651 encoded words that the RFC imposes. It now respects the limit, and
3652 generates multiple encoded words if necessary. To be on the safe side, I
3653 have increased the buffer size for the ${rfc2047: expansion operator from
3656 22. Failure to deliver a bounce message always caused it to be frozen, even if
3657 there was an errors_to setting on the router. The errors_to setting is now
3660 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3661 canonical form (fully expanded) before being placed in
3662 $sender_host_address.
3664 24. Updated eximstats to version 1.33
3666 25. Include certificate and key file names in error message when GnuTLS fails
3667 to set them up, because the GnuTLS error message doesn't include the name
3668 of the failing file when there is a problem reading it.
3670 26. Expand error message when OpenSSL has problems setting up cert/key files.
3673 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3674 (this can affect the format of dates).
3676 28. exim_tidydb, when checking for the continued existence of a message for
3677 which it has found a message-specific retry record, was not finding
3678 messages that were in split spool directories. Consequently, it was
3679 deleting retry records that should have stayed in existence.
3681 29. eximstats updated to version 1.35
3682 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3683 1.35 - bugfix such that pie charts by volume are generated correctly
3685 30. The SPA authentication driver was not abandoning authentication and moving
3686 on to the next authenticator when an expansion was forced to fail,
3687 contradicting the general specification for all authenticators. Instead it
3688 was generating a temporary error. It now behaves as specified.
3690 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3691 (the order specifies the preference for clients). The order is now AES256,
3692 AES128, 3DES, ARCFOUR128.
3694 31. Small patch to Sieve code - explicitly set From: when generating an
3697 32. Exim crashed if a remote delivery caused a very long error message to be
3698 recorded - for instance if somebody sent an entire SpamAssassin report back
3699 as a large number of 550 error lines. This bug was coincidentally fixed by
3700 increasing the size of one of Exim's internal buffers (big_buffer) that
3701 happened as part of the Exiscan merge. However, to be on the safe side, I
3702 have made the code more robust (and fixed the comments that describe what
3705 33. Some experimental protocols are using DNS PTR records for new purposes. The
3706 keys for these records are domain names, not reversed IP addresses. The
3707 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3708 leaves it alone. Component reversal etc. now happens only for IP addresses.
3711 34. The host_aton() function is supposed to be passed a string that is known
3712 to be a valid IP address. However, in the case of IPv6 addresses, it was
3713 not checking this. This is a hostage to fortune. Exim now panics and dies
3714 if the condition is not met. A case was found where this could be provoked
3715 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3716 components; fortuitously, this particular loophole had already been fixed
3717 by change 4.50/55 or 4.44/33 above.
3719 If there are any other similar loopholes, the new check in host_aton()
3720 itself should stop them being exploited. The report I received stated that
3721 data on the command line could provoke the exploit when Exim was running as
3722 exim, but did not say which command line option was involved. All I could
3723 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3724 running as the user.
3727 35. There was a buffer overflow vulnerability in the SPA authentication code
3728 (which came originally from the Samba project). I have added a test to the
3729 spa_base64_to_bits() function which I hope fixes it.
3732 36. The daemon start-up calls getloadavg() while still root for those OS that
3733 need the first call to be done as root, but it missed one case: when
3734 deliver_queue_load_max is set with deliver_drop_privilege. This is
3735 necessary for the benefit of the queue runner, because there is no re-exec
3736 when deliver_drop_privilege is set.
3738 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3739 was in use, was not putting the data itself into the right store pool;
3740 consequently, it could be overwritten for a subsequent message in the same
3741 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3744 38. Sometimes the final signoff response after QUIT could fail to get
3745 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3746 < 0 before doing a fflush(). This bug looks as though it goes back to the
3747 introduction of TLS in release 3.20, but "sometimes" must have been rare
3748 because the tests only now provoked it.
3754 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3755 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3756 Now it returns an integer. A number of calls were still expecting a T/F
3757 return. Fortuitously, in all cases, the tests worked in OK situations,
3758 which is the norm. However, things would have gone wrong on any write
3759 failures on the smtp file descriptor. This function is used when sending
3760 messages over SMTP and also when doing verify callouts.
3762 2. When Exim is called to do synchronous delivery of a locally submitted
3763 message (the -odf or -odi options), it no longer closes stderr before doing
3766 3. Implemented the mua_wrapper option.
3768 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3770 5. Implemented the functions header_remove(), header_testname(),
3771 header_add_at_position(), and receive_remove_recipient(), and exported them
3774 6. If an ACL "warn" statement specified the addition of headers, Exim already
3775 inserted X-ACL-Warn: at the start if there was no header name. However, it
3776 was not making this test for the second and subsequent header lines if
3777 there were newlines in the string. This meant that an invalid header could
3778 be inserted if Exim was badly configured.
3780 7. Allow an ACL "warn" statement to add header lines at the start or after all
3781 the Received: headers, as well as at the end.
3783 8. Added the rcpt_4xx retry error code.
3785 9. Added postmaster_mailfrom=xxx to callout verification option.
3787 10. Added mailfrom=xxxx to the callout verification option, for verify=
3790 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3791 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3793 12. Inserted some casts to stop certain compilers warning when using pointer
3794 differences as field lengths or precisions in printf-type calls (mostly
3795 affecting debugging statements).
3797 13. Added optional readline() support for -be (dynamically loaded).
3799 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3800 same clock tick as a message's arrival, so that its received time was the
3801 same as the "first fail" time on the retry record, and that message
3802 remained on the queue past the ultimate address timeout, every queue runner
3803 would try a delivery (because it was past the ultimate address timeout) but
3804 after another failure, the ultimate address timeout, which should have then
3805 bounced the address, did not kick in. This was a "< instead of <=" error;
3806 in most cases the first failure would have been in the next clock tick
3807 after the received time, and all would be well.
3809 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3810 being recognized when the domain list was tested by the match_domain
3811 condition in an expansion string.
3813 16. Added the ${str2b64: operator.
3815 17. Exim was always calling setrlimit() to set a large limit for the number of
3816 processes, without checking whether the existing limit was already
3817 adequate. (It did check for the limit on file descriptors.) Furthermore,
3818 errors from getrlimit() and setrlimit() were being ignored. Now they are
3819 logged to the main and panic logs, but Exim does carry on, to try to do its
3820 job under whatever limits there are.
3822 18. Imported PCRE 5.0.
3824 19. Trivial typo in log message " temporarily refused connection" (the leading
3827 20. If the log selector return_path_on_delivery was set and an address was
3828 redirected to /dev/null, the delivery process crashed because it assumed
3829 that a return path would always be set for a "successful" delivery. In this
3830 case, the whole delivery is bypassed as an optimization, and therefore no
3833 21. Internal re-arrangement: the function for sending a challenge and reading
3834 a response while authentication was assuming a zero-terminated challenge
3835 string. It's now changed to take a pointer and a length, to allow for
3836 binary data in such strings.
3838 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3840 23. Exim was not respecting finduser_retries when seeking the login of the
3841 uid under which it was called; it was always trying 10 times. (The default
3842 setting of finduser_retries is zero.) Also, it was sleeping after the final
3843 failure, which is pointless.
3845 24. Implemented tls_on_connect_ports.
3847 25. Implemented acl_smtp_predata.
3849 26. If the domain in control=submission is set empty, Exim assumes that the
3850 authenticated id is a complete email address when it generates From: or
3851 Sender: header lines.
3853 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3854 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3855 chown and chgrp in /bin and hostname in /usr/bin.
3857 28. Exim was keeping the "process log" file open after each use, just as it
3858 does for the main log. This opens the possibility of it remaining open for
3859 long periods when the USR1 signal hits a daemon. Occasional processlog
3860 errors were reported, that could have been caused by this. Anyway, it seems
3861 much more sensible not to leave this file open at all, so that is what now
3864 29. The long-running daemon process does not normally write to the log once it
3865 has entered its main loop, and it closes the log before doing so. This is
3866 so that log files can straightforwardly be renamed and moved. However,
3867 there are a couple of unusual error situations where the daemon does write
3868 log entries, and I had neglected to close the log afterwards.
3870 30. The text of an SMTP error response that was received during a remote
3871 delivery was being truncated at 512 bytes. This is too short for some of
3872 the long messages that one sometimes sees. I've increased the limit to
3875 31. It is now possible to make retry rules that apply only when a message has a
3876 specific sender, in particular, an empty sender.
3878 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3879 it possible to be selective about when SMTP synchronization is enforced.
3881 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3883 32. Implemented hosts_connection_nolog.
3885 33. Added an ACL for QUIT.
3887 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3890 35. Added mailbox_size and mailbox_filecount to appendfile.
3892 36. Added control = no_multiline_responses to ACLs.
3894 37. There was a bug in the logic of the code that waits for the clock to tick
3895 in the case where the clock went backwards by a substantial amount such
3896 that the microsecond fraction of "now" was more than the microsecond
3897 fraction of "then" (but the whole seconds number was less).
3899 38. Added support for the libradius Radius client library this is found on
3900 FreeBSD (previously only the radiusclient library was supported).
3906 1. When certain lookups returned multiple values in the form name=value, the
3907 quoting of the values was not always being done properly. Specifically:
3908 (a) If the value started with a double quote, but contained no whitespace,
3910 (b) If the value contained whitespace other than a space character (i.e.
3911 tabs or newlines or carriage returns) it was not quoted.
3912 This fix has been applied to the mysql and pgsql lookups by writing a
3913 separate quoting function and calling it from the lookup code. The fix
3914 should probably also be applied to nisplus, ibase and oracle lookups, but
3915 since I cannot test any of those, I have not disturbed their existing code.
3917 2. A hit in the callout cache for a specific address caused a log line with no
3918 reason for rejecting RCPT. Now it says "Previous (cached) callout
3919 verification failure".
3921 3. There was an off-by-one bug in the queryprogram router. An over-long
3922 return line was truncated at 256 instead of 255 characters, thereby
3923 overflowing its buffer with the terminating zero. As well as fixing this, I
3924 have increased the buffer size to 1024 (and made a note to document this).
3926 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3927 when Exim is waiting for an SMTP response from a remote server, Exim
3928 restarts its select() call on the socket, thereby resetting its timeout.
3929 This is not a problem when such interrupts are rare. Somebody set up a cron
3930 job to run exiwhat every 2 minutes, which is less than the normal select()
3931 timeout (5 or 10 minutes). This meant that the select() timeout never
3932 kicked in because it was always reset. I have fixed this by comparing the
3933 time when an interrupt arrives with the time at the start of the first call
3934 to select(). If more time than the timeout has elapsed, the interrupt is
3935 treated as a timeout.
3937 5. Some internal re-factoring in preparation for the addition of Sieve
3938 extensions (by MH). In particular, the "personal" test is moved to a
3939 separate function, and given an option for scanning Cc: and Bcc: (which is
3940 not set for Exim filters).
3942 6. When Exim created an email address using the login of the caller as the
3943 local part (e.g. when creating a From: or Sender: header line), it was not
3944 quoting the local part when it contained special characters such as @.
3946 7. Installed new OpenBSD configuration files.
3948 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3949 try to make them clearer.
3951 9. Callout options, other than the timeout value, were being ignored when
3952 verifying sender addresses in header lines. For example, when using
3954 verify = header_sender/callout=no_cache
3956 the cache was (incorrectly) being used.
3958 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3959 only the Cygwin environment.
3961 11. When return_path_on_delivery was set as a log selector, if different remote
3962 addresses in the same message used different return paths and parallel
3963 remote delivery occurred, the wrong values would sometimes be logged.
3964 (Whenever a remote delivery process finished, the return path value from
3965 the most recently started remote delivery process was logged.)
3967 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3968 lines when AUTH and/or TLS are in use. This is the "received protocol"
3969 field. Exim used to use "asmtp" for authenticated SMTP, without any
3970 indication (in the protocol name) for TLS use. Now it follows the RFC and
3971 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3972 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3973 names appear in log lines as well as in Received: header lines.
3975 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3976 extensions, and comparison tests, and to fix some bugs.
3978 14. Changes to the "personal" filter test:
3980 (1) The test was buggy in that it was just doing the equivalent of
3981 "contains" tests on header lines. For example, if a user's address was
3982 anne@some.where, the "personal" test would incorrectly be true for
3984 To: susanne@some.where
3986 This test is now done by extracting each address from the header in turn,
3987 and checking the entire address. Other tests that are part of "personal"
3988 are now done using regular expressions (for example, to check local parts
3989 of addresses in From: header lines).
3991 (2) The list of non-personal local parts in From: addresses has been
3992 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3993 taken from the Sieve specification recommendations.
3995 (3) If the message contains any header line starting with "List-" it is
3996 treated as non-personal.
3998 (4) The test for "circular" in the Subject: header line has been removed
3999 because it now seems ill-conceived.
4001 15. Minor typos in src/EDITME comments corrected.
4003 16. Installed latest exipick from John Jetmore.
4005 17. If headers_add on a router specified a text string that was too long for
4006 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4007 of string_sprintf() is now avoided.
4009 18. $message_body_size was not set (it was always zero) when running the DATA
4010 ACL and the local_scan() function.
4012 19. For the "mail" command in an Exim filter, no default was being set for
4013 the once_repeat time, causing a random time value to be used if "once" was
4014 specified. (If the value happened to be <= 0, no repeat happened.) The
4015 default is now 0s, meaning "never repeat". The "vacation" command was OK
4016 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4017 (I found it when inspecting the code).
4019 20. There is now an overall timeout for performing a callout verification. It
4020 defaults to 4 times the callout timeout, which applies to individual SMTP
4021 commands during the callout. The overall timeout applies when there is more
4022 than one host that can be tried. The timeout is checked before trying the
4023 next host. This prevents very long delays if there are a large number of
4024 hosts and all are timing out (e.g. when the network connections are timing
4025 out). The value of the overall timeout can be changed by specifying an
4026 additional sub-option for "callout", called "maxwait". For example:
4028 verify = sender/callout=5s,maxwait=20s
4030 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4031 to the end before writing, but this should make it even safer).
4033 22. Exim was forgetting that it had advertised PIPELINING for the second and
4034 subsequent messages on an SMTP connection. It was also not resetting its
4035 memory on STARTTLS and an internal HELO.
4037 23. When Exim logs an SMTP synchronization error within a session, it now
4038 records whether PIPELINING has been advertised or not.
4040 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4041 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4042 rather than long int.
4044 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4046 26. Added the never_mail option to autoreply.
4052 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4053 crash if the getsockname() call failed; this can happen if a connection is
4054 closed very soon after it is established. The problem was simply in the
4055 order in which certain operations were done, causing Exim to try to write
4056 to the SMTP stream before it had set up the file descriptor. The bug has
4057 been fixed by making things happen in the correct order.
4063 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4064 before the connection was closed, thus losing the rejection response.
4066 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4067 some early Solaris releases, but causes trouble in current releases where
4068 socklen_t is defined.
4070 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4073 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4076 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4077 that releasing the top of it at the end releases what was used for sub-
4078 expansions (unless the block got too big). However, discard this block if
4079 the first thing is a variable or header, so that we can use its block when
4080 it is dynamic (useful for very large $message_headers, for example).
4082 6. Lookups now cache *every* query, not just the most recent. A new, separate
4083 store pool is used for this. It can be recovered when all lookup caches are
4084 flushed. Lookups now release memory at the end of their result strings.
4085 This has involved some general refactoring of the lookup sources.
4087 7. Some code has been added to the store_xxx() functions to reduce the amount
4088 of flapping under certain conditions.
4090 8. log_incoming_interface used to affect only the <= reception log lines. Now
4091 it causes the local interface and port to be added to several more SMTP log
4092 lines, for example "SMTP connection from", and rejection lines.
4094 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4096 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4098 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4099 could be overwritten at the end of the current message (or the start of a
4100 new message if it was set in a HELO ACL). The value is now preserved for
4101 the duration of the SMTP connection.
4103 12. If a transport had a headers_rewrite setting, and a matching header line
4104 contained an unqualified address, that address was qualified, even if it
4105 did not match any rewriting rules. The underlying bug was that the values
4106 of the flags that permit the existence of unqualified sender and recipient
4107 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4108 non-local messages, and by -bnq for local messages) were not being
4109 preserved with the message after it was received.
4111 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4112 "next input=" as part of the text comprising the host identity instead of
4113 the correct text. The code was using the same buffer for two different
4114 strings. However, depending on which order the printing function evaluated
4115 its arguments, the bug did not always show up. Under Linux, for example, my
4116 test suite worked just fine.
4118 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4119 doesn't work with some older versions of Perl. It has been changed to "my",
4120 which in any case is probably the better facility to use.
4122 15. A really picky compiler found some instances of statements for creating
4123 error messages that either had too many or two few arguments for the format
4126 16. The size of the buffer for calls to the DNS resolver has been increased
4127 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4128 for addresses that have a lot of PTR records. This alleviates a problem; it
4129 does not fully solve it.
4131 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4132 buffer now truncates the list and logs the incident, which is the same
4133 action as happens when Exim is looking up a host name and its aliases.
4134 Previously in this situation something unpredictable would happen;
4135 sometimes it was "internal error: store_reset failed".
4137 18. If a server dropped the connection unexpectedly when an Exim client was
4138 using GnuTLS and trying to read a response, the client delivery process
4139 crashed while trying to generate an error log message.
4141 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4142 string, for example:
4144 warn message = H1: something\nH2: something
4146 the text was added as a single header line from Exim's point of view
4147 though it ended up OK in the delivered message. However, searching for the
4148 second and subsequent header lines using $h_h2: did not work. This has been
4149 fixed. Similarly, if a system filter added multiple headers in this way,
4150 the routers could not see them.
4152 20. Expanded the error message when iplsearch is called with an invalid key to
4153 suggest using net-iplsearch in a host list.
4155 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4156 ACLs are no longer actually imposed (and a message to that effect is
4159 22. If a "gecos" field in a passwd entry contained escaped characters, in
4160 particular, if it contained a \" sequence, Exim got it wrong when building
4161 a From: or a Sender: header from that name. A second bug also caused
4162 incorrect handling when an unquoted " was present following a character
4163 that needed quoting.
4165 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4166 was not being matched caselessly.
4168 24. Arranged for all hyphens in the exim.8 source to be escaped with
4171 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4172 properly. Recipient callout cache records were still being keyed to include
4173 the sender, even when use_sender was set false. This led to far more
4174 callouts that were necessary. The sender is no longer included in the key
4175 when use_sender is false.
4177 26. Added "control = submission" modifier to ACLs.
4179 27. Added the ${base62d: operator to decode base 62 numbers.
4181 28. dnsdb lookups can now access SRV records.
4183 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4184 the configuration file.
4186 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4187 (-v) mode. This makes the output for a verbose queue run more intelligible.
4189 31. Added a use_postmaster feature to recipient callouts.
4191 32. Added the $body_zerocount variable, containing the number of binary zero
4192 bytes in the message body.
4194 33. The time of last modification of the "new" subdirectory is now used as the
4195 "mailbox time last read" when there is a quota error for a maildir
4198 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4200 35. Added +ignore_unknown as a special item in host lists.
4202 36. Code for decoding IPv6 addresses in host lists is now included, even if
4203 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4204 address was recognized as an IP address, but was then not correctly decoded
4205 into binary, causing unexpected and incorrect effects when compared with
4212 1. Very minor rewording of debugging text in manualroute to say "list of
4213 hosts" instead of "hostlist".
4215 2. If verify=header_syntax was set, and a header line with an unqualified
4216 address (no domain) and a large number of spaces between the end of the
4217 name and the colon was received, the reception process suffered a buffer
4218 overflow, and (when I tested it) crashed. This was caused by some obsolete
4219 code that should have been removed. The fix is to remove it!
4221 3. When running in the test harness, delay a bit after writing a bounce
4222 message to get a bit more predictability in the log output.
4224 4. Added a call to search_tidyup() just before forking a reception process. In
4225 theory, someone could use a lookup in the expansion of smtp_accept_max_
4226 per_host which, without the tidyup, could leave open a database connection.
4228 5. Added the variables $recipient_data and $sender_data which get set from a
4229 lookup success in an ACL "recipients" or "senders" condition, or a router
4230 "senders" option, similar to $domain_data and $local_part_data.
4232 6. Moved the writing of debug_print from before to after the "senders" test
4235 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4236 problems for message scanning, either using a data ACL, or using
4237 local_scan() because the Received: header was not generated till after they
4238 were called (in order to set the time as the time of reception completion).
4239 I have revised the way this works. The header is now generated after the
4240 body is received, but before the ACL or local_scan() are called. After they
4241 are run, the timestamp in the header is updated.
4247 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4248 before starting a queue runner without re-exec. This happened only when
4249 deliver_drop_privilege was set or when the Exim user was set to root. The
4250 effect of the bug was that timeouts during subsequent deliveries caused
4251 crashes instead of being properly handled. The handler is now left at its
4252 default (and expected) setting.
4254 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4255 message, again when deliver_drop_privilege is set or Exim is run as root.
4256 The bug described in (1) was not present in this case, but the tidying up
4257 of the other signals was missing. I have made the two cases consistent.
4259 3. The ignore_target_hosts setting on a manualroute router was being ignored
4260 for hosts that were looked up using the /MX notation.
4262 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4265 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4266 operated on the sender address. After changing the $sender_address to <>
4267 for the sender address verify, Exim was re-instated it as the original
4268 (before rewriting) address, but remembering that it had rewritten it, so it
4269 wasn't rewriting it again. This bug also had the effect of breaking the
4270 sender address verification caching when the sender address was rewritten.
4272 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4273 This has been changed so that if the ip literal address matches
4274 ignore_target_hosts, the router declines.
4276 7. Added expansion conditions match_domain, match_address, and match_local_
4277 part (NOT match_host).
4279 8. The placeholder for the Received: header didn't have a length field set.
4281 9. Added code to Exim itself and to exim_lock to test for a specific race
4282 condition that could lead to file corruption when using MBX delivery. The
4283 issue is with the lockfile that is created in /tmp. If this file is removed
4284 after a process has opened it but before that process has acquired a lock,
4285 there is the potential for a second process to recreate the file and also
4286 acquire a lock. This could lead to two Exim processes writing to the file
4287 at the same time. The added code performs the same test as UW imapd; it
4288 checks after acquiring the lock that its file descriptor still refers to
4289 the same named file.
4291 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4292 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4295 11. Added the smtp_active_hostname option. If used, this will typically be made
4296 to depend on the incoming interface address. Because $interface_address is
4297 not set up until the daemon has forked a reception process, error responses
4298 that can happen earlier (such as "too many connections") no longer contain
4301 12. If an expansion in a condition on a "warn" statement fails because a lookup
4302 defers, the "warn" statement is abandoned, and the next ACL statement is
4303 processed. Previously this caused the whole ACL to be aborted.
4305 13. Added the iplsearch lookup type.
4307 14. Added ident_timeout as a log selector.
4309 15. Added tls_certificate_verified as a log selector.
4311 16. Added a global option tls_require_ciphers (compare the smtp transport
4312 option of the same name). This controls incoming TLS connections.
4314 17. I finally figured out how to make tls_require_ciphers do a similar thing
4315 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4316 before starting the TLS session.
4318 18. Tabs are now shown as \t in -bP output.
4320 19. If the log selector return_path_on_delivery was set, Exim crashed when
4321 bouncing a message because it had too many Received: header lines.
4323 20. If two routers both had headers_remove settings, and the first one included
4324 a superfluous trailing colon, the final name in the first list and the
4325 first name in the second list were incorrectly joined into one item (with a
4326 colon in the middle).
4332 1. Added -C and -D options to the exinext utility, mainly to make it easier
4333 to include in the automated testing, but these could be helpful when
4334 multiple configurations are in use.
4336 2. The exinext utility was not formatting the output nicely when there was
4337 an alternate port involved in the retry record key, nor when there was a
4338 message id as well (for retries that were specific to a specific message
4339 and a specific host). It was also confused by IPv6 addresses, because of
4340 the additional colons they contain. I have fixed the IPv4 problem, and
4341 patched it up to do a reasonable job for IPv6.
4343 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4344 delivery, the log line now contains "pipelined" if PIPELINING was used.
4346 4. An SMTP transport process used to panic and die if the bind() call to set
4347 an explicit outgoing interface failed. This has been changed; it is now
4348 treated in the same way as a connect() failure.
4350 5. A reference to $sender_host_name in the part of a conditional expansion
4351 that was being skipped was still causing a DNS lookup. This no longer
4354 6. The def: expansion condition was not recognizing references to header lines
4355 that used bh_ and bheader_.
4357 7. Added the _cache feature to named lists.
4359 8. The code for checking quota_filecount in the appendfile transport was
4360 allowing one more file than it should have been.
4362 9. For compatibility with Sendmail, the command line option
4370 and sets the incoming protocol and host name (for trusted callers). The
4371 host name and its colon can be omitted when only the protocol is to be set.
4372 Note the Exim already has two private options, -pd and -ps, that refer to
4373 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4374 "s", but I don't think that's a major issue.
4376 10. A number of refactoring changes to the code, none of which should affect
4379 (a) The number of logging options was getting close to filling up the
4380 32-bit word that was used as a bit map. I have split them into two classes:
4381 those that are passed in the argument to log_write(), and those that are
4382 only ever tested independently outside of that function. These are now in
4383 separate 32-bit words, so there is plenty of room for expansion again.
4384 There is no change in the user interface or the logging behaviour.
4386 (b) When building, for example, log lines, the code previously used a
4387 macro that called string_cat() twice, in order to add two strings. This is
4388 not really sufficiently general. Furthermore, there was one instance where
4389 it was actually wrong because one of the argument was used twice, and in
4390 one call a function was used. (As it happened, calling the function twice
4391 did not affect the overall behaviour.) The macro has been replaced by a
4392 function that can join an arbitrary number of extra strings onto a growing
4395 (c) The code for expansion conditions now uses a table and a binary chop
4396 instead of a serial search (which was left over from when there were very
4397 few conditions). Also, it now recognizes conditions like "pam" even when
4398 the relevant support is not compiled in: a suitably worded error message is
4399 given if an attempt is made to use such a condition.
4401 11. Added ${time_interval:xxxxx}.
4403 12. A bug was causing one of the ddress fields not to be passed back correctly
4404 from remote delivery subprocesses. The field in question was not being
4405 subsequently used, so this caused to problems in practice.
4407 13. Added new log selectors queue_time and deliver_time.
4409 14. Might have fixed a bug in maildirsizefile handling that threw up
4410 "unexpected character" debug warnings, and recalculated the data
4411 unnecessarily. In any case, I expanded the warning message to give more
4414 15. Added the message "Restricted characters in address" to the statements in
4415 the default ACL that block characters like @ and % in local parts.
4417 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4418 Three changes have been made:
4420 (a) There was a serious bug; a negative response to MAIL caused the whole
4421 recipient domain to be cached as invalid, thereby blocking all messages
4422 to all local parts at the same domain, from all senders. This bug has
4423 been fixed. The domain is no longer cached after a negative response to
4424 MAIL if the sender used is not empty.
4426 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4429 (c) A new callout option, "use_sender" has been added for people who want
4430 the modified behaviour.
4436 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4439 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4440 indeed breaks things for older releases.
4442 3. Added additional logging to the case where there is a problem reading data
4443 from a filter that is running in a subprocess using a pipe, in order to
4444 try to track down a specific problem.
4446 4. Testing facility fudge: when running in the test harness and attempting
4447 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4448 getting "No route to host". Convert this to a timeout.
4450 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4453 6. Some OS don't have socklen_t but use size_t instead. This affects the
4454 fifth argument of getsockopt() amongst other things. This is now
4455 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4456 can be set for individual OS. I have set it for SunOS5, OSF1, and
4457 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4458 some earlier ones do not.
4460 7. Change 4.30/15 was not doing the test caselessly.
4462 8. The standard form for an IPv6 address literal was being rejected by address
4463 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4464 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4465 this, as well as the form without the "IPv6" on the front (but only when
4466 address literals are enabled, of course).
4468 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4470 10. Exim crashed if a message with an empty sender address specified by -f
4471 encountered a router with an errors_to setting. This could be provoked only
4472 by a command such as
4476 where an empty string was supplied; "<>" did not hit this bug.
4478 11. Installed PCRE release 4.5.
4480 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4481 remained set. It is now erased.
4483 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4484 times from message ids (which are base 36 rather than the normal 62).
4486 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4487 were being counted as actual protocol errors, and logged if the log
4488 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4489 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4490 and DATA following a set of rejected RCPTs do not count as protocol errors.
4491 In other words, Exim assumes they were pipelined, though this may not
4492 actually be the case. Of course, in all cases the client gets an
4493 appropriate error code.
4495 15. If a lookup fails in an ACL condition, a message about the failure may
4496 be available; it is used if testing the ACL cannot continue, because most
4497 such messages specify what the cause of the deferral is. However, some
4498 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4499 that caused an old message to be retained and used if a later statement
4500 caused a defer, replacing the real cause of the deferral.
4502 16. If an IP address had so many PTR records that the DNS lookup buffer
4503 was not large enough to hold them, Exim could crash while trying to process
4504 the truncated data. It now detects and logs this case.
4506 17. Further to 4.21/58, another change has been made: if (and only if) the
4507 first line of a message (the first header line) ends with CRLF, a bare LF
4508 in a subsequent header line has a space inserted after it, so as not to
4509 terminate the header.
4511 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4512 unnecessarily, used atoi() instead of strtol(), and didn't check the
4513 termination when getting file sizes from file names by regex.
4515 19. Completely re-implemented the support for maildirsize files, in the light
4516 of a number of problems with the previous contributed implementation
4517 (4.30/29). In particular:
4519 . If the quota is zero, the maildirsize file is maintained, but no quota is
4522 . If the maildir directory does not exist, it is created before any attempt
4523 to write a maildirsize file.
4525 . The quota value in the file is just a cache; if the quota is changed in
4526 the transport, the new value overrides.
4528 . A regular expression is available for excluding directories from the
4531 20. The autoreply transport checks the characters in options that define the
4532 message's headers; it allows continued headers, but it was checking with
4533 isspace() after an embedded newline instead of explicitly looking for a
4536 21. If all the "regular" hosts to which an address was routed had passed their
4537 expiry times, and had not reached their retry times, the address was
4538 bounced, even if fallback hosts were defined. Now Exim should go on to try
4541 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4542 equivalent code in the SMTP transport. Some hosts send humungous responses
4543 to HELO/EHLO, more than 1024 it seems.
4545 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4546 gives compiler warnings in some environments. I've now done it "properly",
4549 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4550 (because of problems with the built-in one) was declared to return uschar *
4551 instead of char *, causing compiler failure.
4553 25. Fixed a file descriptor leak when processing alias/forward files.
4555 26. Fixed a minor format string issue in dbfn.c.
4557 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4559 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4560 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4563 29. When a sender address is verified, it is cached, to save repeating the test
4564 when there is more than one recipient in a message. However, when the
4565 verification involves a callout, it is possible for different callout
4566 options to be set for different recipients. It is too complicated to keep
4567 track of this in the cache, so now Exim always runs a verification when a
4568 callout is required, relying on the callout cache for the optimization.
4569 The overhead is duplication of the address routing, but this should not be
4572 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4573 to be verified with callout=postmaster, and the main callout worked but the
4574 postmaster check failed, the verification correctly failed. However, if a
4575 subsequent RCPT command asked for sender verification *without* the
4576 postmaster check, incorrect caching caused this verification also to fail,
4579 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4580 it was not caching the DNS options (qualify_single, search_parents) that
4581 were used when the lookup failed. A subsequent lookup with different
4582 options therefore always gave the same answer, though there were cases
4583 where it should not have. (Example: a "domains = !$mx_any" option on a
4584 dnslookup router: the "domains" option is always processed without any
4585 widening, but the router might have qualify_single set.) Now Exim uses the
4586 cached value only when the same options are set.
4588 32. Added John Jetmore's "exipick" utility to the distribution.
4590 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4591 than a timeout (e.g. a certificate is required, and is not provided), an
4592 Exim server now closes the connection immediately. Previously it waited for
4593 the client to close - but if the client is SSL, it seems that they each
4594 wait for each other, leading to a delay before one of them times out.
4596 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4597 maintained 0.8.x compatibility because I don't think many are using it, and
4598 it is clearly obsolete.
4600 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4603 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4604 client certificate was expired. A simple patch fixes this, though I don't
4605 understand the full logic of why the verify callback is called multiple
4608 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4609 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4610 which causes problems with some clients (such as the Certicom SSL Plus
4611 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4612 disables the coutermeasure allowing Eudora to connect."
4614 38. Exim was not checking that a write() to a log file succeeded. This could
4615 lead to Bad Things if a log got too big, in particular if it hit a file
4616 size limit. Exim now panics and dies if it cannot write to a log file, just
4617 as it does if it cannot open a log file.
4619 39. Modified OS/Makefile-Linux so that it now contains
4621 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4623 The two -D definitions ensure that Exim is compiled with large file
4624 support, which makes it possible to handle log files that are bigger than
4627 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4628 instance) a domain was checked against a named list that involved a lookup,
4629 causing $domain_data to be set, then another domain was checked against the
4630 same list, then the first domain was re-checked, the value of $domain_data
4631 after the final check could be wrong. In particular, if the second check
4632 failed, it could be set empty. This bug probably also applied to
4635 41. The strip_trailing_dot option was not being applied to the address given
4636 with the -f command-line option.
4638 42. The code for reading a message's header from the spool was incrementing
4639 $received_count, but never initializing it. This meant that the value was
4640 incorrect (doubled) while delivering a message in the same process in which
4641 it was received. In the most common configuration of Exim, this never
4642 happens - a fresh exec is done - but it can happen when
4643 deliver_drop_privilege is set.
4645 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4646 it now includes up to 150 characters of the unexpected data in the log
4649 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4650 and building data strings. The size of both of these buffers was 10 000
4651 bytes - far larger than anybody would *ever* want, thought I. Needless to
4652 say, somebody hit the limit. I have increased the maximum line length to
4653 20 000 and the maximum data length of concatenated lines to 100 000. I have
4654 also fixed two bugs, because there was no checking on these buffers. Tsk,
4655 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4656 buffer is too small.
4658 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4659 lsearch lookups. Now it does.
4661 46. When parsing a route_list item in a manualroute router, a fixed-length
4662 buffer was used for the list of hosts. I made this 1024 bytes long,
4663 thinking that nobody would ever have a list of hosts that long. Wrong.
4664 Somebody had a whole pile of complicated expansion conditions, and the
4665 string was silently truncated, leading to an expansion error. It turns out
4666 that it is easier to change to an unlimited length (owing to other changes
4667 that have happened since this code was originally written) than to build
4668 structure for giving a limitation error. The length of the item that
4669 expands into the list of hosts is now unlimited.
4671 47. The lsearch lookup could not handle data where the length of text line was
4672 more than 4095 characters. Such lines were truncated, leading to shortened
4673 data being returned. It should now handle lines of any length.
4675 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4676 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4679 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4680 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4681 no longer happens in comment lines. A list of replacements is now placed
4682 at the head of all of the source files, except those whose only change is
4683 to replace PERL_COMMAND in the very first #! line.
4685 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4686 messages on the queue, with a bottom-up merge sort, using code contributed
4687 by Michael Haardt. This should make operations like -bp somewhat faster on
4688 large queues. It won't affect queue runners, except when queue_run_in_order
4691 51. Installed eximstats 1.31 in the distribution.
4693 52. Added support for SRV lookups to the dnslookup router.
4695 53. If an ACL referred to $message_body or $message_body_end, the value was not
4696 reset for any messages that followed in the same SMTP session.
4698 54. The store-handling optimization for building very long strings was not
4699 differentiating between the different store pools. I don't think this
4700 actually made any difference in practice, but I've tidied it.
4702 55. While running the routers to verify a sender address, $sender_address
4703 was still set to the sender address. This is wrong, because when routing to
4704 send a bounce to the sender, it would be empty. Therefore, I have changed
4705 it so that, while verifying a sender address, $sender_address is set to <>.
4706 (There is no change to what happens when verifying a recipient address.)
4708 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4709 target A or AAAA records (if not already returned) without resetting the
4710 qualify_single or search_parents options of the DNS resolver. These are
4711 inappropriate in this case because the targets of MX and SRV records must
4712 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4713 target that, when qualified, matched something in the local domain. These
4714 two options are now turned off when doing these lookups.
4716 57. It seems that at least some releases of Reiserfs (which does not have the
4717 concept of a fixed number of inodes) returns zero and not -1 for the
4718 number of available inodes. This interacted badly with check_spool_inodes,
4719 which assumed that -1 was the "no such thing" setting. What I have done is
4720 to check that the total number of inodes is greater than zero before doing
4721 the test of how many are available.
4723 58. When a "warn" ACL statement has a log_message modifier, the message is
4724 remembered, and not repeated. This is to avoid a lot of repetition when a
4725 message has many recipients that cause the same warning to be written.
4726 Howewer, Exim was preserving the list of already written lines for an
4727 entire SMTP session, which doesn't seem right. The memory is now reset if a
4728 new message is started.
4730 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4731 debugging output unless log_rewrite was also set.
4733 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4734 of (int)(handle) when we know that handle contains (void *)(-1).
4736 61. The Exim daemon panic-logs an error return when it closes the incoming
4737 connection. However "connection reset by peer" seems to be common, and
4738 isn't really an error worthy of noting specially, so that particular error
4741 62. When Exim is trying to find all the local interfaces, it used to panic and
4742 die if the ioctl to get the interface flags failed. However, it seems that
4743 on at least one OS (Solaris 9) it is possible to have an interface that is
4744 included in the list of interfaces, but for which you get a failure error
4745 for this call. This happens when the interface is not "plumbed" into a
4746 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4747 failure of the "get flags" call assumes that the interface is down.
4749 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4750 makes life easier for people who are doing arithmetic on fields extracted
4751 from dates, where you often get leading zeros that should not be
4752 interpreted as octal.
4754 64. Added qualify_domain to the redirect router, to override the global
4757 65. If a pathologically long header line contained very many addresses (the
4758 report of this problem mentioned 10 000) and each of them was rewritten,
4759 Exim could use up a very large amount of memory. (It kept on making new
4760 copies of the header line as it rewrote, and never released the old ones.)
4761 At the expense of a bit more processing, the header rewriting function has
4762 been changed so that it no longer eats memory in this way.
4764 66. The generation of the Received: header has been moved from the time that a
4765 message starts to be received, to the time that it finishes. The timestamp
4766 in the Received: header should now be very close to that of the <= log
4767 line. There are two side-effects of this change:
4769 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4770 logged header lines no longer include the local Received: line, because
4771 it has not yet been created. The same applies to a copy of the message
4772 that is returned to a non-SMTP sender when a message is rejected.
4774 (b) When a filter file is tested using -bf, no additional Received: header
4775 is added to the test message. After some thought, I decided that this
4778 This change does not affect the value of $received_for. It is still set
4779 after address rewriting, but before local_scan() is called.
4781 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4783 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4784 gave an unhelpful panic error message, and a defer error. I have managed to
4785 change this behaviour so that it now rejects any supplied certificate,
4786 which seems right, as the list of acceptable certificates is empty.
4788 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4789 gave an unhelpful defer error. I have not managed to make this reject any
4790 supplied certificates, but the error message it gives is "no certificate
4791 supplied", which is not helpful.
4793 70. exigrep's output now also includes lines that are not associated with any
4794 message, but which match the given pattern. Implemented by a patch from
4795 Martin Sluka, which also tidied up the Perl a bit.
4797 71. Recipient callout verification, like sender verification, was using <> in
4798 the MAIL FROM command. This isn't really the right thing, since the actual
4799 sender may affect whether the remote host accepts the recipient or not. I
4800 have changed it to use the actual sender in the callout; this means that
4801 the cache record is now keyed on a recipient/sender pair, not just the
4802 recipient address. There doesn't seem to be a real danger of callout loops,
4803 since a callout by the remote host to check the sender would use <>.
4804 [SEE ABOVE: changed after hitting problems.]
4806 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4807 temporary errors. However, in the case of such a code being given after
4808 the end of a data transmission (i.e. after ".") Exim was failing to write
4809 a retry record for the message. (Yes, there was some broken host that was
4810 actually sending 8xx at this point.)
4812 73. An unknown lookup type in a host list could cause Exim to panic-die when
4813 the list was checked. (An example that provoked this was putting <; in the
4814 middle of a list instead of at the start.) If this happened during a DATA
4815 ACL check, a -D file could be left lying around. This kind of configuration
4816 error no longer causes Exim to die; instead it causes a defer errror. The
4817 incident is still logged to the main and panic logs.
4819 74. Buglet left over from Exim 3 conversion. The message "too many messages
4820 in one connection" was written to the rejectlog but not the mainlog, except
4821 when address rewriting (yes!) was being logged.
4823 75. Added write_rejectlog option.
4825 76. When a system filter was run not as root (that is, when system_filter_user
4826 was set), the values of the $n variables were not being returned to the
4827 main process; thus, they were not subsequently available in the $sn
4830 77. Added +return_path_on_delivery log selector.
4832 78. A connection timeout was being treated differently from recipients deferred
4833 when testing hosts_max_try with a message that was older than the host's
4834 retry timeout. (The host should not be counted, thus allowing all hosts to
4835 be tried at least once before bouncing.) This may have been the cause of an
4836 occasionally reported bug whereby a message would remain on the queue
4837 longer than the retry timeout, but would be bounced if a delivery was
4838 forced. I say "may" because I never totally pinned down the problem;
4839 setting up timeout/retry tests is difficult. See also the next item.
4841 79. The ultimate address timeout was not being applied to errors that involved
4842 a combination of host plus message (for example, a timeout on a MAIL
4843 command). When an address resolved to a number of possible hosts, and they
4844 were not all tried for each delivery (e.g. because of hosts_max_try), a
4845 message could remain on the queue longer than the retry timeout.
4847 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4850 81. Fixed an obscure SMTP outgoing bug which required at least the following
4851 conditions: (a) there was another message waiting for the same server;
4852 (b) the server returned 5xx to all RCPT commands in the first message so
4853 that the message was not completed; (c) the server dropped the connection
4854 or gave a negative response to the RSET that Exim sends to abort the
4855 transaction. The observed case was a dropped connection after DATA that had
4856 been sent in pipelining mode. That is, the server had advertised PIPELINING
4857 but was not implementing it correctly. The effect of the bug was incorrect
4858 behaviour, such as trying another host, and this could lead to a crash.
4864 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4865 and daemon.c were passed as pointers to ints; they should have been
4866 pointers to socklen_t variables (which are typically unsigned ints).
4868 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4871 3. Fixed a really odd bug that affected only the testing scheme; patching a
4872 certain fixed string in the binary changed the value of another string that
4873 happened to be identical to the end of the original first string.
4875 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4876 name", it returns that address as the IP address. On some operating
4877 systems (e.g. Solaris), it also passes back the IP address string as the
4878 "host name". However, on others (e.g. Linux), it passes back an empty
4879 string. Exim wasn't checking for this, and was changing the host name to an
4880 empty string, assuming it had been canonicized.
4882 5. Although rare, it is permitted to have more than one PTR record for a given
4883 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4884 all the names associated with an address, because they do in Solaris.
4885 However, it seems that they do not in Linux for data that comes from the
4886 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4887 I found this out when I moved to a new Linux workstation and tried to run
4888 the Exim test suite.
4890 To get round this problem I have changed the code so that it now does its
4891 own call to the DNS to look up PTR records when searching for a host name.
4892 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4893 addresses that are only in /etc/hosts are still found.
4895 This behaviour is, however, controlled by an option called host_lookup_
4896 order, which defaults to "bydns:byaddr". If people want to use the other
4897 order, or indeed, just use one or the other means of lookup, they can
4898 specify it in this variable.
4900 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4901 some operating systems, this comes back from gethostbyaddr() as an empty
4902 string, and this is what Exim used to test for. However, it seems that in
4903 other systems, "." is yielded. Exim now tests for this case too.
4905 7. The values of check_spool_space and check_log_space are now held internally
4906 as a number of kilobytes instead of an absolute number of bytes. If a
4907 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4908 kilobyte. This means that much larger values can be stored.
4910 8. Exim monitor: an attempt to get the action menu when not actually pointing
4911 at a message produces an empty menu entitled "No message selected". This
4912 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4913 no entries in it ("Shell widget menu has zero width and/or height"). So I
4914 have added a single, blank menu entry in this case.
4916 9. Added ${quote_local_part.
4918 10. MIME decoding is now applied to the contents of Subject: header lines when
4921 11. Now that a reference to $sender_host_address automatically causes a reverse
4922 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4923 host lookup before query-style lookups in lists that might use this
4924 variable. This has therefore been abolished, and the "net-" prefix is no
4925 longer necessary for query-style lookups.
4927 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4928 have been a typo for LFLAGS, so it has been changed.
4930 13. The install script calls Exim with "-C /dev/null" in order to find the
4931 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4932 to be output. Howeve, since Exim outputs its version number before the
4933 error, it didn't break the script. It just looked ugly. I fixed this by
4934 always allowing "-C /dev/null" if the caller is root.
4936 14. Ignore overlarge ACL variable number when reading spool file - insurance
4937 against a later release with more variables having written the file.
4939 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4940 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4941 this, as well as the form without the "IPv6" on the front.
4943 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4944 OS/Makefile-Darwin file.
4946 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4947 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4949 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4950 variable. However, some OS use ints for inodes. I've added cast to long int
4951 to get rid of the compiler warning.
4953 19. I had forgotten to lock out "/../" in configuration file names when
4954 ALT_CONFIG_PREFIX was set.
4956 20. Routers used for verification do not need to specify transports. However,
4957 if such a router generated a host list, and callout was configured, Exim
4958 crashed, because it could not find a port number from the (non-existent)
4959 transport. It now assumes port 25 in this circumstance.
4961 21. Added the -t option to exigrep.
4963 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4964 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4965 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4966 them is set, LOOKUP_LSEARCH is forced.
4968 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4970 24. Added sender and host information to the "rejected by local_scan()" log
4971 line; previously there was no indication of these.
4973 25. Added .include_if_exists.
4975 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4976 Linux. It turns out that not all file systems support this. Apparently some
4977 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4978 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4979 not supported on the file descriptor, is now ignored when Exim is trying to
4980 sync a directory. This applies only to Linux.
4982 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4984 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4985 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4988 29. Added support for maildirsize files from supplied patch (modified a bit).
4990 30. The use of :fail: followed by an empty string could lead Exim to respond to
4991 sender verification failures with (e.g.):
4993 550 Verification failed for <xxx>
4994 550 Sender verify failed
4996 where the first response line was missing the '-' that indicates it is not
4997 the final line of the response.
4999 31. The loop for finding the name of the user that called Exim had a hardwired
5000 limit of 10; it now uses the value of finduser_retries, which is used for
5001 all other user lookups.
5003 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5006 33. Exim was neglecting to zero errno before one call of strtol() when
5007 expanding a string and expecting an integer value. On some systems this
5008 resulted in spurious "integer overflow" errors. Also, it was casting the
5009 result into an int without checking.
5011 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5012 did not work. The code looks as if it has *never* worked, though it appears
5013 to have been documented since at least releast 1.62. I have made it work.
5015 35. The "timeout_DNS" error in retry rules, also documented since at least
5016 1.62, also never worked. As it isn't clear exactly what this means, and
5017 clearly it isn't a major issue, I have abolished the feature by treating it
5018 as "timeout", and writing a warning to the main and panic logs.
5020 36. The display of retry rules for -brt wasn't always showing the error code
5023 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5024 timeout_connect_A, timeout_connect_MX.
5026 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5027 to the empty sender.
5029 39. The daemon was not analysing the content of -oX till after it had closed
5030 stderr and disconnected from the controlling terminal. This meant that any
5031 syntax errors were only noted on the panic log, and the return code from
5032 the command was 0. By re-arranging the code a little, I've made the
5033 decoding happen first, so such errors now appear on stderr, and the return
5034 code is 1. However, the actual setting up of the sockets still happens in
5035 the disconnected process, so errors there are still only recorded on the
5038 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5039 connections (as happens on some IP stacks) was logged at start up time as
5040 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5041 it from "IPv6 and IPv4", which means that two separate sockets are being
5044 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5045 says whether AF_INET or AF_INET6 was passed as an argument.
5047 42. Exiwhat output was messed up when time zones were included in log
5050 43. Exiwhat now gives more information about the daemon's listening ports,
5051 and whether -tls-on-connect was used.
5053 44. The "port" option of the smtp transport is now expanded.
5055 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5056 silently ignored. Now an error message is written to the main and panic
5059 46. There's a new ACL modifier called "logwrite" which writes to a log file
5060 as soon as it is encountered.
5062 47. Added $local_user_uid and $local_user_gid at routing time.
5064 48. Exim crashed when trying to verify a sender address that was being
5067 49. Exim was recognizing only a space character after ".include". It now also
5068 recognizes a tab character.
5070 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5071 extracting the relevant information from the specification. The man page no
5072 longer contains scrambled data for the -d option, and I've added a section
5073 at the front about calling Exim under different names.
5075 51. Added "extra_headers" argument to the "mail" command in filter files.
5077 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5080 53. Installed eximstats 1.29.
5082 54. Added transport_filter_timeout as a generic transport option.
5084 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5085 Cc: header lines. This was required by RFC 822, but it not required by RFC
5088 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5089 incoming messages that did not have them. Now it does so only if the
5090 message originates locally, that is, if there is no associated remote host
5091 address. When Resent- header lines are present, this applies to the Resent-
5092 lines rather than the non-Resent- lines.
5094 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5095 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5097 58. Messages for configuration errors now include the name of the main
5098 configuration file - useful now that there may be more than one file in a
5099 list (.included file names were always shown).
5101 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5102 for those rare installations that do not start the daemon as root or run it
5103 setuid root. I've cut out the call to initgroups() if the daemon is not
5106 60. The Exim user and group can now be bound into the binary as text strings
5107 that are looked up at the start of Exim's processing.
5109 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5111 62. Added $mailstore_basename variable.
5113 63. Installed patch to sieve.c from Michael Haardt.
5115 64. When Exim failed to open the panic log after failing to open the main log,
5116 the original message it was trying to log was written to stderr and debug
5117 output, but if they were not available (the usual case in production), it
5118 was lost. Now it is written to syslog before the two lines that record the
5119 failures to open the logs.
5121 65. Users' Exim filters run in subprocesses under the user's uid. It is
5122 possible for a "deliver" command or an alias in a "personal" command to
5123 provoke an address rewrite. If logging of address rewriting is configured,
5124 this fails because the process is not running as root or exim. There may be
5125 a better way of dealing with this, but for the moment (because 4.30 needs
5126 to be released), I have disabled address rewrite logging when running a
5127 filter in a non-root, non-exim process.
5133 1. The buildconfig auxiliary program wasn't quoting the value set for
5134 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5135 not defined. This bug was masked in 4.22 by the effect that was fixed in
5138 2. Some messages that were rejected after a message id was allocated were
5139 shown as "incomplete" by exigrep. It no longer does this for messages that
5140 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5142 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5143 have allow_domain_literals set, the ID did not get logged in the <= line.
5144 Domain literals are now always recognized in Message-ID: header lines.
5146 4. The first argument for a ${extract expansion item is the key name or field
5147 number. Leading and trailing spaces in this item were not being ignored,
5148 causing some misleading effects.
5150 5. When deliver_drop_privilege was set, single queue runner processes started
5151 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5152 same command in the process it spins off) were not dropping privilege.
5154 6. When the daemon running as "exim" started a queue runner, it always
5155 re-executed Exim in the spun-off process. This is a waste of effort when
5156 deliver_drop_privilege is set. The new process now just calls the
5157 queue-runner function directly.
5163 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5166 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5167 ignored. Though the use of -oP was forcing the writing of a pid file, it
5168 was always written to the default place.
5170 3. If the message "no IP address found for host xxxx" is generated during
5171 incoming verification, it is now followed by identification of the incoming
5172 connection (so you can more easily find what provoked it).
5174 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5176 5. Added some features to "harden" Exim a bit more against certain attacks:
5178 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5179 be put in Local/Makefile. This is like the never_users runtime option,
5180 but it cannot be overridden. The default setting is "root".
5182 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5183 prefix string with which any file named in a -C command line option
5186 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5187 is retained for -C and -D only if the caller of Exim is root. Without
5188 it, the exim user may also use -C and -D and retain privilege.
5190 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5191 command line option is disabled.
5193 6. Macro names set by the -D option must start with an upper case letter, just
5194 like macro names defined in the configuration file.
5196 7. Added "dereference=" facility to LDAP.
5198 8. Two instances of the typo "uknown" in the source files are fixed.
5200 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5201 the Configure-Makefile script screwed up while processing it.
5203 10. Incorporated PCRE 4.4.
5205 11. The SMTP synchronization check was not operating right at the start of an
5206 SMTP session. For example, it could not catch a HELO sent before the client
5207 waited for the greeting. There is now a check for outstanding input at the
5208 point when the greeting is written. Because of the duplex, asynchronous
5209 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5210 way, but not yet received, when the check is performed.
5212 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5213 on TCP/IP sockets, because this apparently causes some broken clients to
5216 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5217 unchanged) from the Cygwin maintainer.
5219 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5220 was supported instead of testing for mbx. Effectively a typo.
5222 15. The spa authenticator server code was not checking that the input it
5223 received was valid base64.
5225 16. The debug output line for the "set" modifier in ACLs was not showing the
5226 name of the variable that was being set.
5228 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5230 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5231 lookup on its own. Something else was needed to trigger the lookup. For
5232 example, a match in host_lookup or the need for a host name in a host list.
5233 Now, if $sender_host_name is referenced and the host name has not yet been
5234 looked up, a lookup is performed. If the lookup fails, the variable remains
5235 empty, and $host_lookup_failed is set to "1".
5237 19. Added "eqi" as a case-independent comparison operator.
5239 20. The saslauthd authentication condition could segfault if neither service
5240 nor realm was specified.
5242 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5243 error message that was logged was misleading, and incoming SMTP
5244 connections were dropped. The message is now more accurate, and temporary
5245 errors are given to SMTP connections.
5247 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5248 (except RSET) with a 421 error, until QUIT is received. However, it was
5249 failing to send a response to QUIT.
5251 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5252 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5253 ACL is now run after the helo_try_verify_hosts code.
5255 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5256 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5257 case-independent, so other case variants are also recognized). Apparently
5258 some systems use these upper case variants.
5260 25. If more than two messages were waiting for the same host, and a transport
5261 filter was specified for the transport, Exim sent two messages over the
5262 same TCP/IP connection, and then failed with "socket operation on non-
5263 socket" when it tried to send the third.
5265 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5267 27. The extern definition of crypt16() in expand.c was not being excluded when
5268 the OS had its own crypt16() function.
5270 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5271 as a preferred synonym for return_size_limit, both as an option and as an
5274 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5276 30. Changed the default configuration ACL to relax the local part checking rule
5277 for addresses that are not in any local domains. For these addresses,
5278 slashes and pipe symbols are allowed within local parts, but the sequence
5279 /../ is explicitly forbidden.
5281 31. SPA server authentication was not clearing the challenge buffer before
5284 32. log_message in a "warn" ACL statement was writing to the reject log as
5285 well as to the main log, which contradicts the documentation and doesn't
5286 seem right (because no rejection is happening). So I have stopped it.
5288 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5289 However, I am unable to do any testing of this.
5291 34. Fixed an infelicity in the appendfile transport. When checking directories
5292 for a mailbox, to see if any needed to be created, it was accidentally
5293 using path names with one or more superfluous leading slashes; tracing
5294 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5296 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5297 contents are added to the log line that is written for every discarded
5298 recipient. (Previously a log_message setting was ignored.)
5300 36. The ${quote: operator now quotes the string if it is empty.
5302 37. The install script runs exim in order to find its version number. If for
5303 some reason other than non-existence or emptiness, which it checks, it
5304 could not run './exim', it was installing it with an empty version number,
5305 i.e. as "exim-". This error state is now caught, and the installation is
5308 38. An argument was missing from the function that creates an error message
5309 when Exim fails to connect to the socket for saslauthd authentication.
5310 This could cause Exim to crash, or give a corrupted message.
5312 39. Added isip, isip4, and isip6 to ${if conditions.
5314 40. The ACL variables $acl_xx are now saved with the message, and can be
5315 accessed later in routers, transports, and filters.
5317 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5318 strings in the file are not string-expanded.
5320 42. If a MAIL command specified a SIZE value that was too large to fit into an
5321 int variable, the check against message_size_limit failed. Such values are
5322 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5323 day this will have to be increased, but I don't think I want to be around
5324 when emails are that large.
5331 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5332 iconv() is not standard in FreeBSD.
5334 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5335 IPv6 enabled. The observed symptom was a segmentation fault on return from
5336 the function os_common_find_running_interfaces() in src/os.c.
5338 3. In the check_special_case() function in daemon.c I had used "errno" as an
5339 argument name, which causes warnings on some systems. This was basically a
5340 typo, since it was named "eno" in the comments!
5342 4. The code that waits for the clock to tick (at a resolution of some fraction
5343 of a second) so as to ensure message-id uniqueness was always waiting for
5344 at least one whole tick, when it could have waited for less. [This is
5345 almost certainly not relevant at current processor speeds, where it is
5346 unlikely to ever wait at all. But we try to future-proof.]
5348 5. The function that sleeps for a time interval that includes fractions of a
5349 second contained a race. It did not block SIGALRM between setting the
5350 timer, and suspending (a couple of lines later). If the interval was short
5351 and the sigsuspend() was delayed until after it had expired, the suspension
5352 never ended. On busy systems this could lead to processes getting stuck for
5355 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5356 process, before it forks any delivery processes. The open lookup caching
5357 mechanism meant that the open file or database connection was passed into
5358 the delivery process. The problem was that delivery processes always tidy
5359 up cached lookup data. This could cause a problem for the next delivery
5360 process started by the queue runner, because the external queue runner
5361 process does not know about the closure. So the next delivery process
5362 still has data in the lookup cache. In the case of a file lookup, there was
5363 no problem because closing a file descriptor in a subprocess doesn't affect
5364 the parent. However, if the lookup was caching a connection to a database,
5365 the connection was closed, and the second delivery process was likely to
5366 see errors such as "PGSQL: query failed: server closed the connection
5367 unexpectedly". The problem has been fixed by closing all cached lookups
5368 in a queue runner before running a delivery process.
5370 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5371 seem to have the "const" qualifier which it has on other OS. I've
5374 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5375 *of the same type* (client or server) with the same public name that an
5376 error should be diagnosed.
5378 9. When Exim looked up a host name for an IP address, but failed to find the
5379 original IP address when looking up the host name (a safety check), it
5380 output the message "<ip address> does not match any IP for NULL", which was
5381 confusing, to say the least. The bug was that the host name should have
5382 appeared instead of "NULL".
5384 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5385 user that is built into the binary, and the -C or -D options is used, root
5386 privilege is dropped before the configuration file is read. In addition,
5387 logging is switched to stderr instead of the normal log files. If the
5388 configuration then re-defines the Exim user, the unprivileged environment
5389 is probably not what is expected, so Exim logs a panic warning message (but
5392 However, if deliver_drop_privilege is set, the unprivileged state may well
5393 be exactly what is intended, so the warning has been cut out in that case,
5394 and Exim is allowed to try to write to its normal log files.
5400 1. smtp_return_error_details was not giving details for temporary sender
5401 or receiver verification errors.
5403 2. Diagnose a configuration error if two authenticators have the same public
5406 3. Exim used not to create the message log file for a message until the first
5407 delivery attempt. This could be confusing when incoming messages were held
5408 for policy or load reasons. The message log file is now created at the time
5409 the message is received, and an initial "Received" line is written to it.
5411 4. The automatically generated man page for command line options had a minor
5412 bug that caused no ill effects; however, a more serious problem was that
5413 the procedure for building the man page automatically didn't always
5414 operate. Consequently, release 4.20 contains an out-of-date version. This
5415 shouldn't happen again.
5417 5. When building Exim with embedded Perl support, the script that builds the
5418 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5419 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5421 6. The freeze_tell option was not being used for messages that were frozen on
5422 arrival, either by an ACL or by local_scan().
5424 7. Added the smtp_incomplete_transaction log selector.
5426 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5427 was accepting AUTH without a new EHLO.
5429 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5430 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5433 10. Logging of TCP/IP connections (when configured) now happens in the main
5434 daemon process instead of the child process, so that the TCP/IP connection
5435 count is more accurate (but it can never be perfect).
5437 11. The use of "drop" in a nested ACL was not being handled correctly in the
5438 outer ACL. Now, if condition failure induced by the nested "drop" causes
5439 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5440 or "require"), the connection is dropped.
5442 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5443 that yield "discard" can now be used with an "accept" or a "discard" verb,
5444 but an error is generated for any others (because I can't see a useful way
5445 to define what should happen).
5447 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5448 are now processed in the same way as lines in the Exim configuration file.
5449 In particular, continuation lines are supported.
5451 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5453 15. Added -ti meaning -t -i.
5455 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5456 domains, and warn by logging if others are found.
5458 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5459 returned from GIFCONF when Exim is trying to find the list of interfaces on
5460 a host. The code in os.c has been modified to copy each ifreq to an aligned
5461 structure in all cases.
5463 Also, in some cases, the returned ifreq's were being copied to a 'struct
5464 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5465 means the last couple of bytes of an IPv6 address could be chopped if the
5466 ifreq contained only a normal sockaddr (14 bytes storage).
5468 18. Named domain lists were not supported in the hosts_treat_as_local option.
5469 An entry such as +xxxx was not recognized, and was treated as a literal
5472 19. Ensure that header lines added by a DATA ACL are included in the reject log
5473 if the ACL subsequently rejects the message.
5475 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5476 MD5 (which is deprecated).
5478 21. When testing a filter file using -bf, Exim was writing a message when it
5479 took the sender from a "From " line in the message, but it was not doing so
5480 when it took $return_path from a Return-Path: header line. It now does.
5482 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5483 with a valid header line field name (a series of printing characters
5484 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5486 23. Changed "disc" in the source to "disk" to conform to the documentation and
5487 the book and for uniformity.
5489 24. Ignore Sendmail's -Ooption=value command line item.
5491 25. When execve() failed while trying to run a command in a pipe transport,
5492 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5493 could be confused with a return value of 69 from the command itself. This
5494 has been changed to 127, the value the shell returns if it is asked to run
5495 a non-existent command. The wording for the related log line suggests a
5496 non-existent command as the problem.
5498 26. If received_header_text expands to an empty string, do not add a Received:
5499 header line to the message. (Well, it adds a token one on the spool, but
5500 marks it "old" so that it doesn't get used or transmitted.)
5502 27. Installed eximstats 1.28 (addition of -nt option).
5504 28. There was no check for failure on the call to getsockname() in the daemon
5505 code. This can fail if there is a shortage of resources on the system, with
5506 ENOMEM, for example. A temporary error is now given on failure.
5508 29. Contrary to the C standard, it seems that in some environments, the
5509 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5510 program. Exim now does this explicitly; it affects the formatting of
5511 timestamps using strftime().
5513 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5514 complaints. I've now initialized all the variables, to avoid this.
5516 32. Header lines added by a system filter were not being "seen" during
5517 transport-time rewrites.
5519 33. The info_callback() function passed to OpenSSL is set up with type void
5520 (*)(SSL *, int, int), as described somewhere. However, when calling the
5521 function (actually a macro) that sets it up, the type void(*)() is
5522 expected. I've put in a cast to prevent warnings from picky compilers.
5524 34. If a DNS black list lookup found a CNAME record, but there were no A
5525 records associated with the domain it pointed at, Exim crashed.
5527 35. If a DNS black list lookup returned more than one A record, Exim ignored
5528 all but the first. It now scans all returned addresses if a particular IP
5529 value is being sought. In this situation, the contents of the
5530 $dnslist_value variable are a list of all the addresses, separated by a
5533 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5534 to accept a host name and all its aliases if the forward lookup for any of
5535 them yielded the IP address of the incoming connection. Now it accepts only
5536 those names whose forward lookup yields the correct IP address. Any other
5537 names are discarded. This closes a loophole whereby a rogue DNS
5538 administrator could create reverse DNS records to break through a
5539 wildcarded host restriction in an ACL.
5541 37. If a user filter or a system filter that ran in a subprocess used any of
5542 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5543 the wrong values were passed to the pipe command ($thisaddress had the
5544 value of $0, $0 had the value of $1, etc). This bug was introduced by
5545 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5547 38. Improved the line breaking for long SMTP error messages from ACLs.
5548 Previously, if there was no break point between 40 and 75 characters, Exim
5549 left the rest of the message alone. Two changes have been made: (a) I've
5550 reduced the minimum length to 35 characters; (b) if it can't find a break
5551 point between 35 and 75 characters, it looks ahead and uses the first one
5552 that it finds. This may give the occasional overlong line, but at least the
5553 remaining text gets split now.
5555 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5556 file descriptors might be low, and that setting 1000 would always raise it.
5557 It turns out that in some environments, the limit is already over 1000 and
5558 that lowering it causes trouble. So now Exim takes care not to decrease it.
5560 40. When delivering a message, the value of $return_path is set to $sender_
5561 address at the start of routing (routers may change the value). By an
5562 oversight, this default was not being set up when an address was tested by
5563 -bt or -bv, which affected the outcome if any router or filter referred to
5566 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5567 log only when "message" or "log_message" are set. However, if one of the
5568 conditions was an address verification, or a call to a nested ACL, the
5569 messages generated by the underlying test were being passed through. This
5570 no longer happens. The underlying message is available in $acl_verify_
5571 message for both "message" and "log_message" expansions, so it can be
5572 passed through if needed.
5574 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5575 new expansion $bh_ to give the encoded byte string without charset
5576 translation. Translation happens only if iconv() is available; HAVE_ICONV
5577 indicates this at build time. HEADERS_CHARSET gives the charset to
5578 translate to; headers_charset can change it in the configuration, and
5579 "headers charset" can change it in an individual filter file.
5581 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5582 that creates RFC 2047 encoded "words" labels them as that charset instead
5583 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5584 expansion operator; (ii) when Exim creates a From: line for a local
5585 message; (iii) when a header line is rewritten to include a "phrase" part.
5587 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5588 buggy, causing it to skip the first lines of messages whose message ID
5589 ended in 'D'. This would not have bitten before Exim release 4.14, because
5590 message IDs were unlikely to end in 'D' before then. The effect was to have
5591 incorrect size information for certain domains.
5593 45. #include "config.h" was missing at the start of the crypt16.c module. This
5594 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5597 46. If there was a timeout during a "random" callout check, Exim treated it as
5598 a failure of the random address, and carried on sending RSET and the real
5599 address. If the delay was just some slowness somewhere, the response to the
5600 original RCPT would be taken as a response to RSET and so on, causing
5601 mayhem of various kinds.
5603 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5604 when I implemented it. It didn't allow for the fact that some option values
5605 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5606 the right test for positive values.
5608 48. Domain names in DNS records are case-independent. Exim always looks them up
5609 in lower case. Some resolvers return domain names in exactly the case they
5610 appear in the zone file, that is, they may contain uppercase letters. Not
5611 all resolvers do this - some return always lower case. Exim was treating a
5612 change of case by a resolver as a change of domain, similar to a widening
5613 of a domain abbreviation. This triggered its re-routing code and so it was
5614 trying to route what was effectively the same domain again. This normally
5615 caused routing to fail (because the router wouldn't handle the domain
5616 twice). Now Exim checks for this case specially, and just changes the
5617 casing of the domain that it ultimately uses when it transmits the message
5620 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5623 50. If a filter generated a file delivery with a non-absolute name (possible if
5624 no home directory exists for the router), the forbid_file option was not
5627 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5628 the existing equality matching.
5630 52. Exim was using ints instead of ino_t variables in some places where it was
5631 dealing with inode numbers.
5633 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5634 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5635 called TMPDIR, and if it finds it is different, it changes its value.
5637 54. The smtp_printf() function is now made available to local_scan() so
5638 additional output lines can be written before returning. There is also an
5639 smtp_fflush() function to enable the detection of a dropped connection.
5640 The variables smtp_input and smtp_batched_input are exported to
5643 55. Changed the default runtime configuration: the message "Unknown user"
5644 has been removed from the ACL, and instead placed on the localuser router,
5645 using the cannot_route_message feature. This means that any verification
5646 failures that generate their own messages won't get overridden. Similarly,
5647 the "Unrouteable address" message that was in the ACL for unverifiable
5648 relay addresses has also been removed.
5650 56. Added hosts_avoid_esmtp to the smtp transport.
5652 57. The exicyclog script was not checking for the esoteric option
5653 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5654 will work only if exicyclog is run under the appropriate euid.
5656 58. Following a discussion on the list, the rules by which Exim recognises line
5657 endings on incoming messages have been changed. The -dropcr and drop_cr
5658 options are now no-ops, retained only for backwards compatibility. The
5659 following line terminators are recognized: LF CRLF CR. However, special
5660 processing applies to CR:
5662 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5663 nor a local message in the state where . is a terminator.
5665 (ii) If a bare CR is encountered in a header line, an extra space is added
5666 after the line terminator so as not to end the header. The reasoning
5667 behind this is that bare CRs in header lines are most likely either
5668 to be mistakes, or people trying to play silly games.
5670 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5671 was being incorrectly given as 18 bytes larger than it should have been.
5672 This is a VOB (very old bug).
5674 60. This may never have affected anything current, but just in case it has:
5675 When the local host is found other than at the start of a list of hosts,
5676 the local host, those with the same MX, and any that follow, are discarded.
5677 When the list in question was part of a longer list of hosts, the following
5678 hosts (not currently being processed) were also being discarded. This no
5679 longer happens. I'm not sure if this situation could ever has previously
5682 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5685 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5686 header. This is something that is recommended in a new Internet Draft, and
5687 is something that is documented as being done by Sendmail. There are two
5688 possible values. For messages generated by the autoreply transport, Exim
5691 Auto-Submitted: auto-replied
5693 whereas for all other generated messages (e.g. bounces) it adds
5695 Auto-Submitted: auto-generated
5697 63. The "personal" condition in filters now includes a test for the
5698 Auto-Submitted: header. If it contains the string "auto-" the message it
5699 not considered personal.
5701 64. Added rcpt_include_affixes as a generic transport option.
5703 65. Added queue_only_override (default true).
5705 66. Added the syslog_duplication option.
5707 67. If what should have been the first header line of a message consisted of
5708 a space followed by a colon, Exim was mis-interpreting it as a header line.
5709 It isn't of course - it is syntactically invalid and should therefore be
5710 treated as the start of the message body. The misbehaviour could have
5711 caused a number of strange effects, including loss of data in subsequent
5712 header lines, and spool format errors.
5714 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5715 client host had authenticated. This control can now be exercised by an ACL
5716 for more flexibility.
5718 69. By default, callouts do not happen when testing with -bh. There is now a
5719 variant, -bhc, which does actually run the callout code, including
5720 consulting and updating the callout cache.
5722 70. Added support for saslauthd authentication, courtesy of Alexander
5725 71. If statvfs() failed on the spool or log directories while checking their
5726 size for availability, Exim confusingly gave the error "space shortage".
5727 Furthermore, in debugging mode it crashed with a floating point exception.
5728 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5729 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5730 serious problem, Exim now writes to the main and panic logs when this
5731 happens, with details of the failure. It then refuses to accept the
5732 incoming message, giving the message "spool directory problem" or "log
5733 directory problem" with a 421 code for SMTP messages.
5735 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5736 0, 1, and 2 exist, because some OS complain for execs without them (see
5737 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5738 descriptors. However, the code omitted to check that the open succeeded,
5739 causing mysterious errors if for some reason the permissions on /dev/null
5740 got screwed. Now Exim writes a message to the main and panic logs, and
5741 bombs out if it can't open /dev/null.
5743 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5744 interact so that it is all more flexible. It is supposed to remain
5745 backwards compatible. Also added extra_local_interfaces.
5747 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5748 to bomb out with an assertion failure - to the client this appears as a
5749 connection drop. This problem occurs in the part of the code that was taken
5750 from the Samba project. Fortunately, the assertion is in a very simple
5751 function, so I have fixed this by reproducing the function inline in the
5752 one place where it is called, and arranging for authentication to fail
5753 instead of killing the process with assert().
5755 75. The SPA client code was not working when the server requested OEM rather
5756 than Unicode encoding.
5758 76. Added code to make require_files with a specific uid setting more usable in
5759 the case where statting the file as root fails - usually a non-root-mounted
5760 NFS file system. When this happens and the failure is EACCES, Exim now
5761 forks a subprocess and does the per-uid checking as the relevant uid.
5763 77. Added process_log_path.
5765 78. If log_file_path was not explicitly set, a setting of check_log_space or
5766 check_log_inodes was ignored.
5768 79. If a space check for the spool or log partitions fails, the incident is now
5769 logged. Of course, in the latter case the data may get lost...
5771 80. Added the %p formatting code to string_format() so that it can be used to
5772 print addresses in debug_print(). Adjusted all the address printing in the
5773 debugging in store.c to use %p rather than %d.
5775 81. There was a concern that a line of code in smtp_in.c could overflow a
5776 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5777 initially expressed, the concern was not well-founded, because trailing
5778 spaces are removed early. However, if the trailing spaces were followed by
5779 a NULL, they did not get removed, so the overflow was possible. Two fixes
5782 (a) I re-wrote the offending code in a cleaner fashion.
5783 (b) If an incoming SMTP command contains a NULL character, it is rejected
5786 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5787 runs initgroups(), so that if the Exim user is in any additional groups,
5788 they will be used during message reception.
5794 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob