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 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
38 JH/02 Support "G" suffix to numbers in ${if comparisons.
40 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
42 NM/01 Bugzilla 1197 - Spec typo
43 Bugzilla 1196 - Spec examples corrections
45 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
47 PP/09 Add gnutls_enable_pkcs11 option.
53 PP/01 Handle short writes when writing local log-files.
54 In practice, only affects FreeBSD (8 onwards).
55 Bugzilla 1053, with thanks to Dmitry Isaikin.
57 NM/01 Bugzilla 949 - Documentation tweak
59 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
62 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
64 PP/02 Implemented gsasl authenticator.
66 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
68 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
69 `pkg-config foo` for cflags/libs.
71 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
72 with rest of GSASL and with heimdal_gssapi.
74 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
75 `pkg-config foo` for cflags/libs for the TLS implementation.
77 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
78 properties get this fed in as external SSF. A number of robustness
79 and debugging improvements to the cyrus_sasl authenticator.
81 PP/08 cyrus_sasl server now expands the server_realm option.
83 PP/09 Bugzilla 1214 - Log authentication information in reject log.
84 Patch by Jeremy Harris.
86 PP/10 Added dbmjz lookup type.
88 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
90 PP/12 MAIL args handles TAB as well as SP, for better interop with
91 non-compliant senders.
92 Analysis and variant patch by Todd Lyons.
94 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
95 Bug report from Lars Müller <lars@samba.org> (via SUSE),
96 Patch from Dirk Mueller <dmueller@suse.com>
98 PP/13 tls_peerdn now print-escaped for spool files.
99 Observed some $tls_peerdn in wild which contained \n, which resulted
100 in spool file corruption.
102 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
103 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
104 or write after TLS renegotiation, which otherwise led to messages
107 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
108 as a tracking header (ie: a signed header comes before the signature).
109 Patch from Wolfgang Breyha.
111 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
112 comma-sep list; embedded commas doubled.
114 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
116 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
118 Report and patch from Dmitry Banschikov.
120 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
121 Removed SSL_clear() after SSL_new() which led to protocol negotiation
122 failures. We appear to now support TLS1.1+ with Exim.
124 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
125 lets Exim select keys and certificates based upon TLS SNI from client.
126 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
127 before an outbound SMTP session. New log_selector, +tls_sni.
129 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
130 NULL dereference. Report and patch from Alun Jones.
132 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
133 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
134 Not seeing resolver debug output on NetBSD, but suspect this is a
135 resolver implementation change.
137 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
138 Left warnings. Added "eximon gdb" invocation mode.
140 PP/21 Defaulting "accept_8bitmime" to true, not false.
142 PP/22 Added -bw for inetd wait mode support.
144 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
145 locate the relevant includes and libraries. Made this the default.
147 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
148 Bugzilla 1246, report and most of solution from Tomasz Kusy.
150 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
151 This may cause build issues on older platforms.
153 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
154 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
155 gnutls_require_mac & gnutls_require_protocols (no longer supported).
156 Added SNI support via GnuTLS too.
157 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
159 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
161 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
162 Applied second patch from Janne, implementing suggestion to default
163 multiple-strings-in-record handling to match SPF spec.
165 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
167 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
168 read-only, out of scope).
169 Patch from Wolfgang Breyha, report from Stuart Northfield.
171 PP/29 Fix three issues highlighted by clang analyser static analysis.
172 Only crash-plausible issue would require the Cambridge-specific
173 iplookup router and a misconfiguration.
174 Report from Marcin Mirosław.
176 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
178 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
179 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
180 As part of this, removing so much warning spew let me fix some minor
181 real issues in debug logging.
183 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
184 assignment on my part. Fixed.
186 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
187 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
188 Janne Snabb (who went above and beyond: thank you).
190 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
191 string otherwise requires a connection and a bunch more work and it's
192 relatively easy to get wrong. Should also expose TLS library linkage
195 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
196 64-bit ${eval} (JH/03).
198 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
199 GNU libc to support some of the 64-bit stuff, should not lead to
200 conflicts. Defined before os.h is pulled in, so if a given platform
201 needs to override this, it can.
203 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
204 protection layer was required, which is not implemented.
205 Bugzilla 1254, patch from Wolfgang Breyha.
207 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
208 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
209 tls_dhparam take prime identifiers. Also unbreak combination of
210 OpenSSL+DH_params+TLSSNI.
212 PP/39 Disable SSLv2 by default in OpenSSL support.
218 PP/01 Solaris build fix for Oracle's LDAP libraries.
219 Bugzilla 1109, patch from Stephen Usher.
221 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
223 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
226 TF/02 Fix a couple more cases where we did not log the error message
227 when unlink() failed. See also change 4.74-TF/03.
229 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
230 lock up or crash if it happened to be inside a call to libc when it
231 got a SIGUSR1 from exiwhat.
233 The SIGUSR1 handler appends the current process status to the process
234 log which is later printed by exiwhat. It used to use the general
235 purpose logging code to do this, but several functions it calls are
236 not safe for signals.
238 The new output code in the SIGUSR1 handler is specific to the process
239 log, and simple enough that it's easy to inspect for signal safety.
240 Removing some special cases also simplifies the general logging code.
241 Removing the spurious timestamps from the process log simplifies
244 TF/04 Improved ratelimit ACL condition.
246 The /noupdate option has been deprecated in favour of /readonly which
247 has clearer semantics. The /leaky, /strict, and /readonly update modes
248 are mutually exclusive. The update mode is no longer included in the
249 database key; it just determines when the database is updated. (This
250 means that when you upgrde Exim will forget old rate measurements.)
252 Exim now checks that the per_* options are used with an update mode that
253 makes sense for the current ACL. For example, when Exim is processing a
254 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
255 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
256 must specify per_mail/readonly. If you omit the update mode it defaults to
257 /leaky where that makes sense (as before) or /readonly where required.
259 The /noupdate option is now undocumented but still supported for
260 backwards compatibility. It is equivalent to /readonly except that in
261 ACLs where /readonly is required you may specify /leaky/noupdate or
262 /strict/noupdate which are treated the same as /readonly.
264 A useful new feature is the /count= option. This is a generalization
265 of the per_byte option, so that you can measure the throughput of other
266 aggregate values. For example, the per_byte option is now equivalent
267 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
269 The per_rcpt option has been generalized using the /count= mechanism
270 (though it's more complicated than the per_byte equivalence). When it is
271 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
272 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
273 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
274 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
275 Note that using per_rcpt with a non-readonly update mode in more than
276 one ACL will cause the recipients to be double-counted. (The per_mail
277 and per_byte options don't have this problem.)
279 The handling of very low rates has changed slightly. If the computed rate
280 is less than the event's count (usually one) then this event is the first
281 after a long gap. In this case the rate is set to the same as this event's
282 count, so that the first message of a spam run is counted properly.
284 The major new feature is a mechanism for counting the rate of unique
285 events. The new per_addr option counts the number of different
286 recipients that someone has sent messages to in the last time period. It
287 behaves like per_rcpt if all the recipient addresses are different, but
288 duplicate recipient addresses do not increase the measured rate. Like
289 the /count= option this is a general mechanism, so the per_addr option
290 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
291 example, measure the rate that a client uses different sender addresses
292 with the options per_mail/unique=$sender_address. There are further
293 details in the main documentation.
295 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
297 TF/06 Removed a few PCRE remnants.
299 TF/07 Automatically extract Exim's version number from tags in the git
300 repository when doing development or release builds.
302 PP/02 Raise smtp_cmd_buffer_size to 16kB.
303 Bugzilla 879. Patch from Paul Fisher.
305 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
306 Heavily based on revision 40f9a89a from Simon Arlott's tree.
309 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
311 PP/05 Variable $av_failed, true if the AV scanner deferred.
312 Bugzilla 1078. Patch from John Horne.
314 PP/06 Stop make process more reliably on build failure.
315 Bugzilla 1087. Patch from Heiko Schlittermann.
317 PP/07 Make maildir_use_size_file an _expandable_ boolean.
318 Bugzilla 1089. Patch from Heiko Schlittermann.
320 PP/08 Handle ${run} returning more data than OS pipe buffer size.
321 Bugzilla 1131. Patch from Holger Weiß.
323 PP/09 Handle IPv6 addresses with SPF.
324 Bugzilla 860. Patch from Wolfgang Breyha.
326 PP/10 GnuTLS: support TLS 1.2 & 1.1.
328 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
331 PP/11 match_* no longer expand right-hand-side by default.
332 New compile-time build option, EXPAND_LISTMATCH_RHS.
333 New expansion conditions, "inlist", "inlisti".
335 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
337 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
339 PP/14 fix log_write() format string regression from TF/03.
340 Bugzilla 1152. Patch from Dmitry Isaikin.
346 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
348 PP/02 Harmonised TLS library version reporting; only show if debugging.
349 Layout now matches that introduced for other libraries in 4.74 PP/03.
351 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
353 PP/04 New "dns_use_edns0" global option.
355 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
358 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
359 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
361 TK/01 Updated PolarSSL code to 0.14.2.
362 Bugzilla 1097. Patch from Andreas Metzler.
364 PP/07 Catch divide-by-zero in ${eval:...}.
367 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
370 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
371 format-string attack -- SECURITY: remote arbitrary code execution.
373 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
374 time unintentionally subject to list matching rules, letting the header
375 cause arbitrary Exim lookups (of items which can occur in lists, *not*
376 arbitrary string expansion). This allowed for information disclosure.
378 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
379 INT_MIN/-1 -- value coerced to INT_MAX.
385 NM/01 Workround for PCRE version dependency in version reporting
388 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
389 This fixes portability to compilers other than gcc, notably
390 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
392 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
393 makefiles for portability to HP-UX and POSIX correctness.
395 PP/01 Permit LOOKUP_foo enabling on the make command-line.
396 Also via indented variable definition in the Makefile.
397 (Debugging by Oliver Heesakkers).
399 PP/02 Restore caching of spamd results with expanded spamd_address.
400 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
402 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
403 Improves build reliability. Fix from: Frank Elsner
405 NM/02 Fix wide character breakage in the rfc2047 coding
406 Fixes bug 1064. Patch from Andrey N. Oktyabrski
408 NM/03 Allow underscore in dnslist lookups
409 Fixes bug 1026. Patch from Graeme Fowler
411 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
412 Code patches from Adam Ciarcinski of NetBSD.
414 NM/04 Fixed exiqgrep to cope with mailq missing size issue
417 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
418 is logged, to avoid truncation. Patch from John Horne.
420 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
421 Patch from Jakob Hirsch.
423 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
424 SQL string expansion failure details.
425 Patch from Andrey Oktyabrski.
427 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
428 Patch from Simon Arlott.
430 PP/09 New lookups functionality failed to compile on old gcc which rejects
431 extern declarations in function scope.
432 Patch from Oliver Fleischmann
434 PP/10 Use sig_atomic_t for flags set from signal handlers.
435 Check getgroups() return and improve debugging.
436 Fixed developed for diagnosis in bug 927 (which turned out to be
439 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
440 Patch from Mark Zealey.
442 PP/12 Bugzilla 1056: Improved spamd server selection.
443 Patch from Mark Zealey.
445 PP/13 Bugzilla 1086: Deal with maildir quota file races.
446 Based on patch from Heiko Schlittermann.
448 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
449 Patch from Uwe Doering, sign-off by Michael Haardt.
451 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
452 variable declaration deep within a block. Bug and patch from
455 PP/15 lookups-Makefile IRIX compatibilty coercion.
457 PP/16 Make DISABLE_DKIM build knob functional.
459 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
460 Patch by Simon Arlott
462 TF/03 Fix valgrind.h portability to C89 compilers that do not support
463 variable argument macros. Our copy now differs from upstream.
469 TF/01 Failure to get a lock on a hints database can have serious
470 consequences so log it to the panic log.
472 TF/02 Log LMTP confirmation messages in the same way as SMTP,
473 controlled using the smtp_confirmation log selector.
475 TF/03 Include the error message when we fail to unlink a spool file.
477 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
478 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
479 for maintaining out-of-tree patches for some time.
481 PP/01 Bugzilla 139: Documentation and portability issues.
482 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
483 Handle per-OS dynamic-module compilation flags.
485 PP/02 Let /dev/null have normal permissions.
486 The 4.73 fixes were a little too stringent and complained about the
487 permissions on /dev/null. Exempt it from some checks.
488 Reported by Andreas M. Kirchwitz.
490 PP/03 Report version information for many libraries, including
491 Exim version information for dynamically loaded libraries. Created
492 version.h, now support a version extension string for distributors
493 who patch heavily. Dynamic module ABI change.
495 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
496 privilege escalation vulnerability whereby the Exim run-time user
497 can cause root to append content of the attacker's choosing to
500 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
503 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
504 If dropping privileges for untrusted macros, we disabled normal logging
505 on the basis that it would fail; for the Exim run-time user, this is not
506 the case, and it resulted in successful deliveries going unlogged.
507 Fixed. Reported by Andreas Metzler.
513 PP/01 Date: & Message-Id: revert to normally being appended to a message,
514 only prepend for the Resent-* case. Fixes regression introduced in
515 Exim 4.70 by NM/22 for Bugzilla 607.
517 PP/02 Include check_rfc2047_length in configure.default because we're seeing
518 increasing numbers of administrators be bitten by this.
520 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
522 PP/03 Bugzilla 994: added openssl_options main configuration option.
524 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
526 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
528 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
530 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
531 without a peer certificate, leading to a segfault because of an
532 assumption that peers always have certificates. Be a little more
533 paranoid. Problem reported by Martin Tscholak.
535 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
536 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
537 NB: ClamAV planning to remove STREAM in "middle of 2010".
538 CL also introduces -bmalware, various -d+acl logging additions and
539 more caution in buffer sizes.
541 PP/09 Implemented reverse_ip expansion operator.
543 PP/10 Bugzilla 937: provide a "debug" ACL control.
545 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
547 PP/12 Bugzilla 973: Implement --version.
549 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
551 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
553 PP/15 Bugzilla 816: support multiple condition rules on Routers.
555 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
556 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
557 ignore trailing whitespace.
559 JJ/02 prevent non-panic DKIM error from being sent to paniclog
561 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
564 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
565 Notification from Dr Andrew Aitchison.
567 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
568 ExtendedDetectionInfo response format.
569 Notification from John Horne.
571 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
574 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
575 XSL and documented dependency on system catalogs, with examples of how
578 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
581 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
582 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
583 configuration file which is writeable by the Exim user or group.
585 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
586 of configuration files to cover files specified with the -C option if
587 they are going to be used with root privileges, not just the default
590 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
591 option (effectively making it always true).
593 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
594 files to be used while preserving root privileges.
596 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
597 that rogue child processes cannot use them.
599 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
600 run-time user, instead of root.
602 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
603 Exim run-time user without dropping privileges.
605 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
606 result string, instead of calling string_vformat() twice with the same
609 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
610 for other users. Others should always drop root privileges if they use
611 -C on the command line, even for a whitelisted configure file.
613 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
615 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
621 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
622 $data_path, and $header_path variables; fixed documentation bugs and
625 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
626 exipick to access non-standard spools, including the "frozen" queue
629 NM/01 Bugzilla 965: Support mysql stored procedures.
630 Patch from Alain Williams
632 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
634 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
635 Patch from Andreas Metzler
637 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
638 Patch from Kirill Miazine
640 NM/05 Bugzilla 671: Added umask to procmail example.
642 JJ/03 installed exipick 20100323.0, fixing doc bug
644 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
645 directory. Notification and patch from Dan Rosenberg.
647 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
649 TK/02 Improve log output when DKIM signing operation fails.
651 MH/01 Treat the transport option dkim_domain as a colon separated
652 list, not as a single string, and sign the message with each element,
653 omitting multiple occurences of the same signer.
655 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
656 Bugzilla 985, 986. Patch by Simon Arlott
658 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
659 Patch by Simon Arlott
661 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
662 MBX locking. Notification from Dan Rosenberg.
668 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
670 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
672 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
674 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
676 NM/04 Bugzilla 915: Flag broken perl installation during build.
682 TK/01 Added patch by Johannes Berg that expands the main option
683 "spamd_address" if it starts with a dollar sign.
685 TK/02 Write list of recipients to X-Envelope-Sender header when building
686 the mbox-format spool file for content scanning (suggested by Jakob
689 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
690 (http://www.dcc-servers.net/) support via dccifd. Activated by
691 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
693 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
694 by Mark Daniel Reidel <mr@df.eu>.
696 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
697 When building exim an external PCRE library is now needed -
698 PCRE is a system library on the majority of modern systems.
699 See entry on PCRE_LIBS in EDITME file.
701 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
702 conversation. Added nologin parameter to request.
703 Patch contributed by Kirill Miazine.
705 TF/01 Do not log submission mode rewrites if they do not change the address.
707 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
709 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
710 log files in place. Contributed by Roberto Lima.
712 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
714 TF/03 Bugzilla 615: When checking the local_parts router precondition
715 after a local_part_suffix or local_part_prefix option, Exim now
716 does not use the address's named list lookup cache, since this
717 contains cached lookups for the whole local part.
719 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
720 Robert Millan. Documentation is in experimental-spec.txt.
722 TF/04 Bugzilla 668: Fix parallel build (make -j).
724 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
726 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
727 Patch provided by Jan Srzednicki.
729 TF/05 Leading white space used to be stripped from $spam_report which
730 wrecked the formatting. Now it is preserved.
732 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
733 that they are available at delivery time.
735 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
737 TF/08 TLS error reporting now respects the incoming_interface and
738 incoming_port log selectors.
740 TF/09 Produce a more useful error message if an SMTP transport's hosts
741 setting expands to an empty string.
743 NM/06 Bugzilla 744: EXPN did not work under TLS.
744 Patch provided by Phil Pennock.
746 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
747 Patch provided by Richard Godbee.
749 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
750 acl_smtp_notquit, added index entry.
752 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
753 Patch provided by Eugene Bujak.
755 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
756 accept(). Patch provided by Maxim Dounin.
758 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
759 Patch provided by Phil Pennock.
761 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
763 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
764 Patch provided by Brad "anomie" Jorsch.
766 NM/14 Bugzilla 622: Added timeout setting to transport filter.
767 Patch provided by Dean Brooks.
769 TK/05 Add native DKIM support (does not depend on external libraries).
771 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
772 Patch provided by Graeme Fowler.
774 NM/16 Bugzilla 851: Documentation example syntax fix.
776 NM/17 Changed NOTICE file to remove references to embedded PCRE.
778 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
781 NM/19 Bugzilla 745: TLS version reporting.
782 Patch provided by Phil Pennock.
784 NM/20 Bugzilla 167: bool: condition support.
785 Patch provided by Phil Pennock.
787 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
788 clients. Patch provided by Phil Pennock.
790 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
791 Patch provided by Brad "anomie" Jorsch.
793 NM/23 Bugzilla 687: Fix misparses in eximstats.
794 Patch provided by Heiko Schlittermann.
796 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
797 Patch provided by Heiko Schlittermann.
799 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
800 plus update to original patch.
802 NM/26 Bugzilla 799: Documentation correction for ratelimit.
804 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
805 Patch provided by David Brownlee.
807 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
809 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
811 NM/30 Bugzilla 888: TLS documentation bugfixes.
813 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
815 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
816 Unlike the original bugzilla I have changed all shell scripts in src tree.
818 NM/33 Bugzilla 898: Transport filter timeout fix.
819 Patch by Todd Rinaldo.
821 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
822 Patch by Serge Demonchaux.
824 NM/35 Bugzilla 39: Base64 decode bug fixes.
825 Patch by Jakob Hirsch.
827 NM/36 Bugzilla 909: Correct connect() call in dcc code.
829 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
831 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
833 NM/39 Bugzilla 911: Fixed MakeLinks build script.
839 TK/01 Add preliminary DKIM support. Currently requires a forked version of
840 ALT-N's libdkim that I have put here:
841 http://duncanthrax.net/exim-experimental/
843 Note to Michael Haardt: I had to rename some vars in sieve.c. They
844 were called 'true' and it seems that C99 defines that as a reserved
845 keyword to be used with 'bool' variable types. That means you could
846 not include C99-style headers which use bools without triggering
847 build errors in sieve.c.
849 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
850 as mailq or other aliases. Changed the --help handling significantly
851 to do whats expected. exim_usage() emits usage/help information.
853 SC/01 Added the -bylocaldomain option to eximstats.
855 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
857 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
859 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
865 PH/01 Another patch from the Sieve maintainer.
867 PH/02 When an IPv6 address is converted to a string for single-key lookup
868 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
869 dots are used instead of colons so that keys in lsearch files need not
870 contain colons. This was done some time before quoting was made available
871 in lsearch files. However, iplsearch files do require colons in IPv6 keys
872 (notated using the quote facility) so as to distinguish them from IPv4
873 keys. This meant that lookups for IP addresses in host lists did not work
874 for iplsearch lookups.
876 This has been fixed by arranging for IPv6 addresses to be expressed with
877 colons if the lookup type is iplsearch. This is not incompatible, because
878 previously such lookups could never work.
880 The situation is now rather anomolous, since one *can* have colons in
881 ordinary lsearch keys. However, making the change in all cases is
882 incompatible and would probably break a number of configurations.
884 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
887 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
888 conversion specification without a maximum field width, thereby enabling
889 a rogue spamd server to cause a buffer overflow. While nobody in their
890 right mind would setup Exim to query an untrusted spamd server, an
891 attacker that gains access to a server running spamd could potentially
892 exploit this vulnerability to run arbitrary code as the Exim user.
894 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
895 $primary_hostname instead of what libspf2 thinks the hosts name is.
897 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
898 a directory entry by the name of the lookup key. Previously, if a
899 symlink pointed to a non-existing file or a file in a directory that
900 Exim lacked permissions to read, a lookup for a key matching that
901 symlink would fail. Now it is enough that a matching directory entry
902 exists, symlink or not. (Bugzilla 503.)
904 PH/03 The body_linecount and body_zerocount variables are now exported in the
907 PH/04 Added the $dnslist_matched variable.
909 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
910 This means they are set thereafter only if the connection becomes
913 PH/06 Added the client_condition to authenticators so that some can be skipped
914 by clients under certain conditions.
916 PH/07 The error message for a badly-placed control=no_multiline_responses left
917 "_responses" off the end of the name.
919 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
921 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
922 (without spaces) instead of just copying the configuration text.
924 PH/10 Added the /noupdate option to the ratelimit ACL condition.
926 PH/11 Added $max_received_linelength.
928 PH/12 Added +ignore_defer and +include_defer to host lists.
930 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
931 way in which PCRE > 7.0 is built.
933 PH/14 Implemented queue_only_load_latch.
935 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
936 MAIL command. The effect was to mangle the value on 64-bit systems.
938 PH/16 Another patch from the Sieve maintainer.
940 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
942 PH/18 If a system quota error occurred while trying to create the file for
943 a maildir delivery, the message "Mailbox is full" was not appended to the
944 bounce if the delivery eventually timed out. Change 4.67/27 below applied
945 only to a quota excession during the actual writing of the file.
947 PH/19 It seems that peer DN values may contain newlines (and other non-printing
948 characters?) which causes problems in log lines. The DN values are now
949 passed through string_printing() before being added to log lines.
951 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
952 and InterBase are left for another time.)
954 PH/21 Added message_body_newlines option.
956 PH/22 Guard against possible overflow in moan_check_errorcopy().
958 PH/23 POSIX allows open() to be a macro; guard against that.
960 PH/24 If the recipient of an error message contained an @ in the local part
961 (suitably quoted, of course), incorrect values were put in $domain and
962 $local_part during the evaluation of errors_copy.
968 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
969 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
972 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
973 issue a MAIL command.
975 PH/02 In an ACL statement such as
977 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
979 if a client was not listed at all, or was listed with a value other than
980 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
981 the condition was not true (as it should be), so access was not denied.
982 The bug was that the ! inversion was incorrectly passed on to the second
983 item. This has been fixed.
985 PH/03 Added additional dnslists conditions == and =& which are different from
986 = and & when the dns lookup returns more than one IP address.
988 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
989 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
991 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
992 FSYNC, which compiles an option called disable_fsync that allows for
993 bypassing fsync(). The documentation is heavily laced with warnings.
995 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
997 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
998 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
999 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1000 including adding "make clean"; (3) Added -fPIC when compiling the test
1001 dynamically loaded module, to get rid of a warning.
1003 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1004 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1005 = 0s. The bug is otherwise harmless.
1007 PH/07 There was a bug in the dovecot authenticator such that the value of
1008 $auth1 could be overwritten, and so not correctly preserved, after a
1009 successful authentication. This usually meant that the value preserved by
1010 the server_setid option was incorrect.
1012 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1014 PH/09 Installed PCRE release 7.0.
1016 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1017 run for batched SMTP input. It is now run at the start of every message
1018 in the batch. While fixing this I discovered that the process information
1019 (output by running exiwhat) was not always getting set for -bs and -bS
1020 input. This is fixed, and it now also says "batched" for BSMTP.
1022 PH/11 Added control=no_pipelining.
1024 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1025 patch, slightly modified), and move the expansion of helo_data till after
1026 the connection is made in the smtp transport (so it can use these
1029 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1031 PH/14 Added log_selector = +pid.
1033 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1035 PH/16 Add ${if forany and ${if forall.
1037 PH/17 Added dsn_from option to vary the From: line in DSNs.
1039 PH/18 Flush SMTP output before performing a callout, unless control =
1040 no_callout_flush is set.
1042 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1043 was true (the default) a successful delivery failed to delete the retry
1044 item, thus causing premature timeout of the address. The bug is now
1047 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1049 PH/21 Long custom messages for fakedefer and fakereject are now split up
1050 into multiline reponses in the same way that messages for "deny" and
1051 other ACL rejections are.
1053 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1054 with slight modification.
1056 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1057 draft, changing the syntax and factoring some duplicate code".
1059 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1060 for deliveries of the second and subsequent messages over the same SMTP
1063 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1064 ${reduce, with only minor "tidies".
1066 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1068 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1069 expansion side effects.
1071 PH/27 When a message times out after an over-quota error from an Exim-imposed
1072 quota, the bounce message says "mailbox is full". This message was not
1073 being given when it was a system quota that was exceeded. It now should
1076 MH/03 Made $recipients available in local_scan(). local_scan() already has
1077 better access to the recipient list through recipients_list[], but
1078 $recipients can be useful in postmaster-provided expansion strings.
1080 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1081 in the case of a MAIL command with additional options following the
1082 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1083 were accidentally chopped off.
1085 PH/29 SMTP synchronization checks are implemented when a command is read -
1086 there is a check that no more input is waiting when there shouldn't be
1087 any. However, for some commands, a delay in an ACL can mean that it is
1088 some time before the response is written. In this time, more input might
1089 arrive, invalidly. So now there are extra checks after an ACL has run for
1090 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1091 pipelining has not been advertised.
1093 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1095 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1096 returned in the SMTP response when smtp_return_error_details was false.
1097 This has been fixed.
1099 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1100 instead of the C I/O that was originally supplied, because problems were
1101 reported on Solaris.
1103 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1104 Exim which did not show up earlier: it was assuming that a call to
1105 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1106 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1107 was a macro that became an assignment, so it seemed to work. This has
1108 changed to a proper function call with a void return, hence the compile
1109 error. Exim's code has been fixed.
1111 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1114 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1116 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1118 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1119 because it turns out to be too big or there is a timeout) there is no
1120 "Completed" line in the log. When some messages of this type were
1121 selected by exigrep, they were listed as "not completed". Others were
1122 picked up by some special patterns. I have improved the selection
1123 criteria to be more general.
1125 PH/38 The host_find_failed option in the manualroute router can now be set
1126 to "ignore", to completely ignore a host whose IP address cannot be
1127 found. If all hosts are ignored, the behaviour is controlled by the new
1128 host_all_ignored option.
1130 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1131 homing or because of multiple MX records with /mx) generated more than
1132 one IP address, and the following item turned out to be the local host,
1133 all the secondary addresses of the first item were incorrectly removed
1134 from the list, along with the local host and any following hosts (which
1135 is what is supposed to happen).
1137 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1138 whoever called Exim into the -H file. In the case of the daemon it was
1139 behaving confusingly. When first started, it used values for whoever
1140 started the daemon, but after a SIGHUP it used the Exim user (because it
1141 calls itself on a restart). I have changed the code so that it now always
1144 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1145 message are rejected with the same error (e.g. no authentication or bad
1146 sender address), and a DATA command is nevertheless sent (as can happen
1147 with PIPELINING or a stupid MUA), the error message that was given to the
1148 RCPT commands is included in the rejection of the DATA command. This is
1149 intended to be helpful for MUAs that show only the final error to their
1152 PH/42 Another patch from the Sieve maintainer.
1154 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1155 Eximstats - Fixed some broken HTML links and added missing column headers
1157 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1158 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1160 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1162 PH/43 Yet another patch from the Sieve maintainer.
1164 PH/44 I found a way to check for a TCP/IP connection going away before sending
1165 the response to the final '.' that terminates a message, but only in the
1166 case where the client has not sent further data following the '.'
1167 (unfortunately, this is allowed). However, in many cases there won't be
1168 any further data because there won't be any more messages to send. A call
1169 to select() can be used: if it shows that the input is "ready", there is
1170 either input waiting, or the socket has been closed. An attempt to read
1171 the next input character can distinguish the two cases. Previously, Exim
1172 would have sent an OK response which the client would never have see.
1173 This could lead to message repetition. This fix should cure that, at
1174 least in a lot of common cases.
1176 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1177 advertised in response to EHLO.
1183 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1184 fixed by 4.65/MH/01 (is this a record?) are fixed:
1186 (i) An empty string was always treated as zero by the numeric comparison
1187 operators. This behaviour has been restored.
1189 (ii) It is documented that the numeric comparison operators always treat
1190 their arguments as decimal numbers. This was broken in that numbers
1191 starting with 0 were being interpreted as octal.
1193 While fixing these problems I realized that there was another issue that
1194 hadn't been noticed. Values of message_size_limit (both the global option
1195 and the transport option) were treated as octal if they started with 0.
1196 The documentation was vague. These values are now always treated as
1197 decimal, and I will make that clear in the documentation.
1203 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1204 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1207 MH/01 Don't check that the operands of numeric comparison operators are
1208 integers when their expansion is in "skipping" mode (fixes bug
1209 introduced by 4.64-PH/07).
1211 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1212 child addresses, Exim now panics and dies. Previously, because the count
1213 is held in a short int, deliveries were likely to be lost. As such a
1214 large number of recipients for a single message is ridiculous
1215 (performance will be very, very poor), I have chosen to impose a limit
1216 rather than extend the field.
1222 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1223 leftover -K file (the existence of which was triggered by #402).
1224 While we were at it, introduced process PID as part of the -K
1225 filename. This should rule out race conditions when creating
1228 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1229 processing considerably. Previous code took too long for large mails,
1230 triggering a timeout which in turn triggers #401.
1232 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1233 in the DK code in transports.c. sendfile() is not really portable,
1234 hence the _LINUX specificness.
1236 TF/01 In the add_headers option to the mail command in an Exim filter,
1237 there was a bug that Exim would claim a syntax error in any
1238 header after the first one which had an odd number of characters
1241 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1242 callout verification, Exim cached a "reject" for the entire domain. This
1243 is correct for most verifications, but it is not correct for a recipient
1244 verification with use_sender or use_postmaster set, because in that case
1245 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1246 case of MAIL FROM:<> rejection from other early rejections (e.g.
1247 rejection of HELO). When verifying a recipient using a non-null MAIL
1248 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1249 Whatever the result of the callout, the value of the domain cache is
1250 left unchanged (for any other kind of callout, getting as far as trying
1251 RCPT means that the domain itself is ok).
1253 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1256 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1257 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1258 session, the daemon ignores ECONNECT errors and logs others; it now
1259 ignores EPIPE as well.
1261 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1262 (quoted-printable decoding).
1264 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1265 later the small subsequent patch to fix an introduced bug.
1267 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1269 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1271 PH/08 An error is now given if message_size_limit is specified negative.
1273 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1274 to be given (somewhat) arbitrary names.
1276 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1279 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1280 miscellaneous code fixes
1282 PH/10 Added the log_reject_target ACL modifier to specify where to log
1285 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1286 hostname. This is wrong, because it relates to the incoming message (and
1287 probably the interface on which it is arriving) and not to the outgoing
1288 callout (which could be using a different interface). This has been
1289 changed to use the value of the helo_data option from the smtp transport
1290 instead - this is what is used when a message is actually being sent. If
1291 there is no remote transport (possible with a router that sets up host
1292 addresses), $smtp_active_hostname is used.
1294 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1295 tweaks were necessary in order to get it to work (see also 21 below):
1296 (a) The code assumed that strncpy() returns a negative number on buffer
1297 overflow, which isn't the case. Replaced with Exim's string_format()
1299 (b) There were several signed/unsigned issues. I just did the minimum
1300 hacking in of casts. There is scope for a larger refactoring.
1301 (c) The code used strcasecmp() which is not a standard C function.
1302 Replaced with Exim's strcmpic() function.
1303 (d) The code set only $1; it now sets $auth1 as well.
1304 (e) A simple test gave the error "authentication client didn't specify
1305 service in request". It would seem that Dovecot has changed its
1306 interface. Fortunately there's a specification; I followed it and
1307 changed what the client sends and it appears to be working now.
1309 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1312 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1313 address A is aliased to B and C, where B exists and C does not. Without
1314 -v the output is "A verified" because verification stops after a
1315 successful redirection if more than one address is generated. However,
1316 with -v the child addresses are also verified. Exim was outputting "A
1317 failed to verify" and then showing the successful verification for C,
1318 with its parentage. It now outputs "B failed to verify", showing B's
1319 parentage before showing the successful verification of C.
1321 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1322 look up a TXT record in a specific list after matching in a combined
1325 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1326 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1327 they consult the DNS. I had assumed they would set it the way they
1328 wanted; and indeed my experiments on Linux seem to show that in some
1329 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1330 To be on the safe side, however, I have now made the interface to
1331 host_find_byname() similar to host_find_bydns(), with an argument
1332 containing the DNS resolver options. The host_find_byname() function now
1333 sets these options at its start, just as host_find_bydns() does. The smtp
1334 transport options dns_qualify_single and dns_search_parents are passed to
1335 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1336 of host_find_byname() use the default settings of RES_DEFNAMES
1337 (qualify_single) but not RES_DNSRCH (search_parents).
1339 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1340 spool_read_header() do less string testing, by means of a preliminary
1341 switch on the second character of optional "-foo" lines. (This is
1342 overdue, caused by the large number of possibilities that now exist.
1343 Originally there were few.) While I was there, I also converted the
1344 str(n)cmp tests so they don't re-test the leading "-" and the first
1345 character, in the hope this might squeeze out yet more improvement.
1347 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1348 flag allowing group syntax was set by the header_syntax check but not
1349 turned off, possible causing trouble later; (2) The flag was not being
1350 set at all for the header_verify test, causing "group"-style headers to
1351 be rejected. I have now set it in this case, and also caused header_
1352 verify to ignore an empty address taken from a group. While doing this, I
1353 came across some other cases where the code for allowing group syntax
1354 while scanning a header line wasn't quite right (mostly, not resetting
1355 the flag correctly in the right place). These bugs could have caused
1356 trouble for malformed header lines. I hope it is now all correct.
1358 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1359 with the "reply" argument non-NULL. The code, however (which originally
1360 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1361 but it didn't always do it. This confused somebody who was copying the
1362 code for some other use. I have removed all the tests.
1364 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1365 feature that was used to support insecure browsers during the U.S. crypto
1366 embargo. It requires special client support, and Exim is probably the
1367 only MTA that supported it -- and would never use it because real RSA is
1368 always available. This code has been removed, because it had the bad
1369 effect of slowing Exim down by computing (never used) parameters for the
1370 RSA_EXPORT functionality.
1372 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1373 authenticator to fail if there's a tab character in the incoming data
1374 (there should never be unless someone is messing about, as it's supposed
1375 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1376 if the connection is using TLS or if the remote IP is the same as the
1377 local IP, and the "valid-client-cert option" if a client certificate has
1380 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1381 authenticators. This can be used for authorization after authentication
1382 succeeds. (In the case of plaintext, it servers for both authentication
1385 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1386 if any retry times were supplied.
1388 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1389 connection, where there is no client IP address to check. In this
1390 situation, the verify now always succeeds.
1392 PH/25 Applied John Jetmore's -Mset patch.
1394 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1396 PH/27 In a string expansion for a processed (not raw) header when multiple
1397 headers of the same name were present, leading whitespace was being
1398 removed from all of them, but trailing whitespace was being removed only
1399 from the last one. Now trailing whitespace is removed from each header
1400 before concatenation. Completely empty headers in a concatenation (as
1401 before) are ignored.
1403 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1404 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1406 PH/29 [Removed. This was a change that I later backed out, and forgot to
1407 correct the ChangeLog entry (that I had efficiently created) before
1408 committing the later change.]
1410 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1411 address errors (4xx response to RCPT) over the same connection as other
1412 messages routed to the same hosts. Such deliveries are always "forced",
1413 so retry times are not inspected. This resulted in far too many retries
1414 for the affected addresses. The effect occurred only when there were more
1415 hosts than the hosts_max_try setting in the smtp transport when it had
1416 the 4xx errors. Those hosts that it had tried were not added to the list
1417 of hosts for which the message was waiting, so if all were tried, there
1418 was no problem. Two fixes have been applied:
1420 (i) If there are any address or message errors in an SMTP delivery, none
1421 of the hosts (tried or untried) are now added to the list of hosts
1422 for which the message is waiting, so the message should not be a
1423 candidate for sending over the same connection that was used for a
1424 successful delivery of some other message. This seems entirely
1425 reasonable: after all the message is NOT "waiting for some host".
1426 This is so "obvious" that I'm not sure why it wasn't done
1427 previously. Hope I haven't missed anything, but it can't do any
1428 harm, as the worst effect is to miss an optimization.
1430 (ii) If, despite (i), such a delivery is accidentally attempted, the
1431 routing retry time is respected, so at least it doesn't keep
1432 hammering the server.
1434 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1435 in ${readsocket because some servers need this prod.
1437 PH/32 Added some extra debug output when updating a wait-xxx database.
1439 PH/33 The hint "could be header name not terminated by colon", which has been
1440 given for certain expansion errors for a long time, was not being given
1441 for the ${if def:h_colon_omitted{... case.
1443 PH/34 The spec says: "With one important exception, whenever a domain list is
1444 being scanned, $domain contains the subject domain." There was at least
1445 one case where this was not true.
1447 PH/35 The error "getsockname() failed: connection reset by peer" was being
1448 written to the panic log as well as the main log, but it isn't really
1449 panic-worthy as it just means the connection died rather early on. I have
1450 removed the panic log writing for the ECONNRESET error when getsockname()
1453 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1454 runs only) independently of the message's sender address. This meant
1455 that, if the 4xx error was in fact related to the sender, a different
1456 message to the same recipient with a different sender could confuse
1457 things. In particualar, this can happen when sending to a greylisting
1458 server, but other circumstances could also provoke similar problems.
1459 I have changed the default so that the retry time for these errors is now
1460 based a combination of the sender and recipient addresses. This change
1461 can be overridden by setting address_retry_include_sender=false in the
1464 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1465 remote server are returned as part of bounce messages. This was not
1466 happening for LMTP over a pipe (the lmtp transport), but now it is the
1467 same for both kinds of LMTP.
1469 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1470 in header lines that were in fact CNAMEs. This is no longer the case.
1472 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1473 and queue runs started by the daemon processed all messages. This has
1474 been fixed so that -R and -S can now usefully be given with -q<time>.
1476 PH/40 Import PCRE release 6.7 (fixes some bugs).
1478 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1480 PH/42 Give an error if -q is specified more than once.
1482 PH/43 Renamed the variables $interface_address and $interface_port as
1483 $received_ip_address and $received_port, to make it clear that these
1484 values apply to message reception, and not to the outgoing interface when
1485 a message is delivered. (The old names remain recognized, of course.)
1487 PH/44 There was no timeout on the connect() call when using a Unix domain
1488 socket in the ${readsocket expansion. There now is.
1490 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1491 be meaningful with "accept".
1493 SC/01 Eximstats V1.43
1494 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1496 SC/02 Eximstats V1.44
1497 Use a glob alias rather than an array ref in the generated
1498 parser. This improves both readability and performance.
1500 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1501 Collect SpamAssassin and rejection statistics.
1502 Don't display local sender or destination tables unless
1503 there is data to show.
1504 Added average volumes into the top table text output.
1506 SC/04 Eximstats V1.46
1507 Collect data on the number of addresses (recipients)
1508 as well as the number of messages.
1510 SC/05 Eximstats V1.47
1511 Added 'Message too big' to the list of mail rejection
1512 reasons (thanks to Marco Gaiarin).
1514 SC/06 Eximstats V1.48
1515 Mainlog lines which have GMT offsets and are too short to
1516 have a flag are now skipped.
1518 SC/07 Eximstats V1.49 (Alain Williams)
1519 Added the -emptyok flag.
1521 SC/08 Eximstats V1.50
1522 Fixes for obtaining the IP address from reject messages.
1524 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1525 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1526 whitesspace changes from 4.64-PH/27
1528 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1531 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1532 are found, allow negative numbers in numeric criteria)
1534 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1536 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1537 to match changes made in 4.64-PH/43
1539 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1541 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1542 "domain" is actually an IP address, and force a failure. This locks out
1543 those revolvers/nameservers that support "A-for-A" lookups, in
1544 contravention of the specifications.
1546 PH/48 When a host name was looked up from an IP address, and the subsequent
1547 forward lookup of the name timed out, the host name was left in
1548 $sender_host_name, contrary to the specification.
1550 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1551 restricted to single-key lookups, Exim was not diagnosing an error if
1552 * or *@ was used with a query-style lookup.
1554 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1556 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1557 long ago, but noone interested enough thought of it. Let's just say that
1558 the "1.1" means that there are some new functions that weren't there at
1559 some point in the past.
1561 PH/51 Error processing for expansion failure of helo_data from an smtp
1562 transport during callout processing was broken.
1564 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1565 tested/used via the -bh/-bhc/-bs options.
1567 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1568 bug, fixed in subsequent PCRE releases).
1570 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1571 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1573 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1579 SC/01 Use a glob alias rather than an array ref in eximstats generated
1580 parser. This improves both readability and performance.
1582 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1583 Don't display local sender or destination tables in eximstats unless
1584 there is data to show.
1585 Added average volumes into the eximstats top table text output.
1587 SC/03 Collect data on the number of addresses (recipients) as well
1588 as the number of messages in eximstats.
1590 TF/01 Correct an error in the documentation for the redirect router. Exim
1591 does (usually) call initgroups() when daemonizing.
1593 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1594 with consistent privilege compared to when running as a daemon.
1596 TF/03 Note in the spec that $authenticated_id is not set for local
1597 submissions from trusted users.
1599 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1600 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1602 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1603 by adding some example configuration directives to the default
1604 configuration file. A little bit of work is required to uncomment the
1605 directives and define how usernames and passwords are checked, but
1606 there is now a framework to start from.
1608 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1609 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1610 without this. I don't know how relevant this is to other LDAP libraries.
1612 PH/02 Add the verb name to the "unknown ACL verb" error.
1614 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1616 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1618 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1619 directory not expanded when it should be if an expanded home directory
1620 was set for the address (which is overridden by the transport).
1622 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1625 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1626 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1627 because it is too late at that time, and has no effect.
1629 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1630 security issue with \' (bugzilla #107). I could not use the
1631 PQescapeStringConn() function, because it needs a PGconn value as one of
1634 PH/08 When testing addresses using -bt, indicate those final addresses that
1635 are duplicates that would not cause an additional delivery. At least one
1636 person was confused, thinking that -bt output corresponded to deliveries.
1637 (Suppressing duplicates isn't a good idea as you lose the information
1638 about possibly different redirections that led to the duplicates.)
1640 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1641 systems where poll() doesn't work, in particular OS X.
1643 PH/10 Added more information to debugging output for retry time not reached.
1645 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1646 operations in malware.c.
1648 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1651 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1652 syslog_duplication set false, log lines that would normally be written
1653 both the the main log and to the reject log were not written to syslog at
1656 PH/14 In the default configuration, change the use of "message" in ACL warn
1657 statements to "add_header".
1659 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1660 not followed by a command (e.g. "seen endif").
1662 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1663 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1666 PH/17 Added extra conditions to the default value of delay_warning_condition
1670 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1671 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1672 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1675 The Auto-Submitted: and various List- headers are standardised, whereas I
1676 don't think Precedence: ever was.
1678 PH/18 Refactored debugging code in route_finduser() to show more information,
1679 in particular, the error code if getpwnam() issues one.
1681 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1682 This is apparently needed in addition to the PH/07 change above to avoid
1683 any possible encoding problems.
1685 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1686 but not after initializing Perl.
1688 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1689 output them only if debugging. By default they are written stderr,
1690 apparently, which is not desirable.
1692 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1695 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1698 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1700 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1701 authenticated or an ident call has been made. Suppress the default
1702 values for $authenticated_id and $authenticated_sender (but permit -oMai
1703 and -oMas) when testing with -bh.
1705 PH/24 Re-jigged the order of the tests in the default configuration so that the
1706 tests for valid domains and recipients precede the DNS black list and CSA
1707 tests, on the grounds that those ones are more expensive.
1709 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1710 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1711 This bug exists in every version of Exim that I still have, right back to
1714 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1715 However, an attempt to turn on case-sensitivity in a regex key by
1716 including (?-i) didn't work because the subject string was already
1717 lowercased, and the effects were non-intuitive. It turns out that a
1718 one-line patch can be used to allow (?-i) to work as expected.
1724 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1725 other effects) broke the use of negated acl sub-conditions.
1727 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1730 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1731 "Deny" causes Exim to reject the incoming connection with a 554 error.
1732 Unfortunately, if there is a major crisis, such as a disk failure,
1733 tcp-wrappers gives "deny", whereas what one would like would be some
1734 kind of temporary error. A kludge has been added to help with this.
1735 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1736 554 error is used if errno is still zero or contains ENOENT (which occurs
1737 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1740 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1742 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1743 errors. Otherwise a message that provokes a temporary error (when other
1744 messages do not) can cause a whole host to time out.
1746 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1747 addresses were routed directly to files or pipes from a redirect router.
1748 File deliveries just didn't batch; pipe deliveries might have suffered
1751 PH/06 A failure to get a lock for a hints database would erroneously always say
1752 "Failed to get write lock", even when it was really a read lock.
1754 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1755 of 0600. This has been changed to use the value of the lockfile_mode
1756 option (which defaults to 0600).
1758 PH/08 Applied small patch from the Sieve maintainer.
1760 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1761 folder from quota calculations, a direct delivery into this folder messed
1762 up the contents of the maildirsize file. This was because the regex was
1763 used only to exclude .Trash (or whatever) when the size of the mailbox
1764 was calculated. There was no check that a delivery was happening into an
1765 excluded directory. This bug has been fixed by ignoring all quota
1766 processing for deliveries into excluded directories.
1768 PH/10 Added the maildirfolder_create_regex option to appendfile.
1774 PH/01 The code for finding all the local interface addresses on a FreeBSD
1775 system running IPv6 was broken. This may well have applied to all BSD
1776 systems, as well as to others that have similar system calls. The broken
1777 code found IPv4 interfaces correctly, but gave incorrect values for the
1778 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1779 that it would not match correctly against @[] and not recognize the IPv6
1782 PH/02 The ipliteral router was not recognizing addresses of the form user@
1783 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1785 PH/03 Added disable_ipv6.
1787 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1788 decoded form, because it is most often used to construct To: headers
1789 lines in autoreplies, and the decoded form may well be syntactically
1790 invalid. However, $reply_address has leading white space removed, and all
1791 newlines turned into spaces so that the autoreply transport does not
1794 PH/05 If group was specified without a user on a router, and no group or user
1795 was specified on a transport, the group from the router was ignored.
1797 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1798 for visible compile-time settings that can be used to change these
1799 numbers, for those that want even more. Backwards compatibility with old
1800 spool files has been maintained. However, going back to a previous Exim
1801 release will lost any variables that are in spool files.
1803 PH/07 Two small changes when running in the test harness: increase delay when
1804 passing a TCP/IP connection to a new process, in case the original
1805 process has to generate a bounce, and remove special handling of
1806 127.0.0.2 (sic), which is no longer necessary.
1808 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1809 be the same on different OS.
1811 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1814 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1815 whether --show-vars was specified or not
1817 JJ/02 exipick: Added support for new ACL variable spool format introduced
1820 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1821 syntactically invalid From: or Reply-to: line, and a filter used this to
1822 generate an autoreply, and therefore failed to obtain an address for the
1823 autoreply, Exim could try to deliver to a non-existent relative file
1824 name, causing unrelated and misleading errors. What now happens is that
1825 it logs this as a hard delivery error, but does not attempt to create a
1828 PH/11 The exinext utility has a -C option for testing purposes, but although
1829 the given file was scanned by exinext itself; it wasn't being passed on
1830 when Exim was called.
1832 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1833 an end-of-file indication when reading a command response.
1835 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1836 compiled. In many other places in Exim, IPv6 addresses are always
1837 recognized, so I have changed this. It also means that IPv4 domain
1838 literals of the form [IPV4:n.n.n.n] are now always recognized.
1840 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1841 used if the router is not running as root, for example, when verifying at
1842 ACL time, or when using -bh. The debugging output from this situation was
1843 non-existent - all you got was a failure to exec. I have made two
1846 (a) Failures to set uid/gid, the current directory, or a process leader
1847 in a subprocess such as that created by queryprogram now generate
1848 suitable debugging ouput when -d is set.
1850 (b) The queryprogram router detects when it is not running as root,
1851 outputs suitable debugging information if -d is set, and then runs
1852 the subprocess without attempting to change uid/gid.
1854 PH/15 Minor change to Makefile for building test_host (undocumented testing
1857 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1858 additional section of a DNS packet that returns MX or SRV records.
1859 Instead, it always explicitly searches for A/AAAA records. This avoids
1860 major problems that occur when a DNS server includes only records of one
1861 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1862 fixed another bug: if SRV records were looked up and the corresponding
1863 address records were *not* found in the additional section, the port
1864 values from the SRV records were lost.
1866 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1867 using the correct key (the original address) when searching the retry
1868 rules in order to find which one to use for generating the retry hint.
1870 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1871 adding the default one. Similarly, if it contains a Reply-To: header, the
1872 errors_reply_to option, if set, is not used.
1874 PH/19 When calculating a retry time, Exim used to measure the "time since
1875 failure" by looking at the "first failed" field in the retry record. Now
1876 it does not use this if it is later than than the arrival time of the
1877 message. Instead it uses the arrival time. This makes for better
1878 behaviour in cases where some deliveries succeed, thus re-setting the
1879 "first failed" field. An example is a quota failure for a huge message
1880 when small messages continue to be delivered. Without this change, the
1881 "time since failure" will always be short, possible causing more frequent
1882 delivery attempts for the huge message than are intended.
1883 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1885 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1886 $1, $2, $3) because the numerical variables can be reset during some
1887 expansion items (e.g. "match"), thereby losing the authentication data.
1889 PH/21 Make -bV show the size of off_t variables so that the test suite can
1890 decide whether to run tests for quotas > 2G.
1892 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1893 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1894 filecount value is greater than 2G or if a quota value is greater than 2G
1895 on a system where the size of off_t is not greater than 4, a panic error
1898 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1899 never match. The debug and -bh output now contains an explicit error
1900 message indicating a malformed IPv4 address or mask.
1902 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1903 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1904 PH/23 above applies.
1906 PH/25 Do not write to syslog when running in the test harness. The only
1907 occasion when this arises is a failure to open the main or panic logs
1908 (for which there is an explicit test).
1910 PH/26 Added the /no_tell option to "control=freeze".
1912 PH/27 If a host name lookup failed very early in a connection, for example, if
1913 the IP address matched host_lookup and the reverse lookup yielded a name
1914 that did not have a forward lookup, an error message of the form "no IP
1915 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1916 could be logged. Now it outputs the IP address instead of "NULL".
1918 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1919 allows the sender and the authenticated sender to be set when
1920 submitting a message from within Exim. Since child_open_exim() is
1921 documented for local_scan(), the new function should be too.
1923 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1924 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1925 results in an empty string is now treated as unset.
1927 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1929 PH/31 Added sender_verify_fail logging option.
1931 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1932 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1933 tidied the source and removed it altogether.
1935 PH/33 When a queue run was abandoned because the load average was too high, a
1936 log line was always written; now it is written only if the queue_run log
1937 selector is set. In addition, the log line for abandonment now contains
1938 information about the queue run such as the pid. This is always present
1939 in "start" and "stop" lines but was omitted from the "abandon" line.
1941 PH/34 Omit spaces between a header name and the colon in the error message that
1942 is given when verify = headers_syntax fails (if there are lots of them,
1943 the message gets confusing).
1945 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1946 names, as there are now some PTR records that contain slashes. This check
1947 is only to protect against broken name servers that fall over on strange
1948 characters, so the fact that it applies to all lookups doesn't matter.
1950 PH/36 Now that the new test suite is complete, we can remove some of the
1951 special code in Exim that was needed for the old test suite. For example,
1952 sorting DNS records because real resolvers return them in an arbitrary
1953 order. The new test suite's fake resolver always returns records in the
1956 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1957 bounces) except when queue_only is set, to avoid logging races between
1958 the different processes.
1960 PH/38 Panic-die if .include specifies a non-absolute path.
1962 PH/39 A tweak to the "H" retry rule from its user.
1964 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1965 a label. They prevented compilation on older perls.
1967 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1968 a warning to be raised on newish perls.
1970 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1971 on queue. Changes to match documented behaviour of showing count of
1972 messages matching specified criteria.
1974 PH/40 Changed the default ident timeout from 30s to 5s.
1976 PH/41 Added support for the use of login_cap features, on those BSD systems
1977 that have them, for controlling the resources used by pipe deliveries.
1979 PH/42 The content-scanning code uses fopen() to create files in which to put
1980 message data. Previously it was not paying any attention to the mode of
1981 the files. Exim runs with umask(0) because the rest of the code creates
1982 files with open(), and sets the required mode explicitly. Thus, these
1983 files were ending up world-writeable. This was not a big issue, because,
1984 being within the spool directory, they were not world-accessible. I have
1985 created a function called modefopen, which takes an additional mode
1986 argument. It sets umask(777), creates the file, chmods it to the required
1987 mode, then resets the umask. All the relevant calls to fopen() in the
1988 content scanning code have been changed to use this function.
1990 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1991 to 24 hours. This avoids potential overflow problems when processing G
1992 and H retry rules. I suspect nobody ever tinkers with this value.
1994 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1996 PH/45 When the plaintext authenticator is running as a client, the server's
1997 challenges are checked to ensure they are valid base64 strings. By
1998 default, the authentication attempt is cancelled if an invalid string is
1999 received. Setting client_ignore_invalid_base64 true ignores these errors.
2000 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2001 they are received. Thus, the responses can be made to depend on the
2002 challenges. If an invalid string is ignored, an empty string is placed in
2005 PH/46 Messages that are created by the autoreply transport now contains a
2006 References: header, in accordance with RFCs 2822 and 3834.
2008 PH/47 Added authenticated_sender_force to the smtp transport.
2010 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2012 PH/49 Installed latest patch from the Sieve maintainer.
2014 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2015 was also set, the appendfile transport was unnecessarily scanning a
2016 directory of message files (e.g. for maildir delivery) to find the count
2017 of files (along with the size), even though it did not need this
2018 information. It now does the scan only if it needs to find either the
2019 size of the count of files.
2021 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2023 PH/52 Two bugs concerned with error handling when the smtp transport is
2026 (i) Exim was not creating retry information for temporary errors given
2027 for individual recipients after the DATA command when the smtp transport
2028 was used in LMTP mode. This meant that they could be retried too
2029 frequently, and not timed out correctly.
2031 (ii) Exim was setting the flag that allows error details to be returned
2032 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2033 recipients that were returned after the DATA command.
2035 PH/53 This is related to PH/52, but is more general: for any failing address,
2036 when detailed error information was permitted to be returned to the
2037 sender, but the error was temporary, then after the final timeout, only
2038 "retry timeout exceeded" was returned. Now it returns the full error as
2039 well as "retry timeout exceeded".
2041 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2042 do this, and (what is worse) MTAs that accept it.
2044 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2045 will now be deprecated.
2047 PH/56 New os.c-cygwin from the Cygwin maintainer.
2049 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2050 formats (previously only available in exim formats via -bpr, -bpru,
2051 and -bpra. Now also available in native and exiqgrep formats)
2053 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2054 with very large, slow to parse queues
2056 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2058 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2060 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2061 responses to authentication challenges, though it was showing the
2062 challenges; (ii) I've removed the CR characters from the debug output for
2065 PH/58 Allow for the insertion of a newline as well as a space when a string
2066 is turned into more than one encoded-word during RFC 2047 encoding. The
2067 Sieve code now uses this.
2069 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2070 data_4xx, lost_connection, tls_required.
2072 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2073 message was being sent as an SMTP response.
2075 PH/61 Add -l and -k options to exicyclog.
2077 PH/62 When verifying, if an address was redirected to one new address, so that
2078 verification continued, and the new address failed or deferred after
2079 having set something in $address_data, the value of $address_data was not
2080 passed back to the ACL. This was different to the case when no
2081 redirection occurred. The value is now passed back in both cases.
2083 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2084 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2085 use login_cap.h, so on its own it isn't the distinguishing feature. The
2086 new name refers directly to the setclassresources() function.
2088 PH/65 Added configuration files for NetBSD3.
2090 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2092 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2093 is preferred over IPv4.
2095 PH/68 The bounce_return_message and bounce_return_body options were not being
2096 honoured for bounces generated during the reception of non-SMTP messages.
2097 In particular, this applied to messages rejected by the ACL. This bug has
2098 been fixed. However, if bounce_return_message is true and bounce_return_
2099 body is false, the headers that are returned for a non-SMTP message
2100 include only those that have been read before the error was detected.
2101 (In the case of an ACL rejection, they have all been read.)
2103 PH/69 The HTML version of the specification is now built in a directory called
2104 spec_html instead of spec.html, because the latter looks like a path with
2105 a MIME-type, and this confuses some software.
2107 PH/70 Catch two compiler warnings in sieve.c.
2109 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2110 function verify_get_ident() calls ip_connect() to connect a socket, but
2111 if the "connect()" function timed out, ip_connect() used to close the
2112 socket. However, verify_get_ident() also closes the socket later, and in
2113 between Exim writes to the log, which may get opened at this point. When
2114 the socket was closed in ip_connect(), the log could get the same file
2115 descriptor number as the socket. This naturally causes chaos. The fix is
2116 not to close the socket in ip_connect(); the socket should be closed by
2117 the function that creates it. There was only one place in the code where
2118 this was missing, in the iplookup router, which I don't think anybody now
2119 uses, but I've fixed it anyway.
2121 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2122 well as to direct DNS lookups. Otherwise the handling of names in host
2123 lists is inconsistent and therefore confusing.
2129 PH/01 Two changes to the default runtime configuration:
2131 (1) Move the checks for relay_from_hosts and authenticated clients from
2132 after to before the (commented out) DNS black list checks.
2134 (2) Add control=submission to the relay_from_hosts and authenticated
2135 clients checks, on the grounds that messages accepted by these
2136 statements are most likely to be submissions.
2138 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2140 (1) Generate an error if the third argument for the ${prvs expansion is
2143 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2146 (3) Reset the variables that are obtained from the first argument of
2147 ${prvscheck and used in the second argument before leaving the code,
2148 because their memory is reclaimed, so using them afterwards may do
2151 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2152 one (it's much easier than Tom thought :-).
2154 (5) Because of (4), we can now allow for the use of $prvscheck_result
2155 inside the third argument.
2157 PH/03 For some reason, the default setting of PATH when running a command from
2158 a pipe transport was just "/usr/bin". I have changed it to
2161 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2162 anything to be listed in the output from -bV.
2164 PH/05 When a filter generated an autoreply, the entire To: header line was
2165 quoted in the delivery log line, like this:
2167 => >A.N.Other <ano@some.domain> <original@ddress> ...
2169 This has been changed so that it extracts the operative address. There
2170 may be more than one such address. If so, they are comma-separated, like
2173 => >ano@some.domain,ona@other.domain <original@ddress> ...
2175 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2176 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2177 not being looked up in the rDNS to get a host name, Exim was showing the
2178 IP address twice in Received: lines, even though the IP addresses were
2179 identical. For example:
2181 Received: from [1.2.3.4] (helo=[1.2.3.4])
2183 However, if the real host name was known, it was omitting the HELO data
2184 if it matched the actual IP address. This has been tidied up so that it
2185 doesn't show the same IP address twice.
2187 PH/07 When both +timestamp and +memory debugging was on, the value given by
2188 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2189 called by the debug printing, thereby overwriting the timestamp buffer.
2190 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2192 PH/08 When the original message was included in an autoreply transport, it
2193 always said "this is a copy of the message, including all the headers",
2194 even if body_only or headers_only was set. It now gives an appropriate
2197 PH/09 Applied a patch from the Sieve maintainer which:
2199 o fixes some comments
2200 o adds the (disabled) notify extension core
2201 o adds some debug output for the result of if/elsif tests
2202 o points to the current vacation draft in the documentation
2203 and documents the missing references header update
2207 o fixes a bug in processing the envelope test (when testing
2208 multiple envelope elements, the last element determinted the
2211 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2212 Electronic Mail") by including:
2214 Auto-submitted: auto-generated
2216 in the messages that it generates (bounce messages and others, such as
2217 warnings). In the case of bounce messages for non-SMTP mesages, there was
2218 also a typo: it was using "Auto_submitted" (underscore instead of
2219 hyphen). Since every message generated by Exim is necessarily in response
2220 to another message, thes have all been changed to:
2222 Auto-Submitted: auto-replied
2224 in accordance with these statements in the RFC:
2226 The auto-replied keyword:
2228 - SHOULD be used on messages sent in direct response to another
2229 message by an automatic process,
2231 - MUST NOT be used on manually-generated messages,
2233 - MAY be used on Delivery Status Notifications (DSNs) and Message
2234 Disposition Notifications (MDNs),
2236 - MUST NOT be used on messages generated by automatic or periodic
2237 processes, except for messages which are automatic responses to
2240 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2241 to the default Received: header definition.
2243 PH/12 Added log selector acl_warn_skipped (default on).
2245 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2246 variables because (a) they are in the wrong storage pool and (b) even if
2247 they were copied, it wouldn't work properly because of the caching.
2249 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2250 checking when decoding. Apparently there are clients that generate
2251 overlong encoded strings. Why am I not surprised?
2253 PH/15 If the first argument of "${if match_address" was not empty, but did not
2254 contain an "@" character, Exim crashed. Now it writes a panic log message
2255 and treats the condition as false.
2257 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2259 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2260 extension "envelope-auth". The code is finished and in agreement with
2261 other implementations, but there is no documentation so far and in fact,
2262 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2263 not changing the active code.
2265 Print executed "if" and "elsif" statements when debugging is used. This
2266 helps a great deal to understand what a filter does.
2268 Document more things not specified clearly in RFC3028. I had all this
2269 sorted out, when out of a sudden new issues came to my mind. Oops."
2271 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2274 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2275 canonical form (as documented). However, after a host name lookup from
2276 the IP address, check_host() was doing a simple string comparison with
2277 addresses acquired from the DNS when checking that the found name did
2278 have the original IP as one of its addresses. Since any found IPv6
2279 addresses are likely to be in abbreviated form, the comparison could
2280 fail. Luckily, there already exists a function for doing the comparison
2281 by converting both addresses to binary, so now that is used instead of
2282 the text comparison.
2284 PH/20 There was another similar case to PH/19, when a complete host name was
2285 given in a host list; looking up its IP address could give an abbreviated
2286 form, whereas the current host's name might or might not be abbreviated.
2287 The same fix has been applied.
2293 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2294 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2297 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2298 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2300 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2302 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2303 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2304 Also, for "Auto-Submitted", treat anything other than "no" as
2305 non-personal, in accordance with RFC 3834. (Previously it treated
2306 anything starting "auto-" as non-personal.)
2308 TF/01 The control=submission/name=... option had a problem with syntax
2309 errors if the name included a slash character. The /name= option
2310 now slurps the rest of the string, so it can include any characters
2311 but it must come last in the list of options (after /sender_retain
2314 PH/05 Some modifications to the interface to the fake nameserver for the new
2322 TK/01 Added the "success_on_redirect" address verification option. See
2323 NewStuff for rationale and an example.
2325 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2327 PH/02 Patch to exigrep to allow it to work on syslog lines.
2329 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2330 fread() to skip over the body file's header line, because in Cygwin the
2331 header line is locked and is inaccessible.
2333 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2334 co-exist for some time) to make it clear that it is the Exim ID that is
2335 referenced, not the Message-ID: header line.
2337 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2338 string_format() function, because snprintf() does not exist on all
2341 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2342 ${stat: expansion item.
2344 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2345 protocol synchronization error", to keep the pedants happy.
2347 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2348 well as for IRIX systems, when gcc is being used. See the host.c source
2351 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2353 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2356 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2357 transport and to the smtp transport in LMTP mode.
2359 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2361 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2362 run a filter in a subprocess. This could lead to confusion in subsequent
2363 lookups in the parent process. There should also be a search_tidyup() at
2364 the end of the subprocess.
2366 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2367 only if the host matched helo_try_verify_hosts, which caused the
2368 verification to occur when the EHLO/HELO command was issued. The ACL just
2369 tested the remembered result. Now, if a previous verification attempt has
2370 not happened, "verify = helo" does it there and then.
2372 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2374 TK/03 Fix log output including CR from clamd.
2376 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2377 exist provoked a memory error which could cause a segfault.
2379 PH/15 Installed PCRE 6.2
2381 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2383 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2384 of the problem. Specifically, suggested +O2 rather than +O1 for the
2387 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2389 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2390 record, the port was not correctly logged when the outgoing_port log
2391 selector was set (it logged the transort's default port).
2393 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2394 fallback_hosts, and "hosts" in the smtp transport.
2396 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2397 host errors such as "Connection refused".
2399 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2400 authentication with radiusclient 0.4.9:
2402 - Error returned from rc_read_config was caught wrongly
2403 - Username/password not passed on to radius server due to wrong length.
2405 The presumption is that some radiusclient API changes for 4.51/PH/17
2406 were not taken care of correctly. The code is still untested by me (my
2407 Linux distribution still has 0.3.2 of radiusclient), but it was
2408 contributed by a Radius user.
2410 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2411 expanding the "port" option of the smtp transport.
2413 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2414 while reading a MIME header. Thanks to Tom Hughes for a patch.
2416 PH/24 Include config.h inside local_scan.h so that configuration settings are
2419 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2420 that in an ACL for RCPT (for example), you can examine exactly what was
2423 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2424 commands, but it was not correctly comparing the address with the actual
2425 client host address. Thus, it would show the EHLO address in Received:
2426 header lines when this was not necessary.
2428 PH/27 Added the % operator to ${eval:}.
2430 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2431 it should be ignoring failures (because with -C, for example, it has lost
2432 privilege). It wasn't ignoring creation failures other than "already
2435 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2438 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2439 because an input line was too long, either on its own, or by virtue of
2440 too many continuations, the temporary file was not being removed, and the
2441 return code was incorrect.
2443 PH/31 Missing "BOOL" in function definition in filtertest.c.
2445 PH/32 Applied Sieve patches from the maintainer.
2447 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2449 PH/33 Added "verify = not_blind".
2451 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2452 Local/Makefile (with some defaults set). These are used in built scripts
2453 such as exicyclog, but they have never been used in the exim_install
2454 script (though there are many overriding facilities there). I have
2455 arranged that the exim_install script now takes note of these two
2458 PH/35 Installed configuration files for Dragonfly.
2460 PH/36 When a locally submitted message by a trusted user did not contain a
2461 From: header, and the sender address was obtained from -f or from an SMTP
2462 MAIL command, and the trusted user did not use -F to supply a sender
2463 name, $originator_name was incorrectly used when constructing a From:
2464 header. Furthermore, $originator_name was used for submission mode
2465 messages from external hosts without From: headers in a similar way,
2466 which is clearly wrong.
2468 PH/37 Added control=suppress_local_fixups.
2470 PH/38 When log_selector = +received_sender was set, and the addition of the
2471 sender made the log line's construction buffer exactly full, or one byte
2472 less than full, an overflow happened when the terminating "\n" was
2475 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2476 when the result of a list match is failure because a DNS lookup failed.
2478 PH/40 RM_COMMAND is now used in the building process.
2480 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2481 the "build-* directories that it finds.
2483 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2484 address in a domain literal was a prefix of an interface address.
2486 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2487 when verifying a sender address, unless rewrite_headers is false.
2489 PH/44 Wrote a long comment about why errors_to addresses are verified as
2490 recipients, not senders.
2492 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2493 the ratelimit ACL was added.
2495 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2497 PH/46 Added extra information about PostgreSQL errors to the error string.
2499 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2500 avoiding the need to install special zones in a real server. This is
2501 backwards compatible; if it can't find the fake resolver, it drops back.
2502 Thus, both old and new test suites can be run.
2504 TF/02 Added util/ratelimit.pl
2506 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2507 clock is set back in time.
2509 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2510 Candler <B.Candler@pobox.com>.
2512 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2513 OK for addresses that are the result of redirections.
2515 PH/48 A number of further additions for the benefit of the new test suite,
2516 including a fake gethostbyname() that interfaces to the fake DNS resolver
2519 TF/06 The fix for widen_domains has also been applied to qualify_single and
2520 search_parents which are the other dnslookup options that can cause
2523 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2526 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2528 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2529 service" if it encountered an MX record with an empty target hostname.
2530 The message is now "an MX or SRV record indicated no SMTP service".
2532 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2533 if the DNS of the sending site is misconfigured. This is quite a
2534 common situation. This change restores the behaviour of treating a
2535 helo verification defer as a failure.
2537 PH/51 If self=fail was set on a router, the bounce message did not include the
2538 actual error message.
2544 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2546 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2547 command itself ended in error, the underlying message about the transport
2548 filter timeout was being overwritten with the pipe command error. Now the
2549 underlying error message should be appended to the second error message.
2551 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2553 PH/02 Reduce the amount of output that "make" produces by default. Full output
2554 can still be requested.
2556 PH/03 The warning log line about a condition test deferring for a "warn" verb
2557 was being output only once per connection, rather than after each
2558 occurrence (because it was using the same function as for successful
2559 "warn" verbs). This seems wrong, so I have changed it.
2561 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2562 it should not have, which might have caused a crash in the right
2563 circumstances, but probably never did.
2565 PH/04 Installed a modified version of Tony Finch's patch to make submission
2566 mode fix the return path as well as the Sender: header line, and to
2567 add a /name= option so that you can make the user's friendly name appear
2570 TF/03 Added the control = fakedefer ACL modifier.
2572 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2573 Mark Lowes for thorough testing.
2575 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2577 TK/03 Merged latest SRS patch from Miles Wilton.
2579 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2580 with the definition in sysexits.h (which is #included earlier).
2581 Fortunately, Exim does not actually use EX_OK. The code used to try to
2582 preserve the sysexits.h value, by assumimg that macro definitions were
2583 scanned for macro replacements. I have been disabused of this notion,
2584 so now the code just undefines EX_OK before #including unistd.h.
2586 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2587 in the smtp transport. When a block could not be written in a single
2588 write() function, the timeout was being re-applied to each part-write.
2589 This seems wrong - if the receiver was accepting one byte at a time it
2590 would take for ever. The timeout is now adjusted when this happens. It
2591 doesn't have to be particularly precise.
2593 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2594 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2596 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2597 without a domain if the check to <postmaster@domain> fails.
2599 SC/01 Eximstats: added -xls and the ability to specify output files
2600 (patch written by Frank Heydlauf).
2602 SC/02 Eximstats: use FileHandles for outputing results.
2604 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2606 SC/04 Eximstats: fixed display of large numbers with -nvr option
2608 SC/05 Eximstats: fixed merging of reports with empty tables.
2610 SC/06 Eximstats: added the -include_original_destination flag
2612 SC/07 Eximstats: removed tabs and trailing whitespace.
2614 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2616 TK/06 MBOX spool code: Add real "From " MBOX separator line
2617 so the .eml file is really in mbox format (even though
2618 most programs do not really care). Patch from Alex Miller.
2620 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2621 The latter is generated from $received_to and is only set if the
2622 message has one envelope recipient. SA can use these headers,
2623 obviously out-of-the-box. Patch from Alex Miller.
2625 PH/08 The ${def test on a variable was returning false if the variable's
2626 value was "0", contrary to what the specification has always said!
2627 The result should be true unless the variable is empty.
2629 PH/09 The syntax error of a character other than { following "${if
2630 def:variable_name" (after optional whitespace) was not being diagnosed.
2631 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2632 accidental colon was present, for example, could give incorrect results.
2634 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2635 result is used (not including appendfile, where other changes are about
2638 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2639 This involved changing a lot of size variables from int to off_t. It
2640 should work with maildirs and everything.
2642 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2643 spamd dying while we are connected to it.
2645 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2648 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2649 function for BDB 4.3.
2651 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2653 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2654 the sources for PCRE have been re-arranged and more files are now
2657 PH/15 The code I had for printing potentially long long variables in PH/11
2658 above was not the best (it lost precision). The length of off_t variables
2659 is now inspected at build time, and an appropriate printing format (%ld
2660 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2661 to be "long long int" or "long int". This is needed for the internal
2662 formatting function string_vformat().
2664 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2665 the configuration file to be ":syslog", then the script "guesses" where
2666 the logs files are, rather than using the compiled in default. In our
2667 case the guess is not the same as the compiled default, so the script
2668 suddenly stopped working when I started to use syslog. The patch checks
2669 to see if log_file_path is "". If so, it attempts to read it from exim
2670 with no configuration file to get the compiled in version, before it
2671 falls back to the previous guessing code."
2673 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2674 implementing BATV in an Exim configuration. See NewStuff for the gory
2677 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2678 Makefile that are specific to HP-UX.
2680 PH/18 If the "use_postmaster" option was set for a recipient callout together
2681 with the "random" option, the postmaster address was used as the MAIL
2682 FROM address for the random test, but not for the subsequent recipient
2683 test. It is now used for both.
2685 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2686 patch removes a few documentation additions to RFC 3028, because the
2687 latest draft now contains them. It adds the new en;ascii-case comparator
2688 and a new error check for 8bit text in MIME parts. Comparator and
2689 require names are now matched exactly. I enabled the subaddress
2690 extension, but it is not well tested yet (read: it works for me)."
2692 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2693 rework some of the code of TK/09 above to avoid the hardwired use of
2694 "%lld" and "long long". Replaced the call to snprintf() with a call to
2697 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2698 records point to non-existent hosts", "retry timeout exceeded", and
2699 "retry time not reached for any host after a long failure period".
2701 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2702 experimental DomainKeys support:
2704 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2705 (2) On an error such as an illegally used "control", the wrong name for
2706 the control was given.
2708 These problems did NOT occur unless DomainKeys support was compiled.
2710 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2712 PH/24 Added ${if match_ip condition.
2714 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2715 databases so that it will be absolutely obvious if a crash occurs in the
2716 DB library. This is a regular occurrence (often caused by mis-matched
2719 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2720 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2721 picked up on a user's system that detects such things. There doesn't seem
2722 to be a gcc warning option for this - only an attribute that has to be
2723 put on the function's prototype. It seems that in Fedora Core 4 they have
2724 set this on a number of new functions. No doubt there will be more in due
2727 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2728 specify a transport. However, if an address that was verified by such a
2729 router was the subject of a callout, Exim crashed because it tried to
2730 read the rcpt_include_affixes from the non-existent transport. Now it
2731 just assumes that the setting of that option is false. This bug was
2732 introduced by 4.51/PH/31.
2734 PH/28 Changed -d+all to exclude +memory, because that information is very
2735 rarely of interest, but it makes the output a lot bigger. People tend to
2736 do -d+all out of habit.
2738 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2739 code in os-type was giving problems when libc.so lives in lib64, like on
2742 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2743 aren't the modern standard, and it seems that some systems' include files
2744 don't always have them. Exim was already checking for some of the newer
2745 ones like T_AAAA, and defining it itself. I've added checks for all the
2746 record types that Exim uses.
2748 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2749 not automatically generating a new one, as it is supposed to. This
2750 prevented TLS from working. If the file did exist, but contained invalid
2751 data, a new version was generated, as expected. It was only the case of a
2752 non-existent file that was broken.
2754 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2755 with a change in libdomainkeys > 0.64.
2757 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2758 from DNS. If the selector record carries the flag, it now has
2759 precedence over the domain-wide flag.
2761 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2763 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2764 the use of a port name that isn't defined in /etc/services) occurred, the
2765 message was deferred as in a normal delivery, and thus remained on the
2766 spool, instead of being failed because of the mua_wrapper setting. This
2767 is now fixed, and I tidied up some of the mua_wrapper messages at the
2770 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2771 the messages in a hash of arrays rather than using individual hashes.
2772 This is a bit cleaner and results in dramatic memory savings, albeit
2773 at a slight CPU cost.
2775 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2776 as requested by Marc Sherman.
2778 SC/10 Eximstats: added histograms for user specified patterns as requested
2781 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2783 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2784 fopen() in the content-scanning modules that did not already have it.
2790 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2791 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2793 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2795 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2797 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2798 to test Sieve filters that use "vacation".
2800 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2801 that changes the way the GnuTLS parameters are stored in the cache file.
2802 The new format can be generated externally. For backward compatibility,
2803 if the data in the cache doesn't make sense, Exim assumes it has read an
2804 old-format file, and it generates new data and writes a new file. This
2805 means that you can't go back to an older release without removing the
2808 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2809 work if there are any delivery delays because "one_time" forces the
2810 parent to be marked "delivered", so its unseen clone is never tried
2811 again. For this reason, Exim now forbids the simultaneous setting of
2814 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2815 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2816 entry if you want to know the details. The fix, however, neglected to
2817 consider the case when local delivery batching is involved. The test for
2818 "previously delivered" was not happening when checking to see if an
2819 address could be batched with a previous (undelivered) one; under
2820 certain circumstances this could lead to multiple deliveries to the same
2823 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2824 in its include files, and this causes problems building Exim.
2826 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2827 header_syntax) but Exim was just ignoring anything given after a slash.
2828 In particular, this caused confusion with an attempt to use "verify =
2829 reverse_host_lookup/defer_ok". An error is now given when options are
2830 supplied for verify items that do not have them. (Maybe reverse_host_
2831 lookup should have a defer_ok option, but that's a different point.)
2833 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2834 defined by RFC 821) to 2048, because there were problems with some AUTH
2835 commands, and RFC 1869 says the size should be increased for extended
2836 SMTP commands that take arguments.
2838 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2841 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2842 "unknown" error; now it says that the functionality isn't in the binary.
2844 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2845 an address' error message when a string expansion fails (syntax or
2846 whatever). Otherwise the password may appear in the log. Following change
2847 PH/42 below, there is no longer a chance of it appearing in a bounce
2850 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2852 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2853 of its addresses was ever tried. (Bugzilla bug #2.)
2855 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2856 the result incorrectly in the debug output. (It correctly added a newline
2857 to what was transported.)
2859 TF/01 Added $received_time.
2861 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2862 commented out examples of how to interface to a virus scanner and to
2863 SpamAssassin. Also added commented examples of av_scanner and
2864 spamd_address settings.
2866 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2867 and controls are allowed in which ACLs. There were a couple of minor
2868 errors. Some of the entries in the conditions table (which is a table of
2869 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2870 negation of where the condition IS allowed.
2872 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2874 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2875 header file does not have a version number, so I've had to invent a new
2876 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2877 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2878 radiusclient), but it was contributed by a Radius user.
2880 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2881 files or directories, for OpenSSL.
2883 PH/19 When an Exim process that is running as root has to create an Exim log
2884 file, it does so in a subprocess that runs as exim:exim so as to get the
2885 ownership right at creation (otherwise, other Exim processes might see
2886 the file with the wrong ownership). There was no test for failure of this
2887 fork() call, which would lead to the process getting stuck as it waited
2888 for a non-existent subprocess. Forks do occasionally fail when resources
2889 run out. I reviewed all the other calls to fork(); they all seem to check
2892 PH/20 When checking for unexpected SMTP input at connect time (before writing
2893 the banner), Exim was not dealing correctly with a non-positive return
2894 from the read() function. If the client had disconnected by this time,
2895 the result was a log entry for a synchronization error with an empty
2896 string after "input=" when read() returned zero. If read() returned -1
2897 (an event I could not check), uninitialized data bytes were printed.
2898 There were reports of junk text (parts of files, etc) appearing after
2901 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2903 PH/22 Added support for macro redefinition, and (re)definition in between
2904 driver and ACL definitions.
2906 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2907 forgetting to use the resulting value; it was using the unexpanded value.
2909 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2910 hadn't been configured. The fix is from Juergen Kreileder, who
2911 understands it better than I do:
2913 "Here's what I see happening with three configured cyrus_sasl
2914 authenticators configured (plain, login, cram-md5):
2916 On startup auth_cyrus_sasl_init() gets called for each of these.
2917 This means three calls to sasl_listmech() without a specified mech_list.
2918 => SASL tests which mechs of all available mechs actually work
2919 => three warnings about OTP not working
2920 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2922 With the patch, sasl_listmech() also gets called three times. But now
2923 SASL's mech_list option is set to the server_mech specified in the the
2924 authenticator. Or in other words, the answer from sasl_listmech()
2925 gets limited to just the mech you're testing for (which is different
2927 => the return list contains just 'plain' or 'login', 'cram-md5' or
2928 nothing depending on the value of ob->server_mech.
2930 I've just tested the patch: Authentication still works fine,
2931 unavailable mechs specified in the exim configuration are still
2932 caught, and the auth.log warnings about OTP are gone."
2934 PH/25 When debugging is enabled, the contents of the command line are added
2935 to the debugging output, even when log_selector=+arguments is not
2938 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2939 answer is "GNU", and only if the return is "GNU/something" is the answer
2942 PH/27 $acl_verify_message is now set immediately after the failure of a
2943 verification in an ACL, and so is available in subsequent modifiers. In
2944 particular, the message can be preserved by coding like this:
2946 warn !verify = sender
2947 set acl_m0 = $acl_verify_message
2949 Previously, $acl_verify_message was set only while expanding "message"
2950 and "log_message" when a very denied access.
2952 PH/28 Modified OS/os.c-Linux with
2954 -#ifndef OS_LOAD_AVERAGE
2955 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2957 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2958 nomenclature these days.)
2960 PH/29 Installed patch from the Sieve maintainer that adds the options
2961 sieve_useraddress and sieve_subaddress to the redirect router.
2963 PH/30 In these circumstances:
2964 . Two addresses routed to the same list of hosts;
2965 . First host does not offer TLS;
2966 . First host accepts first address;
2967 . First host gives temporary error to second address;
2968 . Second host offers TLS and a TLS session is established;
2969 . Second host accepts second address.
2970 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2971 and peerdn, if requested) that were in fact used only for the second
2974 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2975 attention to any local part prefix or suffix that was matched by the
2976 router that accepted the address. It now behaves in the same way as it
2977 does for delivery: the affixes are removed from the local part unless
2978 rcpt_include_affixes is set on the transport.
2980 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2981 timeout during the DATA phase of an incoming message.
2983 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2984 applied a patch sanctioned by the Sieve maintainer.
2986 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2987 the uid or gid is negative. A case of a negative gid caused this to be
2988 noticed. The fix allows for either to be negative.
2990 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2991 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2994 PH/36 The change PH/12 above was broken. Fixed it.
2996 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2997 the grounds that routing the same address twice would always produce the
2998 same answer. This might have been true once, but it is certainly no
2999 longer true now. Routing a child address may depend on the previous
3000 routing that produced that child. Some complicated redirection strategies
3001 went wrong when messages had multiple recipients, and made Exim's
3002 behaviour dependent on the order in which the addresses were given.
3004 I have moved the duplicate checking until after the routing is complete.
3005 Exim scans the addresses that are assigned to local and remote
3006 transports, and removes any duplicates. This means that more work will be
3007 done, as duplicates will always all be routed, but duplicates are
3008 presumably rare, so I don't expect this is of any significance.
3010 For deliveries to pipes, files, and autoreplies, the duplicate checking
3011 still happens during the routing process, since they are not going to be
3014 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3015 It corrects a timeout issue with spamd. This is Ian's comment: "The
3016 background is that sometimes spamd either never reads data from a
3017 connection it has accepted, or it never writes response data. The exiscan
3018 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3019 blindly assumes that writes won't block so it may never time out."
3021 PH/39 Allow G after quota size as well as K and M.
3023 PH/40 The value set for $authenticated_id in an authenticator may not contain
3024 binary zeroes or newlines because the value is written to log lines and
3025 to spool files. There was no check on this. Now the value is run through
3026 the string_printing() function so that such characters are converted to
3027 printable escape sequences.
3029 PH/41 $message_linecount is a new variable that contains the total number of
3030 lines in the message. Compare $body_linecount, which is the count for the
3033 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3034 bounce and delay warning messages, except in certain special cases, which
3037 (a) An SMTP error message from a remote host;
3038 (b) A message specified in a :fail: redirection;
3039 (c) A message specified in a "fail" command in a system filter;
3040 (d) A message specified in a FAIL return from the queryprogram router;
3041 (e) A message specified by the cannot_route_message router option.
3043 In these cases only, Exim does include the error details in bounce and
3044 warning messages. There are also a few cases where bland messages such
3045 as "unrouteable address" or "local delivery error" are given.
3047 PH/43 $value is now also set for the "else" part of a ${run expansion.
3049 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3050 being worked on, but at least Exim now implements the latest version to
3053 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3054 process to complete was treated as a delivery failure, a timeout while
3055 writing the message to the pipe was logged, but erroneously treated as a
3056 successful delivery. Such timeouts include transport filter timeouts. For
3057 consistency with the overall process timeout, these timeouts are now
3058 treated as errors, giving rise to delivery failures by default. However,
3059 there is now a new Boolean option for the pipe transport called
3060 timeout_defer, which, if set TRUE, converts the failures into defers for
3061 both kinds of timeout. A transport filter timeout is now identified in
3064 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3065 systems where "make" and "gmake" are different, calling "gmake" at top
3066 level broke things. I've arranged for the value of $(MAKE) to be passed
3067 from the Makefile to this script so that it can call the same version of
3071 A note about Exim versions 4.44 and 4.50
3072 ----------------------------------------
3074 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3075 changes of various kinds. As a consequence, a big documentation update was
3076 needed. This delayed the release for rather longer than seemed good, especially
3077 in the light of a couple of (minor) security issues. Therefore, the changes
3078 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3079 release. So 4.44 and 4.50 are in effect two different branches that both start
3082 I have left the 4.50 change log unchanged; it contains all the changes since
3083 4.43. The change log for 4.44 is below; many of its items are identical to
3084 those for 4.50. This seems to be the most sensible way to preserve the
3085 historical information.
3091 1. Minor wording change to the doc/README.SIEVE file.
3093 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3094 computation of the current number of files was incorrect.
3096 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3097 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3100 4. Give more explanation in the error message when the command for a transport
3101 filter fails to execute.
3103 5. There are several places where Exim runs a non-Exim command in a
3104 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3105 was being done only for the command run by the queryprogram router. It is
3106 now done for all such subprocesses. The other cases are: ${run, transport
3107 filters, and the commands run by the lmtp and pipe transports.
3109 6. Added CONFIGURE_GROUP build-time option.
3111 7. Some older OS have a limit of 256 on the maximum number of file
3112 descriptors. Exim was using setrlimit() to set 1000 as a large value
3113 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3114 systems. I've change it so that if it can't get 1000, it tries for 256.
3116 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3117 was an oversight, and furthermore, ever since the addition of extra
3118 controls (e.g. 4.43/32), the checks on when to allow different forms of
3119 "control" were broken. There should now be diagnostics for all cases when a
3120 control that does not make sense is encountered.
3122 9. Added the /retain_sender option to "control=submission".
3124 10. $recipients is now available in the predata ACL (oversight).
3126 11. Tidy the search cache before the fork to do a delivery from a message
3127 received from the command line. Otherwise the child will trigger a lookup
3128 failure and thereby defer the delivery if it tries to use (for example) a
3129 cached ldap connection that the parent has called unbind on.
3131 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3132 of $address_data from the recipient verification was clobbered by the
3133 sender verification.
3135 13. The value of address_data from a sender verification is now available in
3136 $sender_address_data in subsequent conditions in the ACL statement.
3138 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3140 15. Added a new option "connect=<time>" to callout options, to set a different
3143 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3144 was its contents. (It was OK if the option was not defined at all.)
3146 17. A "Completed" log line is now written for messages that are removed from
3147 the spool by the -Mrm option.
3149 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3150 information about exactly what failed.
3152 19. Added -dd to debug only the daemon process.
3154 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3155 handles timeouts, both on the server side and network timeouts. Renamed the
3156 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3158 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3159 It is now set to "smtps".
3161 22. $host_address is now set to the target address during the checking of
3162 ignore_target_hosts.
3164 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3165 being passed; this would have caused $sender_host_name to have been used if
3166 matching the list had actually called for a host name (not very likely,
3167 since this list is usually IP addresses). A host name is now passed as
3170 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3171 code that specifies a non-restarting handler (typically sigaction() in
3172 modern systems) in an attempt to fix a rare and obscure crash bug.
3174 25. Narrowed the window for a race in the daemon that could cause it to ignore
3175 SIGCHLD signals. This is not a major problem, because they are used only to
3176 wake it up if nothing else does.
3178 26. A malformed maildirsize file could cause Exim to calculate negative values
3179 for the mailbox size or file count. Odd effects could occur as a result.
3180 The maildirsize information is now recalculated if the size or filecount
3183 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3184 support for a long time. Removed HAVE_SYS_VFS_H.
3186 28. Installed the latest version of exipick from John Jetmore.
3188 29. In an address list, if the pattern was not a regular expression, an empty
3189 subject address (from a bounce message) matched only if the pattern was an
3190 empty string. Non-empty patterns were not even tested. This was the wrong
3191 because it is perfectly reasonable to use an empty address as part of a
3192 database query. An empty address is now tested by patterns that are
3193 lookups. However, all the other forms of pattern expect the subject to
3194 contain a local part and a domain, and therefore, for them, an empty
3195 address still always fails if the pattern is not itself empty.
3197 30. Exim went into a mad DNS loop when attempting to do a callout where the
3198 host was specified on an smtp transport, and looking it up yielded more
3199 than one IP address.
3201 31. Re-factored the code for checking spool and log partition space into a
3202 function that finds that data and another that does the check. The former
3203 is then used to implement four new variables: $spool_space, $log_space,
3204 $spool_inodes, and $log_inodes.
3206 32. The RFC2047 encoding function was originally intended for short strings
3207 such as real names; it was not keeping to the 75-character limit for
3208 encoded words that the RFC imposes. It now respects the limit, and
3209 generates multiple encoded words if necessary. To be on the safe side, I
3210 have increased the buffer size for the ${rfc2047: expansion operator from
3213 33. It is now permitted to omit both strings after an "if" condition; if the
3214 condition is true, the result is "true". As before, when the second string
3215 is omitted, a false condition yields an empty string. This makes it less
3216 cumbersome to write custom ACL and router conditions.
3218 34. Failure to deliver a bounce message always caused it to be frozen, even if
3219 there was an errors_to setting on the router. The errors_to setting is now
3222 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3223 canonical form (fully expanded) before being placed in
3224 $sender_host_address.
3226 36. The table in the code that translates DNS record types into text (T_A to
3227 "A" for instance) was missing entries for NS and CNAME. It is just possible
3228 that this could have caused confusion if both these types were looked up
3229 for the same domain, because the text type is used as part of Exim's
3230 per-process caching. But the chance of anyone hitting this buglet seems
3233 37. The dnsdb lookup has been extended in a number of ways.
3235 (1) There is a new type, "zns", which walks up the domain tree until it
3236 finds some nameserver records. It should be used with care.
3238 (2) There is a new type, "mxh", which is like "mx" except that it returns
3239 just the host names, not the priorities.
3241 (3) It is now possible to give a list of domains (or IP addresses) to be
3242 looked up. The behaviour when one of the lookups defers can be
3243 controlled by a keyword.
3245 (4) It is now possible to specify the separator character for use when
3246 multiple records are returned.
3248 38. The dnslists ACL condition has been extended: it is now possible to supply
3249 a list of IP addresses and/or domains to be looked up in a particular DNS
3252 39. Added log_selector=+queue_time_overall.
3254 40. When running the queue in the test harness, wait just a tad after forking a
3255 delivery process, to get repeatability of debugging output.
3257 41. Include certificate and key file names in error message when GnuTLS fails
3258 to set them up, because the GnuTLS error message doesn't include the name
3259 of the failing file when there is a problem reading it.
3261 42. Allow both -bf and -bF in the same test run.
3263 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3265 44. The "Exiscan patch" is now merged into the mainline Exim source.
3267 45. Sometimes the final signoff response after QUIT could fail to get
3268 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3269 < 0 before doing a fflush(). This bug looks as though it goes back to the
3270 introduction of TLS in release 3.20, but "sometimes" must have been rare
3271 because the tests only now provoked it.
3273 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3274 (this can affect the format of dates).
3276 47. exim_tidydb, when checking for the continued existence of a message for
3277 which it has found a message-specific retry record, was not finding
3278 messages that were in split spool directories. Consequently, it was
3279 deleting retry records that should have stayed in existence.
3281 48. Steve fixed some bugs in eximstats.
3283 49. The SPA authentication driver was not abandoning authentication and moving
3284 on to the next authenticator when an expansion was forced to fail,
3285 contradicting the general specification for all authenticators. Instead it
3286 was generating a temporary error. It now behaves as specified.
3288 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3289 (the order specifies the preference for clients). The order is now AES256,
3290 AES128, 3DES, ARCFOUR128.
3292 51. Small patch to Sieve code - explicitly set From: when generating an
3295 52. Exim crashed if a remote delivery caused a very long error message to be
3296 recorded - for instance if somebody sent an entire SpamAssassin report back
3297 as a large number of 550 error lines. This bug was coincidentally fixed by
3298 increasing the size of one of Exim's internal buffers (big_buffer) that
3299 happened as part of the Exiscan merge. However, to be on the safe side, I
3300 have made the code more robust (and fixed the comments that describe what
3303 53. Now that there can be additional text after "Completed" in log lines (if
3304 the queue_time_overall log selector is set), a one-byte patch to exigrep
3305 was needed to allow it to recognize "Completed" as not the last thing in
3308 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3309 patch that reportedly fixes this has been added. I am not expert enough to
3310 create a test for it. This is what the patch creator wrote:
3312 "I found a little strange behaviour of ldap code when working with
3313 Windows 2003 AD Domain, where users was placed in more than one
3314 Organization Units. When I tried to give exim partial DN, the exit code
3315 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3316 But simultaneously result of request was absolutely normal ldap result,
3317 so I produce this patch..."
3319 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3320 REFERENCE, so I have modified the code to exclude the patch when that macro
3323 55. Some experimental protocols are using DNS PTR records for new purposes. The
3324 keys for these records are domain names, not reversed IP addresses. The
3325 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3326 leaves it alone. Component reversal etc. now happens only for IP addresses.
3329 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3331 57. Double the size of the debug message buffer (to 2048) so that more of very
3332 long debug lines gets shown.
3334 58. The exicyclog utility now does better if the number of log files to keep
3335 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3337 59. Two changes related to the smtp_active_hostname option:
3339 (1) $smtp_active_hostname is now available as a variable.
3340 (2) The default for smtp_banner uses $smtp_active_hostname instead
3341 of $primary_hostname.
3343 60. The host_aton() function is supposed to be passed a string that is known
3344 to be a valid IP address. However, in the case of IPv6 addresses, it was
3345 not checking this. This is a hostage to fortune. Exim now panics and dies
3346 if the condition is not met. A case was found where this could be provoked
3347 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3348 components; fortuitously, this particular loophole had already been fixed
3349 by change 4.50/55 above.
3351 If there are any other similar loopholes, the new check in host_aton()
3352 itself should stop them being exploited. The report I received stated that
3353 data on the command line could provoke the exploit when Exim was running as
3354 exim, but did not say which command line option was involved. All I could
3355 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3356 running as the user.
3359 61. There was a buffer overflow vulnerability in the SPA authentication code
3360 (which came originally from the Samba project). I have added a test to the
3361 spa_base64_to_bits() function which I hope fixes it.
3364 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3365 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3367 63. The daemon start-up calls getloadavg() while still root for those OS that
3368 need the first call to be done as root, but it missed one case: when
3369 deliver_queue_load_max is set with deliver_drop_privilege. This is
3370 necessary for the benefit of the queue runner, because there is no re-exec
3371 when deliver_drop_privilege is set.
3373 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3374 This has been fixed.
3376 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3377 was in use, was not putting the data itself into the right store pool;
3378 consequently, it could be overwritten for a subsequent message in the same
3379 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3382 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3384 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3385 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3386 the function were treating the return as a boolean value, which happened to
3387 work because 0=false and not-0=true, but is not correct code.
3389 68. The host_aton() function was not handling scoped IPv6 addresses (those
3390 with, for example, "%eth0" on the end) correctly.
3392 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3393 negated items (that is, ~something) in unsigned ints. Some compilers
3394 apparently mutter when there is no cast.
3396 70. If an address verification called from an ACL failed, and did not produce a
3397 user-specific message (i.e. there was only a "system" message), nothing was
3398 put in $acl_verify_message. In this situation, it now puts the system
3401 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3402 session; change 4.31/43 added the unwanted input to the log line - except
3403 that it did not do this in the start of session case. It now does.
3405 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3406 This is wrong and can cause the other end to generate a synchronization
3407 error if it is another Exim or anything else that does the synchronization
3408 check. A QUIT command is no longer sent after a timeout.
3410 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3411 during host lookups.
3413 74. The defer_ok option of callout verification was not working if it was used
3414 when verifying addresses in header lines, that is, for this case:
3416 verify = header_sender/callout=defer_ok
3418 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3419 those file descriptors could be used for SMTP connections. If anything
3420 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3421 could be sent to the SMTP client, causing chaos. The daemon now opens
3422 stdin, stdout, and stderr to /dev/null when it puts itself into the
3425 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3426 log by default. The user can override this with suitable Perl magic.
3428 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3429 the log message when discard triggers, was not working for the DATA ACL or
3430 for the non-SMTP ACL.
3432 78. Error message wording change in sieve.c.
3434 79. If smtp_accept_max_per_host was set, the number of connections could be
3435 restricted to fewer than expected, because the daemon was trying to set up
3436 a new connection before checking whether the processes handling previous
3437 connections had finished. The check for completed processes is now done
3438 earlier. On busy systems, this bug wouldn't be noticed because something
3439 else would have woken the daemon, and it would have reaped the completed
3442 80. If a message was submitted locally by a user whose login name contained one
3443 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3444 It caused a spool format error. I have fixed the spool reading code. A
3445 related problem was that the "from" clause in the Received: line became
3446 illegal because of the space(s). It is now covered by ${quote_local_part.
3448 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3451 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3453 83. Patch from Sieve maintainer to fix unterminated string problem in
3454 "vacation" handling.
3456 84. Some minor changes to the Linux configuration files to help with other
3457 OS variants using glibc.
3459 85. One more patch for Sieve to update vacation handling to latest spec.
3462 ----------------------------------------------------
3463 See the note above about the 4.44 and 4.50 releases.
3464 ----------------------------------------------------
3470 1. Change 4.43/35 introduced a bug that caused file counts to be
3471 incorrectly computed when quota_filecount was set in an appendfile
3474 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3475 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3478 3. Give more explanation in the error message when the command for a transport
3479 filter fails to execute.
3481 4. There are several places where Exim runs a non-Exim command in a
3482 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3483 was being done only for the command run by the queryprogram router. It is
3484 now done for all such subprocesses. The other cases are: ${run, transport
3485 filters, and the commands run by the lmtp and pipe transports.
3487 5. Some older OS have a limit of 256 on the maximum number of file
3488 descriptors. Exim was using setrlimit() to set 1000 as a large value
3489 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3490 systems. I've change it so that if it can't get 1000, it tries for 256.
3492 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3493 was an oversight, and furthermore, ever since the addition of extra
3494 controls (e.g. 4.43/32), the checks on when to allow different forms of
3495 "control" were broken. There should now be diagnostics for all cases when a
3496 control that does not make sense is encountered.
3498 7. $recipients is now available in the predata ACL (oversight).
3500 8. Tidy the search cache before the fork to do a delivery from a message
3501 received from the command line. Otherwise the child will trigger a lookup
3502 failure and thereby defer the delivery if it tries to use (for example) a
3503 cached ldap connection that the parent has called unbind on.
3505 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3506 of $address_data from the recipient verification was clobbered by the
3507 sender verification.
3509 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3510 was its contents. (It was OK if the option was not defined at all.)
3512 11. A "Completed" log line is now written for messages that are removed from
3513 the spool by the -Mrm option.
3515 12. $host_address is now set to the target address during the checking of
3516 ignore_target_hosts.
3518 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3519 being passed; this would have caused $sender_host_name to have been used if
3520 matching the list had actually called for a host name (not very likely,
3521 since this list is usually IP addresses). A host name is now passed as
3524 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3525 code that specifies a non-restarting handler (typically sigaction() in
3526 modern systems) in an attempt to fix a rare and obscure crash bug.
3528 15. Narrowed the window for a race in the daemon that could cause it to ignore
3529 SIGCHLD signals. This is not a major problem, because they are used only to
3530 wake it up if nothing else does.
3532 16. A malformed maildirsize file could cause Exim to calculate negative values
3533 for the mailbox size or file count. Odd effects could occur as a result.
3534 The maildirsize information is now recalculated if the size or filecount
3537 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3538 support for a long time. Removed HAVE_SYS_VFS_H.
3540 18. Updated exipick to current release from John Jetmore.
3542 19. Allow an empty sender to be matched against a lookup in an address list.
3543 Previously the only cases considered were a regular expression, or an
3546 20. Exim went into a mad DNS lookup loop when doing a callout where the
3547 host was specified on the transport, if the DNS lookup yielded more than
3550 21. The RFC2047 encoding function was originally intended for short strings
3551 such as real names; it was not keeping to the 75-character limit for
3552 encoded words that the RFC imposes. It now respects the limit, and
3553 generates multiple encoded words if necessary. To be on the safe side, I
3554 have increased the buffer size for the ${rfc2047: expansion operator from
3557 22. Failure to deliver a bounce message always caused it to be frozen, even if
3558 there was an errors_to setting on the router. The errors_to setting is now
3561 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3562 canonical form (fully expanded) before being placed in
3563 $sender_host_address.
3565 24. Updated eximstats to version 1.33
3567 25. Include certificate and key file names in error message when GnuTLS fails
3568 to set them up, because the GnuTLS error message doesn't include the name
3569 of the failing file when there is a problem reading it.
3571 26. Expand error message when OpenSSL has problems setting up cert/key files.
3574 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3575 (this can affect the format of dates).
3577 28. exim_tidydb, when checking for the continued existence of a message for
3578 which it has found a message-specific retry record, was not finding
3579 messages that were in split spool directories. Consequently, it was
3580 deleting retry records that should have stayed in existence.
3582 29. eximstats updated to version 1.35
3583 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3584 1.35 - bugfix such that pie charts by volume are generated correctly
3586 30. The SPA authentication driver was not abandoning authentication and moving
3587 on to the next authenticator when an expansion was forced to fail,
3588 contradicting the general specification for all authenticators. Instead it
3589 was generating a temporary error. It now behaves as specified.
3591 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3592 (the order specifies the preference for clients). The order is now AES256,
3593 AES128, 3DES, ARCFOUR128.
3595 31. Small patch to Sieve code - explicitly set From: when generating an
3598 32. Exim crashed if a remote delivery caused a very long error message to be
3599 recorded - for instance if somebody sent an entire SpamAssassin report back
3600 as a large number of 550 error lines. This bug was coincidentally fixed by
3601 increasing the size of one of Exim's internal buffers (big_buffer) that
3602 happened as part of the Exiscan merge. However, to be on the safe side, I
3603 have made the code more robust (and fixed the comments that describe what
3606 33. Some experimental protocols are using DNS PTR records for new purposes. The
3607 keys for these records are domain names, not reversed IP addresses. The
3608 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3609 leaves it alone. Component reversal etc. now happens only for IP addresses.
3612 34. The host_aton() function is supposed to be passed a string that is known
3613 to be a valid IP address. However, in the case of IPv6 addresses, it was
3614 not checking this. This is a hostage to fortune. Exim now panics and dies
3615 if the condition is not met. A case was found where this could be provoked
3616 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3617 components; fortuitously, this particular loophole had already been fixed
3618 by change 4.50/55 or 4.44/33 above.
3620 If there are any other similar loopholes, the new check in host_aton()
3621 itself should stop them being exploited. The report I received stated that
3622 data on the command line could provoke the exploit when Exim was running as
3623 exim, but did not say which command line option was involved. All I could
3624 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3625 running as the user.
3628 35. There was a buffer overflow vulnerability in the SPA authentication code
3629 (which came originally from the Samba project). I have added a test to the
3630 spa_base64_to_bits() function which I hope fixes it.
3633 36. The daemon start-up calls getloadavg() while still root for those OS that
3634 need the first call to be done as root, but it missed one case: when
3635 deliver_queue_load_max is set with deliver_drop_privilege. This is
3636 necessary for the benefit of the queue runner, because there is no re-exec
3637 when deliver_drop_privilege is set.
3639 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3640 was in use, was not putting the data itself into the right store pool;
3641 consequently, it could be overwritten for a subsequent message in the same
3642 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3645 38. Sometimes the final signoff response after QUIT could fail to get
3646 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3647 < 0 before doing a fflush(). This bug looks as though it goes back to the
3648 introduction of TLS in release 3.20, but "sometimes" must have been rare
3649 because the tests only now provoked it.
3655 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3656 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3657 Now it returns an integer. A number of calls were still expecting a T/F
3658 return. Fortuitously, in all cases, the tests worked in OK situations,
3659 which is the norm. However, things would have gone wrong on any write
3660 failures on the smtp file descriptor. This function is used when sending
3661 messages over SMTP and also when doing verify callouts.
3663 2. When Exim is called to do synchronous delivery of a locally submitted
3664 message (the -odf or -odi options), it no longer closes stderr before doing
3667 3. Implemented the mua_wrapper option.
3669 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3671 5. Implemented the functions header_remove(), header_testname(),
3672 header_add_at_position(), and receive_remove_recipient(), and exported them
3675 6. If an ACL "warn" statement specified the addition of headers, Exim already
3676 inserted X-ACL-Warn: at the start if there was no header name. However, it
3677 was not making this test for the second and subsequent header lines if
3678 there were newlines in the string. This meant that an invalid header could
3679 be inserted if Exim was badly configured.
3681 7. Allow an ACL "warn" statement to add header lines at the start or after all
3682 the Received: headers, as well as at the end.
3684 8. Added the rcpt_4xx retry error code.
3686 9. Added postmaster_mailfrom=xxx to callout verification option.
3688 10. Added mailfrom=xxxx to the callout verification option, for verify=
3691 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3692 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3694 12. Inserted some casts to stop certain compilers warning when using pointer
3695 differences as field lengths or precisions in printf-type calls (mostly
3696 affecting debugging statements).
3698 13. Added optional readline() support for -be (dynamically loaded).
3700 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3701 same clock tick as a message's arrival, so that its received time was the
3702 same as the "first fail" time on the retry record, and that message
3703 remained on the queue past the ultimate address timeout, every queue runner
3704 would try a delivery (because it was past the ultimate address timeout) but
3705 after another failure, the ultimate address timeout, which should have then
3706 bounced the address, did not kick in. This was a "< instead of <=" error;
3707 in most cases the first failure would have been in the next clock tick
3708 after the received time, and all would be well.
3710 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3711 being recognized when the domain list was tested by the match_domain
3712 condition in an expansion string.
3714 16. Added the ${str2b64: operator.
3716 17. Exim was always calling setrlimit() to set a large limit for the number of
3717 processes, without checking whether the existing limit was already
3718 adequate. (It did check for the limit on file descriptors.) Furthermore,
3719 errors from getrlimit() and setrlimit() were being ignored. Now they are
3720 logged to the main and panic logs, but Exim does carry on, to try to do its
3721 job under whatever limits there are.
3723 18. Imported PCRE 5.0.
3725 19. Trivial typo in log message " temporarily refused connection" (the leading
3728 20. If the log selector return_path_on_delivery was set and an address was
3729 redirected to /dev/null, the delivery process crashed because it assumed
3730 that a return path would always be set for a "successful" delivery. In this
3731 case, the whole delivery is bypassed as an optimization, and therefore no
3734 21. Internal re-arrangement: the function for sending a challenge and reading
3735 a response while authentication was assuming a zero-terminated challenge
3736 string. It's now changed to take a pointer and a length, to allow for
3737 binary data in such strings.
3739 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3741 23. Exim was not respecting finduser_retries when seeking the login of the
3742 uid under which it was called; it was always trying 10 times. (The default
3743 setting of finduser_retries is zero.) Also, it was sleeping after the final
3744 failure, which is pointless.
3746 24. Implemented tls_on_connect_ports.
3748 25. Implemented acl_smtp_predata.
3750 26. If the domain in control=submission is set empty, Exim assumes that the
3751 authenticated id is a complete email address when it generates From: or
3752 Sender: header lines.
3754 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3755 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3756 chown and chgrp in /bin and hostname in /usr/bin.
3758 28. Exim was keeping the "process log" file open after each use, just as it
3759 does for the main log. This opens the possibility of it remaining open for
3760 long periods when the USR1 signal hits a daemon. Occasional processlog
3761 errors were reported, that could have been caused by this. Anyway, it seems
3762 much more sensible not to leave this file open at all, so that is what now
3765 29. The long-running daemon process does not normally write to the log once it
3766 has entered its main loop, and it closes the log before doing so. This is
3767 so that log files can straightforwardly be renamed and moved. However,
3768 there are a couple of unusual error situations where the daemon does write
3769 log entries, and I had neglected to close the log afterwards.
3771 30. The text of an SMTP error response that was received during a remote
3772 delivery was being truncated at 512 bytes. This is too short for some of
3773 the long messages that one sometimes sees. I've increased the limit to
3776 31. It is now possible to make retry rules that apply only when a message has a
3777 specific sender, in particular, an empty sender.
3779 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3780 it possible to be selective about when SMTP synchronization is enforced.
3782 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3784 32. Implemented hosts_connection_nolog.
3786 33. Added an ACL for QUIT.
3788 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3791 35. Added mailbox_size and mailbox_filecount to appendfile.
3793 36. Added control = no_multiline_responses to ACLs.
3795 37. There was a bug in the logic of the code that waits for the clock to tick
3796 in the case where the clock went backwards by a substantial amount such
3797 that the microsecond fraction of "now" was more than the microsecond
3798 fraction of "then" (but the whole seconds number was less).
3800 38. Added support for the libradius Radius client library this is found on
3801 FreeBSD (previously only the radiusclient library was supported).
3807 1. When certain lookups returned multiple values in the form name=value, the
3808 quoting of the values was not always being done properly. Specifically:
3809 (a) If the value started with a double quote, but contained no whitespace,
3811 (b) If the value contained whitespace other than a space character (i.e.
3812 tabs or newlines or carriage returns) it was not quoted.
3813 This fix has been applied to the mysql and pgsql lookups by writing a
3814 separate quoting function and calling it from the lookup code. The fix
3815 should probably also be applied to nisplus, ibase and oracle lookups, but
3816 since I cannot test any of those, I have not disturbed their existing code.
3818 2. A hit in the callout cache for a specific address caused a log line with no
3819 reason for rejecting RCPT. Now it says "Previous (cached) callout
3820 verification failure".
3822 3. There was an off-by-one bug in the queryprogram router. An over-long
3823 return line was truncated at 256 instead of 255 characters, thereby
3824 overflowing its buffer with the terminating zero. As well as fixing this, I
3825 have increased the buffer size to 1024 (and made a note to document this).
3827 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3828 when Exim is waiting for an SMTP response from a remote server, Exim
3829 restarts its select() call on the socket, thereby resetting its timeout.
3830 This is not a problem when such interrupts are rare. Somebody set up a cron
3831 job to run exiwhat every 2 minutes, which is less than the normal select()
3832 timeout (5 or 10 minutes). This meant that the select() timeout never
3833 kicked in because it was always reset. I have fixed this by comparing the
3834 time when an interrupt arrives with the time at the start of the first call
3835 to select(). If more time than the timeout has elapsed, the interrupt is
3836 treated as a timeout.
3838 5. Some internal re-factoring in preparation for the addition of Sieve
3839 extensions (by MH). In particular, the "personal" test is moved to a
3840 separate function, and given an option for scanning Cc: and Bcc: (which is
3841 not set for Exim filters).
3843 6. When Exim created an email address using the login of the caller as the
3844 local part (e.g. when creating a From: or Sender: header line), it was not
3845 quoting the local part when it contained special characters such as @.
3847 7. Installed new OpenBSD configuration files.
3849 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3850 try to make them clearer.
3852 9. Callout options, other than the timeout value, were being ignored when
3853 verifying sender addresses in header lines. For example, when using
3855 verify = header_sender/callout=no_cache
3857 the cache was (incorrectly) being used.
3859 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3860 only the Cygwin environment.
3862 11. When return_path_on_delivery was set as a log selector, if different remote
3863 addresses in the same message used different return paths and parallel
3864 remote delivery occurred, the wrong values would sometimes be logged.
3865 (Whenever a remote delivery process finished, the return path value from
3866 the most recently started remote delivery process was logged.)
3868 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3869 lines when AUTH and/or TLS are in use. This is the "received protocol"
3870 field. Exim used to use "asmtp" for authenticated SMTP, without any
3871 indication (in the protocol name) for TLS use. Now it follows the RFC and
3872 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3873 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3874 names appear in log lines as well as in Received: header lines.
3876 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3877 extensions, and comparison tests, and to fix some bugs.
3879 14. Changes to the "personal" filter test:
3881 (1) The test was buggy in that it was just doing the equivalent of
3882 "contains" tests on header lines. For example, if a user's address was
3883 anne@some.where, the "personal" test would incorrectly be true for
3885 To: susanne@some.where
3887 This test is now done by extracting each address from the header in turn,
3888 and checking the entire address. Other tests that are part of "personal"
3889 are now done using regular expressions (for example, to check local parts
3890 of addresses in From: header lines).
3892 (2) The list of non-personal local parts in From: addresses has been
3893 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3894 taken from the Sieve specification recommendations.
3896 (3) If the message contains any header line starting with "List-" it is
3897 treated as non-personal.
3899 (4) The test for "circular" in the Subject: header line has been removed
3900 because it now seems ill-conceived.
3902 15. Minor typos in src/EDITME comments corrected.
3904 16. Installed latest exipick from John Jetmore.
3906 17. If headers_add on a router specified a text string that was too long for
3907 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3908 of string_sprintf() is now avoided.
3910 18. $message_body_size was not set (it was always zero) when running the DATA
3911 ACL and the local_scan() function.
3913 19. For the "mail" command in an Exim filter, no default was being set for
3914 the once_repeat time, causing a random time value to be used if "once" was
3915 specified. (If the value happened to be <= 0, no repeat happened.) The
3916 default is now 0s, meaning "never repeat". The "vacation" command was OK
3917 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3918 (I found it when inspecting the code).
3920 20. There is now an overall timeout for performing a callout verification. It
3921 defaults to 4 times the callout timeout, which applies to individual SMTP
3922 commands during the callout. The overall timeout applies when there is more
3923 than one host that can be tried. The timeout is checked before trying the
3924 next host. This prevents very long delays if there are a large number of
3925 hosts and all are timing out (e.g. when the network connections are timing
3926 out). The value of the overall timeout can be changed by specifying an
3927 additional sub-option for "callout", called "maxwait". For example:
3929 verify = sender/callout=5s,maxwait=20s
3931 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3932 to the end before writing, but this should make it even safer).
3934 22. Exim was forgetting that it had advertised PIPELINING for the second and
3935 subsequent messages on an SMTP connection. It was also not resetting its
3936 memory on STARTTLS and an internal HELO.
3938 23. When Exim logs an SMTP synchronization error within a session, it now
3939 records whether PIPELINING has been advertised or not.
3941 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3942 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3943 rather than long int.
3945 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3947 26. Added the never_mail option to autoreply.
3953 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3954 crash if the getsockname() call failed; this can happen if a connection is
3955 closed very soon after it is established. The problem was simply in the
3956 order in which certain operations were done, causing Exim to try to write
3957 to the SMTP stream before it had set up the file descriptor. The bug has
3958 been fixed by making things happen in the correct order.
3964 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3965 before the connection was closed, thus losing the rejection response.
3967 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3968 some early Solaris releases, but causes trouble in current releases where
3969 socklen_t is defined.
3971 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3974 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3977 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3978 that releasing the top of it at the end releases what was used for sub-
3979 expansions (unless the block got too big). However, discard this block if
3980 the first thing is a variable or header, so that we can use its block when
3981 it is dynamic (useful for very large $message_headers, for example).
3983 6. Lookups now cache *every* query, not just the most recent. A new, separate
3984 store pool is used for this. It can be recovered when all lookup caches are
3985 flushed. Lookups now release memory at the end of their result strings.
3986 This has involved some general refactoring of the lookup sources.
3988 7. Some code has been added to the store_xxx() functions to reduce the amount
3989 of flapping under certain conditions.
3991 8. log_incoming_interface used to affect only the <= reception log lines. Now
3992 it causes the local interface and port to be added to several more SMTP log
3993 lines, for example "SMTP connection from", and rejection lines.
3995 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3997 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3999 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4000 could be overwritten at the end of the current message (or the start of a
4001 new message if it was set in a HELO ACL). The value is now preserved for
4002 the duration of the SMTP connection.
4004 12. If a transport had a headers_rewrite setting, and a matching header line
4005 contained an unqualified address, that address was qualified, even if it
4006 did not match any rewriting rules. The underlying bug was that the values
4007 of the flags that permit the existence of unqualified sender and recipient
4008 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4009 non-local messages, and by -bnq for local messages) were not being
4010 preserved with the message after it was received.
4012 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4013 "next input=" as part of the text comprising the host identity instead of
4014 the correct text. The code was using the same buffer for two different
4015 strings. However, depending on which order the printing function evaluated
4016 its arguments, the bug did not always show up. Under Linux, for example, my
4017 test suite worked just fine.
4019 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4020 doesn't work with some older versions of Perl. It has been changed to "my",
4021 which in any case is probably the better facility to use.
4023 15. A really picky compiler found some instances of statements for creating
4024 error messages that either had too many or two few arguments for the format
4027 16. The size of the buffer for calls to the DNS resolver has been increased
4028 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4029 for addresses that have a lot of PTR records. This alleviates a problem; it
4030 does not fully solve it.
4032 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4033 buffer now truncates the list and logs the incident, which is the same
4034 action as happens when Exim is looking up a host name and its aliases.
4035 Previously in this situation something unpredictable would happen;
4036 sometimes it was "internal error: store_reset failed".
4038 18. If a server dropped the connection unexpectedly when an Exim client was
4039 using GnuTLS and trying to read a response, the client delivery process
4040 crashed while trying to generate an error log message.
4042 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4043 string, for example:
4045 warn message = H1: something\nH2: something
4047 the text was added as a single header line from Exim's point of view
4048 though it ended up OK in the delivered message. However, searching for the
4049 second and subsequent header lines using $h_h2: did not work. This has been
4050 fixed. Similarly, if a system filter added multiple headers in this way,
4051 the routers could not see them.
4053 20. Expanded the error message when iplsearch is called with an invalid key to
4054 suggest using net-iplsearch in a host list.
4056 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4057 ACLs are no longer actually imposed (and a message to that effect is
4060 22. If a "gecos" field in a passwd entry contained escaped characters, in
4061 particular, if it contained a \" sequence, Exim got it wrong when building
4062 a From: or a Sender: header from that name. A second bug also caused
4063 incorrect handling when an unquoted " was present following a character
4064 that needed quoting.
4066 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4067 was not being matched caselessly.
4069 24. Arranged for all hyphens in the exim.8 source to be escaped with
4072 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4073 properly. Recipient callout cache records were still being keyed to include
4074 the sender, even when use_sender was set false. This led to far more
4075 callouts that were necessary. The sender is no longer included in the key
4076 when use_sender is false.
4078 26. Added "control = submission" modifier to ACLs.
4080 27. Added the ${base62d: operator to decode base 62 numbers.
4082 28. dnsdb lookups can now access SRV records.
4084 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4085 the configuration file.
4087 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4088 (-v) mode. This makes the output for a verbose queue run more intelligible.
4090 31. Added a use_postmaster feature to recipient callouts.
4092 32. Added the $body_zerocount variable, containing the number of binary zero
4093 bytes in the message body.
4095 33. The time of last modification of the "new" subdirectory is now used as the
4096 "mailbox time last read" when there is a quota error for a maildir
4099 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4101 35. Added +ignore_unknown as a special item in host lists.
4103 36. Code for decoding IPv6 addresses in host lists is now included, even if
4104 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4105 address was recognized as an IP address, but was then not correctly decoded
4106 into binary, causing unexpected and incorrect effects when compared with
4113 1. Very minor rewording of debugging text in manualroute to say "list of
4114 hosts" instead of "hostlist".
4116 2. If verify=header_syntax was set, and a header line with an unqualified
4117 address (no domain) and a large number of spaces between the end of the
4118 name and the colon was received, the reception process suffered a buffer
4119 overflow, and (when I tested it) crashed. This was caused by some obsolete
4120 code that should have been removed. The fix is to remove it!
4122 3. When running in the test harness, delay a bit after writing a bounce
4123 message to get a bit more predictability in the log output.
4125 4. Added a call to search_tidyup() just before forking a reception process. In
4126 theory, someone could use a lookup in the expansion of smtp_accept_max_
4127 per_host which, without the tidyup, could leave open a database connection.
4129 5. Added the variables $recipient_data and $sender_data which get set from a
4130 lookup success in an ACL "recipients" or "senders" condition, or a router
4131 "senders" option, similar to $domain_data and $local_part_data.
4133 6. Moved the writing of debug_print from before to after the "senders" test
4136 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4137 problems for message scanning, either using a data ACL, or using
4138 local_scan() because the Received: header was not generated till after they
4139 were called (in order to set the time as the time of reception completion).
4140 I have revised the way this works. The header is now generated after the
4141 body is received, but before the ACL or local_scan() are called. After they
4142 are run, the timestamp in the header is updated.
4148 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4149 before starting a queue runner without re-exec. This happened only when
4150 deliver_drop_privilege was set or when the Exim user was set to root. The
4151 effect of the bug was that timeouts during subsequent deliveries caused
4152 crashes instead of being properly handled. The handler is now left at its
4153 default (and expected) setting.
4155 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4156 message, again when deliver_drop_privilege is set or Exim is run as root.
4157 The bug described in (1) was not present in this case, but the tidying up
4158 of the other signals was missing. I have made the two cases consistent.
4160 3. The ignore_target_hosts setting on a manualroute router was being ignored
4161 for hosts that were looked up using the /MX notation.
4163 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4166 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4167 operated on the sender address. After changing the $sender_address to <>
4168 for the sender address verify, Exim was re-instated it as the original
4169 (before rewriting) address, but remembering that it had rewritten it, so it
4170 wasn't rewriting it again. This bug also had the effect of breaking the
4171 sender address verification caching when the sender address was rewritten.
4173 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4174 This has been changed so that if the ip literal address matches
4175 ignore_target_hosts, the router declines.
4177 7. Added expansion conditions match_domain, match_address, and match_local_
4178 part (NOT match_host).
4180 8. The placeholder for the Received: header didn't have a length field set.
4182 9. Added code to Exim itself and to exim_lock to test for a specific race
4183 condition that could lead to file corruption when using MBX delivery. The
4184 issue is with the lockfile that is created in /tmp. If this file is removed
4185 after a process has opened it but before that process has acquired a lock,
4186 there is the potential for a second process to recreate the file and also
4187 acquire a lock. This could lead to two Exim processes writing to the file
4188 at the same time. The added code performs the same test as UW imapd; it
4189 checks after acquiring the lock that its file descriptor still refers to
4190 the same named file.
4192 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4193 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4196 11. Added the smtp_active_hostname option. If used, this will typically be made
4197 to depend on the incoming interface address. Because $interface_address is
4198 not set up until the daemon has forked a reception process, error responses
4199 that can happen earlier (such as "too many connections") no longer contain
4202 12. If an expansion in a condition on a "warn" statement fails because a lookup
4203 defers, the "warn" statement is abandoned, and the next ACL statement is
4204 processed. Previously this caused the whole ACL to be aborted.
4206 13. Added the iplsearch lookup type.
4208 14. Added ident_timeout as a log selector.
4210 15. Added tls_certificate_verified as a log selector.
4212 16. Added a global option tls_require_ciphers (compare the smtp transport
4213 option of the same name). This controls incoming TLS connections.
4215 17. I finally figured out how to make tls_require_ciphers do a similar thing
4216 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4217 before starting the TLS session.
4219 18. Tabs are now shown as \t in -bP output.
4221 19. If the log selector return_path_on_delivery was set, Exim crashed when
4222 bouncing a message because it had too many Received: header lines.
4224 20. If two routers both had headers_remove settings, and the first one included
4225 a superfluous trailing colon, the final name in the first list and the
4226 first name in the second list were incorrectly joined into one item (with a
4227 colon in the middle).
4233 1. Added -C and -D options to the exinext utility, mainly to make it easier
4234 to include in the automated testing, but these could be helpful when
4235 multiple configurations are in use.
4237 2. The exinext utility was not formatting the output nicely when there was
4238 an alternate port involved in the retry record key, nor when there was a
4239 message id as well (for retries that were specific to a specific message
4240 and a specific host). It was also confused by IPv6 addresses, because of
4241 the additional colons they contain. I have fixed the IPv4 problem, and
4242 patched it up to do a reasonable job for IPv6.
4244 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4245 delivery, the log line now contains "pipelined" if PIPELINING was used.
4247 4. An SMTP transport process used to panic and die if the bind() call to set
4248 an explicit outgoing interface failed. This has been changed; it is now
4249 treated in the same way as a connect() failure.
4251 5. A reference to $sender_host_name in the part of a conditional expansion
4252 that was being skipped was still causing a DNS lookup. This no longer
4255 6. The def: expansion condition was not recognizing references to header lines
4256 that used bh_ and bheader_.
4258 7. Added the _cache feature to named lists.
4260 8. The code for checking quota_filecount in the appendfile transport was
4261 allowing one more file than it should have been.
4263 9. For compatibility with Sendmail, the command line option
4271 and sets the incoming protocol and host name (for trusted callers). The
4272 host name and its colon can be omitted when only the protocol is to be set.
4273 Note the Exim already has two private options, -pd and -ps, that refer to
4274 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4275 "s", but I don't think that's a major issue.
4277 10. A number of refactoring changes to the code, none of which should affect
4280 (a) The number of logging options was getting close to filling up the
4281 32-bit word that was used as a bit map. I have split them into two classes:
4282 those that are passed in the argument to log_write(), and those that are
4283 only ever tested independently outside of that function. These are now in
4284 separate 32-bit words, so there is plenty of room for expansion again.
4285 There is no change in the user interface or the logging behaviour.
4287 (b) When building, for example, log lines, the code previously used a
4288 macro that called string_cat() twice, in order to add two strings. This is
4289 not really sufficiently general. Furthermore, there was one instance where
4290 it was actually wrong because one of the argument was used twice, and in
4291 one call a function was used. (As it happened, calling the function twice
4292 did not affect the overall behaviour.) The macro has been replaced by a
4293 function that can join an arbitrary number of extra strings onto a growing
4296 (c) The code for expansion conditions now uses a table and a binary chop
4297 instead of a serial search (which was left over from when there were very
4298 few conditions). Also, it now recognizes conditions like "pam" even when
4299 the relevant support is not compiled in: a suitably worded error message is
4300 given if an attempt is made to use such a condition.
4302 11. Added ${time_interval:xxxxx}.
4304 12. A bug was causing one of the ddress fields not to be passed back correctly
4305 from remote delivery subprocesses. The field in question was not being
4306 subsequently used, so this caused to problems in practice.
4308 13. Added new log selectors queue_time and deliver_time.
4310 14. Might have fixed a bug in maildirsizefile handling that threw up
4311 "unexpected character" debug warnings, and recalculated the data
4312 unnecessarily. In any case, I expanded the warning message to give more
4315 15. Added the message "Restricted characters in address" to the statements in
4316 the default ACL that block characters like @ and % in local parts.
4318 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4319 Three changes have been made:
4321 (a) There was a serious bug; a negative response to MAIL caused the whole
4322 recipient domain to be cached as invalid, thereby blocking all messages
4323 to all local parts at the same domain, from all senders. This bug has
4324 been fixed. The domain is no longer cached after a negative response to
4325 MAIL if the sender used is not empty.
4327 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4330 (c) A new callout option, "use_sender" has been added for people who want
4331 the modified behaviour.
4337 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4340 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4341 indeed breaks things for older releases.
4343 3. Added additional logging to the case where there is a problem reading data
4344 from a filter that is running in a subprocess using a pipe, in order to
4345 try to track down a specific problem.
4347 4. Testing facility fudge: when running in the test harness and attempting
4348 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4349 getting "No route to host". Convert this to a timeout.
4351 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4354 6. Some OS don't have socklen_t but use size_t instead. This affects the
4355 fifth argument of getsockopt() amongst other things. This is now
4356 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4357 can be set for individual OS. I have set it for SunOS5, OSF1, and
4358 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4359 some earlier ones do not.
4361 7. Change 4.30/15 was not doing the test caselessly.
4363 8. The standard form for an IPv6 address literal was being rejected by address
4364 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4365 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4366 this, as well as the form without the "IPv6" on the front (but only when
4367 address literals are enabled, of course).
4369 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4371 10. Exim crashed if a message with an empty sender address specified by -f
4372 encountered a router with an errors_to setting. This could be provoked only
4373 by a command such as
4377 where an empty string was supplied; "<>" did not hit this bug.
4379 11. Installed PCRE release 4.5.
4381 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4382 remained set. It is now erased.
4384 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4385 times from message ids (which are base 36 rather than the normal 62).
4387 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4388 were being counted as actual protocol errors, and logged if the log
4389 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4390 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4391 and DATA following a set of rejected RCPTs do not count as protocol errors.
4392 In other words, Exim assumes they were pipelined, though this may not
4393 actually be the case. Of course, in all cases the client gets an
4394 appropriate error code.
4396 15. If a lookup fails in an ACL condition, a message about the failure may
4397 be available; it is used if testing the ACL cannot continue, because most
4398 such messages specify what the cause of the deferral is. However, some
4399 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4400 that caused an old message to be retained and used if a later statement
4401 caused a defer, replacing the real cause of the deferral.
4403 16. If an IP address had so many PTR records that the DNS lookup buffer
4404 was not large enough to hold them, Exim could crash while trying to process
4405 the truncated data. It now detects and logs this case.
4407 17. Further to 4.21/58, another change has been made: if (and only if) the
4408 first line of a message (the first header line) ends with CRLF, a bare LF
4409 in a subsequent header line has a space inserted after it, so as not to
4410 terminate the header.
4412 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4413 unnecessarily, used atoi() instead of strtol(), and didn't check the
4414 termination when getting file sizes from file names by regex.
4416 19. Completely re-implemented the support for maildirsize files, in the light
4417 of a number of problems with the previous contributed implementation
4418 (4.30/29). In particular:
4420 . If the quota is zero, the maildirsize file is maintained, but no quota is
4423 . If the maildir directory does not exist, it is created before any attempt
4424 to write a maildirsize file.
4426 . The quota value in the file is just a cache; if the quota is changed in
4427 the transport, the new value overrides.
4429 . A regular expression is available for excluding directories from the
4432 20. The autoreply transport checks the characters in options that define the
4433 message's headers; it allows continued headers, but it was checking with
4434 isspace() after an embedded newline instead of explicitly looking for a
4437 21. If all the "regular" hosts to which an address was routed had passed their
4438 expiry times, and had not reached their retry times, the address was
4439 bounced, even if fallback hosts were defined. Now Exim should go on to try
4442 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4443 equivalent code in the SMTP transport. Some hosts send humungous responses
4444 to HELO/EHLO, more than 1024 it seems.
4446 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4447 gives compiler warnings in some environments. I've now done it "properly",
4450 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4451 (because of problems with the built-in one) was declared to return uschar *
4452 instead of char *, causing compiler failure.
4454 25. Fixed a file descriptor leak when processing alias/forward files.
4456 26. Fixed a minor format string issue in dbfn.c.
4458 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4460 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4461 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4464 29. When a sender address is verified, it is cached, to save repeating the test
4465 when there is more than one recipient in a message. However, when the
4466 verification involves a callout, it is possible for different callout
4467 options to be set for different recipients. It is too complicated to keep
4468 track of this in the cache, so now Exim always runs a verification when a
4469 callout is required, relying on the callout cache for the optimization.
4470 The overhead is duplication of the address routing, but this should not be
4473 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4474 to be verified with callout=postmaster, and the main callout worked but the
4475 postmaster check failed, the verification correctly failed. However, if a
4476 subsequent RCPT command asked for sender verification *without* the
4477 postmaster check, incorrect caching caused this verification also to fail,
4480 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4481 it was not caching the DNS options (qualify_single, search_parents) that
4482 were used when the lookup failed. A subsequent lookup with different
4483 options therefore always gave the same answer, though there were cases
4484 where it should not have. (Example: a "domains = !$mx_any" option on a
4485 dnslookup router: the "domains" option is always processed without any
4486 widening, but the router might have qualify_single set.) Now Exim uses the
4487 cached value only when the same options are set.
4489 32. Added John Jetmore's "exipick" utility to the distribution.
4491 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4492 than a timeout (e.g. a certificate is required, and is not provided), an
4493 Exim server now closes the connection immediately. Previously it waited for
4494 the client to close - but if the client is SSL, it seems that they each
4495 wait for each other, leading to a delay before one of them times out.
4497 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4498 maintained 0.8.x compatibility because I don't think many are using it, and
4499 it is clearly obsolete.
4501 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4504 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4505 client certificate was expired. A simple patch fixes this, though I don't
4506 understand the full logic of why the verify callback is called multiple
4509 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4510 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4511 which causes problems with some clients (such as the Certicom SSL Plus
4512 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4513 disables the coutermeasure allowing Eudora to connect."
4515 38. Exim was not checking that a write() to a log file succeeded. This could
4516 lead to Bad Things if a log got too big, in particular if it hit a file
4517 size limit. Exim now panics and dies if it cannot write to a log file, just
4518 as it does if it cannot open a log file.
4520 39. Modified OS/Makefile-Linux so that it now contains
4522 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4524 The two -D definitions ensure that Exim is compiled with large file
4525 support, which makes it possible to handle log files that are bigger than
4528 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4529 instance) a domain was checked against a named list that involved a lookup,
4530 causing $domain_data to be set, then another domain was checked against the
4531 same list, then the first domain was re-checked, the value of $domain_data
4532 after the final check could be wrong. In particular, if the second check
4533 failed, it could be set empty. This bug probably also applied to
4536 41. The strip_trailing_dot option was not being applied to the address given
4537 with the -f command-line option.
4539 42. The code for reading a message's header from the spool was incrementing
4540 $received_count, but never initializing it. This meant that the value was
4541 incorrect (doubled) while delivering a message in the same process in which
4542 it was received. In the most common configuration of Exim, this never
4543 happens - a fresh exec is done - but it can happen when
4544 deliver_drop_privilege is set.
4546 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4547 it now includes up to 150 characters of the unexpected data in the log
4550 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4551 and building data strings. The size of both of these buffers was 10 000
4552 bytes - far larger than anybody would *ever* want, thought I. Needless to
4553 say, somebody hit the limit. I have increased the maximum line length to
4554 20 000 and the maximum data length of concatenated lines to 100 000. I have
4555 also fixed two bugs, because there was no checking on these buffers. Tsk,
4556 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4557 buffer is too small.
4559 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4560 lsearch lookups. Now it does.
4562 46. When parsing a route_list item in a manualroute router, a fixed-length
4563 buffer was used for the list of hosts. I made this 1024 bytes long,
4564 thinking that nobody would ever have a list of hosts that long. Wrong.
4565 Somebody had a whole pile of complicated expansion conditions, and the
4566 string was silently truncated, leading to an expansion error. It turns out
4567 that it is easier to change to an unlimited length (owing to other changes
4568 that have happened since this code was originally written) than to build
4569 structure for giving a limitation error. The length of the item that
4570 expands into the list of hosts is now unlimited.
4572 47. The lsearch lookup could not handle data where the length of text line was
4573 more than 4095 characters. Such lines were truncated, leading to shortened
4574 data being returned. It should now handle lines of any length.
4576 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4577 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4580 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4581 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4582 no longer happens in comment lines. A list of replacements is now placed
4583 at the head of all of the source files, except those whose only change is
4584 to replace PERL_COMMAND in the very first #! line.
4586 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4587 messages on the queue, with a bottom-up merge sort, using code contributed
4588 by Michael Haardt. This should make operations like -bp somewhat faster on
4589 large queues. It won't affect queue runners, except when queue_run_in_order
4592 51. Installed eximstats 1.31 in the distribution.
4594 52. Added support for SRV lookups to the dnslookup router.
4596 53. If an ACL referred to $message_body or $message_body_end, the value was not
4597 reset for any messages that followed in the same SMTP session.
4599 54. The store-handling optimization for building very long strings was not
4600 differentiating between the different store pools. I don't think this
4601 actually made any difference in practice, but I've tidied it.
4603 55. While running the routers to verify a sender address, $sender_address
4604 was still set to the sender address. This is wrong, because when routing to
4605 send a bounce to the sender, it would be empty. Therefore, I have changed
4606 it so that, while verifying a sender address, $sender_address is set to <>.
4607 (There is no change to what happens when verifying a recipient address.)
4609 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4610 target A or AAAA records (if not already returned) without resetting the
4611 qualify_single or search_parents options of the DNS resolver. These are
4612 inappropriate in this case because the targets of MX and SRV records must
4613 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4614 target that, when qualified, matched something in the local domain. These
4615 two options are now turned off when doing these lookups.
4617 57. It seems that at least some releases of Reiserfs (which does not have the
4618 concept of a fixed number of inodes) returns zero and not -1 for the
4619 number of available inodes. This interacted badly with check_spool_inodes,
4620 which assumed that -1 was the "no such thing" setting. What I have done is
4621 to check that the total number of inodes is greater than zero before doing
4622 the test of how many are available.
4624 58. When a "warn" ACL statement has a log_message modifier, the message is
4625 remembered, and not repeated. This is to avoid a lot of repetition when a
4626 message has many recipients that cause the same warning to be written.
4627 Howewer, Exim was preserving the list of already written lines for an
4628 entire SMTP session, which doesn't seem right. The memory is now reset if a
4629 new message is started.
4631 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4632 debugging output unless log_rewrite was also set.
4634 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4635 of (int)(handle) when we know that handle contains (void *)(-1).
4637 61. The Exim daemon panic-logs an error return when it closes the incoming
4638 connection. However "connection reset by peer" seems to be common, and
4639 isn't really an error worthy of noting specially, so that particular error
4642 62. When Exim is trying to find all the local interfaces, it used to panic and
4643 die if the ioctl to get the interface flags failed. However, it seems that
4644 on at least one OS (Solaris 9) it is possible to have an interface that is
4645 included in the list of interfaces, but for which you get a failure error
4646 for this call. This happens when the interface is not "plumbed" into a
4647 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4648 failure of the "get flags" call assumes that the interface is down.
4650 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4651 makes life easier for people who are doing arithmetic on fields extracted
4652 from dates, where you often get leading zeros that should not be
4653 interpreted as octal.
4655 64. Added qualify_domain to the redirect router, to override the global
4658 65. If a pathologically long header line contained very many addresses (the
4659 report of this problem mentioned 10 000) and each of them was rewritten,
4660 Exim could use up a very large amount of memory. (It kept on making new
4661 copies of the header line as it rewrote, and never released the old ones.)
4662 At the expense of a bit more processing, the header rewriting function has
4663 been changed so that it no longer eats memory in this way.
4665 66. The generation of the Received: header has been moved from the time that a
4666 message starts to be received, to the time that it finishes. The timestamp
4667 in the Received: header should now be very close to that of the <= log
4668 line. There are two side-effects of this change:
4670 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4671 logged header lines no longer include the local Received: line, because
4672 it has not yet been created. The same applies to a copy of the message
4673 that is returned to a non-SMTP sender when a message is rejected.
4675 (b) When a filter file is tested using -bf, no additional Received: header
4676 is added to the test message. After some thought, I decided that this
4679 This change does not affect the value of $received_for. It is still set
4680 after address rewriting, but before local_scan() is called.
4682 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4684 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4685 gave an unhelpful panic error message, and a defer error. I have managed to
4686 change this behaviour so that it now rejects any supplied certificate,
4687 which seems right, as the list of acceptable certificates is empty.
4689 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4690 gave an unhelpful defer error. I have not managed to make this reject any
4691 supplied certificates, but the error message it gives is "no certificate
4692 supplied", which is not helpful.
4694 70. exigrep's output now also includes lines that are not associated with any
4695 message, but which match the given pattern. Implemented by a patch from
4696 Martin Sluka, which also tidied up the Perl a bit.
4698 71. Recipient callout verification, like sender verification, was using <> in
4699 the MAIL FROM command. This isn't really the right thing, since the actual
4700 sender may affect whether the remote host accepts the recipient or not. I
4701 have changed it to use the actual sender in the callout; this means that
4702 the cache record is now keyed on a recipient/sender pair, not just the
4703 recipient address. There doesn't seem to be a real danger of callout loops,
4704 since a callout by the remote host to check the sender would use <>.
4705 [SEE ABOVE: changed after hitting problems.]
4707 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4708 temporary errors. However, in the case of such a code being given after
4709 the end of a data transmission (i.e. after ".") Exim was failing to write
4710 a retry record for the message. (Yes, there was some broken host that was
4711 actually sending 8xx at this point.)
4713 73. An unknown lookup type in a host list could cause Exim to panic-die when
4714 the list was checked. (An example that provoked this was putting <; in the
4715 middle of a list instead of at the start.) If this happened during a DATA
4716 ACL check, a -D file could be left lying around. This kind of configuration
4717 error no longer causes Exim to die; instead it causes a defer errror. The
4718 incident is still logged to the main and panic logs.
4720 74. Buglet left over from Exim 3 conversion. The message "too many messages
4721 in one connection" was written to the rejectlog but not the mainlog, except
4722 when address rewriting (yes!) was being logged.
4724 75. Added write_rejectlog option.
4726 76. When a system filter was run not as root (that is, when system_filter_user
4727 was set), the values of the $n variables were not being returned to the
4728 main process; thus, they were not subsequently available in the $sn
4731 77. Added +return_path_on_delivery log selector.
4733 78. A connection timeout was being treated differently from recipients deferred
4734 when testing hosts_max_try with a message that was older than the host's
4735 retry timeout. (The host should not be counted, thus allowing all hosts to
4736 be tried at least once before bouncing.) This may have been the cause of an
4737 occasionally reported bug whereby a message would remain on the queue
4738 longer than the retry timeout, but would be bounced if a delivery was
4739 forced. I say "may" because I never totally pinned down the problem;
4740 setting up timeout/retry tests is difficult. See also the next item.
4742 79. The ultimate address timeout was not being applied to errors that involved
4743 a combination of host plus message (for example, a timeout on a MAIL
4744 command). When an address resolved to a number of possible hosts, and they
4745 were not all tried for each delivery (e.g. because of hosts_max_try), a
4746 message could remain on the queue longer than the retry timeout.
4748 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4751 81. Fixed an obscure SMTP outgoing bug which required at least the following
4752 conditions: (a) there was another message waiting for the same server;
4753 (b) the server returned 5xx to all RCPT commands in the first message so
4754 that the message was not completed; (c) the server dropped the connection
4755 or gave a negative response to the RSET that Exim sends to abort the
4756 transaction. The observed case was a dropped connection after DATA that had
4757 been sent in pipelining mode. That is, the server had advertised PIPELINING
4758 but was not implementing it correctly. The effect of the bug was incorrect
4759 behaviour, such as trying another host, and this could lead to a crash.
4765 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4766 and daemon.c were passed as pointers to ints; they should have been
4767 pointers to socklen_t variables (which are typically unsigned ints).
4769 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4772 3. Fixed a really odd bug that affected only the testing scheme; patching a
4773 certain fixed string in the binary changed the value of another string that
4774 happened to be identical to the end of the original first string.
4776 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4777 name", it returns that address as the IP address. On some operating
4778 systems (e.g. Solaris), it also passes back the IP address string as the
4779 "host name". However, on others (e.g. Linux), it passes back an empty
4780 string. Exim wasn't checking for this, and was changing the host name to an
4781 empty string, assuming it had been canonicized.
4783 5. Although rare, it is permitted to have more than one PTR record for a given
4784 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4785 all the names associated with an address, because they do in Solaris.
4786 However, it seems that they do not in Linux for data that comes from the
4787 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4788 I found this out when I moved to a new Linux workstation and tried to run
4789 the Exim test suite.
4791 To get round this problem I have changed the code so that it now does its
4792 own call to the DNS to look up PTR records when searching for a host name.
4793 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4794 addresses that are only in /etc/hosts are still found.
4796 This behaviour is, however, controlled by an option called host_lookup_
4797 order, which defaults to "bydns:byaddr". If people want to use the other
4798 order, or indeed, just use one or the other means of lookup, they can
4799 specify it in this variable.
4801 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4802 some operating systems, this comes back from gethostbyaddr() as an empty
4803 string, and this is what Exim used to test for. However, it seems that in
4804 other systems, "." is yielded. Exim now tests for this case too.
4806 7. The values of check_spool_space and check_log_space are now held internally
4807 as a number of kilobytes instead of an absolute number of bytes. If a
4808 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4809 kilobyte. This means that much larger values can be stored.
4811 8. Exim monitor: an attempt to get the action menu when not actually pointing
4812 at a message produces an empty menu entitled "No message selected". This
4813 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4814 no entries in it ("Shell widget menu has zero width and/or height"). So I
4815 have added a single, blank menu entry in this case.
4817 9. Added ${quote_local_part.
4819 10. MIME decoding is now applied to the contents of Subject: header lines when
4822 11. Now that a reference to $sender_host_address automatically causes a reverse
4823 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4824 host lookup before query-style lookups in lists that might use this
4825 variable. This has therefore been abolished, and the "net-" prefix is no
4826 longer necessary for query-style lookups.
4828 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4829 have been a typo for LFLAGS, so it has been changed.
4831 13. The install script calls Exim with "-C /dev/null" in order to find the
4832 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4833 to be output. Howeve, since Exim outputs its version number before the
4834 error, it didn't break the script. It just looked ugly. I fixed this by
4835 always allowing "-C /dev/null" if the caller is root.
4837 14. Ignore overlarge ACL variable number when reading spool file - insurance
4838 against a later release with more variables having written the file.
4840 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4841 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4842 this, as well as the form without the "IPv6" on the front.
4844 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4845 OS/Makefile-Darwin file.
4847 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4848 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4850 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4851 variable. However, some OS use ints for inodes. I've added cast to long int
4852 to get rid of the compiler warning.
4854 19. I had forgotten to lock out "/../" in configuration file names when
4855 ALT_CONFIG_PREFIX was set.
4857 20. Routers used for verification do not need to specify transports. However,
4858 if such a router generated a host list, and callout was configured, Exim
4859 crashed, because it could not find a port number from the (non-existent)
4860 transport. It now assumes port 25 in this circumstance.
4862 21. Added the -t option to exigrep.
4864 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4865 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4866 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4867 them is set, LOOKUP_LSEARCH is forced.
4869 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4871 24. Added sender and host information to the "rejected by local_scan()" log
4872 line; previously there was no indication of these.
4874 25. Added .include_if_exists.
4876 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4877 Linux. It turns out that not all file systems support this. Apparently some
4878 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4879 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4880 not supported on the file descriptor, is now ignored when Exim is trying to
4881 sync a directory. This applies only to Linux.
4883 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4885 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4886 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4889 29. Added support for maildirsize files from supplied patch (modified a bit).
4891 30. The use of :fail: followed by an empty string could lead Exim to respond to
4892 sender verification failures with (e.g.):
4894 550 Verification failed for <xxx>
4895 550 Sender verify failed
4897 where the first response line was missing the '-' that indicates it is not
4898 the final line of the response.
4900 31. The loop for finding the name of the user that called Exim had a hardwired
4901 limit of 10; it now uses the value of finduser_retries, which is used for
4902 all other user lookups.
4904 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4907 33. Exim was neglecting to zero errno before one call of strtol() when
4908 expanding a string and expecting an integer value. On some systems this
4909 resulted in spurious "integer overflow" errors. Also, it was casting the
4910 result into an int without checking.
4912 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4913 did not work. The code looks as if it has *never* worked, though it appears
4914 to have been documented since at least releast 1.62. I have made it work.
4916 35. The "timeout_DNS" error in retry rules, also documented since at least
4917 1.62, also never worked. As it isn't clear exactly what this means, and
4918 clearly it isn't a major issue, I have abolished the feature by treating it
4919 as "timeout", and writing a warning to the main and panic logs.
4921 36. The display of retry rules for -brt wasn't always showing the error code
4924 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4925 timeout_connect_A, timeout_connect_MX.
4927 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4928 to the empty sender.
4930 39. The daemon was not analysing the content of -oX till after it had closed
4931 stderr and disconnected from the controlling terminal. This meant that any
4932 syntax errors were only noted on the panic log, and the return code from
4933 the command was 0. By re-arranging the code a little, I've made the
4934 decoding happen first, so such errors now appear on stderr, and the return
4935 code is 1. However, the actual setting up of the sockets still happens in
4936 the disconnected process, so errors there are still only recorded on the
4939 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4940 connections (as happens on some IP stacks) was logged at start up time as
4941 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4942 it from "IPv6 and IPv4", which means that two separate sockets are being
4945 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4946 says whether AF_INET or AF_INET6 was passed as an argument.
4948 42. Exiwhat output was messed up when time zones were included in log
4951 43. Exiwhat now gives more information about the daemon's listening ports,
4952 and whether -tls-on-connect was used.
4954 44. The "port" option of the smtp transport is now expanded.
4956 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4957 silently ignored. Now an error message is written to the main and panic
4960 46. There's a new ACL modifier called "logwrite" which writes to a log file
4961 as soon as it is encountered.
4963 47. Added $local_user_uid and $local_user_gid at routing time.
4965 48. Exim crashed when trying to verify a sender address that was being
4968 49. Exim was recognizing only a space character after ".include". It now also
4969 recognizes a tab character.
4971 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4972 extracting the relevant information from the specification. The man page no
4973 longer contains scrambled data for the -d option, and I've added a section
4974 at the front about calling Exim under different names.
4976 51. Added "extra_headers" argument to the "mail" command in filter files.
4978 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4981 53. Installed eximstats 1.29.
4983 54. Added transport_filter_timeout as a generic transport option.
4985 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4986 Cc: header lines. This was required by RFC 822, but it not required by RFC
4989 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4990 incoming messages that did not have them. Now it does so only if the
4991 message originates locally, that is, if there is no associated remote host
4992 address. When Resent- header lines are present, this applies to the Resent-
4993 lines rather than the non-Resent- lines.
4995 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4996 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4998 58. Messages for configuration errors now include the name of the main
4999 configuration file - useful now that there may be more than one file in a
5000 list (.included file names were always shown).
5002 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5003 for those rare installations that do not start the daemon as root or run it
5004 setuid root. I've cut out the call to initgroups() if the daemon is not
5007 60. The Exim user and group can now be bound into the binary as text strings
5008 that are looked up at the start of Exim's processing.
5010 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5012 62. Added $mailstore_basename variable.
5014 63. Installed patch to sieve.c from Michael Haardt.
5016 64. When Exim failed to open the panic log after failing to open the main log,
5017 the original message it was trying to log was written to stderr and debug
5018 output, but if they were not available (the usual case in production), it
5019 was lost. Now it is written to syslog before the two lines that record the
5020 failures to open the logs.
5022 65. Users' Exim filters run in subprocesses under the user's uid. It is
5023 possible for a "deliver" command or an alias in a "personal" command to
5024 provoke an address rewrite. If logging of address rewriting is configured,
5025 this fails because the process is not running as root or exim. There may be
5026 a better way of dealing with this, but for the moment (because 4.30 needs
5027 to be released), I have disabled address rewrite logging when running a
5028 filter in a non-root, non-exim process.
5034 1. The buildconfig auxiliary program wasn't quoting the value set for
5035 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5036 not defined. This bug was masked in 4.22 by the effect that was fixed in
5039 2. Some messages that were rejected after a message id was allocated were
5040 shown as "incomplete" by exigrep. It no longer does this for messages that
5041 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5043 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5044 have allow_domain_literals set, the ID did not get logged in the <= line.
5045 Domain literals are now always recognized in Message-ID: header lines.
5047 4. The first argument for a ${extract expansion item is the key name or field
5048 number. Leading and trailing spaces in this item were not being ignored,
5049 causing some misleading effects.
5051 5. When deliver_drop_privilege was set, single queue runner processes started
5052 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5053 same command in the process it spins off) were not dropping privilege.
5055 6. When the daemon running as "exim" started a queue runner, it always
5056 re-executed Exim in the spun-off process. This is a waste of effort when
5057 deliver_drop_privilege is set. The new process now just calls the
5058 queue-runner function directly.
5064 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5067 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5068 ignored. Though the use of -oP was forcing the writing of a pid file, it
5069 was always written to the default place.
5071 3. If the message "no IP address found for host xxxx" is generated during
5072 incoming verification, it is now followed by identification of the incoming
5073 connection (so you can more easily find what provoked it).
5075 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5077 5. Added some features to "harden" Exim a bit more against certain attacks:
5079 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5080 be put in Local/Makefile. This is like the never_users runtime option,
5081 but it cannot be overridden. The default setting is "root".
5083 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5084 prefix string with which any file named in a -C command line option
5087 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5088 is retained for -C and -D only if the caller of Exim is root. Without
5089 it, the exim user may also use -C and -D and retain privilege.
5091 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5092 command line option is disabled.
5094 6. Macro names set by the -D option must start with an upper case letter, just
5095 like macro names defined in the configuration file.
5097 7. Added "dereference=" facility to LDAP.
5099 8. Two instances of the typo "uknown" in the source files are fixed.
5101 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5102 the Configure-Makefile script screwed up while processing it.
5104 10. Incorporated PCRE 4.4.
5106 11. The SMTP synchronization check was not operating right at the start of an
5107 SMTP session. For example, it could not catch a HELO sent before the client
5108 waited for the greeting. There is now a check for outstanding input at the
5109 point when the greeting is written. Because of the duplex, asynchronous
5110 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5111 way, but not yet received, when the check is performed.
5113 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5114 on TCP/IP sockets, because this apparently causes some broken clients to
5117 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5118 unchanged) from the Cygwin maintainer.
5120 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5121 was supported instead of testing for mbx. Effectively a typo.
5123 15. The spa authenticator server code was not checking that the input it
5124 received was valid base64.
5126 16. The debug output line for the "set" modifier in ACLs was not showing the
5127 name of the variable that was being set.
5129 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5131 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5132 lookup on its own. Something else was needed to trigger the lookup. For
5133 example, a match in host_lookup or the need for a host name in a host list.
5134 Now, if $sender_host_name is referenced and the host name has not yet been
5135 looked up, a lookup is performed. If the lookup fails, the variable remains
5136 empty, and $host_lookup_failed is set to "1".
5138 19. Added "eqi" as a case-independent comparison operator.
5140 20. The saslauthd authentication condition could segfault if neither service
5141 nor realm was specified.
5143 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5144 error message that was logged was misleading, and incoming SMTP
5145 connections were dropped. The message is now more accurate, and temporary
5146 errors are given to SMTP connections.
5148 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5149 (except RSET) with a 421 error, until QUIT is received. However, it was
5150 failing to send a response to QUIT.
5152 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5153 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5154 ACL is now run after the helo_try_verify_hosts code.
5156 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5157 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5158 case-independent, so other case variants are also recognized). Apparently
5159 some systems use these upper case variants.
5161 25. If more than two messages were waiting for the same host, and a transport
5162 filter was specified for the transport, Exim sent two messages over the
5163 same TCP/IP connection, and then failed with "socket operation on non-
5164 socket" when it tried to send the third.
5166 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5168 27. The extern definition of crypt16() in expand.c was not being excluded when
5169 the OS had its own crypt16() function.
5171 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5172 as a preferred synonym for return_size_limit, both as an option and as an
5175 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5177 30. Changed the default configuration ACL to relax the local part checking rule
5178 for addresses that are not in any local domains. For these addresses,
5179 slashes and pipe symbols are allowed within local parts, but the sequence
5180 /../ is explicitly forbidden.
5182 31. SPA server authentication was not clearing the challenge buffer before
5185 32. log_message in a "warn" ACL statement was writing to the reject log as
5186 well as to the main log, which contradicts the documentation and doesn't
5187 seem right (because no rejection is happening). So I have stopped it.
5189 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5190 However, I am unable to do any testing of this.
5192 34. Fixed an infelicity in the appendfile transport. When checking directories
5193 for a mailbox, to see if any needed to be created, it was accidentally
5194 using path names with one or more superfluous leading slashes; tracing
5195 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5197 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5198 contents are added to the log line that is written for every discarded
5199 recipient. (Previously a log_message setting was ignored.)
5201 36. The ${quote: operator now quotes the string if it is empty.
5203 37. The install script runs exim in order to find its version number. If for
5204 some reason other than non-existence or emptiness, which it checks, it
5205 could not run './exim', it was installing it with an empty version number,
5206 i.e. as "exim-". This error state is now caught, and the installation is
5209 38. An argument was missing from the function that creates an error message
5210 when Exim fails to connect to the socket for saslauthd authentication.
5211 This could cause Exim to crash, or give a corrupted message.
5213 39. Added isip, isip4, and isip6 to ${if conditions.
5215 40. The ACL variables $acl_xx are now saved with the message, and can be
5216 accessed later in routers, transports, and filters.
5218 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5219 strings in the file are not string-expanded.
5221 42. If a MAIL command specified a SIZE value that was too large to fit into an
5222 int variable, the check against message_size_limit failed. Such values are
5223 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5224 day this will have to be increased, but I don't think I want to be around
5225 when emails are that large.
5232 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5233 iconv() is not standard in FreeBSD.
5235 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5236 IPv6 enabled. The observed symptom was a segmentation fault on return from
5237 the function os_common_find_running_interfaces() in src/os.c.
5239 3. In the check_special_case() function in daemon.c I had used "errno" as an
5240 argument name, which causes warnings on some systems. This was basically a
5241 typo, since it was named "eno" in the comments!
5243 4. The code that waits for the clock to tick (at a resolution of some fraction
5244 of a second) so as to ensure message-id uniqueness was always waiting for
5245 at least one whole tick, when it could have waited for less. [This is
5246 almost certainly not relevant at current processor speeds, where it is
5247 unlikely to ever wait at all. But we try to future-proof.]
5249 5. The function that sleeps for a time interval that includes fractions of a
5250 second contained a race. It did not block SIGALRM between setting the
5251 timer, and suspending (a couple of lines later). If the interval was short
5252 and the sigsuspend() was delayed until after it had expired, the suspension
5253 never ended. On busy systems this could lead to processes getting stuck for
5256 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5257 process, before it forks any delivery processes. The open lookup caching
5258 mechanism meant that the open file or database connection was passed into
5259 the delivery process. The problem was that delivery processes always tidy
5260 up cached lookup data. This could cause a problem for the next delivery
5261 process started by the queue runner, because the external queue runner
5262 process does not know about the closure. So the next delivery process
5263 still has data in the lookup cache. In the case of a file lookup, there was
5264 no problem because closing a file descriptor in a subprocess doesn't affect
5265 the parent. However, if the lookup was caching a connection to a database,
5266 the connection was closed, and the second delivery process was likely to
5267 see errors such as "PGSQL: query failed: server closed the connection
5268 unexpectedly". The problem has been fixed by closing all cached lookups
5269 in a queue runner before running a delivery process.
5271 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5272 seem to have the "const" qualifier which it has on other OS. I've
5275 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5276 *of the same type* (client or server) with the same public name that an
5277 error should be diagnosed.
5279 9. When Exim looked up a host name for an IP address, but failed to find the
5280 original IP address when looking up the host name (a safety check), it
5281 output the message "<ip address> does not match any IP for NULL", which was
5282 confusing, to say the least. The bug was that the host name should have
5283 appeared instead of "NULL".
5285 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5286 user that is built into the binary, and the -C or -D options is used, root
5287 privilege is dropped before the configuration file is read. In addition,
5288 logging is switched to stderr instead of the normal log files. If the
5289 configuration then re-defines the Exim user, the unprivileged environment
5290 is probably not what is expected, so Exim logs a panic warning message (but
5293 However, if deliver_drop_privilege is set, the unprivileged state may well
5294 be exactly what is intended, so the warning has been cut out in that case,
5295 and Exim is allowed to try to write to its normal log files.
5301 1. smtp_return_error_details was not giving details for temporary sender
5302 or receiver verification errors.
5304 2. Diagnose a configuration error if two authenticators have the same public
5307 3. Exim used not to create the message log file for a message until the first
5308 delivery attempt. This could be confusing when incoming messages were held
5309 for policy or load reasons. The message log file is now created at the time
5310 the message is received, and an initial "Received" line is written to it.
5312 4. The automatically generated man page for command line options had a minor
5313 bug that caused no ill effects; however, a more serious problem was that
5314 the procedure for building the man page automatically didn't always
5315 operate. Consequently, release 4.20 contains an out-of-date version. This
5316 shouldn't happen again.
5318 5. When building Exim with embedded Perl support, the script that builds the
5319 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5320 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5322 6. The freeze_tell option was not being used for messages that were frozen on
5323 arrival, either by an ACL or by local_scan().
5325 7. Added the smtp_incomplete_transaction log selector.
5327 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5328 was accepting AUTH without a new EHLO.
5330 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5331 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5334 10. Logging of TCP/IP connections (when configured) now happens in the main
5335 daemon process instead of the child process, so that the TCP/IP connection
5336 count is more accurate (but it can never be perfect).
5338 11. The use of "drop" in a nested ACL was not being handled correctly in the
5339 outer ACL. Now, if condition failure induced by the nested "drop" causes
5340 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5341 or "require"), the connection is dropped.
5343 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5344 that yield "discard" can now be used with an "accept" or a "discard" verb,
5345 but an error is generated for any others (because I can't see a useful way
5346 to define what should happen).
5348 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5349 are now processed in the same way as lines in the Exim configuration file.
5350 In particular, continuation lines are supported.
5352 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5354 15. Added -ti meaning -t -i.
5356 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5357 domains, and warn by logging if others are found.
5359 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5360 returned from GIFCONF when Exim is trying to find the list of interfaces on
5361 a host. The code in os.c has been modified to copy each ifreq to an aligned
5362 structure in all cases.
5364 Also, in some cases, the returned ifreq's were being copied to a 'struct
5365 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5366 means the last couple of bytes of an IPv6 address could be chopped if the
5367 ifreq contained only a normal sockaddr (14 bytes storage).
5369 18. Named domain lists were not supported in the hosts_treat_as_local option.
5370 An entry such as +xxxx was not recognized, and was treated as a literal
5373 19. Ensure that header lines added by a DATA ACL are included in the reject log
5374 if the ACL subsequently rejects the message.
5376 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5377 MD5 (which is deprecated).
5379 21. When testing a filter file using -bf, Exim was writing a message when it
5380 took the sender from a "From " line in the message, but it was not doing so
5381 when it took $return_path from a Return-Path: header line. It now does.
5383 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5384 with a valid header line field name (a series of printing characters
5385 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5387 23. Changed "disc" in the source to "disk" to conform to the documentation and
5388 the book and for uniformity.
5390 24. Ignore Sendmail's -Ooption=value command line item.
5392 25. When execve() failed while trying to run a command in a pipe transport,
5393 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5394 could be confused with a return value of 69 from the command itself. This
5395 has been changed to 127, the value the shell returns if it is asked to run
5396 a non-existent command. The wording for the related log line suggests a
5397 non-existent command as the problem.
5399 26. If received_header_text expands to an empty string, do not add a Received:
5400 header line to the message. (Well, it adds a token one on the spool, but
5401 marks it "old" so that it doesn't get used or transmitted.)
5403 27. Installed eximstats 1.28 (addition of -nt option).
5405 28. There was no check for failure on the call to getsockname() in the daemon
5406 code. This can fail if there is a shortage of resources on the system, with
5407 ENOMEM, for example. A temporary error is now given on failure.
5409 29. Contrary to the C standard, it seems that in some environments, the
5410 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5411 program. Exim now does this explicitly; it affects the formatting of
5412 timestamps using strftime().
5414 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5415 complaints. I've now initialized all the variables, to avoid this.
5417 32. Header lines added by a system filter were not being "seen" during
5418 transport-time rewrites.
5420 33. The info_callback() function passed to OpenSSL is set up with type void
5421 (*)(SSL *, int, int), as described somewhere. However, when calling the
5422 function (actually a macro) that sets it up, the type void(*)() is
5423 expected. I've put in a cast to prevent warnings from picky compilers.
5425 34. If a DNS black list lookup found a CNAME record, but there were no A
5426 records associated with the domain it pointed at, Exim crashed.
5428 35. If a DNS black list lookup returned more than one A record, Exim ignored
5429 all but the first. It now scans all returned addresses if a particular IP
5430 value is being sought. In this situation, the contents of the
5431 $dnslist_value variable are a list of all the addresses, separated by a
5434 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5435 to accept a host name and all its aliases if the forward lookup for any of
5436 them yielded the IP address of the incoming connection. Now it accepts only
5437 those names whose forward lookup yields the correct IP address. Any other
5438 names are discarded. This closes a loophole whereby a rogue DNS
5439 administrator could create reverse DNS records to break through a
5440 wildcarded host restriction in an ACL.
5442 37. If a user filter or a system filter that ran in a subprocess used any of
5443 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5444 the wrong values were passed to the pipe command ($thisaddress had the
5445 value of $0, $0 had the value of $1, etc). This bug was introduced by
5446 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5448 38. Improved the line breaking for long SMTP error messages from ACLs.
5449 Previously, if there was no break point between 40 and 75 characters, Exim
5450 left the rest of the message alone. Two changes have been made: (a) I've
5451 reduced the minimum length to 35 characters; (b) if it can't find a break
5452 point between 35 and 75 characters, it looks ahead and uses the first one
5453 that it finds. This may give the occasional overlong line, but at least the
5454 remaining text gets split now.
5456 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5457 file descriptors might be low, and that setting 1000 would always raise it.
5458 It turns out that in some environments, the limit is already over 1000 and
5459 that lowering it causes trouble. So now Exim takes care not to decrease it.
5461 40. When delivering a message, the value of $return_path is set to $sender_
5462 address at the start of routing (routers may change the value). By an
5463 oversight, this default was not being set up when an address was tested by
5464 -bt or -bv, which affected the outcome if any router or filter referred to
5467 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5468 log only when "message" or "log_message" are set. However, if one of the
5469 conditions was an address verification, or a call to a nested ACL, the
5470 messages generated by the underlying test were being passed through. This
5471 no longer happens. The underlying message is available in $acl_verify_
5472 message for both "message" and "log_message" expansions, so it can be
5473 passed through if needed.
5475 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5476 new expansion $bh_ to give the encoded byte string without charset
5477 translation. Translation happens only if iconv() is available; HAVE_ICONV
5478 indicates this at build time. HEADERS_CHARSET gives the charset to
5479 translate to; headers_charset can change it in the configuration, and
5480 "headers charset" can change it in an individual filter file.
5482 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5483 that creates RFC 2047 encoded "words" labels them as that charset instead
5484 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5485 expansion operator; (ii) when Exim creates a From: line for a local
5486 message; (iii) when a header line is rewritten to include a "phrase" part.
5488 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5489 buggy, causing it to skip the first lines of messages whose message ID
5490 ended in 'D'. This would not have bitten before Exim release 4.14, because
5491 message IDs were unlikely to end in 'D' before then. The effect was to have
5492 incorrect size information for certain domains.
5494 45. #include "config.h" was missing at the start of the crypt16.c module. This
5495 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5498 46. If there was a timeout during a "random" callout check, Exim treated it as
5499 a failure of the random address, and carried on sending RSET and the real
5500 address. If the delay was just some slowness somewhere, the response to the
5501 original RCPT would be taken as a response to RSET and so on, causing
5502 mayhem of various kinds.
5504 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5505 when I implemented it. It didn't allow for the fact that some option values
5506 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5507 the right test for positive values.
5509 48. Domain names in DNS records are case-independent. Exim always looks them up
5510 in lower case. Some resolvers return domain names in exactly the case they
5511 appear in the zone file, that is, they may contain uppercase letters. Not
5512 all resolvers do this - some return always lower case. Exim was treating a
5513 change of case by a resolver as a change of domain, similar to a widening
5514 of a domain abbreviation. This triggered its re-routing code and so it was
5515 trying to route what was effectively the same domain again. This normally
5516 caused routing to fail (because the router wouldn't handle the domain
5517 twice). Now Exim checks for this case specially, and just changes the
5518 casing of the domain that it ultimately uses when it transmits the message
5521 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5524 50. If a filter generated a file delivery with a non-absolute name (possible if
5525 no home directory exists for the router), the forbid_file option was not
5528 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5529 the existing equality matching.
5531 52. Exim was using ints instead of ino_t variables in some places where it was
5532 dealing with inode numbers.
5534 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5535 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5536 called TMPDIR, and if it finds it is different, it changes its value.
5538 54. The smtp_printf() function is now made available to local_scan() so
5539 additional output lines can be written before returning. There is also an
5540 smtp_fflush() function to enable the detection of a dropped connection.
5541 The variables smtp_input and smtp_batched_input are exported to
5544 55. Changed the default runtime configuration: the message "Unknown user"
5545 has been removed from the ACL, and instead placed on the localuser router,
5546 using the cannot_route_message feature. This means that any verification
5547 failures that generate their own messages won't get overridden. Similarly,
5548 the "Unrouteable address" message that was in the ACL for unverifiable
5549 relay addresses has also been removed.
5551 56. Added hosts_avoid_esmtp to the smtp transport.
5553 57. The exicyclog script was not checking for the esoteric option
5554 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5555 will work only if exicyclog is run under the appropriate euid.
5557 58. Following a discussion on the list, the rules by which Exim recognises line
5558 endings on incoming messages have been changed. The -dropcr and drop_cr
5559 options are now no-ops, retained only for backwards compatibility. The
5560 following line terminators are recognized: LF CRLF CR. However, special
5561 processing applies to CR:
5563 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5564 nor a local message in the state where . is a terminator.
5566 (ii) If a bare CR is encountered in a header line, an extra space is added
5567 after the line terminator so as not to end the header. The reasoning
5568 behind this is that bare CRs in header lines are most likely either
5569 to be mistakes, or people trying to play silly games.
5571 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5572 was being incorrectly given as 18 bytes larger than it should have been.
5573 This is a VOB (very old bug).
5575 60. This may never have affected anything current, but just in case it has:
5576 When the local host is found other than at the start of a list of hosts,
5577 the local host, those with the same MX, and any that follow, are discarded.
5578 When the list in question was part of a longer list of hosts, the following
5579 hosts (not currently being processed) were also being discarded. This no
5580 longer happens. I'm not sure if this situation could ever has previously
5583 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5586 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5587 header. This is something that is recommended in a new Internet Draft, and
5588 is something that is documented as being done by Sendmail. There are two
5589 possible values. For messages generated by the autoreply transport, Exim
5592 Auto-Submitted: auto-replied
5594 whereas for all other generated messages (e.g. bounces) it adds
5596 Auto-Submitted: auto-generated
5598 63. The "personal" condition in filters now includes a test for the
5599 Auto-Submitted: header. If it contains the string "auto-" the message it
5600 not considered personal.
5602 64. Added rcpt_include_affixes as a generic transport option.
5604 65. Added queue_only_override (default true).
5606 66. Added the syslog_duplication option.
5608 67. If what should have been the first header line of a message consisted of
5609 a space followed by a colon, Exim was mis-interpreting it as a header line.
5610 It isn't of course - it is syntactically invalid and should therefore be
5611 treated as the start of the message body. The misbehaviour could have
5612 caused a number of strange effects, including loss of data in subsequent
5613 header lines, and spool format errors.
5615 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5616 client host had authenticated. This control can now be exercised by an ACL
5617 for more flexibility.
5619 69. By default, callouts do not happen when testing with -bh. There is now a
5620 variant, -bhc, which does actually run the callout code, including
5621 consulting and updating the callout cache.
5623 70. Added support for saslauthd authentication, courtesy of Alexander
5626 71. If statvfs() failed on the spool or log directories while checking their
5627 size for availability, Exim confusingly gave the error "space shortage".
5628 Furthermore, in debugging mode it crashed with a floating point exception.
5629 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5630 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5631 serious problem, Exim now writes to the main and panic logs when this
5632 happens, with details of the failure. It then refuses to accept the
5633 incoming message, giving the message "spool directory problem" or "log
5634 directory problem" with a 421 code for SMTP messages.
5636 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5637 0, 1, and 2 exist, because some OS complain for execs without them (see
5638 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5639 descriptors. However, the code omitted to check that the open succeeded,
5640 causing mysterious errors if for some reason the permissions on /dev/null
5641 got screwed. Now Exim writes a message to the main and panic logs, and
5642 bombs out if it can't open /dev/null.
5644 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5645 interact so that it is all more flexible. It is supposed to remain
5646 backwards compatible. Also added extra_local_interfaces.
5648 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5649 to bomb out with an assertion failure - to the client this appears as a
5650 connection drop. This problem occurs in the part of the code that was taken
5651 from the Samba project. Fortunately, the assertion is in a very simple
5652 function, so I have fixed this by reproducing the function inline in the
5653 one place where it is called, and arranging for authentication to fail
5654 instead of killing the process with assert().
5656 75. The SPA client code was not working when the server requested OEM rather
5657 than Unicode encoding.
5659 76. Added code to make require_files with a specific uid setting more usable in
5660 the case where statting the file as root fails - usually a non-root-mounted
5661 NFS file system. When this happens and the failure is EACCES, Exim now
5662 forks a subprocess and does the per-uid checking as the relevant uid.
5664 77. Added process_log_path.
5666 78. If log_file_path was not explicitly set, a setting of check_log_space or
5667 check_log_inodes was ignored.
5669 79. If a space check for the spool or log partitions fails, the incident is now
5670 logged. Of course, in the latter case the data may get lost...
5672 80. Added the %p formatting code to string_format() so that it can be used to
5673 print addresses in debug_print(). Adjusted all the address printing in the
5674 debugging in store.c to use %p rather than %d.
5676 81. There was a concern that a line of code in smtp_in.c could overflow a
5677 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5678 initially expressed, the concern was not well-founded, because trailing
5679 spaces are removed early. However, if the trailing spaces were followed by
5680 a NULL, they did not get removed, so the overflow was possible. Two fixes
5683 (a) I re-wrote the offending code in a cleaner fashion.
5684 (b) If an incoming SMTP command contains a NULL character, it is rejected
5687 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5688 runs initgroups(), so that if the Exim user is in any additional groups,
5689 they will be used during message reception.
5695 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob