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 commit adds a proper retry rule check during routing so that
117 the final cutoff time is checked against the message's age. I also
118 took the opportunity to unify three very similar blocks of code.
120 I suspect this new check makes the old local delivery cutoff check
121 redundant, but I have not verified this so I left the code in place.
127 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
129 This, or similar/improved, will also be change PP/11 of 4.82.
135 PP/01 Handle short writes when writing local log-files.
136 In practice, only affects FreeBSD (8 onwards).
137 Bugzilla 1053, with thanks to Dmitry Isaikin.
139 NM/01 Bugzilla 949 - Documentation tweak
141 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
144 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
146 PP/02 Implemented gsasl authenticator.
148 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
150 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
151 `pkg-config foo` for cflags/libs.
153 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
154 with rest of GSASL and with heimdal_gssapi.
156 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
157 `pkg-config foo` for cflags/libs for the TLS implementation.
159 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
160 properties get this fed in as external SSF. A number of robustness
161 and debugging improvements to the cyrus_sasl authenticator.
163 PP/08 cyrus_sasl server now expands the server_realm option.
165 PP/09 Bugzilla 1214 - Log authentication information in reject log.
166 Patch by Jeremy Harris.
168 PP/10 Added dbmjz lookup type.
170 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
172 PP/12 MAIL args handles TAB as well as SP, for better interop with
173 non-compliant senders.
174 Analysis and variant patch by Todd Lyons.
176 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
177 Bug report from Lars Müller <lars@samba.org> (via SUSE),
178 Patch from Dirk Mueller <dmueller@suse.com>
180 PP/13 tls_peerdn now print-escaped for spool files.
181 Observed some $tls_peerdn in wild which contained \n, which resulted
182 in spool file corruption.
184 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
185 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
186 or write after TLS renegotiation, which otherwise led to messages
189 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
190 as a tracking header (ie: a signed header comes before the signature).
191 Patch from Wolfgang Breyha.
193 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
194 comma-sep list; embedded commas doubled.
196 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
198 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
200 Report and patch from Dmitry Banschikov.
202 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
203 Removed SSL_clear() after SSL_new() which led to protocol negotiation
204 failures. We appear to now support TLS1.1+ with Exim.
206 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
207 lets Exim select keys and certificates based upon TLS SNI from client.
208 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
209 before an outbound SMTP session. New log_selector, +tls_sni.
211 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
212 NULL dereference. Report and patch from Alun Jones.
214 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
215 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
216 Not seeing resolver debug output on NetBSD, but suspect this is a
217 resolver implementation change.
219 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
220 Left warnings. Added "eximon gdb" invocation mode.
222 PP/21 Defaulting "accept_8bitmime" to true, not false.
224 PP/22 Added -bw for inetd wait mode support.
226 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
227 locate the relevant includes and libraries. Made this the default.
229 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
230 Bugzilla 1246, report and most of solution from Tomasz Kusy.
232 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
233 This may cause build issues on older platforms.
235 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
236 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
237 gnutls_require_mac & gnutls_require_protocols (no longer supported).
238 Added SNI support via GnuTLS too.
239 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
241 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
243 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
244 Applied second patch from Janne, implementing suggestion to default
245 multiple-strings-in-record handling to match SPF spec.
247 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
249 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
250 read-only, out of scope).
251 Patch from Wolfgang Breyha, report from Stuart Northfield.
253 PP/29 Fix three issues highlighted by clang analyser static analysis.
254 Only crash-plausible issue would require the Cambridge-specific
255 iplookup router and a misconfiguration.
256 Report from Marcin Mirosław.
258 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
260 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
261 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
262 As part of this, removing so much warning spew let me fix some minor
263 real issues in debug logging.
265 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
266 assignment on my part. Fixed.
268 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
269 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
270 Janne Snabb (who went above and beyond: thank you).
272 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
273 string otherwise requires a connection and a bunch more work and it's
274 relatively easy to get wrong. Should also expose TLS library linkage
277 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
278 64-bit ${eval} (JH/03).
280 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
281 GNU libc to support some of the 64-bit stuff, should not lead to
282 conflicts. Defined before os.h is pulled in, so if a given platform
283 needs to override this, it can.
285 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
286 protection layer was required, which is not implemented.
287 Bugzilla 1254, patch from Wolfgang Breyha.
289 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
290 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
291 tls_dhparam take prime identifiers. Also unbreak combination of
292 OpenSSL+DH_params+TLSSNI.
294 PP/39 Disable SSLv2 by default in OpenSSL support.
300 PP/01 Solaris build fix for Oracle's LDAP libraries.
301 Bugzilla 1109, patch from Stephen Usher.
303 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
305 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
308 TF/02 Fix a couple more cases where we did not log the error message
309 when unlink() failed. See also change 4.74-TF/03.
311 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
312 lock up or crash if it happened to be inside a call to libc when it
313 got a SIGUSR1 from exiwhat.
315 The SIGUSR1 handler appends the current process status to the process
316 log which is later printed by exiwhat. It used to use the general
317 purpose logging code to do this, but several functions it calls are
318 not safe for signals.
320 The new output code in the SIGUSR1 handler is specific to the process
321 log, and simple enough that it's easy to inspect for signal safety.
322 Removing some special cases also simplifies the general logging code.
323 Removing the spurious timestamps from the process log simplifies
326 TF/04 Improved ratelimit ACL condition.
328 The /noupdate option has been deprecated in favour of /readonly which
329 has clearer semantics. The /leaky, /strict, and /readonly update modes
330 are mutually exclusive. The update mode is no longer included in the
331 database key; it just determines when the database is updated. (This
332 means that when you upgrde Exim will forget old rate measurements.)
334 Exim now checks that the per_* options are used with an update mode that
335 makes sense for the current ACL. For example, when Exim is processing a
336 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
337 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
338 must specify per_mail/readonly. If you omit the update mode it defaults to
339 /leaky where that makes sense (as before) or /readonly where required.
341 The /noupdate option is now undocumented but still supported for
342 backwards compatibility. It is equivalent to /readonly except that in
343 ACLs where /readonly is required you may specify /leaky/noupdate or
344 /strict/noupdate which are treated the same as /readonly.
346 A useful new feature is the /count= option. This is a generalization
347 of the per_byte option, so that you can measure the throughput of other
348 aggregate values. For example, the per_byte option is now equivalent
349 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
351 The per_rcpt option has been generalized using the /count= mechanism
352 (though it's more complicated than the per_byte equivalence). When it is
353 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
354 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
355 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
356 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
357 Note that using per_rcpt with a non-readonly update mode in more than
358 one ACL will cause the recipients to be double-counted. (The per_mail
359 and per_byte options don't have this problem.)
361 The handling of very low rates has changed slightly. If the computed rate
362 is less than the event's count (usually one) then this event is the first
363 after a long gap. In this case the rate is set to the same as this event's
364 count, so that the first message of a spam run is counted properly.
366 The major new feature is a mechanism for counting the rate of unique
367 events. The new per_addr option counts the number of different
368 recipients that someone has sent messages to in the last time period. It
369 behaves like per_rcpt if all the recipient addresses are different, but
370 duplicate recipient addresses do not increase the measured rate. Like
371 the /count= option this is a general mechanism, so the per_addr option
372 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
373 example, measure the rate that a client uses different sender addresses
374 with the options per_mail/unique=$sender_address. There are further
375 details in the main documentation.
377 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
379 TF/06 Removed a few PCRE remnants.
381 TF/07 Automatically extract Exim's version number from tags in the git
382 repository when doing development or release builds.
384 PP/02 Raise smtp_cmd_buffer_size to 16kB.
385 Bugzilla 879. Patch from Paul Fisher.
387 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
388 Heavily based on revision 40f9a89a from Simon Arlott's tree.
391 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
393 PP/05 Variable $av_failed, true if the AV scanner deferred.
394 Bugzilla 1078. Patch from John Horne.
396 PP/06 Stop make process more reliably on build failure.
397 Bugzilla 1087. Patch from Heiko Schlittermann.
399 PP/07 Make maildir_use_size_file an _expandable_ boolean.
400 Bugzilla 1089. Patch from Heiko Schlittermann.
402 PP/08 Handle ${run} returning more data than OS pipe buffer size.
403 Bugzilla 1131. Patch from Holger Weiß.
405 PP/09 Handle IPv6 addresses with SPF.
406 Bugzilla 860. Patch from Wolfgang Breyha.
408 PP/10 GnuTLS: support TLS 1.2 & 1.1.
410 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
413 PP/11 match_* no longer expand right-hand-side by default.
414 New compile-time build option, EXPAND_LISTMATCH_RHS.
415 New expansion conditions, "inlist", "inlisti".
417 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
419 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
421 PP/14 fix log_write() format string regression from TF/03.
422 Bugzilla 1152. Patch from Dmitry Isaikin.
428 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
430 PP/02 Harmonised TLS library version reporting; only show if debugging.
431 Layout now matches that introduced for other libraries in 4.74 PP/03.
433 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
435 PP/04 New "dns_use_edns0" global option.
437 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
440 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
441 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
443 TK/01 Updated PolarSSL code to 0.14.2.
444 Bugzilla 1097. Patch from Andreas Metzler.
446 PP/07 Catch divide-by-zero in ${eval:...}.
449 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
452 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
453 format-string attack -- SECURITY: remote arbitrary code execution.
455 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
456 time unintentionally subject to list matching rules, letting the header
457 cause arbitrary Exim lookups (of items which can occur in lists, *not*
458 arbitrary string expansion). This allowed for information disclosure.
460 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
461 INT_MIN/-1 -- value coerced to INT_MAX.
467 NM/01 Workround for PCRE version dependency in version reporting
470 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
471 This fixes portability to compilers other than gcc, notably
472 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
474 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
475 makefiles for portability to HP-UX and POSIX correctness.
477 PP/01 Permit LOOKUP_foo enabling on the make command-line.
478 Also via indented variable definition in the Makefile.
479 (Debugging by Oliver Heesakkers).
481 PP/02 Restore caching of spamd results with expanded spamd_address.
482 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
484 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
485 Improves build reliability. Fix from: Frank Elsner
487 NM/02 Fix wide character breakage in the rfc2047 coding
488 Fixes bug 1064. Patch from Andrey N. Oktyabrski
490 NM/03 Allow underscore in dnslist lookups
491 Fixes bug 1026. Patch from Graeme Fowler
493 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
494 Code patches from Adam Ciarcinski of NetBSD.
496 NM/04 Fixed exiqgrep to cope with mailq missing size issue
499 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
500 is logged, to avoid truncation. Patch from John Horne.
502 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
503 Patch from Jakob Hirsch.
505 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
506 SQL string expansion failure details.
507 Patch from Andrey Oktyabrski.
509 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
510 Patch from Simon Arlott.
512 PP/09 New lookups functionality failed to compile on old gcc which rejects
513 extern declarations in function scope.
514 Patch from Oliver Fleischmann
516 PP/10 Use sig_atomic_t for flags set from signal handlers.
517 Check getgroups() return and improve debugging.
518 Fixed developed for diagnosis in bug 927 (which turned out to be
521 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
522 Patch from Mark Zealey.
524 PP/12 Bugzilla 1056: Improved spamd server selection.
525 Patch from Mark Zealey.
527 PP/13 Bugzilla 1086: Deal with maildir quota file races.
528 Based on patch from Heiko Schlittermann.
530 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
531 Patch from Uwe Doering, sign-off by Michael Haardt.
533 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
534 variable declaration deep within a block. Bug and patch from
537 PP/15 lookups-Makefile IRIX compatibilty coercion.
539 PP/16 Make DISABLE_DKIM build knob functional.
541 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
542 Patch by Simon Arlott
544 TF/03 Fix valgrind.h portability to C89 compilers that do not support
545 variable argument macros. Our copy now differs from upstream.
551 TF/01 Failure to get a lock on a hints database can have serious
552 consequences so log it to the panic log.
554 TF/02 Log LMTP confirmation messages in the same way as SMTP,
555 controlled using the smtp_confirmation log selector.
557 TF/03 Include the error message when we fail to unlink a spool file.
559 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
560 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
561 for maintaining out-of-tree patches for some time.
563 PP/01 Bugzilla 139: Documentation and portability issues.
564 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
565 Handle per-OS dynamic-module compilation flags.
567 PP/02 Let /dev/null have normal permissions.
568 The 4.73 fixes were a little too stringent and complained about the
569 permissions on /dev/null. Exempt it from some checks.
570 Reported by Andreas M. Kirchwitz.
572 PP/03 Report version information for many libraries, including
573 Exim version information for dynamically loaded libraries. Created
574 version.h, now support a version extension string for distributors
575 who patch heavily. Dynamic module ABI change.
577 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
578 privilege escalation vulnerability whereby the Exim run-time user
579 can cause root to append content of the attacker's choosing to
582 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
585 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
586 If dropping privileges for untrusted macros, we disabled normal logging
587 on the basis that it would fail; for the Exim run-time user, this is not
588 the case, and it resulted in successful deliveries going unlogged.
589 Fixed. Reported by Andreas Metzler.
595 PP/01 Date: & Message-Id: revert to normally being appended to a message,
596 only prepend for the Resent-* case. Fixes regression introduced in
597 Exim 4.70 by NM/22 for Bugzilla 607.
599 PP/02 Include check_rfc2047_length in configure.default because we're seeing
600 increasing numbers of administrators be bitten by this.
602 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
604 PP/03 Bugzilla 994: added openssl_options main configuration option.
606 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
608 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
610 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
612 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
613 without a peer certificate, leading to a segfault because of an
614 assumption that peers always have certificates. Be a little more
615 paranoid. Problem reported by Martin Tscholak.
617 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
618 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
619 NB: ClamAV planning to remove STREAM in "middle of 2010".
620 CL also introduces -bmalware, various -d+acl logging additions and
621 more caution in buffer sizes.
623 PP/09 Implemented reverse_ip expansion operator.
625 PP/10 Bugzilla 937: provide a "debug" ACL control.
627 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
629 PP/12 Bugzilla 973: Implement --version.
631 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
633 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
635 PP/15 Bugzilla 816: support multiple condition rules on Routers.
637 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
638 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
639 ignore trailing whitespace.
641 JJ/02 prevent non-panic DKIM error from being sent to paniclog
643 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
646 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
647 Notification from Dr Andrew Aitchison.
649 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
650 ExtendedDetectionInfo response format.
651 Notification from John Horne.
653 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
656 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
657 XSL and documented dependency on system catalogs, with examples of how
660 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
663 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
664 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
665 configuration file which is writeable by the Exim user or group.
667 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
668 of configuration files to cover files specified with the -C option if
669 they are going to be used with root privileges, not just the default
672 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
673 option (effectively making it always true).
675 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
676 files to be used while preserving root privileges.
678 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
679 that rogue child processes cannot use them.
681 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
682 run-time user, instead of root.
684 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
685 Exim run-time user without dropping privileges.
687 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
688 result string, instead of calling string_vformat() twice with the same
691 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
692 for other users. Others should always drop root privileges if they use
693 -C on the command line, even for a whitelisted configure file.
695 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
697 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
703 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
704 $data_path, and $header_path variables; fixed documentation bugs and
707 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
708 exipick to access non-standard spools, including the "frozen" queue
711 NM/01 Bugzilla 965: Support mysql stored procedures.
712 Patch from Alain Williams
714 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
716 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
717 Patch from Andreas Metzler
719 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
720 Patch from Kirill Miazine
722 NM/05 Bugzilla 671: Added umask to procmail example.
724 JJ/03 installed exipick 20100323.0, fixing doc bug
726 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
727 directory. Notification and patch from Dan Rosenberg.
729 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
731 TK/02 Improve log output when DKIM signing operation fails.
733 MH/01 Treat the transport option dkim_domain as a colon separated
734 list, not as a single string, and sign the message with each element,
735 omitting multiple occurences of the same signer.
737 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
738 Bugzilla 985, 986. Patch by Simon Arlott
740 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
741 Patch by Simon Arlott
743 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
744 MBX locking. Notification from Dan Rosenberg.
750 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
752 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
754 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
756 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
758 NM/04 Bugzilla 915: Flag broken perl installation during build.
764 TK/01 Added patch by Johannes Berg that expands the main option
765 "spamd_address" if it starts with a dollar sign.
767 TK/02 Write list of recipients to X-Envelope-Sender header when building
768 the mbox-format spool file for content scanning (suggested by Jakob
771 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
772 (http://www.dcc-servers.net/) support via dccifd. Activated by
773 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
775 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
776 by Mark Daniel Reidel <mr@df.eu>.
778 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
779 When building exim an external PCRE library is now needed -
780 PCRE is a system library on the majority of modern systems.
781 See entry on PCRE_LIBS in EDITME file.
783 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
784 conversation. Added nologin parameter to request.
785 Patch contributed by Kirill Miazine.
787 TF/01 Do not log submission mode rewrites if they do not change the address.
789 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
791 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
792 log files in place. Contributed by Roberto Lima.
794 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
796 TF/03 Bugzilla 615: When checking the local_parts router precondition
797 after a local_part_suffix or local_part_prefix option, Exim now
798 does not use the address's named list lookup cache, since this
799 contains cached lookups for the whole local part.
801 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
802 Robert Millan. Documentation is in experimental-spec.txt.
804 TF/04 Bugzilla 668: Fix parallel build (make -j).
806 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
808 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
809 Patch provided by Jan Srzednicki.
811 TF/05 Leading white space used to be stripped from $spam_report which
812 wrecked the formatting. Now it is preserved.
814 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
815 that they are available at delivery time.
817 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
819 TF/08 TLS error reporting now respects the incoming_interface and
820 incoming_port log selectors.
822 TF/09 Produce a more useful error message if an SMTP transport's hosts
823 setting expands to an empty string.
825 NM/06 Bugzilla 744: EXPN did not work under TLS.
826 Patch provided by Phil Pennock.
828 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
829 Patch provided by Richard Godbee.
831 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
832 acl_smtp_notquit, added index entry.
834 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
835 Patch provided by Eugene Bujak.
837 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
838 accept(). Patch provided by Maxim Dounin.
840 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
841 Patch provided by Phil Pennock.
843 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
845 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
846 Patch provided by Brad "anomie" Jorsch.
848 NM/14 Bugzilla 622: Added timeout setting to transport filter.
849 Patch provided by Dean Brooks.
851 TK/05 Add native DKIM support (does not depend on external libraries).
853 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
854 Patch provided by Graeme Fowler.
856 NM/16 Bugzilla 851: Documentation example syntax fix.
858 NM/17 Changed NOTICE file to remove references to embedded PCRE.
860 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
863 NM/19 Bugzilla 745: TLS version reporting.
864 Patch provided by Phil Pennock.
866 NM/20 Bugzilla 167: bool: condition support.
867 Patch provided by Phil Pennock.
869 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
870 clients. Patch provided by Phil Pennock.
872 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
873 Patch provided by Brad "anomie" Jorsch.
875 NM/23 Bugzilla 687: Fix misparses in eximstats.
876 Patch provided by Heiko Schlittermann.
878 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
879 Patch provided by Heiko Schlittermann.
881 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
882 plus update to original patch.
884 NM/26 Bugzilla 799: Documentation correction for ratelimit.
886 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
887 Patch provided by David Brownlee.
889 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
891 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
893 NM/30 Bugzilla 888: TLS documentation bugfixes.
895 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
897 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
898 Unlike the original bugzilla I have changed all shell scripts in src tree.
900 NM/33 Bugzilla 898: Transport filter timeout fix.
901 Patch by Todd Rinaldo.
903 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
904 Patch by Serge Demonchaux.
906 NM/35 Bugzilla 39: Base64 decode bug fixes.
907 Patch by Jakob Hirsch.
909 NM/36 Bugzilla 909: Correct connect() call in dcc code.
911 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
913 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
915 NM/39 Bugzilla 911: Fixed MakeLinks build script.
921 TK/01 Add preliminary DKIM support. Currently requires a forked version of
922 ALT-N's libdkim that I have put here:
923 http://duncanthrax.net/exim-experimental/
925 Note to Michael Haardt: I had to rename some vars in sieve.c. They
926 were called 'true' and it seems that C99 defines that as a reserved
927 keyword to be used with 'bool' variable types. That means you could
928 not include C99-style headers which use bools without triggering
929 build errors in sieve.c.
931 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
932 as mailq or other aliases. Changed the --help handling significantly
933 to do whats expected. exim_usage() emits usage/help information.
935 SC/01 Added the -bylocaldomain option to eximstats.
937 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
939 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
941 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
947 PH/01 Another patch from the Sieve maintainer.
949 PH/02 When an IPv6 address is converted to a string for single-key lookup
950 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
951 dots are used instead of colons so that keys in lsearch files need not
952 contain colons. This was done some time before quoting was made available
953 in lsearch files. However, iplsearch files do require colons in IPv6 keys
954 (notated using the quote facility) so as to distinguish them from IPv4
955 keys. This meant that lookups for IP addresses in host lists did not work
956 for iplsearch lookups.
958 This has been fixed by arranging for IPv6 addresses to be expressed with
959 colons if the lookup type is iplsearch. This is not incompatible, because
960 previously such lookups could never work.
962 The situation is now rather anomolous, since one *can* have colons in
963 ordinary lsearch keys. However, making the change in all cases is
964 incompatible and would probably break a number of configurations.
966 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
969 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
970 conversion specification without a maximum field width, thereby enabling
971 a rogue spamd server to cause a buffer overflow. While nobody in their
972 right mind would setup Exim to query an untrusted spamd server, an
973 attacker that gains access to a server running spamd could potentially
974 exploit this vulnerability to run arbitrary code as the Exim user.
976 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
977 $primary_hostname instead of what libspf2 thinks the hosts name is.
979 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
980 a directory entry by the name of the lookup key. Previously, if a
981 symlink pointed to a non-existing file or a file in a directory that
982 Exim lacked permissions to read, a lookup for a key matching that
983 symlink would fail. Now it is enough that a matching directory entry
984 exists, symlink or not. (Bugzilla 503.)
986 PH/03 The body_linecount and body_zerocount variables are now exported in the
989 PH/04 Added the $dnslist_matched variable.
991 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
992 This means they are set thereafter only if the connection becomes
995 PH/06 Added the client_condition to authenticators so that some can be skipped
996 by clients under certain conditions.
998 PH/07 The error message for a badly-placed control=no_multiline_responses left
999 "_responses" off the end of the name.
1001 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1003 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1004 (without spaces) instead of just copying the configuration text.
1006 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1008 PH/11 Added $max_received_linelength.
1010 PH/12 Added +ignore_defer and +include_defer to host lists.
1012 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1013 way in which PCRE > 7.0 is built.
1015 PH/14 Implemented queue_only_load_latch.
1017 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1018 MAIL command. The effect was to mangle the value on 64-bit systems.
1020 PH/16 Another patch from the Sieve maintainer.
1022 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1024 PH/18 If a system quota error occurred while trying to create the file for
1025 a maildir delivery, the message "Mailbox is full" was not appended to the
1026 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1027 only to a quota excession during the actual writing of the file.
1029 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1030 characters?) which causes problems in log lines. The DN values are now
1031 passed through string_printing() before being added to log lines.
1033 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1034 and InterBase are left for another time.)
1036 PH/21 Added message_body_newlines option.
1038 PH/22 Guard against possible overflow in moan_check_errorcopy().
1040 PH/23 POSIX allows open() to be a macro; guard against that.
1042 PH/24 If the recipient of an error message contained an @ in the local part
1043 (suitably quoted, of course), incorrect values were put in $domain and
1044 $local_part during the evaluation of errors_copy.
1050 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1051 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1054 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1055 issue a MAIL command.
1057 PH/02 In an ACL statement such as
1059 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1061 if a client was not listed at all, or was listed with a value other than
1062 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1063 the condition was not true (as it should be), so access was not denied.
1064 The bug was that the ! inversion was incorrectly passed on to the second
1065 item. This has been fixed.
1067 PH/03 Added additional dnslists conditions == and =& which are different from
1068 = and & when the dns lookup returns more than one IP address.
1070 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1071 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1073 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1074 FSYNC, which compiles an option called disable_fsync that allows for
1075 bypassing fsync(). The documentation is heavily laced with warnings.
1077 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1079 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1080 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1081 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1082 including adding "make clean"; (3) Added -fPIC when compiling the test
1083 dynamically loaded module, to get rid of a warning.
1085 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1086 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1087 = 0s. The bug is otherwise harmless.
1089 PH/07 There was a bug in the dovecot authenticator such that the value of
1090 $auth1 could be overwritten, and so not correctly preserved, after a
1091 successful authentication. This usually meant that the value preserved by
1092 the server_setid option was incorrect.
1094 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1096 PH/09 Installed PCRE release 7.0.
1098 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1099 run for batched SMTP input. It is now run at the start of every message
1100 in the batch. While fixing this I discovered that the process information
1101 (output by running exiwhat) was not always getting set for -bs and -bS
1102 input. This is fixed, and it now also says "batched" for BSMTP.
1104 PH/11 Added control=no_pipelining.
1106 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1107 patch, slightly modified), and move the expansion of helo_data till after
1108 the connection is made in the smtp transport (so it can use these
1111 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1113 PH/14 Added log_selector = +pid.
1115 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1117 PH/16 Add ${if forany and ${if forall.
1119 PH/17 Added dsn_from option to vary the From: line in DSNs.
1121 PH/18 Flush SMTP output before performing a callout, unless control =
1122 no_callout_flush is set.
1124 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1125 was true (the default) a successful delivery failed to delete the retry
1126 item, thus causing premature timeout of the address. The bug is now
1129 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1131 PH/21 Long custom messages for fakedefer and fakereject are now split up
1132 into multiline reponses in the same way that messages for "deny" and
1133 other ACL rejections are.
1135 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1136 with slight modification.
1138 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1139 draft, changing the syntax and factoring some duplicate code".
1141 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1142 for deliveries of the second and subsequent messages over the same SMTP
1145 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1146 ${reduce, with only minor "tidies".
1148 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1150 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1151 expansion side effects.
1153 PH/27 When a message times out after an over-quota error from an Exim-imposed
1154 quota, the bounce message says "mailbox is full". This message was not
1155 being given when it was a system quota that was exceeded. It now should
1158 MH/03 Made $recipients available in local_scan(). local_scan() already has
1159 better access to the recipient list through recipients_list[], but
1160 $recipients can be useful in postmaster-provided expansion strings.
1162 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1163 in the case of a MAIL command with additional options following the
1164 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1165 were accidentally chopped off.
1167 PH/29 SMTP synchronization checks are implemented when a command is read -
1168 there is a check that no more input is waiting when there shouldn't be
1169 any. However, for some commands, a delay in an ACL can mean that it is
1170 some time before the response is written. In this time, more input might
1171 arrive, invalidly. So now there are extra checks after an ACL has run for
1172 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1173 pipelining has not been advertised.
1175 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1177 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1178 returned in the SMTP response when smtp_return_error_details was false.
1179 This has been fixed.
1181 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1182 instead of the C I/O that was originally supplied, because problems were
1183 reported on Solaris.
1185 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1186 Exim which did not show up earlier: it was assuming that a call to
1187 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1188 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1189 was a macro that became an assignment, so it seemed to work. This has
1190 changed to a proper function call with a void return, hence the compile
1191 error. Exim's code has been fixed.
1193 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1196 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1198 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1200 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1201 because it turns out to be too big or there is a timeout) there is no
1202 "Completed" line in the log. When some messages of this type were
1203 selected by exigrep, they were listed as "not completed". Others were
1204 picked up by some special patterns. I have improved the selection
1205 criteria to be more general.
1207 PH/38 The host_find_failed option in the manualroute router can now be set
1208 to "ignore", to completely ignore a host whose IP address cannot be
1209 found. If all hosts are ignored, the behaviour is controlled by the new
1210 host_all_ignored option.
1212 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1213 homing or because of multiple MX records with /mx) generated more than
1214 one IP address, and the following item turned out to be the local host,
1215 all the secondary addresses of the first item were incorrectly removed
1216 from the list, along with the local host and any following hosts (which
1217 is what is supposed to happen).
1219 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1220 whoever called Exim into the -H file. In the case of the daemon it was
1221 behaving confusingly. When first started, it used values for whoever
1222 started the daemon, but after a SIGHUP it used the Exim user (because it
1223 calls itself on a restart). I have changed the code so that it now always
1226 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1227 message are rejected with the same error (e.g. no authentication or bad
1228 sender address), and a DATA command is nevertheless sent (as can happen
1229 with PIPELINING or a stupid MUA), the error message that was given to the
1230 RCPT commands is included in the rejection of the DATA command. This is
1231 intended to be helpful for MUAs that show only the final error to their
1234 PH/42 Another patch from the Sieve maintainer.
1236 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1237 Eximstats - Fixed some broken HTML links and added missing column headers
1239 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1240 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1242 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1244 PH/43 Yet another patch from the Sieve maintainer.
1246 PH/44 I found a way to check for a TCP/IP connection going away before sending
1247 the response to the final '.' that terminates a message, but only in the
1248 case where the client has not sent further data following the '.'
1249 (unfortunately, this is allowed). However, in many cases there won't be
1250 any further data because there won't be any more messages to send. A call
1251 to select() can be used: if it shows that the input is "ready", there is
1252 either input waiting, or the socket has been closed. An attempt to read
1253 the next input character can distinguish the two cases. Previously, Exim
1254 would have sent an OK response which the client would never have see.
1255 This could lead to message repetition. This fix should cure that, at
1256 least in a lot of common cases.
1258 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1259 advertised in response to EHLO.
1265 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1266 fixed by 4.65/MH/01 (is this a record?) are fixed:
1268 (i) An empty string was always treated as zero by the numeric comparison
1269 operators. This behaviour has been restored.
1271 (ii) It is documented that the numeric comparison operators always treat
1272 their arguments as decimal numbers. This was broken in that numbers
1273 starting with 0 were being interpreted as octal.
1275 While fixing these problems I realized that there was another issue that
1276 hadn't been noticed. Values of message_size_limit (both the global option
1277 and the transport option) were treated as octal if they started with 0.
1278 The documentation was vague. These values are now always treated as
1279 decimal, and I will make that clear in the documentation.
1285 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1286 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1289 MH/01 Don't check that the operands of numeric comparison operators are
1290 integers when their expansion is in "skipping" mode (fixes bug
1291 introduced by 4.64-PH/07).
1293 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1294 child addresses, Exim now panics and dies. Previously, because the count
1295 is held in a short int, deliveries were likely to be lost. As such a
1296 large number of recipients for a single message is ridiculous
1297 (performance will be very, very poor), I have chosen to impose a limit
1298 rather than extend the field.
1304 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1305 leftover -K file (the existence of which was triggered by #402).
1306 While we were at it, introduced process PID as part of the -K
1307 filename. This should rule out race conditions when creating
1310 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1311 processing considerably. Previous code took too long for large mails,
1312 triggering a timeout which in turn triggers #401.
1314 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1315 in the DK code in transports.c. sendfile() is not really portable,
1316 hence the _LINUX specificness.
1318 TF/01 In the add_headers option to the mail command in an Exim filter,
1319 there was a bug that Exim would claim a syntax error in any
1320 header after the first one which had an odd number of characters
1323 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1324 callout verification, Exim cached a "reject" for the entire domain. This
1325 is correct for most verifications, but it is not correct for a recipient
1326 verification with use_sender or use_postmaster set, because in that case
1327 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1328 case of MAIL FROM:<> rejection from other early rejections (e.g.
1329 rejection of HELO). When verifying a recipient using a non-null MAIL
1330 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1331 Whatever the result of the callout, the value of the domain cache is
1332 left unchanged (for any other kind of callout, getting as far as trying
1333 RCPT means that the domain itself is ok).
1335 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1338 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1339 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1340 session, the daemon ignores ECONNECT errors and logs others; it now
1341 ignores EPIPE as well.
1343 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1344 (quoted-printable decoding).
1346 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1347 later the small subsequent patch to fix an introduced bug.
1349 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1351 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1353 PH/08 An error is now given if message_size_limit is specified negative.
1355 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1356 to be given (somewhat) arbitrary names.
1358 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1361 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1362 miscellaneous code fixes
1364 PH/10 Added the log_reject_target ACL modifier to specify where to log
1367 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1368 hostname. This is wrong, because it relates to the incoming message (and
1369 probably the interface on which it is arriving) and not to the outgoing
1370 callout (which could be using a different interface). This has been
1371 changed to use the value of the helo_data option from the smtp transport
1372 instead - this is what is used when a message is actually being sent. If
1373 there is no remote transport (possible with a router that sets up host
1374 addresses), $smtp_active_hostname is used.
1376 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1377 tweaks were necessary in order to get it to work (see also 21 below):
1378 (a) The code assumed that strncpy() returns a negative number on buffer
1379 overflow, which isn't the case. Replaced with Exim's string_format()
1381 (b) There were several signed/unsigned issues. I just did the minimum
1382 hacking in of casts. There is scope for a larger refactoring.
1383 (c) The code used strcasecmp() which is not a standard C function.
1384 Replaced with Exim's strcmpic() function.
1385 (d) The code set only $1; it now sets $auth1 as well.
1386 (e) A simple test gave the error "authentication client didn't specify
1387 service in request". It would seem that Dovecot has changed its
1388 interface. Fortunately there's a specification; I followed it and
1389 changed what the client sends and it appears to be working now.
1391 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1394 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1395 address A is aliased to B and C, where B exists and C does not. Without
1396 -v the output is "A verified" because verification stops after a
1397 successful redirection if more than one address is generated. However,
1398 with -v the child addresses are also verified. Exim was outputting "A
1399 failed to verify" and then showing the successful verification for C,
1400 with its parentage. It now outputs "B failed to verify", showing B's
1401 parentage before showing the successful verification of C.
1403 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1404 look up a TXT record in a specific list after matching in a combined
1407 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1408 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1409 they consult the DNS. I had assumed they would set it the way they
1410 wanted; and indeed my experiments on Linux seem to show that in some
1411 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1412 To be on the safe side, however, I have now made the interface to
1413 host_find_byname() similar to host_find_bydns(), with an argument
1414 containing the DNS resolver options. The host_find_byname() function now
1415 sets these options at its start, just as host_find_bydns() does. The smtp
1416 transport options dns_qualify_single and dns_search_parents are passed to
1417 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1418 of host_find_byname() use the default settings of RES_DEFNAMES
1419 (qualify_single) but not RES_DNSRCH (search_parents).
1421 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1422 spool_read_header() do less string testing, by means of a preliminary
1423 switch on the second character of optional "-foo" lines. (This is
1424 overdue, caused by the large number of possibilities that now exist.
1425 Originally there were few.) While I was there, I also converted the
1426 str(n)cmp tests so they don't re-test the leading "-" and the first
1427 character, in the hope this might squeeze out yet more improvement.
1429 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1430 flag allowing group syntax was set by the header_syntax check but not
1431 turned off, possible causing trouble later; (2) The flag was not being
1432 set at all for the header_verify test, causing "group"-style headers to
1433 be rejected. I have now set it in this case, and also caused header_
1434 verify to ignore an empty address taken from a group. While doing this, I
1435 came across some other cases where the code for allowing group syntax
1436 while scanning a header line wasn't quite right (mostly, not resetting
1437 the flag correctly in the right place). These bugs could have caused
1438 trouble for malformed header lines. I hope it is now all correct.
1440 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1441 with the "reply" argument non-NULL. The code, however (which originally
1442 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1443 but it didn't always do it. This confused somebody who was copying the
1444 code for some other use. I have removed all the tests.
1446 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1447 feature that was used to support insecure browsers during the U.S. crypto
1448 embargo. It requires special client support, and Exim is probably the
1449 only MTA that supported it -- and would never use it because real RSA is
1450 always available. This code has been removed, because it had the bad
1451 effect of slowing Exim down by computing (never used) parameters for the
1452 RSA_EXPORT functionality.
1454 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1455 authenticator to fail if there's a tab character in the incoming data
1456 (there should never be unless someone is messing about, as it's supposed
1457 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1458 if the connection is using TLS or if the remote IP is the same as the
1459 local IP, and the "valid-client-cert option" if a client certificate has
1462 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1463 authenticators. This can be used for authorization after authentication
1464 succeeds. (In the case of plaintext, it servers for both authentication
1467 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1468 if any retry times were supplied.
1470 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1471 connection, where there is no client IP address to check. In this
1472 situation, the verify now always succeeds.
1474 PH/25 Applied John Jetmore's -Mset patch.
1476 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1478 PH/27 In a string expansion for a processed (not raw) header when multiple
1479 headers of the same name were present, leading whitespace was being
1480 removed from all of them, but trailing whitespace was being removed only
1481 from the last one. Now trailing whitespace is removed from each header
1482 before concatenation. Completely empty headers in a concatenation (as
1483 before) are ignored.
1485 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1486 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1488 PH/29 [Removed. This was a change that I later backed out, and forgot to
1489 correct the ChangeLog entry (that I had efficiently created) before
1490 committing the later change.]
1492 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1493 address errors (4xx response to RCPT) over the same connection as other
1494 messages routed to the same hosts. Such deliveries are always "forced",
1495 so retry times are not inspected. This resulted in far too many retries
1496 for the affected addresses. The effect occurred only when there were more
1497 hosts than the hosts_max_try setting in the smtp transport when it had
1498 the 4xx errors. Those hosts that it had tried were not added to the list
1499 of hosts for which the message was waiting, so if all were tried, there
1500 was no problem. Two fixes have been applied:
1502 (i) If there are any address or message errors in an SMTP delivery, none
1503 of the hosts (tried or untried) are now added to the list of hosts
1504 for which the message is waiting, so the message should not be a
1505 candidate for sending over the same connection that was used for a
1506 successful delivery of some other message. This seems entirely
1507 reasonable: after all the message is NOT "waiting for some host".
1508 This is so "obvious" that I'm not sure why it wasn't done
1509 previously. Hope I haven't missed anything, but it can't do any
1510 harm, as the worst effect is to miss an optimization.
1512 (ii) If, despite (i), such a delivery is accidentally attempted, the
1513 routing retry time is respected, so at least it doesn't keep
1514 hammering the server.
1516 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1517 in ${readsocket because some servers need this prod.
1519 PH/32 Added some extra debug output when updating a wait-xxx database.
1521 PH/33 The hint "could be header name not terminated by colon", which has been
1522 given for certain expansion errors for a long time, was not being given
1523 for the ${if def:h_colon_omitted{... case.
1525 PH/34 The spec says: "With one important exception, whenever a domain list is
1526 being scanned, $domain contains the subject domain." There was at least
1527 one case where this was not true.
1529 PH/35 The error "getsockname() failed: connection reset by peer" was being
1530 written to the panic log as well as the main log, but it isn't really
1531 panic-worthy as it just means the connection died rather early on. I have
1532 removed the panic log writing for the ECONNRESET error when getsockname()
1535 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1536 runs only) independently of the message's sender address. This meant
1537 that, if the 4xx error was in fact related to the sender, a different
1538 message to the same recipient with a different sender could confuse
1539 things. In particualar, this can happen when sending to a greylisting
1540 server, but other circumstances could also provoke similar problems.
1541 I have changed the default so that the retry time for these errors is now
1542 based a combination of the sender and recipient addresses. This change
1543 can be overridden by setting address_retry_include_sender=false in the
1546 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1547 remote server are returned as part of bounce messages. This was not
1548 happening for LMTP over a pipe (the lmtp transport), but now it is the
1549 same for both kinds of LMTP.
1551 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1552 in header lines that were in fact CNAMEs. This is no longer the case.
1554 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1555 and queue runs started by the daemon processed all messages. This has
1556 been fixed so that -R and -S can now usefully be given with -q<time>.
1558 PH/40 Import PCRE release 6.7 (fixes some bugs).
1560 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1562 PH/42 Give an error if -q is specified more than once.
1564 PH/43 Renamed the variables $interface_address and $interface_port as
1565 $received_ip_address and $received_port, to make it clear that these
1566 values apply to message reception, and not to the outgoing interface when
1567 a message is delivered. (The old names remain recognized, of course.)
1569 PH/44 There was no timeout on the connect() call when using a Unix domain
1570 socket in the ${readsocket expansion. There now is.
1572 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1573 be meaningful with "accept".
1575 SC/01 Eximstats V1.43
1576 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1578 SC/02 Eximstats V1.44
1579 Use a glob alias rather than an array ref in the generated
1580 parser. This improves both readability and performance.
1582 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1583 Collect SpamAssassin and rejection statistics.
1584 Don't display local sender or destination tables unless
1585 there is data to show.
1586 Added average volumes into the top table text output.
1588 SC/04 Eximstats V1.46
1589 Collect data on the number of addresses (recipients)
1590 as well as the number of messages.
1592 SC/05 Eximstats V1.47
1593 Added 'Message too big' to the list of mail rejection
1594 reasons (thanks to Marco Gaiarin).
1596 SC/06 Eximstats V1.48
1597 Mainlog lines which have GMT offsets and are too short to
1598 have a flag are now skipped.
1600 SC/07 Eximstats V1.49 (Alain Williams)
1601 Added the -emptyok flag.
1603 SC/08 Eximstats V1.50
1604 Fixes for obtaining the IP address from reject messages.
1606 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1607 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1608 whitesspace changes from 4.64-PH/27
1610 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1613 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1614 are found, allow negative numbers in numeric criteria)
1616 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1618 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1619 to match changes made in 4.64-PH/43
1621 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1623 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1624 "domain" is actually an IP address, and force a failure. This locks out
1625 those revolvers/nameservers that support "A-for-A" lookups, in
1626 contravention of the specifications.
1628 PH/48 When a host name was looked up from an IP address, and the subsequent
1629 forward lookup of the name timed out, the host name was left in
1630 $sender_host_name, contrary to the specification.
1632 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1633 restricted to single-key lookups, Exim was not diagnosing an error if
1634 * or *@ was used with a query-style lookup.
1636 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1638 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1639 long ago, but noone interested enough thought of it. Let's just say that
1640 the "1.1" means that there are some new functions that weren't there at
1641 some point in the past.
1643 PH/51 Error processing for expansion failure of helo_data from an smtp
1644 transport during callout processing was broken.
1646 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1647 tested/used via the -bh/-bhc/-bs options.
1649 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1650 bug, fixed in subsequent PCRE releases).
1652 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1653 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1655 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1661 SC/01 Use a glob alias rather than an array ref in eximstats generated
1662 parser. This improves both readability and performance.
1664 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1665 Don't display local sender or destination tables in eximstats unless
1666 there is data to show.
1667 Added average volumes into the eximstats top table text output.
1669 SC/03 Collect data on the number of addresses (recipients) as well
1670 as the number of messages in eximstats.
1672 TF/01 Correct an error in the documentation for the redirect router. Exim
1673 does (usually) call initgroups() when daemonizing.
1675 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1676 with consistent privilege compared to when running as a daemon.
1678 TF/03 Note in the spec that $authenticated_id is not set for local
1679 submissions from trusted users.
1681 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1682 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1684 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1685 by adding some example configuration directives to the default
1686 configuration file. A little bit of work is required to uncomment the
1687 directives and define how usernames and passwords are checked, but
1688 there is now a framework to start from.
1690 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1691 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1692 without this. I don't know how relevant this is to other LDAP libraries.
1694 PH/02 Add the verb name to the "unknown ACL verb" error.
1696 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1698 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1700 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1701 directory not expanded when it should be if an expanded home directory
1702 was set for the address (which is overridden by the transport).
1704 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1707 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1708 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1709 because it is too late at that time, and has no effect.
1711 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1712 security issue with \' (bugzilla #107). I could not use the
1713 PQescapeStringConn() function, because it needs a PGconn value as one of
1716 PH/08 When testing addresses using -bt, indicate those final addresses that
1717 are duplicates that would not cause an additional delivery. At least one
1718 person was confused, thinking that -bt output corresponded to deliveries.
1719 (Suppressing duplicates isn't a good idea as you lose the information
1720 about possibly different redirections that led to the duplicates.)
1722 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1723 systems where poll() doesn't work, in particular OS X.
1725 PH/10 Added more information to debugging output for retry time not reached.
1727 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1728 operations in malware.c.
1730 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1733 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1734 syslog_duplication set false, log lines that would normally be written
1735 both the the main log and to the reject log were not written to syslog at
1738 PH/14 In the default configuration, change the use of "message" in ACL warn
1739 statements to "add_header".
1741 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1742 not followed by a command (e.g. "seen endif").
1744 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1745 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1748 PH/17 Added extra conditions to the default value of delay_warning_condition
1752 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1753 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1754 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1757 The Auto-Submitted: and various List- headers are standardised, whereas I
1758 don't think Precedence: ever was.
1760 PH/18 Refactored debugging code in route_finduser() to show more information,
1761 in particular, the error code if getpwnam() issues one.
1763 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1764 This is apparently needed in addition to the PH/07 change above to avoid
1765 any possible encoding problems.
1767 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1768 but not after initializing Perl.
1770 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1771 output them only if debugging. By default they are written stderr,
1772 apparently, which is not desirable.
1774 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1777 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1780 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1782 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1783 authenticated or an ident call has been made. Suppress the default
1784 values for $authenticated_id and $authenticated_sender (but permit -oMai
1785 and -oMas) when testing with -bh.
1787 PH/24 Re-jigged the order of the tests in the default configuration so that the
1788 tests for valid domains and recipients precede the DNS black list and CSA
1789 tests, on the grounds that those ones are more expensive.
1791 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1792 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1793 This bug exists in every version of Exim that I still have, right back to
1796 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1797 However, an attempt to turn on case-sensitivity in a regex key by
1798 including (?-i) didn't work because the subject string was already
1799 lowercased, and the effects were non-intuitive. It turns out that a
1800 one-line patch can be used to allow (?-i) to work as expected.
1806 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1807 other effects) broke the use of negated acl sub-conditions.
1809 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1812 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1813 "Deny" causes Exim to reject the incoming connection with a 554 error.
1814 Unfortunately, if there is a major crisis, such as a disk failure,
1815 tcp-wrappers gives "deny", whereas what one would like would be some
1816 kind of temporary error. A kludge has been added to help with this.
1817 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1818 554 error is used if errno is still zero or contains ENOENT (which occurs
1819 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1822 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1824 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1825 errors. Otherwise a message that provokes a temporary error (when other
1826 messages do not) can cause a whole host to time out.
1828 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1829 addresses were routed directly to files or pipes from a redirect router.
1830 File deliveries just didn't batch; pipe deliveries might have suffered
1833 PH/06 A failure to get a lock for a hints database would erroneously always say
1834 "Failed to get write lock", even when it was really a read lock.
1836 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1837 of 0600. This has been changed to use the value of the lockfile_mode
1838 option (which defaults to 0600).
1840 PH/08 Applied small patch from the Sieve maintainer.
1842 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1843 folder from quota calculations, a direct delivery into this folder messed
1844 up the contents of the maildirsize file. This was because the regex was
1845 used only to exclude .Trash (or whatever) when the size of the mailbox
1846 was calculated. There was no check that a delivery was happening into an
1847 excluded directory. This bug has been fixed by ignoring all quota
1848 processing for deliveries into excluded directories.
1850 PH/10 Added the maildirfolder_create_regex option to appendfile.
1856 PH/01 The code for finding all the local interface addresses on a FreeBSD
1857 system running IPv6 was broken. This may well have applied to all BSD
1858 systems, as well as to others that have similar system calls. The broken
1859 code found IPv4 interfaces correctly, but gave incorrect values for the
1860 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1861 that it would not match correctly against @[] and not recognize the IPv6
1864 PH/02 The ipliteral router was not recognizing addresses of the form user@
1865 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1867 PH/03 Added disable_ipv6.
1869 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1870 decoded form, because it is most often used to construct To: headers
1871 lines in autoreplies, and the decoded form may well be syntactically
1872 invalid. However, $reply_address has leading white space removed, and all
1873 newlines turned into spaces so that the autoreply transport does not
1876 PH/05 If group was specified without a user on a router, and no group or user
1877 was specified on a transport, the group from the router was ignored.
1879 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1880 for visible compile-time settings that can be used to change these
1881 numbers, for those that want even more. Backwards compatibility with old
1882 spool files has been maintained. However, going back to a previous Exim
1883 release will lost any variables that are in spool files.
1885 PH/07 Two small changes when running in the test harness: increase delay when
1886 passing a TCP/IP connection to a new process, in case the original
1887 process has to generate a bounce, and remove special handling of
1888 127.0.0.2 (sic), which is no longer necessary.
1890 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1891 be the same on different OS.
1893 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1896 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1897 whether --show-vars was specified or not
1899 JJ/02 exipick: Added support for new ACL variable spool format introduced
1902 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1903 syntactically invalid From: or Reply-to: line, and a filter used this to
1904 generate an autoreply, and therefore failed to obtain an address for the
1905 autoreply, Exim could try to deliver to a non-existent relative file
1906 name, causing unrelated and misleading errors. What now happens is that
1907 it logs this as a hard delivery error, but does not attempt to create a
1910 PH/11 The exinext utility has a -C option for testing purposes, but although
1911 the given file was scanned by exinext itself; it wasn't being passed on
1912 when Exim was called.
1914 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1915 an end-of-file indication when reading a command response.
1917 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1918 compiled. In many other places in Exim, IPv6 addresses are always
1919 recognized, so I have changed this. It also means that IPv4 domain
1920 literals of the form [IPV4:n.n.n.n] are now always recognized.
1922 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1923 used if the router is not running as root, for example, when verifying at
1924 ACL time, or when using -bh. The debugging output from this situation was
1925 non-existent - all you got was a failure to exec. I have made two
1928 (a) Failures to set uid/gid, the current directory, or a process leader
1929 in a subprocess such as that created by queryprogram now generate
1930 suitable debugging ouput when -d is set.
1932 (b) The queryprogram router detects when it is not running as root,
1933 outputs suitable debugging information if -d is set, and then runs
1934 the subprocess without attempting to change uid/gid.
1936 PH/15 Minor change to Makefile for building test_host (undocumented testing
1939 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1940 additional section of a DNS packet that returns MX or SRV records.
1941 Instead, it always explicitly searches for A/AAAA records. This avoids
1942 major problems that occur when a DNS server includes only records of one
1943 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1944 fixed another bug: if SRV records were looked up and the corresponding
1945 address records were *not* found in the additional section, the port
1946 values from the SRV records were lost.
1948 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1949 using the correct key (the original address) when searching the retry
1950 rules in order to find which one to use for generating the retry hint.
1952 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1953 adding the default one. Similarly, if it contains a Reply-To: header, the
1954 errors_reply_to option, if set, is not used.
1956 PH/19 When calculating a retry time, Exim used to measure the "time since
1957 failure" by looking at the "first failed" field in the retry record. Now
1958 it does not use this if it is later than than the arrival time of the
1959 message. Instead it uses the arrival time. This makes for better
1960 behaviour in cases where some deliveries succeed, thus re-setting the
1961 "first failed" field. An example is a quota failure for a huge message
1962 when small messages continue to be delivered. Without this change, the
1963 "time since failure" will always be short, possible causing more frequent
1964 delivery attempts for the huge message than are intended.
1965 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1967 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1968 $1, $2, $3) because the numerical variables can be reset during some
1969 expansion items (e.g. "match"), thereby losing the authentication data.
1971 PH/21 Make -bV show the size of off_t variables so that the test suite can
1972 decide whether to run tests for quotas > 2G.
1974 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1975 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1976 filecount value is greater than 2G or if a quota value is greater than 2G
1977 on a system where the size of off_t is not greater than 4, a panic error
1980 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1981 never match. The debug and -bh output now contains an explicit error
1982 message indicating a malformed IPv4 address or mask.
1984 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1985 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1986 PH/23 above applies.
1988 PH/25 Do not write to syslog when running in the test harness. The only
1989 occasion when this arises is a failure to open the main or panic logs
1990 (for which there is an explicit test).
1992 PH/26 Added the /no_tell option to "control=freeze".
1994 PH/27 If a host name lookup failed very early in a connection, for example, if
1995 the IP address matched host_lookup and the reverse lookup yielded a name
1996 that did not have a forward lookup, an error message of the form "no IP
1997 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1998 could be logged. Now it outputs the IP address instead of "NULL".
2000 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2001 allows the sender and the authenticated sender to be set when
2002 submitting a message from within Exim. Since child_open_exim() is
2003 documented for local_scan(), the new function should be too.
2005 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2006 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2007 results in an empty string is now treated as unset.
2009 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2011 PH/31 Added sender_verify_fail logging option.
2013 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2014 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2015 tidied the source and removed it altogether.
2017 PH/33 When a queue run was abandoned because the load average was too high, a
2018 log line was always written; now it is written only if the queue_run log
2019 selector is set. In addition, the log line for abandonment now contains
2020 information about the queue run such as the pid. This is always present
2021 in "start" and "stop" lines but was omitted from the "abandon" line.
2023 PH/34 Omit spaces between a header name and the colon in the error message that
2024 is given when verify = headers_syntax fails (if there are lots of them,
2025 the message gets confusing).
2027 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2028 names, as there are now some PTR records that contain slashes. This check
2029 is only to protect against broken name servers that fall over on strange
2030 characters, so the fact that it applies to all lookups doesn't matter.
2032 PH/36 Now that the new test suite is complete, we can remove some of the
2033 special code in Exim that was needed for the old test suite. For example,
2034 sorting DNS records because real resolvers return them in an arbitrary
2035 order. The new test suite's fake resolver always returns records in the
2038 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2039 bounces) except when queue_only is set, to avoid logging races between
2040 the different processes.
2042 PH/38 Panic-die if .include specifies a non-absolute path.
2044 PH/39 A tweak to the "H" retry rule from its user.
2046 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2047 a label. They prevented compilation on older perls.
2049 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2050 a warning to be raised on newish perls.
2052 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2053 on queue. Changes to match documented behaviour of showing count of
2054 messages matching specified criteria.
2056 PH/40 Changed the default ident timeout from 30s to 5s.
2058 PH/41 Added support for the use of login_cap features, on those BSD systems
2059 that have them, for controlling the resources used by pipe deliveries.
2061 PH/42 The content-scanning code uses fopen() to create files in which to put
2062 message data. Previously it was not paying any attention to the mode of
2063 the files. Exim runs with umask(0) because the rest of the code creates
2064 files with open(), and sets the required mode explicitly. Thus, these
2065 files were ending up world-writeable. This was not a big issue, because,
2066 being within the spool directory, they were not world-accessible. I have
2067 created a function called modefopen, which takes an additional mode
2068 argument. It sets umask(777), creates the file, chmods it to the required
2069 mode, then resets the umask. All the relevant calls to fopen() in the
2070 content scanning code have been changed to use this function.
2072 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2073 to 24 hours. This avoids potential overflow problems when processing G
2074 and H retry rules. I suspect nobody ever tinkers with this value.
2076 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2078 PH/45 When the plaintext authenticator is running as a client, the server's
2079 challenges are checked to ensure they are valid base64 strings. By
2080 default, the authentication attempt is cancelled if an invalid string is
2081 received. Setting client_ignore_invalid_base64 true ignores these errors.
2082 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2083 they are received. Thus, the responses can be made to depend on the
2084 challenges. If an invalid string is ignored, an empty string is placed in
2087 PH/46 Messages that are created by the autoreply transport now contains a
2088 References: header, in accordance with RFCs 2822 and 3834.
2090 PH/47 Added authenticated_sender_force to the smtp transport.
2092 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2094 PH/49 Installed latest patch from the Sieve maintainer.
2096 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2097 was also set, the appendfile transport was unnecessarily scanning a
2098 directory of message files (e.g. for maildir delivery) to find the count
2099 of files (along with the size), even though it did not need this
2100 information. It now does the scan only if it needs to find either the
2101 size of the count of files.
2103 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2105 PH/52 Two bugs concerned with error handling when the smtp transport is
2108 (i) Exim was not creating retry information for temporary errors given
2109 for individual recipients after the DATA command when the smtp transport
2110 was used in LMTP mode. This meant that they could be retried too
2111 frequently, and not timed out correctly.
2113 (ii) Exim was setting the flag that allows error details to be returned
2114 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2115 recipients that were returned after the DATA command.
2117 PH/53 This is related to PH/52, but is more general: for any failing address,
2118 when detailed error information was permitted to be returned to the
2119 sender, but the error was temporary, then after the final timeout, only
2120 "retry timeout exceeded" was returned. Now it returns the full error as
2121 well as "retry timeout exceeded".
2123 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2124 do this, and (what is worse) MTAs that accept it.
2126 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2127 will now be deprecated.
2129 PH/56 New os.c-cygwin from the Cygwin maintainer.
2131 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2132 formats (previously only available in exim formats via -bpr, -bpru,
2133 and -bpra. Now also available in native and exiqgrep formats)
2135 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2136 with very large, slow to parse queues
2138 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2140 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2142 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2143 responses to authentication challenges, though it was showing the
2144 challenges; (ii) I've removed the CR characters from the debug output for
2147 PH/58 Allow for the insertion of a newline as well as a space when a string
2148 is turned into more than one encoded-word during RFC 2047 encoding. The
2149 Sieve code now uses this.
2151 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2152 data_4xx, lost_connection, tls_required.
2154 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2155 message was being sent as an SMTP response.
2157 PH/61 Add -l and -k options to exicyclog.
2159 PH/62 When verifying, if an address was redirected to one new address, so that
2160 verification continued, and the new address failed or deferred after
2161 having set something in $address_data, the value of $address_data was not
2162 passed back to the ACL. This was different to the case when no
2163 redirection occurred. The value is now passed back in both cases.
2165 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2166 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2167 use login_cap.h, so on its own it isn't the distinguishing feature. The
2168 new name refers directly to the setclassresources() function.
2170 PH/65 Added configuration files for NetBSD3.
2172 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2174 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2175 is preferred over IPv4.
2177 PH/68 The bounce_return_message and bounce_return_body options were not being
2178 honoured for bounces generated during the reception of non-SMTP messages.
2179 In particular, this applied to messages rejected by the ACL. This bug has
2180 been fixed. However, if bounce_return_message is true and bounce_return_
2181 body is false, the headers that are returned for a non-SMTP message
2182 include only those that have been read before the error was detected.
2183 (In the case of an ACL rejection, they have all been read.)
2185 PH/69 The HTML version of the specification is now built in a directory called
2186 spec_html instead of spec.html, because the latter looks like a path with
2187 a MIME-type, and this confuses some software.
2189 PH/70 Catch two compiler warnings in sieve.c.
2191 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2192 function verify_get_ident() calls ip_connect() to connect a socket, but
2193 if the "connect()" function timed out, ip_connect() used to close the
2194 socket. However, verify_get_ident() also closes the socket later, and in
2195 between Exim writes to the log, which may get opened at this point. When
2196 the socket was closed in ip_connect(), the log could get the same file
2197 descriptor number as the socket. This naturally causes chaos. The fix is
2198 not to close the socket in ip_connect(); the socket should be closed by
2199 the function that creates it. There was only one place in the code where
2200 this was missing, in the iplookup router, which I don't think anybody now
2201 uses, but I've fixed it anyway.
2203 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2204 well as to direct DNS lookups. Otherwise the handling of names in host
2205 lists is inconsistent and therefore confusing.
2211 PH/01 Two changes to the default runtime configuration:
2213 (1) Move the checks for relay_from_hosts and authenticated clients from
2214 after to before the (commented out) DNS black list checks.
2216 (2) Add control=submission to the relay_from_hosts and authenticated
2217 clients checks, on the grounds that messages accepted by these
2218 statements are most likely to be submissions.
2220 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2222 (1) Generate an error if the third argument for the ${prvs expansion is
2225 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2228 (3) Reset the variables that are obtained from the first argument of
2229 ${prvscheck and used in the second argument before leaving the code,
2230 because their memory is reclaimed, so using them afterwards may do
2233 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2234 one (it's much easier than Tom thought :-).
2236 (5) Because of (4), we can now allow for the use of $prvscheck_result
2237 inside the third argument.
2239 PH/03 For some reason, the default setting of PATH when running a command from
2240 a pipe transport was just "/usr/bin". I have changed it to
2243 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2244 anything to be listed in the output from -bV.
2246 PH/05 When a filter generated an autoreply, the entire To: header line was
2247 quoted in the delivery log line, like this:
2249 => >A.N.Other <ano@some.domain> <original@ddress> ...
2251 This has been changed so that it extracts the operative address. There
2252 may be more than one such address. If so, they are comma-separated, like
2255 => >ano@some.domain,ona@other.domain <original@ddress> ...
2257 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2258 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2259 not being looked up in the rDNS to get a host name, Exim was showing the
2260 IP address twice in Received: lines, even though the IP addresses were
2261 identical. For example:
2263 Received: from [1.2.3.4] (helo=[1.2.3.4])
2265 However, if the real host name was known, it was omitting the HELO data
2266 if it matched the actual IP address. This has been tidied up so that it
2267 doesn't show the same IP address twice.
2269 PH/07 When both +timestamp and +memory debugging was on, the value given by
2270 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2271 called by the debug printing, thereby overwriting the timestamp buffer.
2272 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2274 PH/08 When the original message was included in an autoreply transport, it
2275 always said "this is a copy of the message, including all the headers",
2276 even if body_only or headers_only was set. It now gives an appropriate
2279 PH/09 Applied a patch from the Sieve maintainer which:
2281 o fixes some comments
2282 o adds the (disabled) notify extension core
2283 o adds some debug output for the result of if/elsif tests
2284 o points to the current vacation draft in the documentation
2285 and documents the missing references header update
2289 o fixes a bug in processing the envelope test (when testing
2290 multiple envelope elements, the last element determinted the
2293 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2294 Electronic Mail") by including:
2296 Auto-submitted: auto-generated
2298 in the messages that it generates (bounce messages and others, such as
2299 warnings). In the case of bounce messages for non-SMTP mesages, there was
2300 also a typo: it was using "Auto_submitted" (underscore instead of
2301 hyphen). Since every message generated by Exim is necessarily in response
2302 to another message, thes have all been changed to:
2304 Auto-Submitted: auto-replied
2306 in accordance with these statements in the RFC:
2308 The auto-replied keyword:
2310 - SHOULD be used on messages sent in direct response to another
2311 message by an automatic process,
2313 - MUST NOT be used on manually-generated messages,
2315 - MAY be used on Delivery Status Notifications (DSNs) and Message
2316 Disposition Notifications (MDNs),
2318 - MUST NOT be used on messages generated by automatic or periodic
2319 processes, except for messages which are automatic responses to
2322 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2323 to the default Received: header definition.
2325 PH/12 Added log selector acl_warn_skipped (default on).
2327 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2328 variables because (a) they are in the wrong storage pool and (b) even if
2329 they were copied, it wouldn't work properly because of the caching.
2331 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2332 checking when decoding. Apparently there are clients that generate
2333 overlong encoded strings. Why am I not surprised?
2335 PH/15 If the first argument of "${if match_address" was not empty, but did not
2336 contain an "@" character, Exim crashed. Now it writes a panic log message
2337 and treats the condition as false.
2339 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2341 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2342 extension "envelope-auth". The code is finished and in agreement with
2343 other implementations, but there is no documentation so far and in fact,
2344 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2345 not changing the active code.
2347 Print executed "if" and "elsif" statements when debugging is used. This
2348 helps a great deal to understand what a filter does.
2350 Document more things not specified clearly in RFC3028. I had all this
2351 sorted out, when out of a sudden new issues came to my mind. Oops."
2353 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2356 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2357 canonical form (as documented). However, after a host name lookup from
2358 the IP address, check_host() was doing a simple string comparison with
2359 addresses acquired from the DNS when checking that the found name did
2360 have the original IP as one of its addresses. Since any found IPv6
2361 addresses are likely to be in abbreviated form, the comparison could
2362 fail. Luckily, there already exists a function for doing the comparison
2363 by converting both addresses to binary, so now that is used instead of
2364 the text comparison.
2366 PH/20 There was another similar case to PH/19, when a complete host name was
2367 given in a host list; looking up its IP address could give an abbreviated
2368 form, whereas the current host's name might or might not be abbreviated.
2369 The same fix has been applied.
2375 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2376 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2379 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2380 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2382 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2384 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2385 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2386 Also, for "Auto-Submitted", treat anything other than "no" as
2387 non-personal, in accordance with RFC 3834. (Previously it treated
2388 anything starting "auto-" as non-personal.)
2390 TF/01 The control=submission/name=... option had a problem with syntax
2391 errors if the name included a slash character. The /name= option
2392 now slurps the rest of the string, so it can include any characters
2393 but it must come last in the list of options (after /sender_retain
2396 PH/05 Some modifications to the interface to the fake nameserver for the new
2404 TK/01 Added the "success_on_redirect" address verification option. See
2405 NewStuff for rationale and an example.
2407 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2409 PH/02 Patch to exigrep to allow it to work on syslog lines.
2411 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2412 fread() to skip over the body file's header line, because in Cygwin the
2413 header line is locked and is inaccessible.
2415 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2416 co-exist for some time) to make it clear that it is the Exim ID that is
2417 referenced, not the Message-ID: header line.
2419 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2420 string_format() function, because snprintf() does not exist on all
2423 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2424 ${stat: expansion item.
2426 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2427 protocol synchronization error", to keep the pedants happy.
2429 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2430 well as for IRIX systems, when gcc is being used. See the host.c source
2433 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2435 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2438 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2439 transport and to the smtp transport in LMTP mode.
2441 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2443 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2444 run a filter in a subprocess. This could lead to confusion in subsequent
2445 lookups in the parent process. There should also be a search_tidyup() at
2446 the end of the subprocess.
2448 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2449 only if the host matched helo_try_verify_hosts, which caused the
2450 verification to occur when the EHLO/HELO command was issued. The ACL just
2451 tested the remembered result. Now, if a previous verification attempt has
2452 not happened, "verify = helo" does it there and then.
2454 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2456 TK/03 Fix log output including CR from clamd.
2458 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2459 exist provoked a memory error which could cause a segfault.
2461 PH/15 Installed PCRE 6.2
2463 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2465 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2466 of the problem. Specifically, suggested +O2 rather than +O1 for the
2469 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2471 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2472 record, the port was not correctly logged when the outgoing_port log
2473 selector was set (it logged the transort's default port).
2475 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2476 fallback_hosts, and "hosts" in the smtp transport.
2478 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2479 host errors such as "Connection refused".
2481 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2482 authentication with radiusclient 0.4.9:
2484 - Error returned from rc_read_config was caught wrongly
2485 - Username/password not passed on to radius server due to wrong length.
2487 The presumption is that some radiusclient API changes for 4.51/PH/17
2488 were not taken care of correctly. The code is still untested by me (my
2489 Linux distribution still has 0.3.2 of radiusclient), but it was
2490 contributed by a Radius user.
2492 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2493 expanding the "port" option of the smtp transport.
2495 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2496 while reading a MIME header. Thanks to Tom Hughes for a patch.
2498 PH/24 Include config.h inside local_scan.h so that configuration settings are
2501 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2502 that in an ACL for RCPT (for example), you can examine exactly what was
2505 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2506 commands, but it was not correctly comparing the address with the actual
2507 client host address. Thus, it would show the EHLO address in Received:
2508 header lines when this was not necessary.
2510 PH/27 Added the % operator to ${eval:}.
2512 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2513 it should be ignoring failures (because with -C, for example, it has lost
2514 privilege). It wasn't ignoring creation failures other than "already
2517 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2520 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2521 because an input line was too long, either on its own, or by virtue of
2522 too many continuations, the temporary file was not being removed, and the
2523 return code was incorrect.
2525 PH/31 Missing "BOOL" in function definition in filtertest.c.
2527 PH/32 Applied Sieve patches from the maintainer.
2529 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2531 PH/33 Added "verify = not_blind".
2533 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2534 Local/Makefile (with some defaults set). These are used in built scripts
2535 such as exicyclog, but they have never been used in the exim_install
2536 script (though there are many overriding facilities there). I have
2537 arranged that the exim_install script now takes note of these two
2540 PH/35 Installed configuration files for Dragonfly.
2542 PH/36 When a locally submitted message by a trusted user did not contain a
2543 From: header, and the sender address was obtained from -f or from an SMTP
2544 MAIL command, and the trusted user did not use -F to supply a sender
2545 name, $originator_name was incorrectly used when constructing a From:
2546 header. Furthermore, $originator_name was used for submission mode
2547 messages from external hosts without From: headers in a similar way,
2548 which is clearly wrong.
2550 PH/37 Added control=suppress_local_fixups.
2552 PH/38 When log_selector = +received_sender was set, and the addition of the
2553 sender made the log line's construction buffer exactly full, or one byte
2554 less than full, an overflow happened when the terminating "\n" was
2557 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2558 when the result of a list match is failure because a DNS lookup failed.
2560 PH/40 RM_COMMAND is now used in the building process.
2562 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2563 the "build-* directories that it finds.
2565 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2566 address in a domain literal was a prefix of an interface address.
2568 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2569 when verifying a sender address, unless rewrite_headers is false.
2571 PH/44 Wrote a long comment about why errors_to addresses are verified as
2572 recipients, not senders.
2574 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2575 the ratelimit ACL was added.
2577 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2579 PH/46 Added extra information about PostgreSQL errors to the error string.
2581 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2582 avoiding the need to install special zones in a real server. This is
2583 backwards compatible; if it can't find the fake resolver, it drops back.
2584 Thus, both old and new test suites can be run.
2586 TF/02 Added util/ratelimit.pl
2588 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2589 clock is set back in time.
2591 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2592 Candler <B.Candler@pobox.com>.
2594 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2595 OK for addresses that are the result of redirections.
2597 PH/48 A number of further additions for the benefit of the new test suite,
2598 including a fake gethostbyname() that interfaces to the fake DNS resolver
2601 TF/06 The fix for widen_domains has also been applied to qualify_single and
2602 search_parents which are the other dnslookup options that can cause
2605 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2608 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2610 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2611 service" if it encountered an MX record with an empty target hostname.
2612 The message is now "an MX or SRV record indicated no SMTP service".
2614 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2615 if the DNS of the sending site is misconfigured. This is quite a
2616 common situation. This change restores the behaviour of treating a
2617 helo verification defer as a failure.
2619 PH/51 If self=fail was set on a router, the bounce message did not include the
2620 actual error message.
2626 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2628 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2629 command itself ended in error, the underlying message about the transport
2630 filter timeout was being overwritten with the pipe command error. Now the
2631 underlying error message should be appended to the second error message.
2633 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2635 PH/02 Reduce the amount of output that "make" produces by default. Full output
2636 can still be requested.
2638 PH/03 The warning log line about a condition test deferring for a "warn" verb
2639 was being output only once per connection, rather than after each
2640 occurrence (because it was using the same function as for successful
2641 "warn" verbs). This seems wrong, so I have changed it.
2643 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2644 it should not have, which might have caused a crash in the right
2645 circumstances, but probably never did.
2647 PH/04 Installed a modified version of Tony Finch's patch to make submission
2648 mode fix the return path as well as the Sender: header line, and to
2649 add a /name= option so that you can make the user's friendly name appear
2652 TF/03 Added the control = fakedefer ACL modifier.
2654 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2655 Mark Lowes for thorough testing.
2657 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2659 TK/03 Merged latest SRS patch from Miles Wilton.
2661 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2662 with the definition in sysexits.h (which is #included earlier).
2663 Fortunately, Exim does not actually use EX_OK. The code used to try to
2664 preserve the sysexits.h value, by assumimg that macro definitions were
2665 scanned for macro replacements. I have been disabused of this notion,
2666 so now the code just undefines EX_OK before #including unistd.h.
2668 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2669 in the smtp transport. When a block could not be written in a single
2670 write() function, the timeout was being re-applied to each part-write.
2671 This seems wrong - if the receiver was accepting one byte at a time it
2672 would take for ever. The timeout is now adjusted when this happens. It
2673 doesn't have to be particularly precise.
2675 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2676 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2678 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2679 without a domain if the check to <postmaster@domain> fails.
2681 SC/01 Eximstats: added -xls and the ability to specify output files
2682 (patch written by Frank Heydlauf).
2684 SC/02 Eximstats: use FileHandles for outputing results.
2686 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2688 SC/04 Eximstats: fixed display of large numbers with -nvr option
2690 SC/05 Eximstats: fixed merging of reports with empty tables.
2692 SC/06 Eximstats: added the -include_original_destination flag
2694 SC/07 Eximstats: removed tabs and trailing whitespace.
2696 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2698 TK/06 MBOX spool code: Add real "From " MBOX separator line
2699 so the .eml file is really in mbox format (even though
2700 most programs do not really care). Patch from Alex Miller.
2702 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2703 The latter is generated from $received_to and is only set if the
2704 message has one envelope recipient. SA can use these headers,
2705 obviously out-of-the-box. Patch from Alex Miller.
2707 PH/08 The ${def test on a variable was returning false if the variable's
2708 value was "0", contrary to what the specification has always said!
2709 The result should be true unless the variable is empty.
2711 PH/09 The syntax error of a character other than { following "${if
2712 def:variable_name" (after optional whitespace) was not being diagnosed.
2713 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2714 accidental colon was present, for example, could give incorrect results.
2716 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2717 result is used (not including appendfile, where other changes are about
2720 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2721 This involved changing a lot of size variables from int to off_t. It
2722 should work with maildirs and everything.
2724 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2725 spamd dying while we are connected to it.
2727 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2730 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2731 function for BDB 4.3.
2733 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2735 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2736 the sources for PCRE have been re-arranged and more files are now
2739 PH/15 The code I had for printing potentially long long variables in PH/11
2740 above was not the best (it lost precision). The length of off_t variables
2741 is now inspected at build time, and an appropriate printing format (%ld
2742 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2743 to be "long long int" or "long int". This is needed for the internal
2744 formatting function string_vformat().
2746 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2747 the configuration file to be ":syslog", then the script "guesses" where
2748 the logs files are, rather than using the compiled in default. In our
2749 case the guess is not the same as the compiled default, so the script
2750 suddenly stopped working when I started to use syslog. The patch checks
2751 to see if log_file_path is "". If so, it attempts to read it from exim
2752 with no configuration file to get the compiled in version, before it
2753 falls back to the previous guessing code."
2755 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2756 implementing BATV in an Exim configuration. See NewStuff for the gory
2759 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2760 Makefile that are specific to HP-UX.
2762 PH/18 If the "use_postmaster" option was set for a recipient callout together
2763 with the "random" option, the postmaster address was used as the MAIL
2764 FROM address for the random test, but not for the subsequent recipient
2765 test. It is now used for both.
2767 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2768 patch removes a few documentation additions to RFC 3028, because the
2769 latest draft now contains them. It adds the new en;ascii-case comparator
2770 and a new error check for 8bit text in MIME parts. Comparator and
2771 require names are now matched exactly. I enabled the subaddress
2772 extension, but it is not well tested yet (read: it works for me)."
2774 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2775 rework some of the code of TK/09 above to avoid the hardwired use of
2776 "%lld" and "long long". Replaced the call to snprintf() with a call to
2779 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2780 records point to non-existent hosts", "retry timeout exceeded", and
2781 "retry time not reached for any host after a long failure period".
2783 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2784 experimental DomainKeys support:
2786 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2787 (2) On an error such as an illegally used "control", the wrong name for
2788 the control was given.
2790 These problems did NOT occur unless DomainKeys support was compiled.
2792 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2794 PH/24 Added ${if match_ip condition.
2796 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2797 databases so that it will be absolutely obvious if a crash occurs in the
2798 DB library. This is a regular occurrence (often caused by mis-matched
2801 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2802 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2803 picked up on a user's system that detects such things. There doesn't seem
2804 to be a gcc warning option for this - only an attribute that has to be
2805 put on the function's prototype. It seems that in Fedora Core 4 they have
2806 set this on a number of new functions. No doubt there will be more in due
2809 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2810 specify a transport. However, if an address that was verified by such a
2811 router was the subject of a callout, Exim crashed because it tried to
2812 read the rcpt_include_affixes from the non-existent transport. Now it
2813 just assumes that the setting of that option is false. This bug was
2814 introduced by 4.51/PH/31.
2816 PH/28 Changed -d+all to exclude +memory, because that information is very
2817 rarely of interest, but it makes the output a lot bigger. People tend to
2818 do -d+all out of habit.
2820 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2821 code in os-type was giving problems when libc.so lives in lib64, like on
2824 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2825 aren't the modern standard, and it seems that some systems' include files
2826 don't always have them. Exim was already checking for some of the newer
2827 ones like T_AAAA, and defining it itself. I've added checks for all the
2828 record types that Exim uses.
2830 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2831 not automatically generating a new one, as it is supposed to. This
2832 prevented TLS from working. If the file did exist, but contained invalid
2833 data, a new version was generated, as expected. It was only the case of a
2834 non-existent file that was broken.
2836 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2837 with a change in libdomainkeys > 0.64.
2839 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2840 from DNS. If the selector record carries the flag, it now has
2841 precedence over the domain-wide flag.
2843 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2845 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2846 the use of a port name that isn't defined in /etc/services) occurred, the
2847 message was deferred as in a normal delivery, and thus remained on the
2848 spool, instead of being failed because of the mua_wrapper setting. This
2849 is now fixed, and I tidied up some of the mua_wrapper messages at the
2852 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2853 the messages in a hash of arrays rather than using individual hashes.
2854 This is a bit cleaner and results in dramatic memory savings, albeit
2855 at a slight CPU cost.
2857 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2858 as requested by Marc Sherman.
2860 SC/10 Eximstats: added histograms for user specified patterns as requested
2863 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2865 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2866 fopen() in the content-scanning modules that did not already have it.
2872 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2873 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2875 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2877 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2879 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2880 to test Sieve filters that use "vacation".
2882 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2883 that changes the way the GnuTLS parameters are stored in the cache file.
2884 The new format can be generated externally. For backward compatibility,
2885 if the data in the cache doesn't make sense, Exim assumes it has read an
2886 old-format file, and it generates new data and writes a new file. This
2887 means that you can't go back to an older release without removing the
2890 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2891 work if there are any delivery delays because "one_time" forces the
2892 parent to be marked "delivered", so its unseen clone is never tried
2893 again. For this reason, Exim now forbids the simultaneous setting of
2896 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2897 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2898 entry if you want to know the details. The fix, however, neglected to
2899 consider the case when local delivery batching is involved. The test for
2900 "previously delivered" was not happening when checking to see if an
2901 address could be batched with a previous (undelivered) one; under
2902 certain circumstances this could lead to multiple deliveries to the same
2905 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2906 in its include files, and this causes problems building Exim.
2908 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2909 header_syntax) but Exim was just ignoring anything given after a slash.
2910 In particular, this caused confusion with an attempt to use "verify =
2911 reverse_host_lookup/defer_ok". An error is now given when options are
2912 supplied for verify items that do not have them. (Maybe reverse_host_
2913 lookup should have a defer_ok option, but that's a different point.)
2915 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2916 defined by RFC 821) to 2048, because there were problems with some AUTH
2917 commands, and RFC 1869 says the size should be increased for extended
2918 SMTP commands that take arguments.
2920 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2923 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2924 "unknown" error; now it says that the functionality isn't in the binary.
2926 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2927 an address' error message when a string expansion fails (syntax or
2928 whatever). Otherwise the password may appear in the log. Following change
2929 PH/42 below, there is no longer a chance of it appearing in a bounce
2932 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2934 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2935 of its addresses was ever tried. (Bugzilla bug #2.)
2937 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2938 the result incorrectly in the debug output. (It correctly added a newline
2939 to what was transported.)
2941 TF/01 Added $received_time.
2943 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2944 commented out examples of how to interface to a virus scanner and to
2945 SpamAssassin. Also added commented examples of av_scanner and
2946 spamd_address settings.
2948 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2949 and controls are allowed in which ACLs. There were a couple of minor
2950 errors. Some of the entries in the conditions table (which is a table of
2951 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2952 negation of where the condition IS allowed.
2954 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2956 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2957 header file does not have a version number, so I've had to invent a new
2958 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2959 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2960 radiusclient), but it was contributed by a Radius user.
2962 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2963 files or directories, for OpenSSL.
2965 PH/19 When an Exim process that is running as root has to create an Exim log
2966 file, it does so in a subprocess that runs as exim:exim so as to get the
2967 ownership right at creation (otherwise, other Exim processes might see
2968 the file with the wrong ownership). There was no test for failure of this
2969 fork() call, which would lead to the process getting stuck as it waited
2970 for a non-existent subprocess. Forks do occasionally fail when resources
2971 run out. I reviewed all the other calls to fork(); they all seem to check
2974 PH/20 When checking for unexpected SMTP input at connect time (before writing
2975 the banner), Exim was not dealing correctly with a non-positive return
2976 from the read() function. If the client had disconnected by this time,
2977 the result was a log entry for a synchronization error with an empty
2978 string after "input=" when read() returned zero. If read() returned -1
2979 (an event I could not check), uninitialized data bytes were printed.
2980 There were reports of junk text (parts of files, etc) appearing after
2983 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2985 PH/22 Added support for macro redefinition, and (re)definition in between
2986 driver and ACL definitions.
2988 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2989 forgetting to use the resulting value; it was using the unexpanded value.
2991 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2992 hadn't been configured. The fix is from Juergen Kreileder, who
2993 understands it better than I do:
2995 "Here's what I see happening with three configured cyrus_sasl
2996 authenticators configured (plain, login, cram-md5):
2998 On startup auth_cyrus_sasl_init() gets called for each of these.
2999 This means three calls to sasl_listmech() without a specified mech_list.
3000 => SASL tests which mechs of all available mechs actually work
3001 => three warnings about OTP not working
3002 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3004 With the patch, sasl_listmech() also gets called three times. But now
3005 SASL's mech_list option is set to the server_mech specified in the the
3006 authenticator. Or in other words, the answer from sasl_listmech()
3007 gets limited to just the mech you're testing for (which is different
3009 => the return list contains just 'plain' or 'login', 'cram-md5' or
3010 nothing depending on the value of ob->server_mech.
3012 I've just tested the patch: Authentication still works fine,
3013 unavailable mechs specified in the exim configuration are still
3014 caught, and the auth.log warnings about OTP are gone."
3016 PH/25 When debugging is enabled, the contents of the command line are added
3017 to the debugging output, even when log_selector=+arguments is not
3020 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3021 answer is "GNU", and only if the return is "GNU/something" is the answer
3024 PH/27 $acl_verify_message is now set immediately after the failure of a
3025 verification in an ACL, and so is available in subsequent modifiers. In
3026 particular, the message can be preserved by coding like this:
3028 warn !verify = sender
3029 set acl_m0 = $acl_verify_message
3031 Previously, $acl_verify_message was set only while expanding "message"
3032 and "log_message" when a very denied access.
3034 PH/28 Modified OS/os.c-Linux with
3036 -#ifndef OS_LOAD_AVERAGE
3037 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3039 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3040 nomenclature these days.)
3042 PH/29 Installed patch from the Sieve maintainer that adds the options
3043 sieve_useraddress and sieve_subaddress to the redirect router.
3045 PH/30 In these circumstances:
3046 . Two addresses routed to the same list of hosts;
3047 . First host does not offer TLS;
3048 . First host accepts first address;
3049 . First host gives temporary error to second address;
3050 . Second host offers TLS and a TLS session is established;
3051 . Second host accepts second address.
3052 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3053 and peerdn, if requested) that were in fact used only for the second
3056 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3057 attention to any local part prefix or suffix that was matched by the
3058 router that accepted the address. It now behaves in the same way as it
3059 does for delivery: the affixes are removed from the local part unless
3060 rcpt_include_affixes is set on the transport.
3062 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3063 timeout during the DATA phase of an incoming message.
3065 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3066 applied a patch sanctioned by the Sieve maintainer.
3068 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3069 the uid or gid is negative. A case of a negative gid caused this to be
3070 noticed. The fix allows for either to be negative.
3072 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3073 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3076 PH/36 The change PH/12 above was broken. Fixed it.
3078 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3079 the grounds that routing the same address twice would always produce the
3080 same answer. This might have been true once, but it is certainly no
3081 longer true now. Routing a child address may depend on the previous
3082 routing that produced that child. Some complicated redirection strategies
3083 went wrong when messages had multiple recipients, and made Exim's
3084 behaviour dependent on the order in which the addresses were given.
3086 I have moved the duplicate checking until after the routing is complete.
3087 Exim scans the addresses that are assigned to local and remote
3088 transports, and removes any duplicates. This means that more work will be
3089 done, as duplicates will always all be routed, but duplicates are
3090 presumably rare, so I don't expect this is of any significance.
3092 For deliveries to pipes, files, and autoreplies, the duplicate checking
3093 still happens during the routing process, since they are not going to be
3096 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3097 It corrects a timeout issue with spamd. This is Ian's comment: "The
3098 background is that sometimes spamd either never reads data from a
3099 connection it has accepted, or it never writes response data. The exiscan
3100 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3101 blindly assumes that writes won't block so it may never time out."
3103 PH/39 Allow G after quota size as well as K and M.
3105 PH/40 The value set for $authenticated_id in an authenticator may not contain
3106 binary zeroes or newlines because the value is written to log lines and
3107 to spool files. There was no check on this. Now the value is run through
3108 the string_printing() function so that such characters are converted to
3109 printable escape sequences.
3111 PH/41 $message_linecount is a new variable that contains the total number of
3112 lines in the message. Compare $body_linecount, which is the count for the
3115 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3116 bounce and delay warning messages, except in certain special cases, which
3119 (a) An SMTP error message from a remote host;
3120 (b) A message specified in a :fail: redirection;
3121 (c) A message specified in a "fail" command in a system filter;
3122 (d) A message specified in a FAIL return from the queryprogram router;
3123 (e) A message specified by the cannot_route_message router option.
3125 In these cases only, Exim does include the error details in bounce and
3126 warning messages. There are also a few cases where bland messages such
3127 as "unrouteable address" or "local delivery error" are given.
3129 PH/43 $value is now also set for the "else" part of a ${run expansion.
3131 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3132 being worked on, but at least Exim now implements the latest version to
3135 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3136 process to complete was treated as a delivery failure, a timeout while
3137 writing the message to the pipe was logged, but erroneously treated as a
3138 successful delivery. Such timeouts include transport filter timeouts. For
3139 consistency with the overall process timeout, these timeouts are now
3140 treated as errors, giving rise to delivery failures by default. However,
3141 there is now a new Boolean option for the pipe transport called
3142 timeout_defer, which, if set TRUE, converts the failures into defers for
3143 both kinds of timeout. A transport filter timeout is now identified in
3146 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3147 systems where "make" and "gmake" are different, calling "gmake" at top
3148 level broke things. I've arranged for the value of $(MAKE) to be passed
3149 from the Makefile to this script so that it can call the same version of
3153 A note about Exim versions 4.44 and 4.50
3154 ----------------------------------------
3156 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3157 changes of various kinds. As a consequence, a big documentation update was
3158 needed. This delayed the release for rather longer than seemed good, especially
3159 in the light of a couple of (minor) security issues. Therefore, the changes
3160 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3161 release. So 4.44 and 4.50 are in effect two different branches that both start
3164 I have left the 4.50 change log unchanged; it contains all the changes since
3165 4.43. The change log for 4.44 is below; many of its items are identical to
3166 those for 4.50. This seems to be the most sensible way to preserve the
3167 historical information.
3173 1. Minor wording change to the doc/README.SIEVE file.
3175 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3176 computation of the current number of files was incorrect.
3178 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3179 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3182 4. Give more explanation in the error message when the command for a transport
3183 filter fails to execute.
3185 5. There are several places where Exim runs a non-Exim command in a
3186 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3187 was being done only for the command run by the queryprogram router. It is
3188 now done for all such subprocesses. The other cases are: ${run, transport
3189 filters, and the commands run by the lmtp and pipe transports.
3191 6. Added CONFIGURE_GROUP build-time option.
3193 7. Some older OS have a limit of 256 on the maximum number of file
3194 descriptors. Exim was using setrlimit() to set 1000 as a large value
3195 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3196 systems. I've change it so that if it can't get 1000, it tries for 256.
3198 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3199 was an oversight, and furthermore, ever since the addition of extra
3200 controls (e.g. 4.43/32), the checks on when to allow different forms of
3201 "control" were broken. There should now be diagnostics for all cases when a
3202 control that does not make sense is encountered.
3204 9. Added the /retain_sender option to "control=submission".
3206 10. $recipients is now available in the predata ACL (oversight).
3208 11. Tidy the search cache before the fork to do a delivery from a message
3209 received from the command line. Otherwise the child will trigger a lookup
3210 failure and thereby defer the delivery if it tries to use (for example) a
3211 cached ldap connection that the parent has called unbind on.
3213 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3214 of $address_data from the recipient verification was clobbered by the
3215 sender verification.
3217 13. The value of address_data from a sender verification is now available in
3218 $sender_address_data in subsequent conditions in the ACL statement.
3220 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3222 15. Added a new option "connect=<time>" to callout options, to set a different
3225 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3226 was its contents. (It was OK if the option was not defined at all.)
3228 17. A "Completed" log line is now written for messages that are removed from
3229 the spool by the -Mrm option.
3231 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3232 information about exactly what failed.
3234 19. Added -dd to debug only the daemon process.
3236 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3237 handles timeouts, both on the server side and network timeouts. Renamed the
3238 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3240 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3241 It is now set to "smtps".
3243 22. $host_address is now set to the target address during the checking of
3244 ignore_target_hosts.
3246 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3247 being passed; this would have caused $sender_host_name to have been used if
3248 matching the list had actually called for a host name (not very likely,
3249 since this list is usually IP addresses). A host name is now passed as
3252 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3253 code that specifies a non-restarting handler (typically sigaction() in
3254 modern systems) in an attempt to fix a rare and obscure crash bug.
3256 25. Narrowed the window for a race in the daemon that could cause it to ignore
3257 SIGCHLD signals. This is not a major problem, because they are used only to
3258 wake it up if nothing else does.
3260 26. A malformed maildirsize file could cause Exim to calculate negative values
3261 for the mailbox size or file count. Odd effects could occur as a result.
3262 The maildirsize information is now recalculated if the size or filecount
3265 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3266 support for a long time. Removed HAVE_SYS_VFS_H.
3268 28. Installed the latest version of exipick from John Jetmore.
3270 29. In an address list, if the pattern was not a regular expression, an empty
3271 subject address (from a bounce message) matched only if the pattern was an
3272 empty string. Non-empty patterns were not even tested. This was the wrong
3273 because it is perfectly reasonable to use an empty address as part of a
3274 database query. An empty address is now tested by patterns that are
3275 lookups. However, all the other forms of pattern expect the subject to
3276 contain a local part and a domain, and therefore, for them, an empty
3277 address still always fails if the pattern is not itself empty.
3279 30. Exim went into a mad DNS loop when attempting to do a callout where the
3280 host was specified on an smtp transport, and looking it up yielded more
3281 than one IP address.
3283 31. Re-factored the code for checking spool and log partition space into a
3284 function that finds that data and another that does the check. The former
3285 is then used to implement four new variables: $spool_space, $log_space,
3286 $spool_inodes, and $log_inodes.
3288 32. The RFC2047 encoding function was originally intended for short strings
3289 such as real names; it was not keeping to the 75-character limit for
3290 encoded words that the RFC imposes. It now respects the limit, and
3291 generates multiple encoded words if necessary. To be on the safe side, I
3292 have increased the buffer size for the ${rfc2047: expansion operator from
3295 33. It is now permitted to omit both strings after an "if" condition; if the
3296 condition is true, the result is "true". As before, when the second string
3297 is omitted, a false condition yields an empty string. This makes it less
3298 cumbersome to write custom ACL and router conditions.
3300 34. Failure to deliver a bounce message always caused it to be frozen, even if
3301 there was an errors_to setting on the router. The errors_to setting is now
3304 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3305 canonical form (fully expanded) before being placed in
3306 $sender_host_address.
3308 36. The table in the code that translates DNS record types into text (T_A to
3309 "A" for instance) was missing entries for NS and CNAME. It is just possible
3310 that this could have caused confusion if both these types were looked up
3311 for the same domain, because the text type is used as part of Exim's
3312 per-process caching. But the chance of anyone hitting this buglet seems
3315 37. The dnsdb lookup has been extended in a number of ways.
3317 (1) There is a new type, "zns", which walks up the domain tree until it
3318 finds some nameserver records. It should be used with care.
3320 (2) There is a new type, "mxh", which is like "mx" except that it returns
3321 just the host names, not the priorities.
3323 (3) It is now possible to give a list of domains (or IP addresses) to be
3324 looked up. The behaviour when one of the lookups defers can be
3325 controlled by a keyword.
3327 (4) It is now possible to specify the separator character for use when
3328 multiple records are returned.
3330 38. The dnslists ACL condition has been extended: it is now possible to supply
3331 a list of IP addresses and/or domains to be looked up in a particular DNS
3334 39. Added log_selector=+queue_time_overall.
3336 40. When running the queue in the test harness, wait just a tad after forking a
3337 delivery process, to get repeatability of debugging output.
3339 41. Include certificate and key file names in error message when GnuTLS fails
3340 to set them up, because the GnuTLS error message doesn't include the name
3341 of the failing file when there is a problem reading it.
3343 42. Allow both -bf and -bF in the same test run.
3345 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3347 44. The "Exiscan patch" is now merged into the mainline Exim source.
3349 45. Sometimes the final signoff response after QUIT could fail to get
3350 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3351 < 0 before doing a fflush(). This bug looks as though it goes back to the
3352 introduction of TLS in release 3.20, but "sometimes" must have been rare
3353 because the tests only now provoked it.
3355 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3356 (this can affect the format of dates).
3358 47. exim_tidydb, when checking for the continued existence of a message for
3359 which it has found a message-specific retry record, was not finding
3360 messages that were in split spool directories. Consequently, it was
3361 deleting retry records that should have stayed in existence.
3363 48. Steve fixed some bugs in eximstats.
3365 49. The SPA authentication driver was not abandoning authentication and moving
3366 on to the next authenticator when an expansion was forced to fail,
3367 contradicting the general specification for all authenticators. Instead it
3368 was generating a temporary error. It now behaves as specified.
3370 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3371 (the order specifies the preference for clients). The order is now AES256,
3372 AES128, 3DES, ARCFOUR128.
3374 51. Small patch to Sieve code - explicitly set From: when generating an
3377 52. Exim crashed if a remote delivery caused a very long error message to be
3378 recorded - for instance if somebody sent an entire SpamAssassin report back
3379 as a large number of 550 error lines. This bug was coincidentally fixed by
3380 increasing the size of one of Exim's internal buffers (big_buffer) that
3381 happened as part of the Exiscan merge. However, to be on the safe side, I
3382 have made the code more robust (and fixed the comments that describe what
3385 53. Now that there can be additional text after "Completed" in log lines (if
3386 the queue_time_overall log selector is set), a one-byte patch to exigrep
3387 was needed to allow it to recognize "Completed" as not the last thing in
3390 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3391 patch that reportedly fixes this has been added. I am not expert enough to
3392 create a test for it. This is what the patch creator wrote:
3394 "I found a little strange behaviour of ldap code when working with
3395 Windows 2003 AD Domain, where users was placed in more than one
3396 Organization Units. When I tried to give exim partial DN, the exit code
3397 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3398 But simultaneously result of request was absolutely normal ldap result,
3399 so I produce this patch..."
3401 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3402 REFERENCE, so I have modified the code to exclude the patch when that macro
3405 55. Some experimental protocols are using DNS PTR records for new purposes. The
3406 keys for these records are domain names, not reversed IP addresses. The
3407 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3408 leaves it alone. Component reversal etc. now happens only for IP addresses.
3411 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3413 57. Double the size of the debug message buffer (to 2048) so that more of very
3414 long debug lines gets shown.
3416 58. The exicyclog utility now does better if the number of log files to keep
3417 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3419 59. Two changes related to the smtp_active_hostname option:
3421 (1) $smtp_active_hostname is now available as a variable.
3422 (2) The default for smtp_banner uses $smtp_active_hostname instead
3423 of $primary_hostname.
3425 60. The host_aton() function is supposed to be passed a string that is known
3426 to be a valid IP address. However, in the case of IPv6 addresses, it was
3427 not checking this. This is a hostage to fortune. Exim now panics and dies
3428 if the condition is not met. A case was found where this could be provoked
3429 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3430 components; fortuitously, this particular loophole had already been fixed
3431 by change 4.50/55 above.
3433 If there are any other similar loopholes, the new check in host_aton()
3434 itself should stop them being exploited. The report I received stated that
3435 data on the command line could provoke the exploit when Exim was running as
3436 exim, but did not say which command line option was involved. All I could
3437 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3438 running as the user.
3441 61. There was a buffer overflow vulnerability in the SPA authentication code
3442 (which came originally from the Samba project). I have added a test to the
3443 spa_base64_to_bits() function which I hope fixes it.
3446 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3447 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3449 63. The daemon start-up calls getloadavg() while still root for those OS that
3450 need the first call to be done as root, but it missed one case: when
3451 deliver_queue_load_max is set with deliver_drop_privilege. This is
3452 necessary for the benefit of the queue runner, because there is no re-exec
3453 when deliver_drop_privilege is set.
3455 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3456 This has been fixed.
3458 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3459 was in use, was not putting the data itself into the right store pool;
3460 consequently, it could be overwritten for a subsequent message in the same
3461 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3464 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3466 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3467 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3468 the function were treating the return as a boolean value, which happened to
3469 work because 0=false and not-0=true, but is not correct code.
3471 68. The host_aton() function was not handling scoped IPv6 addresses (those
3472 with, for example, "%eth0" on the end) correctly.
3474 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3475 negated items (that is, ~something) in unsigned ints. Some compilers
3476 apparently mutter when there is no cast.
3478 70. If an address verification called from an ACL failed, and did not produce a
3479 user-specific message (i.e. there was only a "system" message), nothing was
3480 put in $acl_verify_message. In this situation, it now puts the system
3483 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3484 session; change 4.31/43 added the unwanted input to the log line - except
3485 that it did not do this in the start of session case. It now does.
3487 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3488 This is wrong and can cause the other end to generate a synchronization
3489 error if it is another Exim or anything else that does the synchronization
3490 check. A QUIT command is no longer sent after a timeout.
3492 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3493 during host lookups.
3495 74. The defer_ok option of callout verification was not working if it was used
3496 when verifying addresses in header lines, that is, for this case:
3498 verify = header_sender/callout=defer_ok
3500 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3501 those file descriptors could be used for SMTP connections. If anything
3502 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3503 could be sent to the SMTP client, causing chaos. The daemon now opens
3504 stdin, stdout, and stderr to /dev/null when it puts itself into the
3507 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3508 log by default. The user can override this with suitable Perl magic.
3510 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3511 the log message when discard triggers, was not working for the DATA ACL or
3512 for the non-SMTP ACL.
3514 78. Error message wording change in sieve.c.
3516 79. If smtp_accept_max_per_host was set, the number of connections could be
3517 restricted to fewer than expected, because the daemon was trying to set up
3518 a new connection before checking whether the processes handling previous
3519 connections had finished. The check for completed processes is now done
3520 earlier. On busy systems, this bug wouldn't be noticed because something
3521 else would have woken the daemon, and it would have reaped the completed
3524 80. If a message was submitted locally by a user whose login name contained one
3525 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3526 It caused a spool format error. I have fixed the spool reading code. A
3527 related problem was that the "from" clause in the Received: line became
3528 illegal because of the space(s). It is now covered by ${quote_local_part.
3530 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3533 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3535 83. Patch from Sieve maintainer to fix unterminated string problem in
3536 "vacation" handling.
3538 84. Some minor changes to the Linux configuration files to help with other
3539 OS variants using glibc.
3541 85. One more patch for Sieve to update vacation handling to latest spec.
3544 ----------------------------------------------------
3545 See the note above about the 4.44 and 4.50 releases.
3546 ----------------------------------------------------
3552 1. Change 4.43/35 introduced a bug that caused file counts to be
3553 incorrectly computed when quota_filecount was set in an appendfile
3556 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3557 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3560 3. Give more explanation in the error message when the command for a transport
3561 filter fails to execute.
3563 4. There are several places where Exim runs a non-Exim command in a
3564 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3565 was being done only for the command run by the queryprogram router. It is
3566 now done for all such subprocesses. The other cases are: ${run, transport
3567 filters, and the commands run by the lmtp and pipe transports.
3569 5. Some older OS have a limit of 256 on the maximum number of file
3570 descriptors. Exim was using setrlimit() to set 1000 as a large value
3571 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3572 systems. I've change it so that if it can't get 1000, it tries for 256.
3574 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3575 was an oversight, and furthermore, ever since the addition of extra
3576 controls (e.g. 4.43/32), the checks on when to allow different forms of
3577 "control" were broken. There should now be diagnostics for all cases when a
3578 control that does not make sense is encountered.
3580 7. $recipients is now available in the predata ACL (oversight).
3582 8. Tidy the search cache before the fork to do a delivery from a message
3583 received from the command line. Otherwise the child will trigger a lookup
3584 failure and thereby defer the delivery if it tries to use (for example) a
3585 cached ldap connection that the parent has called unbind on.
3587 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3588 of $address_data from the recipient verification was clobbered by the
3589 sender verification.
3591 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3592 was its contents. (It was OK if the option was not defined at all.)
3594 11. A "Completed" log line is now written for messages that are removed from
3595 the spool by the -Mrm option.
3597 12. $host_address is now set to the target address during the checking of
3598 ignore_target_hosts.
3600 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3601 being passed; this would have caused $sender_host_name to have been used if
3602 matching the list had actually called for a host name (not very likely,
3603 since this list is usually IP addresses). A host name is now passed as
3606 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3607 code that specifies a non-restarting handler (typically sigaction() in
3608 modern systems) in an attempt to fix a rare and obscure crash bug.
3610 15. Narrowed the window for a race in the daemon that could cause it to ignore
3611 SIGCHLD signals. This is not a major problem, because they are used only to
3612 wake it up if nothing else does.
3614 16. A malformed maildirsize file could cause Exim to calculate negative values
3615 for the mailbox size or file count. Odd effects could occur as a result.
3616 The maildirsize information is now recalculated if the size or filecount
3619 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3620 support for a long time. Removed HAVE_SYS_VFS_H.
3622 18. Updated exipick to current release from John Jetmore.
3624 19. Allow an empty sender to be matched against a lookup in an address list.
3625 Previously the only cases considered were a regular expression, or an
3628 20. Exim went into a mad DNS lookup loop when doing a callout where the
3629 host was specified on the transport, if the DNS lookup yielded more than
3632 21. The RFC2047 encoding function was originally intended for short strings
3633 such as real names; it was not keeping to the 75-character limit for
3634 encoded words that the RFC imposes. It now respects the limit, and
3635 generates multiple encoded words if necessary. To be on the safe side, I
3636 have increased the buffer size for the ${rfc2047: expansion operator from
3639 22. Failure to deliver a bounce message always caused it to be frozen, even if
3640 there was an errors_to setting on the router. The errors_to setting is now
3643 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3644 canonical form (fully expanded) before being placed in
3645 $sender_host_address.
3647 24. Updated eximstats to version 1.33
3649 25. Include certificate and key file names in error message when GnuTLS fails
3650 to set them up, because the GnuTLS error message doesn't include the name
3651 of the failing file when there is a problem reading it.
3653 26. Expand error message when OpenSSL has problems setting up cert/key files.
3656 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3657 (this can affect the format of dates).
3659 28. exim_tidydb, when checking for the continued existence of a message for
3660 which it has found a message-specific retry record, was not finding
3661 messages that were in split spool directories. Consequently, it was
3662 deleting retry records that should have stayed in existence.
3664 29. eximstats updated to version 1.35
3665 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3666 1.35 - bugfix such that pie charts by volume are generated correctly
3668 30. The SPA authentication driver was not abandoning authentication and moving
3669 on to the next authenticator when an expansion was forced to fail,
3670 contradicting the general specification for all authenticators. Instead it
3671 was generating a temporary error. It now behaves as specified.
3673 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3674 (the order specifies the preference for clients). The order is now AES256,
3675 AES128, 3DES, ARCFOUR128.
3677 31. Small patch to Sieve code - explicitly set From: when generating an
3680 32. Exim crashed if a remote delivery caused a very long error message to be
3681 recorded - for instance if somebody sent an entire SpamAssassin report back
3682 as a large number of 550 error lines. This bug was coincidentally fixed by
3683 increasing the size of one of Exim's internal buffers (big_buffer) that
3684 happened as part of the Exiscan merge. However, to be on the safe side, I
3685 have made the code more robust (and fixed the comments that describe what
3688 33. Some experimental protocols are using DNS PTR records for new purposes. The
3689 keys for these records are domain names, not reversed IP addresses. The
3690 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3691 leaves it alone. Component reversal etc. now happens only for IP addresses.
3694 34. The host_aton() function is supposed to be passed a string that is known
3695 to be a valid IP address. However, in the case of IPv6 addresses, it was
3696 not checking this. This is a hostage to fortune. Exim now panics and dies
3697 if the condition is not met. A case was found where this could be provoked
3698 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3699 components; fortuitously, this particular loophole had already been fixed
3700 by change 4.50/55 or 4.44/33 above.
3702 If there are any other similar loopholes, the new check in host_aton()
3703 itself should stop them being exploited. The report I received stated that
3704 data on the command line could provoke the exploit when Exim was running as
3705 exim, but did not say which command line option was involved. All I could
3706 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3707 running as the user.
3710 35. There was a buffer overflow vulnerability in the SPA authentication code
3711 (which came originally from the Samba project). I have added a test to the
3712 spa_base64_to_bits() function which I hope fixes it.
3715 36. The daemon start-up calls getloadavg() while still root for those OS that
3716 need the first call to be done as root, but it missed one case: when
3717 deliver_queue_load_max is set with deliver_drop_privilege. This is
3718 necessary for the benefit of the queue runner, because there is no re-exec
3719 when deliver_drop_privilege is set.
3721 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3722 was in use, was not putting the data itself into the right store pool;
3723 consequently, it could be overwritten for a subsequent message in the same
3724 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3727 38. Sometimes the final signoff response after QUIT could fail to get
3728 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3729 < 0 before doing a fflush(). This bug looks as though it goes back to the
3730 introduction of TLS in release 3.20, but "sometimes" must have been rare
3731 because the tests only now provoked it.
3737 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3738 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3739 Now it returns an integer. A number of calls were still expecting a T/F
3740 return. Fortuitously, in all cases, the tests worked in OK situations,
3741 which is the norm. However, things would have gone wrong on any write
3742 failures on the smtp file descriptor. This function is used when sending
3743 messages over SMTP and also when doing verify callouts.
3745 2. When Exim is called to do synchronous delivery of a locally submitted
3746 message (the -odf or -odi options), it no longer closes stderr before doing
3749 3. Implemented the mua_wrapper option.
3751 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3753 5. Implemented the functions header_remove(), header_testname(),
3754 header_add_at_position(), and receive_remove_recipient(), and exported them
3757 6. If an ACL "warn" statement specified the addition of headers, Exim already
3758 inserted X-ACL-Warn: at the start if there was no header name. However, it
3759 was not making this test for the second and subsequent header lines if
3760 there were newlines in the string. This meant that an invalid header could
3761 be inserted if Exim was badly configured.
3763 7. Allow an ACL "warn" statement to add header lines at the start or after all
3764 the Received: headers, as well as at the end.
3766 8. Added the rcpt_4xx retry error code.
3768 9. Added postmaster_mailfrom=xxx to callout verification option.
3770 10. Added mailfrom=xxxx to the callout verification option, for verify=
3773 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3774 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3776 12. Inserted some casts to stop certain compilers warning when using pointer
3777 differences as field lengths or precisions in printf-type calls (mostly
3778 affecting debugging statements).
3780 13. Added optional readline() support for -be (dynamically loaded).
3782 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3783 same clock tick as a message's arrival, so that its received time was the
3784 same as the "first fail" time on the retry record, and that message
3785 remained on the queue past the ultimate address timeout, every queue runner
3786 would try a delivery (because it was past the ultimate address timeout) but
3787 after another failure, the ultimate address timeout, which should have then
3788 bounced the address, did not kick in. This was a "< instead of <=" error;
3789 in most cases the first failure would have been in the next clock tick
3790 after the received time, and all would be well.
3792 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3793 being recognized when the domain list was tested by the match_domain
3794 condition in an expansion string.
3796 16. Added the ${str2b64: operator.
3798 17. Exim was always calling setrlimit() to set a large limit for the number of
3799 processes, without checking whether the existing limit was already
3800 adequate. (It did check for the limit on file descriptors.) Furthermore,
3801 errors from getrlimit() and setrlimit() were being ignored. Now they are
3802 logged to the main and panic logs, but Exim does carry on, to try to do its
3803 job under whatever limits there are.
3805 18. Imported PCRE 5.0.
3807 19. Trivial typo in log message " temporarily refused connection" (the leading
3810 20. If the log selector return_path_on_delivery was set and an address was
3811 redirected to /dev/null, the delivery process crashed because it assumed
3812 that a return path would always be set for a "successful" delivery. In this
3813 case, the whole delivery is bypassed as an optimization, and therefore no
3816 21. Internal re-arrangement: the function for sending a challenge and reading
3817 a response while authentication was assuming a zero-terminated challenge
3818 string. It's now changed to take a pointer and a length, to allow for
3819 binary data in such strings.
3821 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3823 23. Exim was not respecting finduser_retries when seeking the login of the
3824 uid under which it was called; it was always trying 10 times. (The default
3825 setting of finduser_retries is zero.) Also, it was sleeping after the final
3826 failure, which is pointless.
3828 24. Implemented tls_on_connect_ports.
3830 25. Implemented acl_smtp_predata.
3832 26. If the domain in control=submission is set empty, Exim assumes that the
3833 authenticated id is a complete email address when it generates From: or
3834 Sender: header lines.
3836 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3837 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3838 chown and chgrp in /bin and hostname in /usr/bin.
3840 28. Exim was keeping the "process log" file open after each use, just as it
3841 does for the main log. This opens the possibility of it remaining open for
3842 long periods when the USR1 signal hits a daemon. Occasional processlog
3843 errors were reported, that could have been caused by this. Anyway, it seems
3844 much more sensible not to leave this file open at all, so that is what now
3847 29. The long-running daemon process does not normally write to the log once it
3848 has entered its main loop, and it closes the log before doing so. This is
3849 so that log files can straightforwardly be renamed and moved. However,
3850 there are a couple of unusual error situations where the daemon does write
3851 log entries, and I had neglected to close the log afterwards.
3853 30. The text of an SMTP error response that was received during a remote
3854 delivery was being truncated at 512 bytes. This is too short for some of
3855 the long messages that one sometimes sees. I've increased the limit to
3858 31. It is now possible to make retry rules that apply only when a message has a
3859 specific sender, in particular, an empty sender.
3861 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3862 it possible to be selective about when SMTP synchronization is enforced.
3864 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3866 32. Implemented hosts_connection_nolog.
3868 33. Added an ACL for QUIT.
3870 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3873 35. Added mailbox_size and mailbox_filecount to appendfile.
3875 36. Added control = no_multiline_responses to ACLs.
3877 37. There was a bug in the logic of the code that waits for the clock to tick
3878 in the case where the clock went backwards by a substantial amount such
3879 that the microsecond fraction of "now" was more than the microsecond
3880 fraction of "then" (but the whole seconds number was less).
3882 38. Added support for the libradius Radius client library this is found on
3883 FreeBSD (previously only the radiusclient library was supported).
3889 1. When certain lookups returned multiple values in the form name=value, the
3890 quoting of the values was not always being done properly. Specifically:
3891 (a) If the value started with a double quote, but contained no whitespace,
3893 (b) If the value contained whitespace other than a space character (i.e.
3894 tabs or newlines or carriage returns) it was not quoted.
3895 This fix has been applied to the mysql and pgsql lookups by writing a
3896 separate quoting function and calling it from the lookup code. The fix
3897 should probably also be applied to nisplus, ibase and oracle lookups, but
3898 since I cannot test any of those, I have not disturbed their existing code.
3900 2. A hit in the callout cache for a specific address caused a log line with no
3901 reason for rejecting RCPT. Now it says "Previous (cached) callout
3902 verification failure".
3904 3. There was an off-by-one bug in the queryprogram router. An over-long
3905 return line was truncated at 256 instead of 255 characters, thereby
3906 overflowing its buffer with the terminating zero. As well as fixing this, I
3907 have increased the buffer size to 1024 (and made a note to document this).
3909 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3910 when Exim is waiting for an SMTP response from a remote server, Exim
3911 restarts its select() call on the socket, thereby resetting its timeout.
3912 This is not a problem when such interrupts are rare. Somebody set up a cron
3913 job to run exiwhat every 2 minutes, which is less than the normal select()
3914 timeout (5 or 10 minutes). This meant that the select() timeout never
3915 kicked in because it was always reset. I have fixed this by comparing the
3916 time when an interrupt arrives with the time at the start of the first call
3917 to select(). If more time than the timeout has elapsed, the interrupt is
3918 treated as a timeout.
3920 5. Some internal re-factoring in preparation for the addition of Sieve
3921 extensions (by MH). In particular, the "personal" test is moved to a
3922 separate function, and given an option for scanning Cc: and Bcc: (which is
3923 not set for Exim filters).
3925 6. When Exim created an email address using the login of the caller as the
3926 local part (e.g. when creating a From: or Sender: header line), it was not
3927 quoting the local part when it contained special characters such as @.
3929 7. Installed new OpenBSD configuration files.
3931 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3932 try to make them clearer.
3934 9. Callout options, other than the timeout value, were being ignored when
3935 verifying sender addresses in header lines. For example, when using
3937 verify = header_sender/callout=no_cache
3939 the cache was (incorrectly) being used.
3941 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3942 only the Cygwin environment.
3944 11. When return_path_on_delivery was set as a log selector, if different remote
3945 addresses in the same message used different return paths and parallel
3946 remote delivery occurred, the wrong values would sometimes be logged.
3947 (Whenever a remote delivery process finished, the return path value from
3948 the most recently started remote delivery process was logged.)
3950 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3951 lines when AUTH and/or TLS are in use. This is the "received protocol"
3952 field. Exim used to use "asmtp" for authenticated SMTP, without any
3953 indication (in the protocol name) for TLS use. Now it follows the RFC and
3954 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3955 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3956 names appear in log lines as well as in Received: header lines.
3958 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3959 extensions, and comparison tests, and to fix some bugs.
3961 14. Changes to the "personal" filter test:
3963 (1) The test was buggy in that it was just doing the equivalent of
3964 "contains" tests on header lines. For example, if a user's address was
3965 anne@some.where, the "personal" test would incorrectly be true for
3967 To: susanne@some.where
3969 This test is now done by extracting each address from the header in turn,
3970 and checking the entire address. Other tests that are part of "personal"
3971 are now done using regular expressions (for example, to check local parts
3972 of addresses in From: header lines).
3974 (2) The list of non-personal local parts in From: addresses has been
3975 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3976 taken from the Sieve specification recommendations.
3978 (3) If the message contains any header line starting with "List-" it is
3979 treated as non-personal.
3981 (4) The test for "circular" in the Subject: header line has been removed
3982 because it now seems ill-conceived.
3984 15. Minor typos in src/EDITME comments corrected.
3986 16. Installed latest exipick from John Jetmore.
3988 17. If headers_add on a router specified a text string that was too long for
3989 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3990 of string_sprintf() is now avoided.
3992 18. $message_body_size was not set (it was always zero) when running the DATA
3993 ACL and the local_scan() function.
3995 19. For the "mail" command in an Exim filter, no default was being set for
3996 the once_repeat time, causing a random time value to be used if "once" was
3997 specified. (If the value happened to be <= 0, no repeat happened.) The
3998 default is now 0s, meaning "never repeat". The "vacation" command was OK
3999 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4000 (I found it when inspecting the code).
4002 20. There is now an overall timeout for performing a callout verification. It
4003 defaults to 4 times the callout timeout, which applies to individual SMTP
4004 commands during the callout. The overall timeout applies when there is more
4005 than one host that can be tried. The timeout is checked before trying the
4006 next host. This prevents very long delays if there are a large number of
4007 hosts and all are timing out (e.g. when the network connections are timing
4008 out). The value of the overall timeout can be changed by specifying an
4009 additional sub-option for "callout", called "maxwait". For example:
4011 verify = sender/callout=5s,maxwait=20s
4013 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4014 to the end before writing, but this should make it even safer).
4016 22. Exim was forgetting that it had advertised PIPELINING for the second and
4017 subsequent messages on an SMTP connection. It was also not resetting its
4018 memory on STARTTLS and an internal HELO.
4020 23. When Exim logs an SMTP synchronization error within a session, it now
4021 records whether PIPELINING has been advertised or not.
4023 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4024 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4025 rather than long int.
4027 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4029 26. Added the never_mail option to autoreply.
4035 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4036 crash if the getsockname() call failed; this can happen if a connection is
4037 closed very soon after it is established. The problem was simply in the
4038 order in which certain operations were done, causing Exim to try to write
4039 to the SMTP stream before it had set up the file descriptor. The bug has
4040 been fixed by making things happen in the correct order.
4046 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4047 before the connection was closed, thus losing the rejection response.
4049 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4050 some early Solaris releases, but causes trouble in current releases where
4051 socklen_t is defined.
4053 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4056 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4059 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4060 that releasing the top of it at the end releases what was used for sub-
4061 expansions (unless the block got too big). However, discard this block if
4062 the first thing is a variable or header, so that we can use its block when
4063 it is dynamic (useful for very large $message_headers, for example).
4065 6. Lookups now cache *every* query, not just the most recent. A new, separate
4066 store pool is used for this. It can be recovered when all lookup caches are
4067 flushed. Lookups now release memory at the end of their result strings.
4068 This has involved some general refactoring of the lookup sources.
4070 7. Some code has been added to the store_xxx() functions to reduce the amount
4071 of flapping under certain conditions.
4073 8. log_incoming_interface used to affect only the <= reception log lines. Now
4074 it causes the local interface and port to be added to several more SMTP log
4075 lines, for example "SMTP connection from", and rejection lines.
4077 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4079 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4081 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4082 could be overwritten at the end of the current message (or the start of a
4083 new message if it was set in a HELO ACL). The value is now preserved for
4084 the duration of the SMTP connection.
4086 12. If a transport had a headers_rewrite setting, and a matching header line
4087 contained an unqualified address, that address was qualified, even if it
4088 did not match any rewriting rules. The underlying bug was that the values
4089 of the flags that permit the existence of unqualified sender and recipient
4090 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4091 non-local messages, and by -bnq for local messages) were not being
4092 preserved with the message after it was received.
4094 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4095 "next input=" as part of the text comprising the host identity instead of
4096 the correct text. The code was using the same buffer for two different
4097 strings. However, depending on which order the printing function evaluated
4098 its arguments, the bug did not always show up. Under Linux, for example, my
4099 test suite worked just fine.
4101 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4102 doesn't work with some older versions of Perl. It has been changed to "my",
4103 which in any case is probably the better facility to use.
4105 15. A really picky compiler found some instances of statements for creating
4106 error messages that either had too many or two few arguments for the format
4109 16. The size of the buffer for calls to the DNS resolver has been increased
4110 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4111 for addresses that have a lot of PTR records. This alleviates a problem; it
4112 does not fully solve it.
4114 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4115 buffer now truncates the list and logs the incident, which is the same
4116 action as happens when Exim is looking up a host name and its aliases.
4117 Previously in this situation something unpredictable would happen;
4118 sometimes it was "internal error: store_reset failed".
4120 18. If a server dropped the connection unexpectedly when an Exim client was
4121 using GnuTLS and trying to read a response, the client delivery process
4122 crashed while trying to generate an error log message.
4124 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4125 string, for example:
4127 warn message = H1: something\nH2: something
4129 the text was added as a single header line from Exim's point of view
4130 though it ended up OK in the delivered message. However, searching for the
4131 second and subsequent header lines using $h_h2: did not work. This has been
4132 fixed. Similarly, if a system filter added multiple headers in this way,
4133 the routers could not see them.
4135 20. Expanded the error message when iplsearch is called with an invalid key to
4136 suggest using net-iplsearch in a host list.
4138 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4139 ACLs are no longer actually imposed (and a message to that effect is
4142 22. If a "gecos" field in a passwd entry contained escaped characters, in
4143 particular, if it contained a \" sequence, Exim got it wrong when building
4144 a From: or a Sender: header from that name. A second bug also caused
4145 incorrect handling when an unquoted " was present following a character
4146 that needed quoting.
4148 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4149 was not being matched caselessly.
4151 24. Arranged for all hyphens in the exim.8 source to be escaped with
4154 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4155 properly. Recipient callout cache records were still being keyed to include
4156 the sender, even when use_sender was set false. This led to far more
4157 callouts that were necessary. The sender is no longer included in the key
4158 when use_sender is false.
4160 26. Added "control = submission" modifier to ACLs.
4162 27. Added the ${base62d: operator to decode base 62 numbers.
4164 28. dnsdb lookups can now access SRV records.
4166 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4167 the configuration file.
4169 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4170 (-v) mode. This makes the output for a verbose queue run more intelligible.
4172 31. Added a use_postmaster feature to recipient callouts.
4174 32. Added the $body_zerocount variable, containing the number of binary zero
4175 bytes in the message body.
4177 33. The time of last modification of the "new" subdirectory is now used as the
4178 "mailbox time last read" when there is a quota error for a maildir
4181 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4183 35. Added +ignore_unknown as a special item in host lists.
4185 36. Code for decoding IPv6 addresses in host lists is now included, even if
4186 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4187 address was recognized as an IP address, but was then not correctly decoded
4188 into binary, causing unexpected and incorrect effects when compared with
4195 1. Very minor rewording of debugging text in manualroute to say "list of
4196 hosts" instead of "hostlist".
4198 2. If verify=header_syntax was set, and a header line with an unqualified
4199 address (no domain) and a large number of spaces between the end of the
4200 name and the colon was received, the reception process suffered a buffer
4201 overflow, and (when I tested it) crashed. This was caused by some obsolete
4202 code that should have been removed. The fix is to remove it!
4204 3. When running in the test harness, delay a bit after writing a bounce
4205 message to get a bit more predictability in the log output.
4207 4. Added a call to search_tidyup() just before forking a reception process. In
4208 theory, someone could use a lookup in the expansion of smtp_accept_max_
4209 per_host which, without the tidyup, could leave open a database connection.
4211 5. Added the variables $recipient_data and $sender_data which get set from a
4212 lookup success in an ACL "recipients" or "senders" condition, or a router
4213 "senders" option, similar to $domain_data and $local_part_data.
4215 6. Moved the writing of debug_print from before to after the "senders" test
4218 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4219 problems for message scanning, either using a data ACL, or using
4220 local_scan() because the Received: header was not generated till after they
4221 were called (in order to set the time as the time of reception completion).
4222 I have revised the way this works. The header is now generated after the
4223 body is received, but before the ACL or local_scan() are called. After they
4224 are run, the timestamp in the header is updated.
4230 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4231 before starting a queue runner without re-exec. This happened only when
4232 deliver_drop_privilege was set or when the Exim user was set to root. The
4233 effect of the bug was that timeouts during subsequent deliveries caused
4234 crashes instead of being properly handled. The handler is now left at its
4235 default (and expected) setting.
4237 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4238 message, again when deliver_drop_privilege is set or Exim is run as root.
4239 The bug described in (1) was not present in this case, but the tidying up
4240 of the other signals was missing. I have made the two cases consistent.
4242 3. The ignore_target_hosts setting on a manualroute router was being ignored
4243 for hosts that were looked up using the /MX notation.
4245 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4248 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4249 operated on the sender address. After changing the $sender_address to <>
4250 for the sender address verify, Exim was re-instated it as the original
4251 (before rewriting) address, but remembering that it had rewritten it, so it
4252 wasn't rewriting it again. This bug also had the effect of breaking the
4253 sender address verification caching when the sender address was rewritten.
4255 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4256 This has been changed so that if the ip literal address matches
4257 ignore_target_hosts, the router declines.
4259 7. Added expansion conditions match_domain, match_address, and match_local_
4260 part (NOT match_host).
4262 8. The placeholder for the Received: header didn't have a length field set.
4264 9. Added code to Exim itself and to exim_lock to test for a specific race
4265 condition that could lead to file corruption when using MBX delivery. The
4266 issue is with the lockfile that is created in /tmp. If this file is removed
4267 after a process has opened it but before that process has acquired a lock,
4268 there is the potential for a second process to recreate the file and also
4269 acquire a lock. This could lead to two Exim processes writing to the file
4270 at the same time. The added code performs the same test as UW imapd; it
4271 checks after acquiring the lock that its file descriptor still refers to
4272 the same named file.
4274 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4275 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4278 11. Added the smtp_active_hostname option. If used, this will typically be made
4279 to depend on the incoming interface address. Because $interface_address is
4280 not set up until the daemon has forked a reception process, error responses
4281 that can happen earlier (such as "too many connections") no longer contain
4284 12. If an expansion in a condition on a "warn" statement fails because a lookup
4285 defers, the "warn" statement is abandoned, and the next ACL statement is
4286 processed. Previously this caused the whole ACL to be aborted.
4288 13. Added the iplsearch lookup type.
4290 14. Added ident_timeout as a log selector.
4292 15. Added tls_certificate_verified as a log selector.
4294 16. Added a global option tls_require_ciphers (compare the smtp transport
4295 option of the same name). This controls incoming TLS connections.
4297 17. I finally figured out how to make tls_require_ciphers do a similar thing
4298 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4299 before starting the TLS session.
4301 18. Tabs are now shown as \t in -bP output.
4303 19. If the log selector return_path_on_delivery was set, Exim crashed when
4304 bouncing a message because it had too many Received: header lines.
4306 20. If two routers both had headers_remove settings, and the first one included
4307 a superfluous trailing colon, the final name in the first list and the
4308 first name in the second list were incorrectly joined into one item (with a
4309 colon in the middle).
4315 1. Added -C and -D options to the exinext utility, mainly to make it easier
4316 to include in the automated testing, but these could be helpful when
4317 multiple configurations are in use.
4319 2. The exinext utility was not formatting the output nicely when there was
4320 an alternate port involved in the retry record key, nor when there was a
4321 message id as well (for retries that were specific to a specific message
4322 and a specific host). It was also confused by IPv6 addresses, because of
4323 the additional colons they contain. I have fixed the IPv4 problem, and
4324 patched it up to do a reasonable job for IPv6.
4326 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4327 delivery, the log line now contains "pipelined" if PIPELINING was used.
4329 4. An SMTP transport process used to panic and die if the bind() call to set
4330 an explicit outgoing interface failed. This has been changed; it is now
4331 treated in the same way as a connect() failure.
4333 5. A reference to $sender_host_name in the part of a conditional expansion
4334 that was being skipped was still causing a DNS lookup. This no longer
4337 6. The def: expansion condition was not recognizing references to header lines
4338 that used bh_ and bheader_.
4340 7. Added the _cache feature to named lists.
4342 8. The code for checking quota_filecount in the appendfile transport was
4343 allowing one more file than it should have been.
4345 9. For compatibility with Sendmail, the command line option
4353 and sets the incoming protocol and host name (for trusted callers). The
4354 host name and its colon can be omitted when only the protocol is to be set.
4355 Note the Exim already has two private options, -pd and -ps, that refer to
4356 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4357 "s", but I don't think that's a major issue.
4359 10. A number of refactoring changes to the code, none of which should affect
4362 (a) The number of logging options was getting close to filling up the
4363 32-bit word that was used as a bit map. I have split them into two classes:
4364 those that are passed in the argument to log_write(), and those that are
4365 only ever tested independently outside of that function. These are now in
4366 separate 32-bit words, so there is plenty of room for expansion again.
4367 There is no change in the user interface or the logging behaviour.
4369 (b) When building, for example, log lines, the code previously used a
4370 macro that called string_cat() twice, in order to add two strings. This is
4371 not really sufficiently general. Furthermore, there was one instance where
4372 it was actually wrong because one of the argument was used twice, and in
4373 one call a function was used. (As it happened, calling the function twice
4374 did not affect the overall behaviour.) The macro has been replaced by a
4375 function that can join an arbitrary number of extra strings onto a growing
4378 (c) The code for expansion conditions now uses a table and a binary chop
4379 instead of a serial search (which was left over from when there were very
4380 few conditions). Also, it now recognizes conditions like "pam" even when
4381 the relevant support is not compiled in: a suitably worded error message is
4382 given if an attempt is made to use such a condition.
4384 11. Added ${time_interval:xxxxx}.
4386 12. A bug was causing one of the ddress fields not to be passed back correctly
4387 from remote delivery subprocesses. The field in question was not being
4388 subsequently used, so this caused to problems in practice.
4390 13. Added new log selectors queue_time and deliver_time.
4392 14. Might have fixed a bug in maildirsizefile handling that threw up
4393 "unexpected character" debug warnings, and recalculated the data
4394 unnecessarily. In any case, I expanded the warning message to give more
4397 15. Added the message "Restricted characters in address" to the statements in
4398 the default ACL that block characters like @ and % in local parts.
4400 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4401 Three changes have been made:
4403 (a) There was a serious bug; a negative response to MAIL caused the whole
4404 recipient domain to be cached as invalid, thereby blocking all messages
4405 to all local parts at the same domain, from all senders. This bug has
4406 been fixed. The domain is no longer cached after a negative response to
4407 MAIL if the sender used is not empty.
4409 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4412 (c) A new callout option, "use_sender" has been added for people who want
4413 the modified behaviour.
4419 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4422 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4423 indeed breaks things for older releases.
4425 3. Added additional logging to the case where there is a problem reading data
4426 from a filter that is running in a subprocess using a pipe, in order to
4427 try to track down a specific problem.
4429 4. Testing facility fudge: when running in the test harness and attempting
4430 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4431 getting "No route to host". Convert this to a timeout.
4433 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4436 6. Some OS don't have socklen_t but use size_t instead. This affects the
4437 fifth argument of getsockopt() amongst other things. This is now
4438 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4439 can be set for individual OS. I have set it for SunOS5, OSF1, and
4440 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4441 some earlier ones do not.
4443 7. Change 4.30/15 was not doing the test caselessly.
4445 8. The standard form for an IPv6 address literal was being rejected by address
4446 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4447 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4448 this, as well as the form without the "IPv6" on the front (but only when
4449 address literals are enabled, of course).
4451 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4453 10. Exim crashed if a message with an empty sender address specified by -f
4454 encountered a router with an errors_to setting. This could be provoked only
4455 by a command such as
4459 where an empty string was supplied; "<>" did not hit this bug.
4461 11. Installed PCRE release 4.5.
4463 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4464 remained set. It is now erased.
4466 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4467 times from message ids (which are base 36 rather than the normal 62).
4469 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4470 were being counted as actual protocol errors, and logged if the log
4471 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4472 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4473 and DATA following a set of rejected RCPTs do not count as protocol errors.
4474 In other words, Exim assumes they were pipelined, though this may not
4475 actually be the case. Of course, in all cases the client gets an
4476 appropriate error code.
4478 15. If a lookup fails in an ACL condition, a message about the failure may
4479 be available; it is used if testing the ACL cannot continue, because most
4480 such messages specify what the cause of the deferral is. However, some
4481 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4482 that caused an old message to be retained and used if a later statement
4483 caused a defer, replacing the real cause of the deferral.
4485 16. If an IP address had so many PTR records that the DNS lookup buffer
4486 was not large enough to hold them, Exim could crash while trying to process
4487 the truncated data. It now detects and logs this case.
4489 17. Further to 4.21/58, another change has been made: if (and only if) the
4490 first line of a message (the first header line) ends with CRLF, a bare LF
4491 in a subsequent header line has a space inserted after it, so as not to
4492 terminate the header.
4494 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4495 unnecessarily, used atoi() instead of strtol(), and didn't check the
4496 termination when getting file sizes from file names by regex.
4498 19. Completely re-implemented the support for maildirsize files, in the light
4499 of a number of problems with the previous contributed implementation
4500 (4.30/29). In particular:
4502 . If the quota is zero, the maildirsize file is maintained, but no quota is
4505 . If the maildir directory does not exist, it is created before any attempt
4506 to write a maildirsize file.
4508 . The quota value in the file is just a cache; if the quota is changed in
4509 the transport, the new value overrides.
4511 . A regular expression is available for excluding directories from the
4514 20. The autoreply transport checks the characters in options that define the
4515 message's headers; it allows continued headers, but it was checking with
4516 isspace() after an embedded newline instead of explicitly looking for a
4519 21. If all the "regular" hosts to which an address was routed had passed their
4520 expiry times, and had not reached their retry times, the address was
4521 bounced, even if fallback hosts were defined. Now Exim should go on to try
4524 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4525 equivalent code in the SMTP transport. Some hosts send humungous responses
4526 to HELO/EHLO, more than 1024 it seems.
4528 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4529 gives compiler warnings in some environments. I've now done it "properly",
4532 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4533 (because of problems with the built-in one) was declared to return uschar *
4534 instead of char *, causing compiler failure.
4536 25. Fixed a file descriptor leak when processing alias/forward files.
4538 26. Fixed a minor format string issue in dbfn.c.
4540 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4542 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4543 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4546 29. When a sender address is verified, it is cached, to save repeating the test
4547 when there is more than one recipient in a message. However, when the
4548 verification involves a callout, it is possible for different callout
4549 options to be set for different recipients. It is too complicated to keep
4550 track of this in the cache, so now Exim always runs a verification when a
4551 callout is required, relying on the callout cache for the optimization.
4552 The overhead is duplication of the address routing, but this should not be
4555 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4556 to be verified with callout=postmaster, and the main callout worked but the
4557 postmaster check failed, the verification correctly failed. However, if a
4558 subsequent RCPT command asked for sender verification *without* the
4559 postmaster check, incorrect caching caused this verification also to fail,
4562 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4563 it was not caching the DNS options (qualify_single, search_parents) that
4564 were used when the lookup failed. A subsequent lookup with different
4565 options therefore always gave the same answer, though there were cases
4566 where it should not have. (Example: a "domains = !$mx_any" option on a
4567 dnslookup router: the "domains" option is always processed without any
4568 widening, but the router might have qualify_single set.) Now Exim uses the
4569 cached value only when the same options are set.
4571 32. Added John Jetmore's "exipick" utility to the distribution.
4573 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4574 than a timeout (e.g. a certificate is required, and is not provided), an
4575 Exim server now closes the connection immediately. Previously it waited for
4576 the client to close - but if the client is SSL, it seems that they each
4577 wait for each other, leading to a delay before one of them times out.
4579 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4580 maintained 0.8.x compatibility because I don't think many are using it, and
4581 it is clearly obsolete.
4583 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4586 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4587 client certificate was expired. A simple patch fixes this, though I don't
4588 understand the full logic of why the verify callback is called multiple
4591 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4592 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4593 which causes problems with some clients (such as the Certicom SSL Plus
4594 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4595 disables the coutermeasure allowing Eudora to connect."
4597 38. Exim was not checking that a write() to a log file succeeded. This could
4598 lead to Bad Things if a log got too big, in particular if it hit a file
4599 size limit. Exim now panics and dies if it cannot write to a log file, just
4600 as it does if it cannot open a log file.
4602 39. Modified OS/Makefile-Linux so that it now contains
4604 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4606 The two -D definitions ensure that Exim is compiled with large file
4607 support, which makes it possible to handle log files that are bigger than
4610 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4611 instance) a domain was checked against a named list that involved a lookup,
4612 causing $domain_data to be set, then another domain was checked against the
4613 same list, then the first domain was re-checked, the value of $domain_data
4614 after the final check could be wrong. In particular, if the second check
4615 failed, it could be set empty. This bug probably also applied to
4618 41. The strip_trailing_dot option was not being applied to the address given
4619 with the -f command-line option.
4621 42. The code for reading a message's header from the spool was incrementing
4622 $received_count, but never initializing it. This meant that the value was
4623 incorrect (doubled) while delivering a message in the same process in which
4624 it was received. In the most common configuration of Exim, this never
4625 happens - a fresh exec is done - but it can happen when
4626 deliver_drop_privilege is set.
4628 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4629 it now includes up to 150 characters of the unexpected data in the log
4632 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4633 and building data strings. The size of both of these buffers was 10 000
4634 bytes - far larger than anybody would *ever* want, thought I. Needless to
4635 say, somebody hit the limit. I have increased the maximum line length to
4636 20 000 and the maximum data length of concatenated lines to 100 000. I have
4637 also fixed two bugs, because there was no checking on these buffers. Tsk,
4638 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4639 buffer is too small.
4641 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4642 lsearch lookups. Now it does.
4644 46. When parsing a route_list item in a manualroute router, a fixed-length
4645 buffer was used for the list of hosts. I made this 1024 bytes long,
4646 thinking that nobody would ever have a list of hosts that long. Wrong.
4647 Somebody had a whole pile of complicated expansion conditions, and the
4648 string was silently truncated, leading to an expansion error. It turns out
4649 that it is easier to change to an unlimited length (owing to other changes
4650 that have happened since this code was originally written) than to build
4651 structure for giving a limitation error. The length of the item that
4652 expands into the list of hosts is now unlimited.
4654 47. The lsearch lookup could not handle data where the length of text line was
4655 more than 4095 characters. Such lines were truncated, leading to shortened
4656 data being returned. It should now handle lines of any length.
4658 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4659 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4662 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4663 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4664 no longer happens in comment lines. A list of replacements is now placed
4665 at the head of all of the source files, except those whose only change is
4666 to replace PERL_COMMAND in the very first #! line.
4668 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4669 messages on the queue, with a bottom-up merge sort, using code contributed
4670 by Michael Haardt. This should make operations like -bp somewhat faster on
4671 large queues. It won't affect queue runners, except when queue_run_in_order
4674 51. Installed eximstats 1.31 in the distribution.
4676 52. Added support for SRV lookups to the dnslookup router.
4678 53. If an ACL referred to $message_body or $message_body_end, the value was not
4679 reset for any messages that followed in the same SMTP session.
4681 54. The store-handling optimization for building very long strings was not
4682 differentiating between the different store pools. I don't think this
4683 actually made any difference in practice, but I've tidied it.
4685 55. While running the routers to verify a sender address, $sender_address
4686 was still set to the sender address. This is wrong, because when routing to
4687 send a bounce to the sender, it would be empty. Therefore, I have changed
4688 it so that, while verifying a sender address, $sender_address is set to <>.
4689 (There is no change to what happens when verifying a recipient address.)
4691 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4692 target A or AAAA records (if not already returned) without resetting the
4693 qualify_single or search_parents options of the DNS resolver. These are
4694 inappropriate in this case because the targets of MX and SRV records must
4695 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4696 target that, when qualified, matched something in the local domain. These
4697 two options are now turned off when doing these lookups.
4699 57. It seems that at least some releases of Reiserfs (which does not have the
4700 concept of a fixed number of inodes) returns zero and not -1 for the
4701 number of available inodes. This interacted badly with check_spool_inodes,
4702 which assumed that -1 was the "no such thing" setting. What I have done is
4703 to check that the total number of inodes is greater than zero before doing
4704 the test of how many are available.
4706 58. When a "warn" ACL statement has a log_message modifier, the message is
4707 remembered, and not repeated. This is to avoid a lot of repetition when a
4708 message has many recipients that cause the same warning to be written.
4709 Howewer, Exim was preserving the list of already written lines for an
4710 entire SMTP session, which doesn't seem right. The memory is now reset if a
4711 new message is started.
4713 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4714 debugging output unless log_rewrite was also set.
4716 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4717 of (int)(handle) when we know that handle contains (void *)(-1).
4719 61. The Exim daemon panic-logs an error return when it closes the incoming
4720 connection. However "connection reset by peer" seems to be common, and
4721 isn't really an error worthy of noting specially, so that particular error
4724 62. When Exim is trying to find all the local interfaces, it used to panic and
4725 die if the ioctl to get the interface flags failed. However, it seems that
4726 on at least one OS (Solaris 9) it is possible to have an interface that is
4727 included in the list of interfaces, but for which you get a failure error
4728 for this call. This happens when the interface is not "plumbed" into a
4729 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4730 failure of the "get flags" call assumes that the interface is down.
4732 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4733 makes life easier for people who are doing arithmetic on fields extracted
4734 from dates, where you often get leading zeros that should not be
4735 interpreted as octal.
4737 64. Added qualify_domain to the redirect router, to override the global
4740 65. If a pathologically long header line contained very many addresses (the
4741 report of this problem mentioned 10 000) and each of them was rewritten,
4742 Exim could use up a very large amount of memory. (It kept on making new
4743 copies of the header line as it rewrote, and never released the old ones.)
4744 At the expense of a bit more processing, the header rewriting function has
4745 been changed so that it no longer eats memory in this way.
4747 66. The generation of the Received: header has been moved from the time that a
4748 message starts to be received, to the time that it finishes. The timestamp
4749 in the Received: header should now be very close to that of the <= log
4750 line. There are two side-effects of this change:
4752 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4753 logged header lines no longer include the local Received: line, because
4754 it has not yet been created. The same applies to a copy of the message
4755 that is returned to a non-SMTP sender when a message is rejected.
4757 (b) When a filter file is tested using -bf, no additional Received: header
4758 is added to the test message. After some thought, I decided that this
4761 This change does not affect the value of $received_for. It is still set
4762 after address rewriting, but before local_scan() is called.
4764 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4766 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4767 gave an unhelpful panic error message, and a defer error. I have managed to
4768 change this behaviour so that it now rejects any supplied certificate,
4769 which seems right, as the list of acceptable certificates is empty.
4771 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4772 gave an unhelpful defer error. I have not managed to make this reject any
4773 supplied certificates, but the error message it gives is "no certificate
4774 supplied", which is not helpful.
4776 70. exigrep's output now also includes lines that are not associated with any
4777 message, but which match the given pattern. Implemented by a patch from
4778 Martin Sluka, which also tidied up the Perl a bit.
4780 71. Recipient callout verification, like sender verification, was using <> in
4781 the MAIL FROM command. This isn't really the right thing, since the actual
4782 sender may affect whether the remote host accepts the recipient or not. I
4783 have changed it to use the actual sender in the callout; this means that
4784 the cache record is now keyed on a recipient/sender pair, not just the
4785 recipient address. There doesn't seem to be a real danger of callout loops,
4786 since a callout by the remote host to check the sender would use <>.
4787 [SEE ABOVE: changed after hitting problems.]
4789 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4790 temporary errors. However, in the case of such a code being given after
4791 the end of a data transmission (i.e. after ".") Exim was failing to write
4792 a retry record for the message. (Yes, there was some broken host that was
4793 actually sending 8xx at this point.)
4795 73. An unknown lookup type in a host list could cause Exim to panic-die when
4796 the list was checked. (An example that provoked this was putting <; in the
4797 middle of a list instead of at the start.) If this happened during a DATA
4798 ACL check, a -D file could be left lying around. This kind of configuration
4799 error no longer causes Exim to die; instead it causes a defer errror. The
4800 incident is still logged to the main and panic logs.
4802 74. Buglet left over from Exim 3 conversion. The message "too many messages
4803 in one connection" was written to the rejectlog but not the mainlog, except
4804 when address rewriting (yes!) was being logged.
4806 75. Added write_rejectlog option.
4808 76. When a system filter was run not as root (that is, when system_filter_user
4809 was set), the values of the $n variables were not being returned to the
4810 main process; thus, they were not subsequently available in the $sn
4813 77. Added +return_path_on_delivery log selector.
4815 78. A connection timeout was being treated differently from recipients deferred
4816 when testing hosts_max_try with a message that was older than the host's
4817 retry timeout. (The host should not be counted, thus allowing all hosts to
4818 be tried at least once before bouncing.) This may have been the cause of an
4819 occasionally reported bug whereby a message would remain on the queue
4820 longer than the retry timeout, but would be bounced if a delivery was
4821 forced. I say "may" because I never totally pinned down the problem;
4822 setting up timeout/retry tests is difficult. See also the next item.
4824 79. The ultimate address timeout was not being applied to errors that involved
4825 a combination of host plus message (for example, a timeout on a MAIL
4826 command). When an address resolved to a number of possible hosts, and they
4827 were not all tried for each delivery (e.g. because of hosts_max_try), a
4828 message could remain on the queue longer than the retry timeout.
4830 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4833 81. Fixed an obscure SMTP outgoing bug which required at least the following
4834 conditions: (a) there was another message waiting for the same server;
4835 (b) the server returned 5xx to all RCPT commands in the first message so
4836 that the message was not completed; (c) the server dropped the connection
4837 or gave a negative response to the RSET that Exim sends to abort the
4838 transaction. The observed case was a dropped connection after DATA that had
4839 been sent in pipelining mode. That is, the server had advertised PIPELINING
4840 but was not implementing it correctly. The effect of the bug was incorrect
4841 behaviour, such as trying another host, and this could lead to a crash.
4847 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4848 and daemon.c were passed as pointers to ints; they should have been
4849 pointers to socklen_t variables (which are typically unsigned ints).
4851 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4854 3. Fixed a really odd bug that affected only the testing scheme; patching a
4855 certain fixed string in the binary changed the value of another string that
4856 happened to be identical to the end of the original first string.
4858 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4859 name", it returns that address as the IP address. On some operating
4860 systems (e.g. Solaris), it also passes back the IP address string as the
4861 "host name". However, on others (e.g. Linux), it passes back an empty
4862 string. Exim wasn't checking for this, and was changing the host name to an
4863 empty string, assuming it had been canonicized.
4865 5. Although rare, it is permitted to have more than one PTR record for a given
4866 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4867 all the names associated with an address, because they do in Solaris.
4868 However, it seems that they do not in Linux for data that comes from the
4869 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4870 I found this out when I moved to a new Linux workstation and tried to run
4871 the Exim test suite.
4873 To get round this problem I have changed the code so that it now does its
4874 own call to the DNS to look up PTR records when searching for a host name.
4875 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4876 addresses that are only in /etc/hosts are still found.
4878 This behaviour is, however, controlled by an option called host_lookup_
4879 order, which defaults to "bydns:byaddr". If people want to use the other
4880 order, or indeed, just use one or the other means of lookup, they can
4881 specify it in this variable.
4883 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4884 some operating systems, this comes back from gethostbyaddr() as an empty
4885 string, and this is what Exim used to test for. However, it seems that in
4886 other systems, "." is yielded. Exim now tests for this case too.
4888 7. The values of check_spool_space and check_log_space are now held internally
4889 as a number of kilobytes instead of an absolute number of bytes. If a
4890 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4891 kilobyte. This means that much larger values can be stored.
4893 8. Exim monitor: an attempt to get the action menu when not actually pointing
4894 at a message produces an empty menu entitled "No message selected". This
4895 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4896 no entries in it ("Shell widget menu has zero width and/or height"). So I
4897 have added a single, blank menu entry in this case.
4899 9. Added ${quote_local_part.
4901 10. MIME decoding is now applied to the contents of Subject: header lines when
4904 11. Now that a reference to $sender_host_address automatically causes a reverse
4905 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4906 host lookup before query-style lookups in lists that might use this
4907 variable. This has therefore been abolished, and the "net-" prefix is no
4908 longer necessary for query-style lookups.
4910 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4911 have been a typo for LFLAGS, so it has been changed.
4913 13. The install script calls Exim with "-C /dev/null" in order to find the
4914 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4915 to be output. Howeve, since Exim outputs its version number before the
4916 error, it didn't break the script. It just looked ugly. I fixed this by
4917 always allowing "-C /dev/null" if the caller is root.
4919 14. Ignore overlarge ACL variable number when reading spool file - insurance
4920 against a later release with more variables having written the file.
4922 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4923 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4924 this, as well as the form without the "IPv6" on the front.
4926 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4927 OS/Makefile-Darwin file.
4929 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4930 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4932 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4933 variable. However, some OS use ints for inodes. I've added cast to long int
4934 to get rid of the compiler warning.
4936 19. I had forgotten to lock out "/../" in configuration file names when
4937 ALT_CONFIG_PREFIX was set.
4939 20. Routers used for verification do not need to specify transports. However,
4940 if such a router generated a host list, and callout was configured, Exim
4941 crashed, because it could not find a port number from the (non-existent)
4942 transport. It now assumes port 25 in this circumstance.
4944 21. Added the -t option to exigrep.
4946 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4947 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4948 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4949 them is set, LOOKUP_LSEARCH is forced.
4951 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4953 24. Added sender and host information to the "rejected by local_scan()" log
4954 line; previously there was no indication of these.
4956 25. Added .include_if_exists.
4958 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4959 Linux. It turns out that not all file systems support this. Apparently some
4960 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4961 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4962 not supported on the file descriptor, is now ignored when Exim is trying to
4963 sync a directory. This applies only to Linux.
4965 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4967 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4968 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4971 29. Added support for maildirsize files from supplied patch (modified a bit).
4973 30. The use of :fail: followed by an empty string could lead Exim to respond to
4974 sender verification failures with (e.g.):
4976 550 Verification failed for <xxx>
4977 550 Sender verify failed
4979 where the first response line was missing the '-' that indicates it is not
4980 the final line of the response.
4982 31. The loop for finding the name of the user that called Exim had a hardwired
4983 limit of 10; it now uses the value of finduser_retries, which is used for
4984 all other user lookups.
4986 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4989 33. Exim was neglecting to zero errno before one call of strtol() when
4990 expanding a string and expecting an integer value. On some systems this
4991 resulted in spurious "integer overflow" errors. Also, it was casting the
4992 result into an int without checking.
4994 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4995 did not work. The code looks as if it has *never* worked, though it appears
4996 to have been documented since at least releast 1.62. I have made it work.
4998 35. The "timeout_DNS" error in retry rules, also documented since at least
4999 1.62, also never worked. As it isn't clear exactly what this means, and
5000 clearly it isn't a major issue, I have abolished the feature by treating it
5001 as "timeout", and writing a warning to the main and panic logs.
5003 36. The display of retry rules for -brt wasn't always showing the error code
5006 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5007 timeout_connect_A, timeout_connect_MX.
5009 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5010 to the empty sender.
5012 39. The daemon was not analysing the content of -oX till after it had closed
5013 stderr and disconnected from the controlling terminal. This meant that any
5014 syntax errors were only noted on the panic log, and the return code from
5015 the command was 0. By re-arranging the code a little, I've made the
5016 decoding happen first, so such errors now appear on stderr, and the return
5017 code is 1. However, the actual setting up of the sockets still happens in
5018 the disconnected process, so errors there are still only recorded on the
5021 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5022 connections (as happens on some IP stacks) was logged at start up time as
5023 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5024 it from "IPv6 and IPv4", which means that two separate sockets are being
5027 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5028 says whether AF_INET or AF_INET6 was passed as an argument.
5030 42. Exiwhat output was messed up when time zones were included in log
5033 43. Exiwhat now gives more information about the daemon's listening ports,
5034 and whether -tls-on-connect was used.
5036 44. The "port" option of the smtp transport is now expanded.
5038 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5039 silently ignored. Now an error message is written to the main and panic
5042 46. There's a new ACL modifier called "logwrite" which writes to a log file
5043 as soon as it is encountered.
5045 47. Added $local_user_uid and $local_user_gid at routing time.
5047 48. Exim crashed when trying to verify a sender address that was being
5050 49. Exim was recognizing only a space character after ".include". It now also
5051 recognizes a tab character.
5053 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5054 extracting the relevant information from the specification. The man page no
5055 longer contains scrambled data for the -d option, and I've added a section
5056 at the front about calling Exim under different names.
5058 51. Added "extra_headers" argument to the "mail" command in filter files.
5060 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5063 53. Installed eximstats 1.29.
5065 54. Added transport_filter_timeout as a generic transport option.
5067 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5068 Cc: header lines. This was required by RFC 822, but it not required by RFC
5071 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5072 incoming messages that did not have them. Now it does so only if the
5073 message originates locally, that is, if there is no associated remote host
5074 address. When Resent- header lines are present, this applies to the Resent-
5075 lines rather than the non-Resent- lines.
5077 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5078 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5080 58. Messages for configuration errors now include the name of the main
5081 configuration file - useful now that there may be more than one file in a
5082 list (.included file names were always shown).
5084 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5085 for those rare installations that do not start the daemon as root or run it
5086 setuid root. I've cut out the call to initgroups() if the daemon is not
5089 60. The Exim user and group can now be bound into the binary as text strings
5090 that are looked up at the start of Exim's processing.
5092 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5094 62. Added $mailstore_basename variable.
5096 63. Installed patch to sieve.c from Michael Haardt.
5098 64. When Exim failed to open the panic log after failing to open the main log,
5099 the original message it was trying to log was written to stderr and debug
5100 output, but if they were not available (the usual case in production), it
5101 was lost. Now it is written to syslog before the two lines that record the
5102 failures to open the logs.
5104 65. Users' Exim filters run in subprocesses under the user's uid. It is
5105 possible for a "deliver" command or an alias in a "personal" command to
5106 provoke an address rewrite. If logging of address rewriting is configured,
5107 this fails because the process is not running as root or exim. There may be
5108 a better way of dealing with this, but for the moment (because 4.30 needs
5109 to be released), I have disabled address rewrite logging when running a
5110 filter in a non-root, non-exim process.
5116 1. The buildconfig auxiliary program wasn't quoting the value set for
5117 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5118 not defined. This bug was masked in 4.22 by the effect that was fixed in
5121 2. Some messages that were rejected after a message id was allocated were
5122 shown as "incomplete" by exigrep. It no longer does this for messages that
5123 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5125 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5126 have allow_domain_literals set, the ID did not get logged in the <= line.
5127 Domain literals are now always recognized in Message-ID: header lines.
5129 4. The first argument for a ${extract expansion item is the key name or field
5130 number. Leading and trailing spaces in this item were not being ignored,
5131 causing some misleading effects.
5133 5. When deliver_drop_privilege was set, single queue runner processes started
5134 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5135 same command in the process it spins off) were not dropping privilege.
5137 6. When the daemon running as "exim" started a queue runner, it always
5138 re-executed Exim in the spun-off process. This is a waste of effort when
5139 deliver_drop_privilege is set. The new process now just calls the
5140 queue-runner function directly.
5146 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5149 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5150 ignored. Though the use of -oP was forcing the writing of a pid file, it
5151 was always written to the default place.
5153 3. If the message "no IP address found for host xxxx" is generated during
5154 incoming verification, it is now followed by identification of the incoming
5155 connection (so you can more easily find what provoked it).
5157 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5159 5. Added some features to "harden" Exim a bit more against certain attacks:
5161 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5162 be put in Local/Makefile. This is like the never_users runtime option,
5163 but it cannot be overridden. The default setting is "root".
5165 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5166 prefix string with which any file named in a -C command line option
5169 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5170 is retained for -C and -D only if the caller of Exim is root. Without
5171 it, the exim user may also use -C and -D and retain privilege.
5173 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5174 command line option is disabled.
5176 6. Macro names set by the -D option must start with an upper case letter, just
5177 like macro names defined in the configuration file.
5179 7. Added "dereference=" facility to LDAP.
5181 8. Two instances of the typo "uknown" in the source files are fixed.
5183 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5184 the Configure-Makefile script screwed up while processing it.
5186 10. Incorporated PCRE 4.4.
5188 11. The SMTP synchronization check was not operating right at the start of an
5189 SMTP session. For example, it could not catch a HELO sent before the client
5190 waited for the greeting. There is now a check for outstanding input at the
5191 point when the greeting is written. Because of the duplex, asynchronous
5192 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5193 way, but not yet received, when the check is performed.
5195 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5196 on TCP/IP sockets, because this apparently causes some broken clients to
5199 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5200 unchanged) from the Cygwin maintainer.
5202 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5203 was supported instead of testing for mbx. Effectively a typo.
5205 15. The spa authenticator server code was not checking that the input it
5206 received was valid base64.
5208 16. The debug output line for the "set" modifier in ACLs was not showing the
5209 name of the variable that was being set.
5211 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5213 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5214 lookup on its own. Something else was needed to trigger the lookup. For
5215 example, a match in host_lookup or the need for a host name in a host list.
5216 Now, if $sender_host_name is referenced and the host name has not yet been
5217 looked up, a lookup is performed. If the lookup fails, the variable remains
5218 empty, and $host_lookup_failed is set to "1".
5220 19. Added "eqi" as a case-independent comparison operator.
5222 20. The saslauthd authentication condition could segfault if neither service
5223 nor realm was specified.
5225 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5226 error message that was logged was misleading, and incoming SMTP
5227 connections were dropped. The message is now more accurate, and temporary
5228 errors are given to SMTP connections.
5230 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5231 (except RSET) with a 421 error, until QUIT is received. However, it was
5232 failing to send a response to QUIT.
5234 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5235 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5236 ACL is now run after the helo_try_verify_hosts code.
5238 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5239 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5240 case-independent, so other case variants are also recognized). Apparently
5241 some systems use these upper case variants.
5243 25. If more than two messages were waiting for the same host, and a transport
5244 filter was specified for the transport, Exim sent two messages over the
5245 same TCP/IP connection, and then failed with "socket operation on non-
5246 socket" when it tried to send the third.
5248 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5250 27. The extern definition of crypt16() in expand.c was not being excluded when
5251 the OS had its own crypt16() function.
5253 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5254 as a preferred synonym for return_size_limit, both as an option and as an
5257 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5259 30. Changed the default configuration ACL to relax the local part checking rule
5260 for addresses that are not in any local domains. For these addresses,
5261 slashes and pipe symbols are allowed within local parts, but the sequence
5262 /../ is explicitly forbidden.
5264 31. SPA server authentication was not clearing the challenge buffer before
5267 32. log_message in a "warn" ACL statement was writing to the reject log as
5268 well as to the main log, which contradicts the documentation and doesn't
5269 seem right (because no rejection is happening). So I have stopped it.
5271 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5272 However, I am unable to do any testing of this.
5274 34. Fixed an infelicity in the appendfile transport. When checking directories
5275 for a mailbox, to see if any needed to be created, it was accidentally
5276 using path names with one or more superfluous leading slashes; tracing
5277 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5279 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5280 contents are added to the log line that is written for every discarded
5281 recipient. (Previously a log_message setting was ignored.)
5283 36. The ${quote: operator now quotes the string if it is empty.
5285 37. The install script runs exim in order to find its version number. If for
5286 some reason other than non-existence or emptiness, which it checks, it
5287 could not run './exim', it was installing it with an empty version number,
5288 i.e. as "exim-". This error state is now caught, and the installation is
5291 38. An argument was missing from the function that creates an error message
5292 when Exim fails to connect to the socket for saslauthd authentication.
5293 This could cause Exim to crash, or give a corrupted message.
5295 39. Added isip, isip4, and isip6 to ${if conditions.
5297 40. The ACL variables $acl_xx are now saved with the message, and can be
5298 accessed later in routers, transports, and filters.
5300 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5301 strings in the file are not string-expanded.
5303 42. If a MAIL command specified a SIZE value that was too large to fit into an
5304 int variable, the check against message_size_limit failed. Such values are
5305 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5306 day this will have to be increased, but I don't think I want to be around
5307 when emails are that large.
5314 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5315 iconv() is not standard in FreeBSD.
5317 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5318 IPv6 enabled. The observed symptom was a segmentation fault on return from
5319 the function os_common_find_running_interfaces() in src/os.c.
5321 3. In the check_special_case() function in daemon.c I had used "errno" as an
5322 argument name, which causes warnings on some systems. This was basically a
5323 typo, since it was named "eno" in the comments!
5325 4. The code that waits for the clock to tick (at a resolution of some fraction
5326 of a second) so as to ensure message-id uniqueness was always waiting for
5327 at least one whole tick, when it could have waited for less. [This is
5328 almost certainly not relevant at current processor speeds, where it is
5329 unlikely to ever wait at all. But we try to future-proof.]
5331 5. The function that sleeps for a time interval that includes fractions of a
5332 second contained a race. It did not block SIGALRM between setting the
5333 timer, and suspending (a couple of lines later). If the interval was short
5334 and the sigsuspend() was delayed until after it had expired, the suspension
5335 never ended. On busy systems this could lead to processes getting stuck for
5338 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5339 process, before it forks any delivery processes. The open lookup caching
5340 mechanism meant that the open file or database connection was passed into
5341 the delivery process. The problem was that delivery processes always tidy
5342 up cached lookup data. This could cause a problem for the next delivery
5343 process started by the queue runner, because the external queue runner
5344 process does not know about the closure. So the next delivery process
5345 still has data in the lookup cache. In the case of a file lookup, there was
5346 no problem because closing a file descriptor in a subprocess doesn't affect
5347 the parent. However, if the lookup was caching a connection to a database,
5348 the connection was closed, and the second delivery process was likely to
5349 see errors such as "PGSQL: query failed: server closed the connection
5350 unexpectedly". The problem has been fixed by closing all cached lookups
5351 in a queue runner before running a delivery process.
5353 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5354 seem to have the "const" qualifier which it has on other OS. I've
5357 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5358 *of the same type* (client or server) with the same public name that an
5359 error should be diagnosed.
5361 9. When Exim looked up a host name for an IP address, but failed to find the
5362 original IP address when looking up the host name (a safety check), it
5363 output the message "<ip address> does not match any IP for NULL", which was
5364 confusing, to say the least. The bug was that the host name should have
5365 appeared instead of "NULL".
5367 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5368 user that is built into the binary, and the -C or -D options is used, root
5369 privilege is dropped before the configuration file is read. In addition,
5370 logging is switched to stderr instead of the normal log files. If the
5371 configuration then re-defines the Exim user, the unprivileged environment
5372 is probably not what is expected, so Exim logs a panic warning message (but
5375 However, if deliver_drop_privilege is set, the unprivileged state may well
5376 be exactly what is intended, so the warning has been cut out in that case,
5377 and Exim is allowed to try to write to its normal log files.
5383 1. smtp_return_error_details was not giving details for temporary sender
5384 or receiver verification errors.
5386 2. Diagnose a configuration error if two authenticators have the same public
5389 3. Exim used not to create the message log file for a message until the first
5390 delivery attempt. This could be confusing when incoming messages were held
5391 for policy or load reasons. The message log file is now created at the time
5392 the message is received, and an initial "Received" line is written to it.
5394 4. The automatically generated man page for command line options had a minor
5395 bug that caused no ill effects; however, a more serious problem was that
5396 the procedure for building the man page automatically didn't always
5397 operate. Consequently, release 4.20 contains an out-of-date version. This
5398 shouldn't happen again.
5400 5. When building Exim with embedded Perl support, the script that builds the
5401 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5402 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5404 6. The freeze_tell option was not being used for messages that were frozen on
5405 arrival, either by an ACL or by local_scan().
5407 7. Added the smtp_incomplete_transaction log selector.
5409 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5410 was accepting AUTH without a new EHLO.
5412 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5413 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5416 10. Logging of TCP/IP connections (when configured) now happens in the main
5417 daemon process instead of the child process, so that the TCP/IP connection
5418 count is more accurate (but it can never be perfect).
5420 11. The use of "drop" in a nested ACL was not being handled correctly in the
5421 outer ACL. Now, if condition failure induced by the nested "drop" causes
5422 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5423 or "require"), the connection is dropped.
5425 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5426 that yield "discard" can now be used with an "accept" or a "discard" verb,
5427 but an error is generated for any others (because I can't see a useful way
5428 to define what should happen).
5430 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5431 are now processed in the same way as lines in the Exim configuration file.
5432 In particular, continuation lines are supported.
5434 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5436 15. Added -ti meaning -t -i.
5438 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5439 domains, and warn by logging if others are found.
5441 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5442 returned from GIFCONF when Exim is trying to find the list of interfaces on
5443 a host. The code in os.c has been modified to copy each ifreq to an aligned
5444 structure in all cases.
5446 Also, in some cases, the returned ifreq's were being copied to a 'struct
5447 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5448 means the last couple of bytes of an IPv6 address could be chopped if the
5449 ifreq contained only a normal sockaddr (14 bytes storage).
5451 18. Named domain lists were not supported in the hosts_treat_as_local option.
5452 An entry such as +xxxx was not recognized, and was treated as a literal
5455 19. Ensure that header lines added by a DATA ACL are included in the reject log
5456 if the ACL subsequently rejects the message.
5458 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5459 MD5 (which is deprecated).
5461 21. When testing a filter file using -bf, Exim was writing a message when it
5462 took the sender from a "From " line in the message, but it was not doing so
5463 when it took $return_path from a Return-Path: header line. It now does.
5465 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5466 with a valid header line field name (a series of printing characters
5467 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5469 23. Changed "disc" in the source to "disk" to conform to the documentation and
5470 the book and for uniformity.
5472 24. Ignore Sendmail's -Ooption=value command line item.
5474 25. When execve() failed while trying to run a command in a pipe transport,
5475 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5476 could be confused with a return value of 69 from the command itself. This
5477 has been changed to 127, the value the shell returns if it is asked to run
5478 a non-existent command. The wording for the related log line suggests a
5479 non-existent command as the problem.
5481 26. If received_header_text expands to an empty string, do not add a Received:
5482 header line to the message. (Well, it adds a token one on the spool, but
5483 marks it "old" so that it doesn't get used or transmitted.)
5485 27. Installed eximstats 1.28 (addition of -nt option).
5487 28. There was no check for failure on the call to getsockname() in the daemon
5488 code. This can fail if there is a shortage of resources on the system, with
5489 ENOMEM, for example. A temporary error is now given on failure.
5491 29. Contrary to the C standard, it seems that in some environments, the
5492 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5493 program. Exim now does this explicitly; it affects the formatting of
5494 timestamps using strftime().
5496 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5497 complaints. I've now initialized all the variables, to avoid this.
5499 32. Header lines added by a system filter were not being "seen" during
5500 transport-time rewrites.
5502 33. The info_callback() function passed to OpenSSL is set up with type void
5503 (*)(SSL *, int, int), as described somewhere. However, when calling the
5504 function (actually a macro) that sets it up, the type void(*)() is
5505 expected. I've put in a cast to prevent warnings from picky compilers.
5507 34. If a DNS black list lookup found a CNAME record, but there were no A
5508 records associated with the domain it pointed at, Exim crashed.
5510 35. If a DNS black list lookup returned more than one A record, Exim ignored
5511 all but the first. It now scans all returned addresses if a particular IP
5512 value is being sought. In this situation, the contents of the
5513 $dnslist_value variable are a list of all the addresses, separated by a
5516 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5517 to accept a host name and all its aliases if the forward lookup for any of
5518 them yielded the IP address of the incoming connection. Now it accepts only
5519 those names whose forward lookup yields the correct IP address. Any other
5520 names are discarded. This closes a loophole whereby a rogue DNS
5521 administrator could create reverse DNS records to break through a
5522 wildcarded host restriction in an ACL.
5524 37. If a user filter or a system filter that ran in a subprocess used any of
5525 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5526 the wrong values were passed to the pipe command ($thisaddress had the
5527 value of $0, $0 had the value of $1, etc). This bug was introduced by
5528 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5530 38. Improved the line breaking for long SMTP error messages from ACLs.
5531 Previously, if there was no break point between 40 and 75 characters, Exim
5532 left the rest of the message alone. Two changes have been made: (a) I've
5533 reduced the minimum length to 35 characters; (b) if it can't find a break
5534 point between 35 and 75 characters, it looks ahead and uses the first one
5535 that it finds. This may give the occasional overlong line, but at least the
5536 remaining text gets split now.
5538 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5539 file descriptors might be low, and that setting 1000 would always raise it.
5540 It turns out that in some environments, the limit is already over 1000 and
5541 that lowering it causes trouble. So now Exim takes care not to decrease it.
5543 40. When delivering a message, the value of $return_path is set to $sender_
5544 address at the start of routing (routers may change the value). By an
5545 oversight, this default was not being set up when an address was tested by
5546 -bt or -bv, which affected the outcome if any router or filter referred to
5549 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5550 log only when "message" or "log_message" are set. However, if one of the
5551 conditions was an address verification, or a call to a nested ACL, the
5552 messages generated by the underlying test were being passed through. This
5553 no longer happens. The underlying message is available in $acl_verify_
5554 message for both "message" and "log_message" expansions, so it can be
5555 passed through if needed.
5557 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5558 new expansion $bh_ to give the encoded byte string without charset
5559 translation. Translation happens only if iconv() is available; HAVE_ICONV
5560 indicates this at build time. HEADERS_CHARSET gives the charset to
5561 translate to; headers_charset can change it in the configuration, and
5562 "headers charset" can change it in an individual filter file.
5564 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5565 that creates RFC 2047 encoded "words" labels them as that charset instead
5566 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5567 expansion operator; (ii) when Exim creates a From: line for a local
5568 message; (iii) when a header line is rewritten to include a "phrase" part.
5570 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5571 buggy, causing it to skip the first lines of messages whose message ID
5572 ended in 'D'. This would not have bitten before Exim release 4.14, because
5573 message IDs were unlikely to end in 'D' before then. The effect was to have
5574 incorrect size information for certain domains.
5576 45. #include "config.h" was missing at the start of the crypt16.c module. This
5577 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5580 46. If there was a timeout during a "random" callout check, Exim treated it as
5581 a failure of the random address, and carried on sending RSET and the real
5582 address. If the delay was just some slowness somewhere, the response to the
5583 original RCPT would be taken as a response to RSET and so on, causing
5584 mayhem of various kinds.
5586 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5587 when I implemented it. It didn't allow for the fact that some option values
5588 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5589 the right test for positive values.
5591 48. Domain names in DNS records are case-independent. Exim always looks them up
5592 in lower case. Some resolvers return domain names in exactly the case they
5593 appear in the zone file, that is, they may contain uppercase letters. Not
5594 all resolvers do this - some return always lower case. Exim was treating a
5595 change of case by a resolver as a change of domain, similar to a widening
5596 of a domain abbreviation. This triggered its re-routing code and so it was
5597 trying to route what was effectively the same domain again. This normally
5598 caused routing to fail (because the router wouldn't handle the domain
5599 twice). Now Exim checks for this case specially, and just changes the
5600 casing of the domain that it ultimately uses when it transmits the message
5603 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5606 50. If a filter generated a file delivery with a non-absolute name (possible if
5607 no home directory exists for the router), the forbid_file option was not
5610 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5611 the existing equality matching.
5613 52. Exim was using ints instead of ino_t variables in some places where it was
5614 dealing with inode numbers.
5616 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5617 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5618 called TMPDIR, and if it finds it is different, it changes its value.
5620 54. The smtp_printf() function is now made available to local_scan() so
5621 additional output lines can be written before returning. There is also an
5622 smtp_fflush() function to enable the detection of a dropped connection.
5623 The variables smtp_input and smtp_batched_input are exported to
5626 55. Changed the default runtime configuration: the message "Unknown user"
5627 has been removed from the ACL, and instead placed on the localuser router,
5628 using the cannot_route_message feature. This means that any verification
5629 failures that generate their own messages won't get overridden. Similarly,
5630 the "Unrouteable address" message that was in the ACL for unverifiable
5631 relay addresses has also been removed.
5633 56. Added hosts_avoid_esmtp to the smtp transport.
5635 57. The exicyclog script was not checking for the esoteric option
5636 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5637 will work only if exicyclog is run under the appropriate euid.
5639 58. Following a discussion on the list, the rules by which Exim recognises line
5640 endings on incoming messages have been changed. The -dropcr and drop_cr
5641 options are now no-ops, retained only for backwards compatibility. The
5642 following line terminators are recognized: LF CRLF CR. However, special
5643 processing applies to CR:
5645 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5646 nor a local message in the state where . is a terminator.
5648 (ii) If a bare CR is encountered in a header line, an extra space is added
5649 after the line terminator so as not to end the header. The reasoning
5650 behind this is that bare CRs in header lines are most likely either
5651 to be mistakes, or people trying to play silly games.
5653 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5654 was being incorrectly given as 18 bytes larger than it should have been.
5655 This is a VOB (very old bug).
5657 60. This may never have affected anything current, but just in case it has:
5658 When the local host is found other than at the start of a list of hosts,
5659 the local host, those with the same MX, and any that follow, are discarded.
5660 When the list in question was part of a longer list of hosts, the following
5661 hosts (not currently being processed) were also being discarded. This no
5662 longer happens. I'm not sure if this situation could ever has previously
5665 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5668 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5669 header. This is something that is recommended in a new Internet Draft, and
5670 is something that is documented as being done by Sendmail. There are two
5671 possible values. For messages generated by the autoreply transport, Exim
5674 Auto-Submitted: auto-replied
5676 whereas for all other generated messages (e.g. bounces) it adds
5678 Auto-Submitted: auto-generated
5680 63. The "personal" condition in filters now includes a test for the
5681 Auto-Submitted: header. If it contains the string "auto-" the message it
5682 not considered personal.
5684 64. Added rcpt_include_affixes as a generic transport option.
5686 65. Added queue_only_override (default true).
5688 66. Added the syslog_duplication option.
5690 67. If what should have been the first header line of a message consisted of
5691 a space followed by a colon, Exim was mis-interpreting it as a header line.
5692 It isn't of course - it is syntactically invalid and should therefore be
5693 treated as the start of the message body. The misbehaviour could have
5694 caused a number of strange effects, including loss of data in subsequent
5695 header lines, and spool format errors.
5697 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5698 client host had authenticated. This control can now be exercised by an ACL
5699 for more flexibility.
5701 69. By default, callouts do not happen when testing with -bh. There is now a
5702 variant, -bhc, which does actually run the callout code, including
5703 consulting and updating the callout cache.
5705 70. Added support for saslauthd authentication, courtesy of Alexander
5708 71. If statvfs() failed on the spool or log directories while checking their
5709 size for availability, Exim confusingly gave the error "space shortage".
5710 Furthermore, in debugging mode it crashed with a floating point exception.
5711 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5712 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5713 serious problem, Exim now writes to the main and panic logs when this
5714 happens, with details of the failure. It then refuses to accept the
5715 incoming message, giving the message "spool directory problem" or "log
5716 directory problem" with a 421 code for SMTP messages.
5718 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5719 0, 1, and 2 exist, because some OS complain for execs without them (see
5720 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5721 descriptors. However, the code omitted to check that the open succeeded,
5722 causing mysterious errors if for some reason the permissions on /dev/null
5723 got screwed. Now Exim writes a message to the main and panic logs, and
5724 bombs out if it can't open /dev/null.
5726 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5727 interact so that it is all more flexible. It is supposed to remain
5728 backwards compatible. Also added extra_local_interfaces.
5730 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5731 to bomb out with an assertion failure - to the client this appears as a
5732 connection drop. This problem occurs in the part of the code that was taken
5733 from the Samba project. Fortunately, the assertion is in a very simple
5734 function, so I have fixed this by reproducing the function inline in the
5735 one place where it is called, and arranging for authentication to fail
5736 instead of killing the process with assert().
5738 75. The SPA client code was not working when the server requested OEM rather
5739 than Unicode encoding.
5741 76. Added code to make require_files with a specific uid setting more usable in
5742 the case where statting the file as root fails - usually a non-root-mounted
5743 NFS file system. When this happens and the failure is EACCES, Exim now
5744 forks a subprocess and does the per-uid checking as the relevant uid.
5746 77. Added process_log_path.
5748 78. If log_file_path was not explicitly set, a setting of check_log_space or
5749 check_log_inodes was ignored.
5751 79. If a space check for the spool or log partitions fails, the incident is now
5752 logged. Of course, in the latter case the data may get lost...
5754 80. Added the %p formatting code to string_format() so that it can be used to
5755 print addresses in debug_print(). Adjusted all the address printing in the
5756 debugging in store.c to use %p rather than %d.
5758 81. There was a concern that a line of code in smtp_in.c could overflow a
5759 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5760 initially expressed, the concern was not well-founded, because trailing
5761 spaces are removed early. However, if the trailing spaces were followed by
5762 a NULL, they did not get removed, so the overflow was possible. Two fixes
5765 (a) I re-wrote the offending code in a cleaner fashion.
5766 (b) If an incoming SMTP command contains a NULL character, it is rejected
5769 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5770 runs initgroups(), so that if the Exim user is in any additional groups,
5771 they will be used during message reception.
5777 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob