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.
46 PP/01 Handle short writes when writing local log-files.
47 In practice, only affects FreeBSD (8 onwards).
48 Bugzilla 1053, with thanks to Dmitry Isaikin.
50 NM/01 Bugzilla 949 - Documentation tweak
52 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
55 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
57 PP/02 Implemented gsasl authenticator.
59 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
61 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
62 `pkg-config foo` for cflags/libs.
64 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
65 with rest of GSASL and with heimdal_gssapi.
67 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
68 `pkg-config foo` for cflags/libs for the TLS implementation.
70 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
71 properties get this fed in as external SSF. A number of robustness
72 and debugging improvements to the cyrus_sasl authenticator.
74 PP/08 cyrus_sasl server now expands the server_realm option.
76 PP/09 Bugzilla 1214 - Log authentication information in reject log.
77 Patch by Jeremy Harris.
79 PP/10 Added dbmjz lookup type.
81 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
83 PP/12 MAIL args handles TAB as well as SP, for better interop with
84 non-compliant senders.
85 Analysis and variant patch by Todd Lyons.
87 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
88 Bug report from Lars Müller <lars@samba.org> (via SUSE),
89 Patch from Dirk Mueller <dmueller@suse.com>
91 PP/13 tls_peerdn now print-escaped for spool files.
92 Observed some $tls_peerdn in wild which contained \n, which resulted
93 in spool file corruption.
95 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
96 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
97 or write after TLS renegotiation, which otherwise led to messages
100 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
101 as a tracking header (ie: a signed header comes before the signature).
102 Patch from Wolfgang Breyha.
104 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
105 comma-sep list; embedded commas doubled.
107 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
109 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
111 Report and patch from Dmitry Banschikov.
113 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
114 Removed SSL_clear() after SSL_new() which led to protocol negotiation
115 failures. We appear to now support TLS1.1+ with Exim.
117 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
118 lets Exim select keys and certificates based upon TLS SNI from client.
119 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
120 before an outbound SMTP session. New log_selector, +tls_sni.
122 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
123 NULL dereference. Report and patch from Alun Jones.
125 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
126 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
127 Not seeing resolver debug output on NetBSD, but suspect this is a
128 resolver implementation change.
130 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
131 Left warnings. Added "eximon gdb" invocation mode.
133 PP/21 Defaulting "accept_8bitmime" to true, not false.
135 PP/22 Added -bw for inetd wait mode support.
137 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
138 locate the relevant includes and libraries. Made this the default.
140 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
141 Bugzilla 1246, report and most of solution from Tomasz Kusy.
143 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
144 This may cause build issues on older platforms.
146 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
147 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
148 gnutls_require_mac & gnutls_require_protocols (no longer supported).
149 Added SNI support via GnuTLS too.
150 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
152 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
154 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
155 Applied second patch from Janne, implementing suggestion to default
156 multiple-strings-in-record handling to match SPF spec.
158 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
160 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
161 read-only, out of scope).
162 Patch from Wolfgang Breyha, report from Stuart Northfield.
164 PP/29 Fix three issues highlighted by clang analyser static analysis.
165 Only crash-plausible issue would require the Cambridge-specific
166 iplookup router and a misconfiguration.
167 Report from Marcin Mirosław.
169 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
171 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
172 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
173 As part of this, removing so much warning spew let me fix some minor
174 real issues in debug logging.
176 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
177 assignment on my part. Fixed.
179 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
180 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
181 Janne Snabb (who went above and beyond: thank you).
183 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
184 string otherwise requires a connection and a bunch more work and it's
185 relatively easy to get wrong. Should also expose TLS library linkage
188 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
189 64-bit ${eval} (JH/03).
191 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
192 GNU libc to support some of the 64-bit stuff, should not lead to
193 conflicts. Defined before os.h is pulled in, so if a given platform
194 needs to override this, it can.
196 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
197 protection layer was required, which is not implemented.
198 Bugzilla 1254, patch from Wolfgang Breyha.
200 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
201 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
202 tls_dhparam take prime identifiers. Also unbreak combination of
203 OpenSSL+DH_params+TLSSNI.
205 PP/39 Disable SSLv2 by default in OpenSSL support.
211 PP/01 Solaris build fix for Oracle's LDAP libraries.
212 Bugzilla 1109, patch from Stephen Usher.
214 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
216 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
219 TF/02 Fix a couple more cases where we did not log the error message
220 when unlink() failed. See also change 4.74-TF/03.
222 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
223 lock up or crash if it happened to be inside a call to libc when it
224 got a SIGUSR1 from exiwhat.
226 The SIGUSR1 handler appends the current process status to the process
227 log which is later printed by exiwhat. It used to use the general
228 purpose logging code to do this, but several functions it calls are
229 not safe for signals.
231 The new output code in the SIGUSR1 handler is specific to the process
232 log, and simple enough that it's easy to inspect for signal safety.
233 Removing some special cases also simplifies the general logging code.
234 Removing the spurious timestamps from the process log simplifies
237 TF/04 Improved ratelimit ACL condition.
239 The /noupdate option has been deprecated in favour of /readonly which
240 has clearer semantics. The /leaky, /strict, and /readonly update modes
241 are mutually exclusive. The update mode is no longer included in the
242 database key; it just determines when the database is updated. (This
243 means that when you upgrde Exim will forget old rate measurements.)
245 Exim now checks that the per_* options are used with an update mode that
246 makes sense for the current ACL. For example, when Exim is processing a
247 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
248 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
249 must specify per_mail/readonly. If you omit the update mode it defaults to
250 /leaky where that makes sense (as before) or /readonly where required.
252 The /noupdate option is now undocumented but still supported for
253 backwards compatibility. It is equivalent to /readonly except that in
254 ACLs where /readonly is required you may specify /leaky/noupdate or
255 /strict/noupdate which are treated the same as /readonly.
257 A useful new feature is the /count= option. This is a generalization
258 of the per_byte option, so that you can measure the throughput of other
259 aggregate values. For example, the per_byte option is now equivalent
260 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
262 The per_rcpt option has been generalized using the /count= mechanism
263 (though it's more complicated than the per_byte equivalence). When it is
264 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
265 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
266 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
267 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
268 Note that using per_rcpt with a non-readonly update mode in more than
269 one ACL will cause the recipients to be double-counted. (The per_mail
270 and per_byte options don't have this problem.)
272 The handling of very low rates has changed slightly. If the computed rate
273 is less than the event's count (usually one) then this event is the first
274 after a long gap. In this case the rate is set to the same as this event's
275 count, so that the first message of a spam run is counted properly.
277 The major new feature is a mechanism for counting the rate of unique
278 events. The new per_addr option counts the number of different
279 recipients that someone has sent messages to in the last time period. It
280 behaves like per_rcpt if all the recipient addresses are different, but
281 duplicate recipient addresses do not increase the measured rate. Like
282 the /count= option this is a general mechanism, so the per_addr option
283 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
284 example, measure the rate that a client uses different sender addresses
285 with the options per_mail/unique=$sender_address. There are further
286 details in the main documentation.
288 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
290 TF/06 Removed a few PCRE remnants.
292 TF/07 Automatically extract Exim's version number from tags in the git
293 repository when doing development or release builds.
295 PP/02 Raise smtp_cmd_buffer_size to 16kB.
296 Bugzilla 879. Patch from Paul Fisher.
298 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
299 Heavily based on revision 40f9a89a from Simon Arlott's tree.
302 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
304 PP/05 Variable $av_failed, true if the AV scanner deferred.
305 Bugzilla 1078. Patch from John Horne.
307 PP/06 Stop make process more reliably on build failure.
308 Bugzilla 1087. Patch from Heiko Schlittermann.
310 PP/07 Make maildir_use_size_file an _expandable_ boolean.
311 Bugzilla 1089. Patch from Heiko Schlittermann.
313 PP/08 Handle ${run} returning more data than OS pipe buffer size.
314 Bugzilla 1131. Patch from Holger Weiß.
316 PP/09 Handle IPv6 addresses with SPF.
317 Bugzilla 860. Patch from Wolfgang Breyha.
319 PP/10 GnuTLS: support TLS 1.2 & 1.1.
321 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
324 PP/11 match_* no longer expand right-hand-side by default.
325 New compile-time build option, EXPAND_LISTMATCH_RHS.
326 New expansion conditions, "inlist", "inlisti".
328 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
330 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
332 PP/14 fix log_write() format string regression from TF/03.
333 Bugzilla 1152. Patch from Dmitry Isaikin.
339 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
341 PP/02 Harmonised TLS library version reporting; only show if debugging.
342 Layout now matches that introduced for other libraries in 4.74 PP/03.
344 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
346 PP/04 New "dns_use_edns0" global option.
348 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
351 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
352 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
354 TK/01 Updated PolarSSL code to 0.14.2.
355 Bugzilla 1097. Patch from Andreas Metzler.
357 PP/07 Catch divide-by-zero in ${eval:...}.
360 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
363 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
364 format-string attack -- SECURITY: remote arbitrary code execution.
366 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
367 time unintentionally subject to list matching rules, letting the header
368 cause arbitrary Exim lookups (of items which can occur in lists, *not*
369 arbitrary string expansion). This allowed for information disclosure.
371 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
372 INT_MIN/-1 -- value coerced to INT_MAX.
378 NM/01 Workround for PCRE version dependency in version reporting
381 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
382 This fixes portability to compilers other than gcc, notably
383 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
385 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
386 makefiles for portability to HP-UX and POSIX correctness.
388 PP/01 Permit LOOKUP_foo enabling on the make command-line.
389 Also via indented variable definition in the Makefile.
390 (Debugging by Oliver Heesakkers).
392 PP/02 Restore caching of spamd results with expanded spamd_address.
393 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
395 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
396 Improves build reliability. Fix from: Frank Elsner
398 NM/02 Fix wide character breakage in the rfc2047 coding
399 Fixes bug 1064. Patch from Andrey N. Oktyabrski
401 NM/03 Allow underscore in dnslist lookups
402 Fixes bug 1026. Patch from Graeme Fowler
404 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
405 Code patches from Adam Ciarcinski of NetBSD.
407 NM/04 Fixed exiqgrep to cope with mailq missing size issue
410 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
411 is logged, to avoid truncation. Patch from John Horne.
413 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
414 Patch from Jakob Hirsch.
416 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
417 SQL string expansion failure details.
418 Patch from Andrey Oktyabrski.
420 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
421 Patch from Simon Arlott.
423 PP/09 New lookups functionality failed to compile on old gcc which rejects
424 extern declarations in function scope.
425 Patch from Oliver Fleischmann
427 PP/10 Use sig_atomic_t for flags set from signal handlers.
428 Check getgroups() return and improve debugging.
429 Fixed developed for diagnosis in bug 927 (which turned out to be
432 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
433 Patch from Mark Zealey.
435 PP/12 Bugzilla 1056: Improved spamd server selection.
436 Patch from Mark Zealey.
438 PP/13 Bugzilla 1086: Deal with maildir quota file races.
439 Based on patch from Heiko Schlittermann.
441 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
442 Patch from Uwe Doering, sign-off by Michael Haardt.
444 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
445 variable declaration deep within a block. Bug and patch from
448 PP/15 lookups-Makefile IRIX compatibilty coercion.
450 PP/16 Make DISABLE_DKIM build knob functional.
452 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
453 Patch by Simon Arlott
455 TF/03 Fix valgrind.h portability to C89 compilers that do not support
456 variable argument macros. Our copy now differs from upstream.
462 TF/01 Failure to get a lock on a hints database can have serious
463 consequences so log it to the panic log.
465 TF/02 Log LMTP confirmation messages in the same way as SMTP,
466 controlled using the smtp_confirmation log selector.
468 TF/03 Include the error message when we fail to unlink a spool file.
470 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
471 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
472 for maintaining out-of-tree patches for some time.
474 PP/01 Bugzilla 139: Documentation and portability issues.
475 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
476 Handle per-OS dynamic-module compilation flags.
478 PP/02 Let /dev/null have normal permissions.
479 The 4.73 fixes were a little too stringent and complained about the
480 permissions on /dev/null. Exempt it from some checks.
481 Reported by Andreas M. Kirchwitz.
483 PP/03 Report version information for many libraries, including
484 Exim version information for dynamically loaded libraries. Created
485 version.h, now support a version extension string for distributors
486 who patch heavily. Dynamic module ABI change.
488 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
489 privilege escalation vulnerability whereby the Exim run-time user
490 can cause root to append content of the attacker's choosing to
493 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
496 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
497 If dropping privileges for untrusted macros, we disabled normal logging
498 on the basis that it would fail; for the Exim run-time user, this is not
499 the case, and it resulted in successful deliveries going unlogged.
500 Fixed. Reported by Andreas Metzler.
506 PP/01 Date: & Message-Id: revert to normally being appended to a message,
507 only prepend for the Resent-* case. Fixes regression introduced in
508 Exim 4.70 by NM/22 for Bugzilla 607.
510 PP/02 Include check_rfc2047_length in configure.default because we're seeing
511 increasing numbers of administrators be bitten by this.
513 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
515 PP/03 Bugzilla 994: added openssl_options main configuration option.
517 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
519 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
521 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
523 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
524 without a peer certificate, leading to a segfault because of an
525 assumption that peers always have certificates. Be a little more
526 paranoid. Problem reported by Martin Tscholak.
528 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
529 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
530 NB: ClamAV planning to remove STREAM in "middle of 2010".
531 CL also introduces -bmalware, various -d+acl logging additions and
532 more caution in buffer sizes.
534 PP/09 Implemented reverse_ip expansion operator.
536 PP/10 Bugzilla 937: provide a "debug" ACL control.
538 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
540 PP/12 Bugzilla 973: Implement --version.
542 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
544 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
546 PP/15 Bugzilla 816: support multiple condition rules on Routers.
548 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
549 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
550 ignore trailing whitespace.
552 JJ/02 prevent non-panic DKIM error from being sent to paniclog
554 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
557 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
558 Notification from Dr Andrew Aitchison.
560 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
561 ExtendedDetectionInfo response format.
562 Notification from John Horne.
564 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
567 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
568 XSL and documented dependency on system catalogs, with examples of how
571 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
574 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
575 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
576 configuration file which is writeable by the Exim user or group.
578 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
579 of configuration files to cover files specified with the -C option if
580 they are going to be used with root privileges, not just the default
583 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
584 option (effectively making it always true).
586 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
587 files to be used while preserving root privileges.
589 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
590 that rogue child processes cannot use them.
592 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
593 run-time user, instead of root.
595 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
596 Exim run-time user without dropping privileges.
598 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
599 result string, instead of calling string_vformat() twice with the same
602 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
603 for other users. Others should always drop root privileges if they use
604 -C on the command line, even for a whitelisted configure file.
606 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
608 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
614 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
615 $data_path, and $header_path variables; fixed documentation bugs and
618 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
619 exipick to access non-standard spools, including the "frozen" queue
622 NM/01 Bugzilla 965: Support mysql stored procedures.
623 Patch from Alain Williams
625 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
627 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
628 Patch from Andreas Metzler
630 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
631 Patch from Kirill Miazine
633 NM/05 Bugzilla 671: Added umask to procmail example.
635 JJ/03 installed exipick 20100323.0, fixing doc bug
637 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
638 directory. Notification and patch from Dan Rosenberg.
640 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
642 TK/02 Improve log output when DKIM signing operation fails.
644 MH/01 Treat the transport option dkim_domain as a colon separated
645 list, not as a single string, and sign the message with each element,
646 omitting multiple occurences of the same signer.
648 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
649 Bugzilla 985, 986. Patch by Simon Arlott
651 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
652 Patch by Simon Arlott
654 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
655 MBX locking. Notification from Dan Rosenberg.
661 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
663 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
665 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
667 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
669 NM/04 Bugzilla 915: Flag broken perl installation during build.
675 TK/01 Added patch by Johannes Berg that expands the main option
676 "spamd_address" if it starts with a dollar sign.
678 TK/02 Write list of recipients to X-Envelope-Sender header when building
679 the mbox-format spool file for content scanning (suggested by Jakob
682 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
683 (http://www.dcc-servers.net/) support via dccifd. Activated by
684 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
686 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
687 by Mark Daniel Reidel <mr@df.eu>.
689 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
690 When building exim an external PCRE library is now needed -
691 PCRE is a system library on the majority of modern systems.
692 See entry on PCRE_LIBS in EDITME file.
694 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
695 conversation. Added nologin parameter to request.
696 Patch contributed by Kirill Miazine.
698 TF/01 Do not log submission mode rewrites if they do not change the address.
700 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
702 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
703 log files in place. Contributed by Roberto Lima.
705 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
707 TF/03 Bugzilla 615: When checking the local_parts router precondition
708 after a local_part_suffix or local_part_prefix option, Exim now
709 does not use the address's named list lookup cache, since this
710 contains cached lookups for the whole local part.
712 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
713 Robert Millan. Documentation is in experimental-spec.txt.
715 TF/04 Bugzilla 668: Fix parallel build (make -j).
717 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
719 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
720 Patch provided by Jan Srzednicki.
722 TF/05 Leading white space used to be stripped from $spam_report which
723 wrecked the formatting. Now it is preserved.
725 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
726 that they are available at delivery time.
728 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
730 TF/08 TLS error reporting now respects the incoming_interface and
731 incoming_port log selectors.
733 TF/09 Produce a more useful error message if an SMTP transport's hosts
734 setting expands to an empty string.
736 NM/06 Bugzilla 744: EXPN did not work under TLS.
737 Patch provided by Phil Pennock.
739 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
740 Patch provided by Richard Godbee.
742 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
743 acl_smtp_notquit, added index entry.
745 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
746 Patch provided by Eugene Bujak.
748 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
749 accept(). Patch provided by Maxim Dounin.
751 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
752 Patch provided by Phil Pennock.
754 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
756 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
757 Patch provided by Brad "anomie" Jorsch.
759 NM/14 Bugzilla 622: Added timeout setting to transport filter.
760 Patch provided by Dean Brooks.
762 TK/05 Add native DKIM support (does not depend on external libraries).
764 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
765 Patch provided by Graeme Fowler.
767 NM/16 Bugzilla 851: Documentation example syntax fix.
769 NM/17 Changed NOTICE file to remove references to embedded PCRE.
771 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
774 NM/19 Bugzilla 745: TLS version reporting.
775 Patch provided by Phil Pennock.
777 NM/20 Bugzilla 167: bool: condition support.
778 Patch provided by Phil Pennock.
780 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
781 clients. Patch provided by Phil Pennock.
783 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
784 Patch provided by Brad "anomie" Jorsch.
786 NM/23 Bugzilla 687: Fix misparses in eximstats.
787 Patch provided by Heiko Schlittermann.
789 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
790 Patch provided by Heiko Schlittermann.
792 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
793 plus update to original patch.
795 NM/26 Bugzilla 799: Documentation correction for ratelimit.
797 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
798 Patch provided by David Brownlee.
800 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
802 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
804 NM/30 Bugzilla 888: TLS documentation bugfixes.
806 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
808 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
809 Unlike the original bugzilla I have changed all shell scripts in src tree.
811 NM/33 Bugzilla 898: Transport filter timeout fix.
812 Patch by Todd Rinaldo.
814 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
815 Patch by Serge Demonchaux.
817 NM/35 Bugzilla 39: Base64 decode bug fixes.
818 Patch by Jakob Hirsch.
820 NM/36 Bugzilla 909: Correct connect() call in dcc code.
822 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
824 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
826 NM/39 Bugzilla 911: Fixed MakeLinks build script.
832 TK/01 Add preliminary DKIM support. Currently requires a forked version of
833 ALT-N's libdkim that I have put here:
834 http://duncanthrax.net/exim-experimental/
836 Note to Michael Haardt: I had to rename some vars in sieve.c. They
837 were called 'true' and it seems that C99 defines that as a reserved
838 keyword to be used with 'bool' variable types. That means you could
839 not include C99-style headers which use bools without triggering
840 build errors in sieve.c.
842 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
843 as mailq or other aliases. Changed the --help handling significantly
844 to do whats expected. exim_usage() emits usage/help information.
846 SC/01 Added the -bylocaldomain option to eximstats.
848 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
850 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
852 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
858 PH/01 Another patch from the Sieve maintainer.
860 PH/02 When an IPv6 address is converted to a string for single-key lookup
861 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
862 dots are used instead of colons so that keys in lsearch files need not
863 contain colons. This was done some time before quoting was made available
864 in lsearch files. However, iplsearch files do require colons in IPv6 keys
865 (notated using the quote facility) so as to distinguish them from IPv4
866 keys. This meant that lookups for IP addresses in host lists did not work
867 for iplsearch lookups.
869 This has been fixed by arranging for IPv6 addresses to be expressed with
870 colons if the lookup type is iplsearch. This is not incompatible, because
871 previously such lookups could never work.
873 The situation is now rather anomolous, since one *can* have colons in
874 ordinary lsearch keys. However, making the change in all cases is
875 incompatible and would probably break a number of configurations.
877 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
880 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
881 conversion specification without a maximum field width, thereby enabling
882 a rogue spamd server to cause a buffer overflow. While nobody in their
883 right mind would setup Exim to query an untrusted spamd server, an
884 attacker that gains access to a server running spamd could potentially
885 exploit this vulnerability to run arbitrary code as the Exim user.
887 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
888 $primary_hostname instead of what libspf2 thinks the hosts name is.
890 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
891 a directory entry by the name of the lookup key. Previously, if a
892 symlink pointed to a non-existing file or a file in a directory that
893 Exim lacked permissions to read, a lookup for a key matching that
894 symlink would fail. Now it is enough that a matching directory entry
895 exists, symlink or not. (Bugzilla 503.)
897 PH/03 The body_linecount and body_zerocount variables are now exported in the
900 PH/04 Added the $dnslist_matched variable.
902 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
903 This means they are set thereafter only if the connection becomes
906 PH/06 Added the client_condition to authenticators so that some can be skipped
907 by clients under certain conditions.
909 PH/07 The error message for a badly-placed control=no_multiline_responses left
910 "_responses" off the end of the name.
912 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
914 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
915 (without spaces) instead of just copying the configuration text.
917 PH/10 Added the /noupdate option to the ratelimit ACL condition.
919 PH/11 Added $max_received_linelength.
921 PH/12 Added +ignore_defer and +include_defer to host lists.
923 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
924 way in which PCRE > 7.0 is built.
926 PH/14 Implemented queue_only_load_latch.
928 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
929 MAIL command. The effect was to mangle the value on 64-bit systems.
931 PH/16 Another patch from the Sieve maintainer.
933 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
935 PH/18 If a system quota error occurred while trying to create the file for
936 a maildir delivery, the message "Mailbox is full" was not appended to the
937 bounce if the delivery eventually timed out. Change 4.67/27 below applied
938 only to a quota excession during the actual writing of the file.
940 PH/19 It seems that peer DN values may contain newlines (and other non-printing
941 characters?) which causes problems in log lines. The DN values are now
942 passed through string_printing() before being added to log lines.
944 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
945 and InterBase are left for another time.)
947 PH/21 Added message_body_newlines option.
949 PH/22 Guard against possible overflow in moan_check_errorcopy().
951 PH/23 POSIX allows open() to be a macro; guard against that.
953 PH/24 If the recipient of an error message contained an @ in the local part
954 (suitably quoted, of course), incorrect values were put in $domain and
955 $local_part during the evaluation of errors_copy.
961 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
962 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
965 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
966 issue a MAIL command.
968 PH/02 In an ACL statement such as
970 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
972 if a client was not listed at all, or was listed with a value other than
973 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
974 the condition was not true (as it should be), so access was not denied.
975 The bug was that the ! inversion was incorrectly passed on to the second
976 item. This has been fixed.
978 PH/03 Added additional dnslists conditions == and =& which are different from
979 = and & when the dns lookup returns more than one IP address.
981 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
982 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
984 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
985 FSYNC, which compiles an option called disable_fsync that allows for
986 bypassing fsync(). The documentation is heavily laced with warnings.
988 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
990 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
991 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
992 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
993 including adding "make clean"; (3) Added -fPIC when compiling the test
994 dynamically loaded module, to get rid of a warning.
996 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
997 message fails, move_frozen_messages = true and ignore_bounce_errors_after
998 = 0s. The bug is otherwise harmless.
1000 PH/07 There was a bug in the dovecot authenticator such that the value of
1001 $auth1 could be overwritten, and so not correctly preserved, after a
1002 successful authentication. This usually meant that the value preserved by
1003 the server_setid option was incorrect.
1005 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1007 PH/09 Installed PCRE release 7.0.
1009 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1010 run for batched SMTP input. It is now run at the start of every message
1011 in the batch. While fixing this I discovered that the process information
1012 (output by running exiwhat) was not always getting set for -bs and -bS
1013 input. This is fixed, and it now also says "batched" for BSMTP.
1015 PH/11 Added control=no_pipelining.
1017 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1018 patch, slightly modified), and move the expansion of helo_data till after
1019 the connection is made in the smtp transport (so it can use these
1022 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1024 PH/14 Added log_selector = +pid.
1026 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1028 PH/16 Add ${if forany and ${if forall.
1030 PH/17 Added dsn_from option to vary the From: line in DSNs.
1032 PH/18 Flush SMTP output before performing a callout, unless control =
1033 no_callout_flush is set.
1035 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1036 was true (the default) a successful delivery failed to delete the retry
1037 item, thus causing premature timeout of the address. The bug is now
1040 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1042 PH/21 Long custom messages for fakedefer and fakereject are now split up
1043 into multiline reponses in the same way that messages for "deny" and
1044 other ACL rejections are.
1046 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1047 with slight modification.
1049 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1050 draft, changing the syntax and factoring some duplicate code".
1052 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1053 for deliveries of the second and subsequent messages over the same SMTP
1056 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1057 ${reduce, with only minor "tidies".
1059 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1061 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1062 expansion side effects.
1064 PH/27 When a message times out after an over-quota error from an Exim-imposed
1065 quota, the bounce message says "mailbox is full". This message was not
1066 being given when it was a system quota that was exceeded. It now should
1069 MH/03 Made $recipients available in local_scan(). local_scan() already has
1070 better access to the recipient list through recipients_list[], but
1071 $recipients can be useful in postmaster-provided expansion strings.
1073 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1074 in the case of a MAIL command with additional options following the
1075 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1076 were accidentally chopped off.
1078 PH/29 SMTP synchronization checks are implemented when a command is read -
1079 there is a check that no more input is waiting when there shouldn't be
1080 any. However, for some commands, a delay in an ACL can mean that it is
1081 some time before the response is written. In this time, more input might
1082 arrive, invalidly. So now there are extra checks after an ACL has run for
1083 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1084 pipelining has not been advertised.
1086 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1088 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1089 returned in the SMTP response when smtp_return_error_details was false.
1090 This has been fixed.
1092 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1093 instead of the C I/O that was originally supplied, because problems were
1094 reported on Solaris.
1096 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1097 Exim which did not show up earlier: it was assuming that a call to
1098 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1099 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1100 was a macro that became an assignment, so it seemed to work. This has
1101 changed to a proper function call with a void return, hence the compile
1102 error. Exim's code has been fixed.
1104 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1107 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1109 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1111 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1112 because it turns out to be too big or there is a timeout) there is no
1113 "Completed" line in the log. When some messages of this type were
1114 selected by exigrep, they were listed as "not completed". Others were
1115 picked up by some special patterns. I have improved the selection
1116 criteria to be more general.
1118 PH/38 The host_find_failed option in the manualroute router can now be set
1119 to "ignore", to completely ignore a host whose IP address cannot be
1120 found. If all hosts are ignored, the behaviour is controlled by the new
1121 host_all_ignored option.
1123 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1124 homing or because of multiple MX records with /mx) generated more than
1125 one IP address, and the following item turned out to be the local host,
1126 all the secondary addresses of the first item were incorrectly removed
1127 from the list, along with the local host and any following hosts (which
1128 is what is supposed to happen).
1130 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1131 whoever called Exim into the -H file. In the case of the daemon it was
1132 behaving confusingly. When first started, it used values for whoever
1133 started the daemon, but after a SIGHUP it used the Exim user (because it
1134 calls itself on a restart). I have changed the code so that it now always
1137 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1138 message are rejected with the same error (e.g. no authentication or bad
1139 sender address), and a DATA command is nevertheless sent (as can happen
1140 with PIPELINING or a stupid MUA), the error message that was given to the
1141 RCPT commands is included in the rejection of the DATA command. This is
1142 intended to be helpful for MUAs that show only the final error to their
1145 PH/42 Another patch from the Sieve maintainer.
1147 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1148 Eximstats - Fixed some broken HTML links and added missing column headers
1150 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1151 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1153 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1155 PH/43 Yet another patch from the Sieve maintainer.
1157 PH/44 I found a way to check for a TCP/IP connection going away before sending
1158 the response to the final '.' that terminates a message, but only in the
1159 case where the client has not sent further data following the '.'
1160 (unfortunately, this is allowed). However, in many cases there won't be
1161 any further data because there won't be any more messages to send. A call
1162 to select() can be used: if it shows that the input is "ready", there is
1163 either input waiting, or the socket has been closed. An attempt to read
1164 the next input character can distinguish the two cases. Previously, Exim
1165 would have sent an OK response which the client would never have see.
1166 This could lead to message repetition. This fix should cure that, at
1167 least in a lot of common cases.
1169 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1170 advertised in response to EHLO.
1176 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1177 fixed by 4.65/MH/01 (is this a record?) are fixed:
1179 (i) An empty string was always treated as zero by the numeric comparison
1180 operators. This behaviour has been restored.
1182 (ii) It is documented that the numeric comparison operators always treat
1183 their arguments as decimal numbers. This was broken in that numbers
1184 starting with 0 were being interpreted as octal.
1186 While fixing these problems I realized that there was another issue that
1187 hadn't been noticed. Values of message_size_limit (both the global option
1188 and the transport option) were treated as octal if they started with 0.
1189 The documentation was vague. These values are now always treated as
1190 decimal, and I will make that clear in the documentation.
1196 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1197 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1200 MH/01 Don't check that the operands of numeric comparison operators are
1201 integers when their expansion is in "skipping" mode (fixes bug
1202 introduced by 4.64-PH/07).
1204 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1205 child addresses, Exim now panics and dies. Previously, because the count
1206 is held in a short int, deliveries were likely to be lost. As such a
1207 large number of recipients for a single message is ridiculous
1208 (performance will be very, very poor), I have chosen to impose a limit
1209 rather than extend the field.
1215 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1216 leftover -K file (the existence of which was triggered by #402).
1217 While we were at it, introduced process PID as part of the -K
1218 filename. This should rule out race conditions when creating
1221 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1222 processing considerably. Previous code took too long for large mails,
1223 triggering a timeout which in turn triggers #401.
1225 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1226 in the DK code in transports.c. sendfile() is not really portable,
1227 hence the _LINUX specificness.
1229 TF/01 In the add_headers option to the mail command in an Exim filter,
1230 there was a bug that Exim would claim a syntax error in any
1231 header after the first one which had an odd number of characters
1234 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1235 callout verification, Exim cached a "reject" for the entire domain. This
1236 is correct for most verifications, but it is not correct for a recipient
1237 verification with use_sender or use_postmaster set, because in that case
1238 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1239 case of MAIL FROM:<> rejection from other early rejections (e.g.
1240 rejection of HELO). When verifying a recipient using a non-null MAIL
1241 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1242 Whatever the result of the callout, the value of the domain cache is
1243 left unchanged (for any other kind of callout, getting as far as trying
1244 RCPT means that the domain itself is ok).
1246 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1249 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1250 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1251 session, the daemon ignores ECONNECT errors and logs others; it now
1252 ignores EPIPE as well.
1254 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1255 (quoted-printable decoding).
1257 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1258 later the small subsequent patch to fix an introduced bug.
1260 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1262 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1264 PH/08 An error is now given if message_size_limit is specified negative.
1266 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1267 to be given (somewhat) arbitrary names.
1269 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1272 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1273 miscellaneous code fixes
1275 PH/10 Added the log_reject_target ACL modifier to specify where to log
1278 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1279 hostname. This is wrong, because it relates to the incoming message (and
1280 probably the interface on which it is arriving) and not to the outgoing
1281 callout (which could be using a different interface). This has been
1282 changed to use the value of the helo_data option from the smtp transport
1283 instead - this is what is used when a message is actually being sent. If
1284 there is no remote transport (possible with a router that sets up host
1285 addresses), $smtp_active_hostname is used.
1287 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1288 tweaks were necessary in order to get it to work (see also 21 below):
1289 (a) The code assumed that strncpy() returns a negative number on buffer
1290 overflow, which isn't the case. Replaced with Exim's string_format()
1292 (b) There were several signed/unsigned issues. I just did the minimum
1293 hacking in of casts. There is scope for a larger refactoring.
1294 (c) The code used strcasecmp() which is not a standard C function.
1295 Replaced with Exim's strcmpic() function.
1296 (d) The code set only $1; it now sets $auth1 as well.
1297 (e) A simple test gave the error "authentication client didn't specify
1298 service in request". It would seem that Dovecot has changed its
1299 interface. Fortunately there's a specification; I followed it and
1300 changed what the client sends and it appears to be working now.
1302 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1305 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1306 address A is aliased to B and C, where B exists and C does not. Without
1307 -v the output is "A verified" because verification stops after a
1308 successful redirection if more than one address is generated. However,
1309 with -v the child addresses are also verified. Exim was outputting "A
1310 failed to verify" and then showing the successful verification for C,
1311 with its parentage. It now outputs "B failed to verify", showing B's
1312 parentage before showing the successful verification of C.
1314 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1315 look up a TXT record in a specific list after matching in a combined
1318 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1319 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1320 they consult the DNS. I had assumed they would set it the way they
1321 wanted; and indeed my experiments on Linux seem to show that in some
1322 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1323 To be on the safe side, however, I have now made the interface to
1324 host_find_byname() similar to host_find_bydns(), with an argument
1325 containing the DNS resolver options. The host_find_byname() function now
1326 sets these options at its start, just as host_find_bydns() does. The smtp
1327 transport options dns_qualify_single and dns_search_parents are passed to
1328 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1329 of host_find_byname() use the default settings of RES_DEFNAMES
1330 (qualify_single) but not RES_DNSRCH (search_parents).
1332 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1333 spool_read_header() do less string testing, by means of a preliminary
1334 switch on the second character of optional "-foo" lines. (This is
1335 overdue, caused by the large number of possibilities that now exist.
1336 Originally there were few.) While I was there, I also converted the
1337 str(n)cmp tests so they don't re-test the leading "-" and the first
1338 character, in the hope this might squeeze out yet more improvement.
1340 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1341 flag allowing group syntax was set by the header_syntax check but not
1342 turned off, possible causing trouble later; (2) The flag was not being
1343 set at all for the header_verify test, causing "group"-style headers to
1344 be rejected. I have now set it in this case, and also caused header_
1345 verify to ignore an empty address taken from a group. While doing this, I
1346 came across some other cases where the code for allowing group syntax
1347 while scanning a header line wasn't quite right (mostly, not resetting
1348 the flag correctly in the right place). These bugs could have caused
1349 trouble for malformed header lines. I hope it is now all correct.
1351 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1352 with the "reply" argument non-NULL. The code, however (which originally
1353 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1354 but it didn't always do it. This confused somebody who was copying the
1355 code for some other use. I have removed all the tests.
1357 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1358 feature that was used to support insecure browsers during the U.S. crypto
1359 embargo. It requires special client support, and Exim is probably the
1360 only MTA that supported it -- and would never use it because real RSA is
1361 always available. This code has been removed, because it had the bad
1362 effect of slowing Exim down by computing (never used) parameters for the
1363 RSA_EXPORT functionality.
1365 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1366 authenticator to fail if there's a tab character in the incoming data
1367 (there should never be unless someone is messing about, as it's supposed
1368 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1369 if the connection is using TLS or if the remote IP is the same as the
1370 local IP, and the "valid-client-cert option" if a client certificate has
1373 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1374 authenticators. This can be used for authorization after authentication
1375 succeeds. (In the case of plaintext, it servers for both authentication
1378 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1379 if any retry times were supplied.
1381 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1382 connection, where there is no client IP address to check. In this
1383 situation, the verify now always succeeds.
1385 PH/25 Applied John Jetmore's -Mset patch.
1387 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1389 PH/27 In a string expansion for a processed (not raw) header when multiple
1390 headers of the same name were present, leading whitespace was being
1391 removed from all of them, but trailing whitespace was being removed only
1392 from the last one. Now trailing whitespace is removed from each header
1393 before concatenation. Completely empty headers in a concatenation (as
1394 before) are ignored.
1396 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1397 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1399 PH/29 [Removed. This was a change that I later backed out, and forgot to
1400 correct the ChangeLog entry (that I had efficiently created) before
1401 committing the later change.]
1403 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1404 address errors (4xx response to RCPT) over the same connection as other
1405 messages routed to the same hosts. Such deliveries are always "forced",
1406 so retry times are not inspected. This resulted in far too many retries
1407 for the affected addresses. The effect occurred only when there were more
1408 hosts than the hosts_max_try setting in the smtp transport when it had
1409 the 4xx errors. Those hosts that it had tried were not added to the list
1410 of hosts for which the message was waiting, so if all were tried, there
1411 was no problem. Two fixes have been applied:
1413 (i) If there are any address or message errors in an SMTP delivery, none
1414 of the hosts (tried or untried) are now added to the list of hosts
1415 for which the message is waiting, so the message should not be a
1416 candidate for sending over the same connection that was used for a
1417 successful delivery of some other message. This seems entirely
1418 reasonable: after all the message is NOT "waiting for some host".
1419 This is so "obvious" that I'm not sure why it wasn't done
1420 previously. Hope I haven't missed anything, but it can't do any
1421 harm, as the worst effect is to miss an optimization.
1423 (ii) If, despite (i), such a delivery is accidentally attempted, the
1424 routing retry time is respected, so at least it doesn't keep
1425 hammering the server.
1427 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1428 in ${readsocket because some servers need this prod.
1430 PH/32 Added some extra debug output when updating a wait-xxx database.
1432 PH/33 The hint "could be header name not terminated by colon", which has been
1433 given for certain expansion errors for a long time, was not being given
1434 for the ${if def:h_colon_omitted{... case.
1436 PH/34 The spec says: "With one important exception, whenever a domain list is
1437 being scanned, $domain contains the subject domain." There was at least
1438 one case where this was not true.
1440 PH/35 The error "getsockname() failed: connection reset by peer" was being
1441 written to the panic log as well as the main log, but it isn't really
1442 panic-worthy as it just means the connection died rather early on. I have
1443 removed the panic log writing for the ECONNRESET error when getsockname()
1446 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1447 runs only) independently of the message's sender address. This meant
1448 that, if the 4xx error was in fact related to the sender, a different
1449 message to the same recipient with a different sender could confuse
1450 things. In particualar, this can happen when sending to a greylisting
1451 server, but other circumstances could also provoke similar problems.
1452 I have changed the default so that the retry time for these errors is now
1453 based a combination of the sender and recipient addresses. This change
1454 can be overridden by setting address_retry_include_sender=false in the
1457 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1458 remote server are returned as part of bounce messages. This was not
1459 happening for LMTP over a pipe (the lmtp transport), but now it is the
1460 same for both kinds of LMTP.
1462 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1463 in header lines that were in fact CNAMEs. This is no longer the case.
1465 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1466 and queue runs started by the daemon processed all messages. This has
1467 been fixed so that -R and -S can now usefully be given with -q<time>.
1469 PH/40 Import PCRE release 6.7 (fixes some bugs).
1471 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1473 PH/42 Give an error if -q is specified more than once.
1475 PH/43 Renamed the variables $interface_address and $interface_port as
1476 $received_ip_address and $received_port, to make it clear that these
1477 values apply to message reception, and not to the outgoing interface when
1478 a message is delivered. (The old names remain recognized, of course.)
1480 PH/44 There was no timeout on the connect() call when using a Unix domain
1481 socket in the ${readsocket expansion. There now is.
1483 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1484 be meaningful with "accept".
1486 SC/01 Eximstats V1.43
1487 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1489 SC/02 Eximstats V1.44
1490 Use a glob alias rather than an array ref in the generated
1491 parser. This improves both readability and performance.
1493 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1494 Collect SpamAssassin and rejection statistics.
1495 Don't display local sender or destination tables unless
1496 there is data to show.
1497 Added average volumes into the top table text output.
1499 SC/04 Eximstats V1.46
1500 Collect data on the number of addresses (recipients)
1501 as well as the number of messages.
1503 SC/05 Eximstats V1.47
1504 Added 'Message too big' to the list of mail rejection
1505 reasons (thanks to Marco Gaiarin).
1507 SC/06 Eximstats V1.48
1508 Mainlog lines which have GMT offsets and are too short to
1509 have a flag are now skipped.
1511 SC/07 Eximstats V1.49 (Alain Williams)
1512 Added the -emptyok flag.
1514 SC/08 Eximstats V1.50
1515 Fixes for obtaining the IP address from reject messages.
1517 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1518 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1519 whitesspace changes from 4.64-PH/27
1521 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1524 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1525 are found, allow negative numbers in numeric criteria)
1527 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1529 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1530 to match changes made in 4.64-PH/43
1532 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1534 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1535 "domain" is actually an IP address, and force a failure. This locks out
1536 those revolvers/nameservers that support "A-for-A" lookups, in
1537 contravention of the specifications.
1539 PH/48 When a host name was looked up from an IP address, and the subsequent
1540 forward lookup of the name timed out, the host name was left in
1541 $sender_host_name, contrary to the specification.
1543 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1544 restricted to single-key lookups, Exim was not diagnosing an error if
1545 * or *@ was used with a query-style lookup.
1547 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1549 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1550 long ago, but noone interested enough thought of it. Let's just say that
1551 the "1.1" means that there are some new functions that weren't there at
1552 some point in the past.
1554 PH/51 Error processing for expansion failure of helo_data from an smtp
1555 transport during callout processing was broken.
1557 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1558 tested/used via the -bh/-bhc/-bs options.
1560 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1561 bug, fixed in subsequent PCRE releases).
1563 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1564 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1566 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1572 SC/01 Use a glob alias rather than an array ref in eximstats generated
1573 parser. This improves both readability and performance.
1575 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1576 Don't display local sender or destination tables in eximstats unless
1577 there is data to show.
1578 Added average volumes into the eximstats top table text output.
1580 SC/03 Collect data on the number of addresses (recipients) as well
1581 as the number of messages in eximstats.
1583 TF/01 Correct an error in the documentation for the redirect router. Exim
1584 does (usually) call initgroups() when daemonizing.
1586 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1587 with consistent privilege compared to when running as a daemon.
1589 TF/03 Note in the spec that $authenticated_id is not set for local
1590 submissions from trusted users.
1592 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1593 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1595 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1596 by adding some example configuration directives to the default
1597 configuration file. A little bit of work is required to uncomment the
1598 directives and define how usernames and passwords are checked, but
1599 there is now a framework to start from.
1601 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1602 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1603 without this. I don't know how relevant this is to other LDAP libraries.
1605 PH/02 Add the verb name to the "unknown ACL verb" error.
1607 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1609 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1611 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1612 directory not expanded when it should be if an expanded home directory
1613 was set for the address (which is overridden by the transport).
1615 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1618 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1619 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1620 because it is too late at that time, and has no effect.
1622 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1623 security issue with \' (bugzilla #107). I could not use the
1624 PQescapeStringConn() function, because it needs a PGconn value as one of
1627 PH/08 When testing addresses using -bt, indicate those final addresses that
1628 are duplicates that would not cause an additional delivery. At least one
1629 person was confused, thinking that -bt output corresponded to deliveries.
1630 (Suppressing duplicates isn't a good idea as you lose the information
1631 about possibly different redirections that led to the duplicates.)
1633 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1634 systems where poll() doesn't work, in particular OS X.
1636 PH/10 Added more information to debugging output for retry time not reached.
1638 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1639 operations in malware.c.
1641 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1644 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1645 syslog_duplication set false, log lines that would normally be written
1646 both the the main log and to the reject log were not written to syslog at
1649 PH/14 In the default configuration, change the use of "message" in ACL warn
1650 statements to "add_header".
1652 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1653 not followed by a command (e.g. "seen endif").
1655 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1656 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1659 PH/17 Added extra conditions to the default value of delay_warning_condition
1663 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1664 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1665 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1668 The Auto-Submitted: and various List- headers are standardised, whereas I
1669 don't think Precedence: ever was.
1671 PH/18 Refactored debugging code in route_finduser() to show more information,
1672 in particular, the error code if getpwnam() issues one.
1674 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1675 This is apparently needed in addition to the PH/07 change above to avoid
1676 any possible encoding problems.
1678 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1679 but not after initializing Perl.
1681 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1682 output them only if debugging. By default they are written stderr,
1683 apparently, which is not desirable.
1685 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1688 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1691 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1693 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1694 authenticated or an ident call has been made. Suppress the default
1695 values for $authenticated_id and $authenticated_sender (but permit -oMai
1696 and -oMas) when testing with -bh.
1698 PH/24 Re-jigged the order of the tests in the default configuration so that the
1699 tests for valid domains and recipients precede the DNS black list and CSA
1700 tests, on the grounds that those ones are more expensive.
1702 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1703 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1704 This bug exists in every version of Exim that I still have, right back to
1707 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1708 However, an attempt to turn on case-sensitivity in a regex key by
1709 including (?-i) didn't work because the subject string was already
1710 lowercased, and the effects were non-intuitive. It turns out that a
1711 one-line patch can be used to allow (?-i) to work as expected.
1717 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1718 other effects) broke the use of negated acl sub-conditions.
1720 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1723 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1724 "Deny" causes Exim to reject the incoming connection with a 554 error.
1725 Unfortunately, if there is a major crisis, such as a disk failure,
1726 tcp-wrappers gives "deny", whereas what one would like would be some
1727 kind of temporary error. A kludge has been added to help with this.
1728 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1729 554 error is used if errno is still zero or contains ENOENT (which occurs
1730 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1733 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1735 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1736 errors. Otherwise a message that provokes a temporary error (when other
1737 messages do not) can cause a whole host to time out.
1739 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1740 addresses were routed directly to files or pipes from a redirect router.
1741 File deliveries just didn't batch; pipe deliveries might have suffered
1744 PH/06 A failure to get a lock for a hints database would erroneously always say
1745 "Failed to get write lock", even when it was really a read lock.
1747 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1748 of 0600. This has been changed to use the value of the lockfile_mode
1749 option (which defaults to 0600).
1751 PH/08 Applied small patch from the Sieve maintainer.
1753 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1754 folder from quota calculations, a direct delivery into this folder messed
1755 up the contents of the maildirsize file. This was because the regex was
1756 used only to exclude .Trash (or whatever) when the size of the mailbox
1757 was calculated. There was no check that a delivery was happening into an
1758 excluded directory. This bug has been fixed by ignoring all quota
1759 processing for deliveries into excluded directories.
1761 PH/10 Added the maildirfolder_create_regex option to appendfile.
1767 PH/01 The code for finding all the local interface addresses on a FreeBSD
1768 system running IPv6 was broken. This may well have applied to all BSD
1769 systems, as well as to others that have similar system calls. The broken
1770 code found IPv4 interfaces correctly, but gave incorrect values for the
1771 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1772 that it would not match correctly against @[] and not recognize the IPv6
1775 PH/02 The ipliteral router was not recognizing addresses of the form user@
1776 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1778 PH/03 Added disable_ipv6.
1780 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1781 decoded form, because it is most often used to construct To: headers
1782 lines in autoreplies, and the decoded form may well be syntactically
1783 invalid. However, $reply_address has leading white space removed, and all
1784 newlines turned into spaces so that the autoreply transport does not
1787 PH/05 If group was specified without a user on a router, and no group or user
1788 was specified on a transport, the group from the router was ignored.
1790 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1791 for visible compile-time settings that can be used to change these
1792 numbers, for those that want even more. Backwards compatibility with old
1793 spool files has been maintained. However, going back to a previous Exim
1794 release will lost any variables that are in spool files.
1796 PH/07 Two small changes when running in the test harness: increase delay when
1797 passing a TCP/IP connection to a new process, in case the original
1798 process has to generate a bounce, and remove special handling of
1799 127.0.0.2 (sic), which is no longer necessary.
1801 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1802 be the same on different OS.
1804 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1807 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1808 whether --show-vars was specified or not
1810 JJ/02 exipick: Added support for new ACL variable spool format introduced
1813 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1814 syntactically invalid From: or Reply-to: line, and a filter used this to
1815 generate an autoreply, and therefore failed to obtain an address for the
1816 autoreply, Exim could try to deliver to a non-existent relative file
1817 name, causing unrelated and misleading errors. What now happens is that
1818 it logs this as a hard delivery error, but does not attempt to create a
1821 PH/11 The exinext utility has a -C option for testing purposes, but although
1822 the given file was scanned by exinext itself; it wasn't being passed on
1823 when Exim was called.
1825 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1826 an end-of-file indication when reading a command response.
1828 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1829 compiled. In many other places in Exim, IPv6 addresses are always
1830 recognized, so I have changed this. It also means that IPv4 domain
1831 literals of the form [IPV4:n.n.n.n] are now always recognized.
1833 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1834 used if the router is not running as root, for example, when verifying at
1835 ACL time, or when using -bh. The debugging output from this situation was
1836 non-existent - all you got was a failure to exec. I have made two
1839 (a) Failures to set uid/gid, the current directory, or a process leader
1840 in a subprocess such as that created by queryprogram now generate
1841 suitable debugging ouput when -d is set.
1843 (b) The queryprogram router detects when it is not running as root,
1844 outputs suitable debugging information if -d is set, and then runs
1845 the subprocess without attempting to change uid/gid.
1847 PH/15 Minor change to Makefile for building test_host (undocumented testing
1850 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1851 additional section of a DNS packet that returns MX or SRV records.
1852 Instead, it always explicitly searches for A/AAAA records. This avoids
1853 major problems that occur when a DNS server includes only records of one
1854 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1855 fixed another bug: if SRV records were looked up and the corresponding
1856 address records were *not* found in the additional section, the port
1857 values from the SRV records were lost.
1859 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1860 using the correct key (the original address) when searching the retry
1861 rules in order to find which one to use for generating the retry hint.
1863 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1864 adding the default one. Similarly, if it contains a Reply-To: header, the
1865 errors_reply_to option, if set, is not used.
1867 PH/19 When calculating a retry time, Exim used to measure the "time since
1868 failure" by looking at the "first failed" field in the retry record. Now
1869 it does not use this if it is later than than the arrival time of the
1870 message. Instead it uses the arrival time. This makes for better
1871 behaviour in cases where some deliveries succeed, thus re-setting the
1872 "first failed" field. An example is a quota failure for a huge message
1873 when small messages continue to be delivered. Without this change, the
1874 "time since failure" will always be short, possible causing more frequent
1875 delivery attempts for the huge message than are intended.
1876 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1878 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1879 $1, $2, $3) because the numerical variables can be reset during some
1880 expansion items (e.g. "match"), thereby losing the authentication data.
1882 PH/21 Make -bV show the size of off_t variables so that the test suite can
1883 decide whether to run tests for quotas > 2G.
1885 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1886 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1887 filecount value is greater than 2G or if a quota value is greater than 2G
1888 on a system where the size of off_t is not greater than 4, a panic error
1891 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1892 never match. The debug and -bh output now contains an explicit error
1893 message indicating a malformed IPv4 address or mask.
1895 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1896 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1897 PH/23 above applies.
1899 PH/25 Do not write to syslog when running in the test harness. The only
1900 occasion when this arises is a failure to open the main or panic logs
1901 (for which there is an explicit test).
1903 PH/26 Added the /no_tell option to "control=freeze".
1905 PH/27 If a host name lookup failed very early in a connection, for example, if
1906 the IP address matched host_lookup and the reverse lookup yielded a name
1907 that did not have a forward lookup, an error message of the form "no IP
1908 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1909 could be logged. Now it outputs the IP address instead of "NULL".
1911 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1912 allows the sender and the authenticated sender to be set when
1913 submitting a message from within Exim. Since child_open_exim() is
1914 documented for local_scan(), the new function should be too.
1916 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1917 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1918 results in an empty string is now treated as unset.
1920 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1922 PH/31 Added sender_verify_fail logging option.
1924 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1925 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1926 tidied the source and removed it altogether.
1928 PH/33 When a queue run was abandoned because the load average was too high, a
1929 log line was always written; now it is written only if the queue_run log
1930 selector is set. In addition, the log line for abandonment now contains
1931 information about the queue run such as the pid. This is always present
1932 in "start" and "stop" lines but was omitted from the "abandon" line.
1934 PH/34 Omit spaces between a header name and the colon in the error message that
1935 is given when verify = headers_syntax fails (if there are lots of them,
1936 the message gets confusing).
1938 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1939 names, as there are now some PTR records that contain slashes. This check
1940 is only to protect against broken name servers that fall over on strange
1941 characters, so the fact that it applies to all lookups doesn't matter.
1943 PH/36 Now that the new test suite is complete, we can remove some of the
1944 special code in Exim that was needed for the old test suite. For example,
1945 sorting DNS records because real resolvers return them in an arbitrary
1946 order. The new test suite's fake resolver always returns records in the
1949 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1950 bounces) except when queue_only is set, to avoid logging races between
1951 the different processes.
1953 PH/38 Panic-die if .include specifies a non-absolute path.
1955 PH/39 A tweak to the "H" retry rule from its user.
1957 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1958 a label. They prevented compilation on older perls.
1960 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1961 a warning to be raised on newish perls.
1963 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1964 on queue. Changes to match documented behaviour of showing count of
1965 messages matching specified criteria.
1967 PH/40 Changed the default ident timeout from 30s to 5s.
1969 PH/41 Added support for the use of login_cap features, on those BSD systems
1970 that have them, for controlling the resources used by pipe deliveries.
1972 PH/42 The content-scanning code uses fopen() to create files in which to put
1973 message data. Previously it was not paying any attention to the mode of
1974 the files. Exim runs with umask(0) because the rest of the code creates
1975 files with open(), and sets the required mode explicitly. Thus, these
1976 files were ending up world-writeable. This was not a big issue, because,
1977 being within the spool directory, they were not world-accessible. I have
1978 created a function called modefopen, which takes an additional mode
1979 argument. It sets umask(777), creates the file, chmods it to the required
1980 mode, then resets the umask. All the relevant calls to fopen() in the
1981 content scanning code have been changed to use this function.
1983 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1984 to 24 hours. This avoids potential overflow problems when processing G
1985 and H retry rules. I suspect nobody ever tinkers with this value.
1987 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1989 PH/45 When the plaintext authenticator is running as a client, the server's
1990 challenges are checked to ensure they are valid base64 strings. By
1991 default, the authentication attempt is cancelled if an invalid string is
1992 received. Setting client_ignore_invalid_base64 true ignores these errors.
1993 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1994 they are received. Thus, the responses can be made to depend on the
1995 challenges. If an invalid string is ignored, an empty string is placed in
1998 PH/46 Messages that are created by the autoreply transport now contains a
1999 References: header, in accordance with RFCs 2822 and 3834.
2001 PH/47 Added authenticated_sender_force to the smtp transport.
2003 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2005 PH/49 Installed latest patch from the Sieve maintainer.
2007 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2008 was also set, the appendfile transport was unnecessarily scanning a
2009 directory of message files (e.g. for maildir delivery) to find the count
2010 of files (along with the size), even though it did not need this
2011 information. It now does the scan only if it needs to find either the
2012 size of the count of files.
2014 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2016 PH/52 Two bugs concerned with error handling when the smtp transport is
2019 (i) Exim was not creating retry information for temporary errors given
2020 for individual recipients after the DATA command when the smtp transport
2021 was used in LMTP mode. This meant that they could be retried too
2022 frequently, and not timed out correctly.
2024 (ii) Exim was setting the flag that allows error details to be returned
2025 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2026 recipients that were returned after the DATA command.
2028 PH/53 This is related to PH/52, but is more general: for any failing address,
2029 when detailed error information was permitted to be returned to the
2030 sender, but the error was temporary, then after the final timeout, only
2031 "retry timeout exceeded" was returned. Now it returns the full error as
2032 well as "retry timeout exceeded".
2034 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2035 do this, and (what is worse) MTAs that accept it.
2037 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2038 will now be deprecated.
2040 PH/56 New os.c-cygwin from the Cygwin maintainer.
2042 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2043 formats (previously only available in exim formats via -bpr, -bpru,
2044 and -bpra. Now also available in native and exiqgrep formats)
2046 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2047 with very large, slow to parse queues
2049 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2051 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2053 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2054 responses to authentication challenges, though it was showing the
2055 challenges; (ii) I've removed the CR characters from the debug output for
2058 PH/58 Allow for the insertion of a newline as well as a space when a string
2059 is turned into more than one encoded-word during RFC 2047 encoding. The
2060 Sieve code now uses this.
2062 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2063 data_4xx, lost_connection, tls_required.
2065 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2066 message was being sent as an SMTP response.
2068 PH/61 Add -l and -k options to exicyclog.
2070 PH/62 When verifying, if an address was redirected to one new address, so that
2071 verification continued, and the new address failed or deferred after
2072 having set something in $address_data, the value of $address_data was not
2073 passed back to the ACL. This was different to the case when no
2074 redirection occurred. The value is now passed back in both cases.
2076 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2077 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2078 use login_cap.h, so on its own it isn't the distinguishing feature. The
2079 new name refers directly to the setclassresources() function.
2081 PH/65 Added configuration files for NetBSD3.
2083 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2085 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2086 is preferred over IPv4.
2088 PH/68 The bounce_return_message and bounce_return_body options were not being
2089 honoured for bounces generated during the reception of non-SMTP messages.
2090 In particular, this applied to messages rejected by the ACL. This bug has
2091 been fixed. However, if bounce_return_message is true and bounce_return_
2092 body is false, the headers that are returned for a non-SMTP message
2093 include only those that have been read before the error was detected.
2094 (In the case of an ACL rejection, they have all been read.)
2096 PH/69 The HTML version of the specification is now built in a directory called
2097 spec_html instead of spec.html, because the latter looks like a path with
2098 a MIME-type, and this confuses some software.
2100 PH/70 Catch two compiler warnings in sieve.c.
2102 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2103 function verify_get_ident() calls ip_connect() to connect a socket, but
2104 if the "connect()" function timed out, ip_connect() used to close the
2105 socket. However, verify_get_ident() also closes the socket later, and in
2106 between Exim writes to the log, which may get opened at this point. When
2107 the socket was closed in ip_connect(), the log could get the same file
2108 descriptor number as the socket. This naturally causes chaos. The fix is
2109 not to close the socket in ip_connect(); the socket should be closed by
2110 the function that creates it. There was only one place in the code where
2111 this was missing, in the iplookup router, which I don't think anybody now
2112 uses, but I've fixed it anyway.
2114 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2115 well as to direct DNS lookups. Otherwise the handling of names in host
2116 lists is inconsistent and therefore confusing.
2122 PH/01 Two changes to the default runtime configuration:
2124 (1) Move the checks for relay_from_hosts and authenticated clients from
2125 after to before the (commented out) DNS black list checks.
2127 (2) Add control=submission to the relay_from_hosts and authenticated
2128 clients checks, on the grounds that messages accepted by these
2129 statements are most likely to be submissions.
2131 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2133 (1) Generate an error if the third argument for the ${prvs expansion is
2136 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2139 (3) Reset the variables that are obtained from the first argument of
2140 ${prvscheck and used in the second argument before leaving the code,
2141 because their memory is reclaimed, so using them afterwards may do
2144 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2145 one (it's much easier than Tom thought :-).
2147 (5) Because of (4), we can now allow for the use of $prvscheck_result
2148 inside the third argument.
2150 PH/03 For some reason, the default setting of PATH when running a command from
2151 a pipe transport was just "/usr/bin". I have changed it to
2154 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2155 anything to be listed in the output from -bV.
2157 PH/05 When a filter generated an autoreply, the entire To: header line was
2158 quoted in the delivery log line, like this:
2160 => >A.N.Other <ano@some.domain> <original@ddress> ...
2162 This has been changed so that it extracts the operative address. There
2163 may be more than one such address. If so, they are comma-separated, like
2166 => >ano@some.domain,ona@other.domain <original@ddress> ...
2168 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2169 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2170 not being looked up in the rDNS to get a host name, Exim was showing the
2171 IP address twice in Received: lines, even though the IP addresses were
2172 identical. For example:
2174 Received: from [1.2.3.4] (helo=[1.2.3.4])
2176 However, if the real host name was known, it was omitting the HELO data
2177 if it matched the actual IP address. This has been tidied up so that it
2178 doesn't show the same IP address twice.
2180 PH/07 When both +timestamp and +memory debugging was on, the value given by
2181 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2182 called by the debug printing, thereby overwriting the timestamp buffer.
2183 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2185 PH/08 When the original message was included in an autoreply transport, it
2186 always said "this is a copy of the message, including all the headers",
2187 even if body_only or headers_only was set. It now gives an appropriate
2190 PH/09 Applied a patch from the Sieve maintainer which:
2192 o fixes some comments
2193 o adds the (disabled) notify extension core
2194 o adds some debug output for the result of if/elsif tests
2195 o points to the current vacation draft in the documentation
2196 and documents the missing references header update
2200 o fixes a bug in processing the envelope test (when testing
2201 multiple envelope elements, the last element determinted the
2204 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2205 Electronic Mail") by including:
2207 Auto-submitted: auto-generated
2209 in the messages that it generates (bounce messages and others, such as
2210 warnings). In the case of bounce messages for non-SMTP mesages, there was
2211 also a typo: it was using "Auto_submitted" (underscore instead of
2212 hyphen). Since every message generated by Exim is necessarily in response
2213 to another message, thes have all been changed to:
2215 Auto-Submitted: auto-replied
2217 in accordance with these statements in the RFC:
2219 The auto-replied keyword:
2221 - SHOULD be used on messages sent in direct response to another
2222 message by an automatic process,
2224 - MUST NOT be used on manually-generated messages,
2226 - MAY be used on Delivery Status Notifications (DSNs) and Message
2227 Disposition Notifications (MDNs),
2229 - MUST NOT be used on messages generated by automatic or periodic
2230 processes, except for messages which are automatic responses to
2233 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2234 to the default Received: header definition.
2236 PH/12 Added log selector acl_warn_skipped (default on).
2238 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2239 variables because (a) they are in the wrong storage pool and (b) even if
2240 they were copied, it wouldn't work properly because of the caching.
2242 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2243 checking when decoding. Apparently there are clients that generate
2244 overlong encoded strings. Why am I not surprised?
2246 PH/15 If the first argument of "${if match_address" was not empty, but did not
2247 contain an "@" character, Exim crashed. Now it writes a panic log message
2248 and treats the condition as false.
2250 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2252 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2253 extension "envelope-auth". The code is finished and in agreement with
2254 other implementations, but there is no documentation so far and in fact,
2255 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2256 not changing the active code.
2258 Print executed "if" and "elsif" statements when debugging is used. This
2259 helps a great deal to understand what a filter does.
2261 Document more things not specified clearly in RFC3028. I had all this
2262 sorted out, when out of a sudden new issues came to my mind. Oops."
2264 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2267 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2268 canonical form (as documented). However, after a host name lookup from
2269 the IP address, check_host() was doing a simple string comparison with
2270 addresses acquired from the DNS when checking that the found name did
2271 have the original IP as one of its addresses. Since any found IPv6
2272 addresses are likely to be in abbreviated form, the comparison could
2273 fail. Luckily, there already exists a function for doing the comparison
2274 by converting both addresses to binary, so now that is used instead of
2275 the text comparison.
2277 PH/20 There was another similar case to PH/19, when a complete host name was
2278 given in a host list; looking up its IP address could give an abbreviated
2279 form, whereas the current host's name might or might not be abbreviated.
2280 The same fix has been applied.
2286 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2287 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2290 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2291 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2293 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2295 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2296 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2297 Also, for "Auto-Submitted", treat anything other than "no" as
2298 non-personal, in accordance with RFC 3834. (Previously it treated
2299 anything starting "auto-" as non-personal.)
2301 TF/01 The control=submission/name=... option had a problem with syntax
2302 errors if the name included a slash character. The /name= option
2303 now slurps the rest of the string, so it can include any characters
2304 but it must come last in the list of options (after /sender_retain
2307 PH/05 Some modifications to the interface to the fake nameserver for the new
2315 TK/01 Added the "success_on_redirect" address verification option. See
2316 NewStuff for rationale and an example.
2318 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2320 PH/02 Patch to exigrep to allow it to work on syslog lines.
2322 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2323 fread() to skip over the body file's header line, because in Cygwin the
2324 header line is locked and is inaccessible.
2326 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2327 co-exist for some time) to make it clear that it is the Exim ID that is
2328 referenced, not the Message-ID: header line.
2330 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2331 string_format() function, because snprintf() does not exist on all
2334 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2335 ${stat: expansion item.
2337 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2338 protocol synchronization error", to keep the pedants happy.
2340 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2341 well as for IRIX systems, when gcc is being used. See the host.c source
2344 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2346 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2349 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2350 transport and to the smtp transport in LMTP mode.
2352 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2354 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2355 run a filter in a subprocess. This could lead to confusion in subsequent
2356 lookups in the parent process. There should also be a search_tidyup() at
2357 the end of the subprocess.
2359 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2360 only if the host matched helo_try_verify_hosts, which caused the
2361 verification to occur when the EHLO/HELO command was issued. The ACL just
2362 tested the remembered result. Now, if a previous verification attempt has
2363 not happened, "verify = helo" does it there and then.
2365 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2367 TK/03 Fix log output including CR from clamd.
2369 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2370 exist provoked a memory error which could cause a segfault.
2372 PH/15 Installed PCRE 6.2
2374 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2376 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2377 of the problem. Specifically, suggested +O2 rather than +O1 for the
2380 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2382 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2383 record, the port was not correctly logged when the outgoing_port log
2384 selector was set (it logged the transort's default port).
2386 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2387 fallback_hosts, and "hosts" in the smtp transport.
2389 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2390 host errors such as "Connection refused".
2392 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2393 authentication with radiusclient 0.4.9:
2395 - Error returned from rc_read_config was caught wrongly
2396 - Username/password not passed on to radius server due to wrong length.
2398 The presumption is that some radiusclient API changes for 4.51/PH/17
2399 were not taken care of correctly. The code is still untested by me (my
2400 Linux distribution still has 0.3.2 of radiusclient), but it was
2401 contributed by a Radius user.
2403 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2404 expanding the "port" option of the smtp transport.
2406 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2407 while reading a MIME header. Thanks to Tom Hughes for a patch.
2409 PH/24 Include config.h inside local_scan.h so that configuration settings are
2412 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2413 that in an ACL for RCPT (for example), you can examine exactly what was
2416 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2417 commands, but it was not correctly comparing the address with the actual
2418 client host address. Thus, it would show the EHLO address in Received:
2419 header lines when this was not necessary.
2421 PH/27 Added the % operator to ${eval:}.
2423 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2424 it should be ignoring failures (because with -C, for example, it has lost
2425 privilege). It wasn't ignoring creation failures other than "already
2428 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2431 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2432 because an input line was too long, either on its own, or by virtue of
2433 too many continuations, the temporary file was not being removed, and the
2434 return code was incorrect.
2436 PH/31 Missing "BOOL" in function definition in filtertest.c.
2438 PH/32 Applied Sieve patches from the maintainer.
2440 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2442 PH/33 Added "verify = not_blind".
2444 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2445 Local/Makefile (with some defaults set). These are used in built scripts
2446 such as exicyclog, but they have never been used in the exim_install
2447 script (though there are many overriding facilities there). I have
2448 arranged that the exim_install script now takes note of these two
2451 PH/35 Installed configuration files for Dragonfly.
2453 PH/36 When a locally submitted message by a trusted user did not contain a
2454 From: header, and the sender address was obtained from -f or from an SMTP
2455 MAIL command, and the trusted user did not use -F to supply a sender
2456 name, $originator_name was incorrectly used when constructing a From:
2457 header. Furthermore, $originator_name was used for submission mode
2458 messages from external hosts without From: headers in a similar way,
2459 which is clearly wrong.
2461 PH/37 Added control=suppress_local_fixups.
2463 PH/38 When log_selector = +received_sender was set, and the addition of the
2464 sender made the log line's construction buffer exactly full, or one byte
2465 less than full, an overflow happened when the terminating "\n" was
2468 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2469 when the result of a list match is failure because a DNS lookup failed.
2471 PH/40 RM_COMMAND is now used in the building process.
2473 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2474 the "build-* directories that it finds.
2476 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2477 address in a domain literal was a prefix of an interface address.
2479 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2480 when verifying a sender address, unless rewrite_headers is false.
2482 PH/44 Wrote a long comment about why errors_to addresses are verified as
2483 recipients, not senders.
2485 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2486 the ratelimit ACL was added.
2488 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2490 PH/46 Added extra information about PostgreSQL errors to the error string.
2492 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2493 avoiding the need to install special zones in a real server. This is
2494 backwards compatible; if it can't find the fake resolver, it drops back.
2495 Thus, both old and new test suites can be run.
2497 TF/02 Added util/ratelimit.pl
2499 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2500 clock is set back in time.
2502 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2503 Candler <B.Candler@pobox.com>.
2505 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2506 OK for addresses that are the result of redirections.
2508 PH/48 A number of further additions for the benefit of the new test suite,
2509 including a fake gethostbyname() that interfaces to the fake DNS resolver
2512 TF/06 The fix for widen_domains has also been applied to qualify_single and
2513 search_parents which are the other dnslookup options that can cause
2516 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2519 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2521 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2522 service" if it encountered an MX record with an empty target hostname.
2523 The message is now "an MX or SRV record indicated no SMTP service".
2525 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2526 if the DNS of the sending site is misconfigured. This is quite a
2527 common situation. This change restores the behaviour of treating a
2528 helo verification defer as a failure.
2530 PH/51 If self=fail was set on a router, the bounce message did not include the
2531 actual error message.
2537 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2539 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2540 command itself ended in error, the underlying message about the transport
2541 filter timeout was being overwritten with the pipe command error. Now the
2542 underlying error message should be appended to the second error message.
2544 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2546 PH/02 Reduce the amount of output that "make" produces by default. Full output
2547 can still be requested.
2549 PH/03 The warning log line about a condition test deferring for a "warn" verb
2550 was being output only once per connection, rather than after each
2551 occurrence (because it was using the same function as for successful
2552 "warn" verbs). This seems wrong, so I have changed it.
2554 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2555 it should not have, which might have caused a crash in the right
2556 circumstances, but probably never did.
2558 PH/04 Installed a modified version of Tony Finch's patch to make submission
2559 mode fix the return path as well as the Sender: header line, and to
2560 add a /name= option so that you can make the user's friendly name appear
2563 TF/03 Added the control = fakedefer ACL modifier.
2565 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2566 Mark Lowes for thorough testing.
2568 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2570 TK/03 Merged latest SRS patch from Miles Wilton.
2572 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2573 with the definition in sysexits.h (which is #included earlier).
2574 Fortunately, Exim does not actually use EX_OK. The code used to try to
2575 preserve the sysexits.h value, by assumimg that macro definitions were
2576 scanned for macro replacements. I have been disabused of this notion,
2577 so now the code just undefines EX_OK before #including unistd.h.
2579 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2580 in the smtp transport. When a block could not be written in a single
2581 write() function, the timeout was being re-applied to each part-write.
2582 This seems wrong - if the receiver was accepting one byte at a time it
2583 would take for ever. The timeout is now adjusted when this happens. It
2584 doesn't have to be particularly precise.
2586 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2587 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2589 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2590 without a domain if the check to <postmaster@domain> fails.
2592 SC/01 Eximstats: added -xls and the ability to specify output files
2593 (patch written by Frank Heydlauf).
2595 SC/02 Eximstats: use FileHandles for outputing results.
2597 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2599 SC/04 Eximstats: fixed display of large numbers with -nvr option
2601 SC/05 Eximstats: fixed merging of reports with empty tables.
2603 SC/06 Eximstats: added the -include_original_destination flag
2605 SC/07 Eximstats: removed tabs and trailing whitespace.
2607 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2609 TK/06 MBOX spool code: Add real "From " MBOX separator line
2610 so the .eml file is really in mbox format (even though
2611 most programs do not really care). Patch from Alex Miller.
2613 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2614 The latter is generated from $received_to and is only set if the
2615 message has one envelope recipient. SA can use these headers,
2616 obviously out-of-the-box. Patch from Alex Miller.
2618 PH/08 The ${def test on a variable was returning false if the variable's
2619 value was "0", contrary to what the specification has always said!
2620 The result should be true unless the variable is empty.
2622 PH/09 The syntax error of a character other than { following "${if
2623 def:variable_name" (after optional whitespace) was not being diagnosed.
2624 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2625 accidental colon was present, for example, could give incorrect results.
2627 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2628 result is used (not including appendfile, where other changes are about
2631 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2632 This involved changing a lot of size variables from int to off_t. It
2633 should work with maildirs and everything.
2635 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2636 spamd dying while we are connected to it.
2638 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2641 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2642 function for BDB 4.3.
2644 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2646 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2647 the sources for PCRE have been re-arranged and more files are now
2650 PH/15 The code I had for printing potentially long long variables in PH/11
2651 above was not the best (it lost precision). The length of off_t variables
2652 is now inspected at build time, and an appropriate printing format (%ld
2653 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2654 to be "long long int" or "long int". This is needed for the internal
2655 formatting function string_vformat().
2657 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2658 the configuration file to be ":syslog", then the script "guesses" where
2659 the logs files are, rather than using the compiled in default. In our
2660 case the guess is not the same as the compiled default, so the script
2661 suddenly stopped working when I started to use syslog. The patch checks
2662 to see if log_file_path is "". If so, it attempts to read it from exim
2663 with no configuration file to get the compiled in version, before it
2664 falls back to the previous guessing code."
2666 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2667 implementing BATV in an Exim configuration. See NewStuff for the gory
2670 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2671 Makefile that are specific to HP-UX.
2673 PH/18 If the "use_postmaster" option was set for a recipient callout together
2674 with the "random" option, the postmaster address was used as the MAIL
2675 FROM address for the random test, but not for the subsequent recipient
2676 test. It is now used for both.
2678 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2679 patch removes a few documentation additions to RFC 3028, because the
2680 latest draft now contains them. It adds the new en;ascii-case comparator
2681 and a new error check for 8bit text in MIME parts. Comparator and
2682 require names are now matched exactly. I enabled the subaddress
2683 extension, but it is not well tested yet (read: it works for me)."
2685 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2686 rework some of the code of TK/09 above to avoid the hardwired use of
2687 "%lld" and "long long". Replaced the call to snprintf() with a call to
2690 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2691 records point to non-existent hosts", "retry timeout exceeded", and
2692 "retry time not reached for any host after a long failure period".
2694 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2695 experimental DomainKeys support:
2697 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2698 (2) On an error such as an illegally used "control", the wrong name for
2699 the control was given.
2701 These problems did NOT occur unless DomainKeys support was compiled.
2703 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2705 PH/24 Added ${if match_ip condition.
2707 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2708 databases so that it will be absolutely obvious if a crash occurs in the
2709 DB library. This is a regular occurrence (often caused by mis-matched
2712 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2713 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2714 picked up on a user's system that detects such things. There doesn't seem
2715 to be a gcc warning option for this - only an attribute that has to be
2716 put on the function's prototype. It seems that in Fedora Core 4 they have
2717 set this on a number of new functions. No doubt there will be more in due
2720 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2721 specify a transport. However, if an address that was verified by such a
2722 router was the subject of a callout, Exim crashed because it tried to
2723 read the rcpt_include_affixes from the non-existent transport. Now it
2724 just assumes that the setting of that option is false. This bug was
2725 introduced by 4.51/PH/31.
2727 PH/28 Changed -d+all to exclude +memory, because that information is very
2728 rarely of interest, but it makes the output a lot bigger. People tend to
2729 do -d+all out of habit.
2731 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2732 code in os-type was giving problems when libc.so lives in lib64, like on
2735 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2736 aren't the modern standard, and it seems that some systems' include files
2737 don't always have them. Exim was already checking for some of the newer
2738 ones like T_AAAA, and defining it itself. I've added checks for all the
2739 record types that Exim uses.
2741 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2742 not automatically generating a new one, as it is supposed to. This
2743 prevented TLS from working. If the file did exist, but contained invalid
2744 data, a new version was generated, as expected. It was only the case of a
2745 non-existent file that was broken.
2747 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2748 with a change in libdomainkeys > 0.64.
2750 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2751 from DNS. If the selector record carries the flag, it now has
2752 precedence over the domain-wide flag.
2754 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2756 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2757 the use of a port name that isn't defined in /etc/services) occurred, the
2758 message was deferred as in a normal delivery, and thus remained on the
2759 spool, instead of being failed because of the mua_wrapper setting. This
2760 is now fixed, and I tidied up some of the mua_wrapper messages at the
2763 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2764 the messages in a hash of arrays rather than using individual hashes.
2765 This is a bit cleaner and results in dramatic memory savings, albeit
2766 at a slight CPU cost.
2768 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2769 as requested by Marc Sherman.
2771 SC/10 Eximstats: added histograms for user specified patterns as requested
2774 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2776 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2777 fopen() in the content-scanning modules that did not already have it.
2783 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2784 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2786 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2788 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2790 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2791 to test Sieve filters that use "vacation".
2793 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2794 that changes the way the GnuTLS parameters are stored in the cache file.
2795 The new format can be generated externally. For backward compatibility,
2796 if the data in the cache doesn't make sense, Exim assumes it has read an
2797 old-format file, and it generates new data and writes a new file. This
2798 means that you can't go back to an older release without removing the
2801 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2802 work if there are any delivery delays because "one_time" forces the
2803 parent to be marked "delivered", so its unseen clone is never tried
2804 again. For this reason, Exim now forbids the simultaneous setting of
2807 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2808 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2809 entry if you want to know the details. The fix, however, neglected to
2810 consider the case when local delivery batching is involved. The test for
2811 "previously delivered" was not happening when checking to see if an
2812 address could be batched with a previous (undelivered) one; under
2813 certain circumstances this could lead to multiple deliveries to the same
2816 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2817 in its include files, and this causes problems building Exim.
2819 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2820 header_syntax) but Exim was just ignoring anything given after a slash.
2821 In particular, this caused confusion with an attempt to use "verify =
2822 reverse_host_lookup/defer_ok". An error is now given when options are
2823 supplied for verify items that do not have them. (Maybe reverse_host_
2824 lookup should have a defer_ok option, but that's a different point.)
2826 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2827 defined by RFC 821) to 2048, because there were problems with some AUTH
2828 commands, and RFC 1869 says the size should be increased for extended
2829 SMTP commands that take arguments.
2831 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2834 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2835 "unknown" error; now it says that the functionality isn't in the binary.
2837 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2838 an address' error message when a string expansion fails (syntax or
2839 whatever). Otherwise the password may appear in the log. Following change
2840 PH/42 below, there is no longer a chance of it appearing in a bounce
2843 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2845 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2846 of its addresses was ever tried. (Bugzilla bug #2.)
2848 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2849 the result incorrectly in the debug output. (It correctly added a newline
2850 to what was transported.)
2852 TF/01 Added $received_time.
2854 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2855 commented out examples of how to interface to a virus scanner and to
2856 SpamAssassin. Also added commented examples of av_scanner and
2857 spamd_address settings.
2859 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2860 and controls are allowed in which ACLs. There were a couple of minor
2861 errors. Some of the entries in the conditions table (which is a table of
2862 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2863 negation of where the condition IS allowed.
2865 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2867 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2868 header file does not have a version number, so I've had to invent a new
2869 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2870 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2871 radiusclient), but it was contributed by a Radius user.
2873 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2874 files or directories, for OpenSSL.
2876 PH/19 When an Exim process that is running as root has to create an Exim log
2877 file, it does so in a subprocess that runs as exim:exim so as to get the
2878 ownership right at creation (otherwise, other Exim processes might see
2879 the file with the wrong ownership). There was no test for failure of this
2880 fork() call, which would lead to the process getting stuck as it waited
2881 for a non-existent subprocess. Forks do occasionally fail when resources
2882 run out. I reviewed all the other calls to fork(); they all seem to check
2885 PH/20 When checking for unexpected SMTP input at connect time (before writing
2886 the banner), Exim was not dealing correctly with a non-positive return
2887 from the read() function. If the client had disconnected by this time,
2888 the result was a log entry for a synchronization error with an empty
2889 string after "input=" when read() returned zero. If read() returned -1
2890 (an event I could not check), uninitialized data bytes were printed.
2891 There were reports of junk text (parts of files, etc) appearing after
2894 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2896 PH/22 Added support for macro redefinition, and (re)definition in between
2897 driver and ACL definitions.
2899 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2900 forgetting to use the resulting value; it was using the unexpanded value.
2902 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2903 hadn't been configured. The fix is from Juergen Kreileder, who
2904 understands it better than I do:
2906 "Here's what I see happening with three configured cyrus_sasl
2907 authenticators configured (plain, login, cram-md5):
2909 On startup auth_cyrus_sasl_init() gets called for each of these.
2910 This means three calls to sasl_listmech() without a specified mech_list.
2911 => SASL tests which mechs of all available mechs actually work
2912 => three warnings about OTP not working
2913 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2915 With the patch, sasl_listmech() also gets called three times. But now
2916 SASL's mech_list option is set to the server_mech specified in the the
2917 authenticator. Or in other words, the answer from sasl_listmech()
2918 gets limited to just the mech you're testing for (which is different
2920 => the return list contains just 'plain' or 'login', 'cram-md5' or
2921 nothing depending on the value of ob->server_mech.
2923 I've just tested the patch: Authentication still works fine,
2924 unavailable mechs specified in the exim configuration are still
2925 caught, and the auth.log warnings about OTP are gone."
2927 PH/25 When debugging is enabled, the contents of the command line are added
2928 to the debugging output, even when log_selector=+arguments is not
2931 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2932 answer is "GNU", and only if the return is "GNU/something" is the answer
2935 PH/27 $acl_verify_message is now set immediately after the failure of a
2936 verification in an ACL, and so is available in subsequent modifiers. In
2937 particular, the message can be preserved by coding like this:
2939 warn !verify = sender
2940 set acl_m0 = $acl_verify_message
2942 Previously, $acl_verify_message was set only while expanding "message"
2943 and "log_message" when a very denied access.
2945 PH/28 Modified OS/os.c-Linux with
2947 -#ifndef OS_LOAD_AVERAGE
2948 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2950 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2951 nomenclature these days.)
2953 PH/29 Installed patch from the Sieve maintainer that adds the options
2954 sieve_useraddress and sieve_subaddress to the redirect router.
2956 PH/30 In these circumstances:
2957 . Two addresses routed to the same list of hosts;
2958 . First host does not offer TLS;
2959 . First host accepts first address;
2960 . First host gives temporary error to second address;
2961 . Second host offers TLS and a TLS session is established;
2962 . Second host accepts second address.
2963 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2964 and peerdn, if requested) that were in fact used only for the second
2967 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2968 attention to any local part prefix or suffix that was matched by the
2969 router that accepted the address. It now behaves in the same way as it
2970 does for delivery: the affixes are removed from the local part unless
2971 rcpt_include_affixes is set on the transport.
2973 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2974 timeout during the DATA phase of an incoming message.
2976 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2977 applied a patch sanctioned by the Sieve maintainer.
2979 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2980 the uid or gid is negative. A case of a negative gid caused this to be
2981 noticed. The fix allows for either to be negative.
2983 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2984 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2987 PH/36 The change PH/12 above was broken. Fixed it.
2989 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2990 the grounds that routing the same address twice would always produce the
2991 same answer. This might have been true once, but it is certainly no
2992 longer true now. Routing a child address may depend on the previous
2993 routing that produced that child. Some complicated redirection strategies
2994 went wrong when messages had multiple recipients, and made Exim's
2995 behaviour dependent on the order in which the addresses were given.
2997 I have moved the duplicate checking until after the routing is complete.
2998 Exim scans the addresses that are assigned to local and remote
2999 transports, and removes any duplicates. This means that more work will be
3000 done, as duplicates will always all be routed, but duplicates are
3001 presumably rare, so I don't expect this is of any significance.
3003 For deliveries to pipes, files, and autoreplies, the duplicate checking
3004 still happens during the routing process, since they are not going to be
3007 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3008 It corrects a timeout issue with spamd. This is Ian's comment: "The
3009 background is that sometimes spamd either never reads data from a
3010 connection it has accepted, or it never writes response data. The exiscan
3011 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3012 blindly assumes that writes won't block so it may never time out."
3014 PH/39 Allow G after quota size as well as K and M.
3016 PH/40 The value set for $authenticated_id in an authenticator may not contain
3017 binary zeroes or newlines because the value is written to log lines and
3018 to spool files. There was no check on this. Now the value is run through
3019 the string_printing() function so that such characters are converted to
3020 printable escape sequences.
3022 PH/41 $message_linecount is a new variable that contains the total number of
3023 lines in the message. Compare $body_linecount, which is the count for the
3026 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3027 bounce and delay warning messages, except in certain special cases, which
3030 (a) An SMTP error message from a remote host;
3031 (b) A message specified in a :fail: redirection;
3032 (c) A message specified in a "fail" command in a system filter;
3033 (d) A message specified in a FAIL return from the queryprogram router;
3034 (e) A message specified by the cannot_route_message router option.
3036 In these cases only, Exim does include the error details in bounce and
3037 warning messages. There are also a few cases where bland messages such
3038 as "unrouteable address" or "local delivery error" are given.
3040 PH/43 $value is now also set for the "else" part of a ${run expansion.
3042 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3043 being worked on, but at least Exim now implements the latest version to
3046 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3047 process to complete was treated as a delivery failure, a timeout while
3048 writing the message to the pipe was logged, but erroneously treated as a
3049 successful delivery. Such timeouts include transport filter timeouts. For
3050 consistency with the overall process timeout, these timeouts are now
3051 treated as errors, giving rise to delivery failures by default. However,
3052 there is now a new Boolean option for the pipe transport called
3053 timeout_defer, which, if set TRUE, converts the failures into defers for
3054 both kinds of timeout. A transport filter timeout is now identified in
3057 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3058 systems where "make" and "gmake" are different, calling "gmake" at top
3059 level broke things. I've arranged for the value of $(MAKE) to be passed
3060 from the Makefile to this script so that it can call the same version of
3064 A note about Exim versions 4.44 and 4.50
3065 ----------------------------------------
3067 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3068 changes of various kinds. As a consequence, a big documentation update was
3069 needed. This delayed the release for rather longer than seemed good, especially
3070 in the light of a couple of (minor) security issues. Therefore, the changes
3071 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3072 release. So 4.44 and 4.50 are in effect two different branches that both start
3075 I have left the 4.50 change log unchanged; it contains all the changes since
3076 4.43. The change log for 4.44 is below; many of its items are identical to
3077 those for 4.50. This seems to be the most sensible way to preserve the
3078 historical information.
3084 1. Minor wording change to the doc/README.SIEVE file.
3086 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3087 computation of the current number of files was incorrect.
3089 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3090 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3093 4. Give more explanation in the error message when the command for a transport
3094 filter fails to execute.
3096 5. There are several places where Exim runs a non-Exim command in a
3097 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3098 was being done only for the command run by the queryprogram router. It is
3099 now done for all such subprocesses. The other cases are: ${run, transport
3100 filters, and the commands run by the lmtp and pipe transports.
3102 6. Added CONFIGURE_GROUP build-time option.
3104 7. Some older OS have a limit of 256 on the maximum number of file
3105 descriptors. Exim was using setrlimit() to set 1000 as a large value
3106 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3107 systems. I've change it so that if it can't get 1000, it tries for 256.
3109 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3110 was an oversight, and furthermore, ever since the addition of extra
3111 controls (e.g. 4.43/32), the checks on when to allow different forms of
3112 "control" were broken. There should now be diagnostics for all cases when a
3113 control that does not make sense is encountered.
3115 9. Added the /retain_sender option to "control=submission".
3117 10. $recipients is now available in the predata ACL (oversight).
3119 11. Tidy the search cache before the fork to do a delivery from a message
3120 received from the command line. Otherwise the child will trigger a lookup
3121 failure and thereby defer the delivery if it tries to use (for example) a
3122 cached ldap connection that the parent has called unbind on.
3124 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3125 of $address_data from the recipient verification was clobbered by the
3126 sender verification.
3128 13. The value of address_data from a sender verification is now available in
3129 $sender_address_data in subsequent conditions in the ACL statement.
3131 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3133 15. Added a new option "connect=<time>" to callout options, to set a different
3136 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3137 was its contents. (It was OK if the option was not defined at all.)
3139 17. A "Completed" log line is now written for messages that are removed from
3140 the spool by the -Mrm option.
3142 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3143 information about exactly what failed.
3145 19. Added -dd to debug only the daemon process.
3147 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3148 handles timeouts, both on the server side and network timeouts. Renamed the
3149 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3151 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3152 It is now set to "smtps".
3154 22. $host_address is now set to the target address during the checking of
3155 ignore_target_hosts.
3157 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3158 being passed; this would have caused $sender_host_name to have been used if
3159 matching the list had actually called for a host name (not very likely,
3160 since this list is usually IP addresses). A host name is now passed as
3163 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3164 code that specifies a non-restarting handler (typically sigaction() in
3165 modern systems) in an attempt to fix a rare and obscure crash bug.
3167 25. Narrowed the window for a race in the daemon that could cause it to ignore
3168 SIGCHLD signals. This is not a major problem, because they are used only to
3169 wake it up if nothing else does.
3171 26. A malformed maildirsize file could cause Exim to calculate negative values
3172 for the mailbox size or file count. Odd effects could occur as a result.
3173 The maildirsize information is now recalculated if the size or filecount
3176 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3177 support for a long time. Removed HAVE_SYS_VFS_H.
3179 28. Installed the latest version of exipick from John Jetmore.
3181 29. In an address list, if the pattern was not a regular expression, an empty
3182 subject address (from a bounce message) matched only if the pattern was an
3183 empty string. Non-empty patterns were not even tested. This was the wrong
3184 because it is perfectly reasonable to use an empty address as part of a
3185 database query. An empty address is now tested by patterns that are
3186 lookups. However, all the other forms of pattern expect the subject to
3187 contain a local part and a domain, and therefore, for them, an empty
3188 address still always fails if the pattern is not itself empty.
3190 30. Exim went into a mad DNS loop when attempting to do a callout where the
3191 host was specified on an smtp transport, and looking it up yielded more
3192 than one IP address.
3194 31. Re-factored the code for checking spool and log partition space into a
3195 function that finds that data and another that does the check. The former
3196 is then used to implement four new variables: $spool_space, $log_space,
3197 $spool_inodes, and $log_inodes.
3199 32. The RFC2047 encoding function was originally intended for short strings
3200 such as real names; it was not keeping to the 75-character limit for
3201 encoded words that the RFC imposes. It now respects the limit, and
3202 generates multiple encoded words if necessary. To be on the safe side, I
3203 have increased the buffer size for the ${rfc2047: expansion operator from
3206 33. It is now permitted to omit both strings after an "if" condition; if the
3207 condition is true, the result is "true". As before, when the second string
3208 is omitted, a false condition yields an empty string. This makes it less
3209 cumbersome to write custom ACL and router conditions.
3211 34. Failure to deliver a bounce message always caused it to be frozen, even if
3212 there was an errors_to setting on the router. The errors_to setting is now
3215 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3216 canonical form (fully expanded) before being placed in
3217 $sender_host_address.
3219 36. The table in the code that translates DNS record types into text (T_A to
3220 "A" for instance) was missing entries for NS and CNAME. It is just possible
3221 that this could have caused confusion if both these types were looked up
3222 for the same domain, because the text type is used as part of Exim's
3223 per-process caching. But the chance of anyone hitting this buglet seems
3226 37. The dnsdb lookup has been extended in a number of ways.
3228 (1) There is a new type, "zns", which walks up the domain tree until it
3229 finds some nameserver records. It should be used with care.
3231 (2) There is a new type, "mxh", which is like "mx" except that it returns
3232 just the host names, not the priorities.
3234 (3) It is now possible to give a list of domains (or IP addresses) to be
3235 looked up. The behaviour when one of the lookups defers can be
3236 controlled by a keyword.
3238 (4) It is now possible to specify the separator character for use when
3239 multiple records are returned.
3241 38. The dnslists ACL condition has been extended: it is now possible to supply
3242 a list of IP addresses and/or domains to be looked up in a particular DNS
3245 39. Added log_selector=+queue_time_overall.
3247 40. When running the queue in the test harness, wait just a tad after forking a
3248 delivery process, to get repeatability of debugging output.
3250 41. Include certificate and key file names in error message when GnuTLS fails
3251 to set them up, because the GnuTLS error message doesn't include the name
3252 of the failing file when there is a problem reading it.
3254 42. Allow both -bf and -bF in the same test run.
3256 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3258 44. The "Exiscan patch" is now merged into the mainline Exim source.
3260 45. Sometimes the final signoff response after QUIT could fail to get
3261 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3262 < 0 before doing a fflush(). This bug looks as though it goes back to the
3263 introduction of TLS in release 3.20, but "sometimes" must have been rare
3264 because the tests only now provoked it.
3266 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3267 (this can affect the format of dates).
3269 47. exim_tidydb, when checking for the continued existence of a message for
3270 which it has found a message-specific retry record, was not finding
3271 messages that were in split spool directories. Consequently, it was
3272 deleting retry records that should have stayed in existence.
3274 48. Steve fixed some bugs in eximstats.
3276 49. The SPA authentication driver was not abandoning authentication and moving
3277 on to the next authenticator when an expansion was forced to fail,
3278 contradicting the general specification for all authenticators. Instead it
3279 was generating a temporary error. It now behaves as specified.
3281 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3282 (the order specifies the preference for clients). The order is now AES256,
3283 AES128, 3DES, ARCFOUR128.
3285 51. Small patch to Sieve code - explicitly set From: when generating an
3288 52. Exim crashed if a remote delivery caused a very long error message to be
3289 recorded - for instance if somebody sent an entire SpamAssassin report back
3290 as a large number of 550 error lines. This bug was coincidentally fixed by
3291 increasing the size of one of Exim's internal buffers (big_buffer) that
3292 happened as part of the Exiscan merge. However, to be on the safe side, I
3293 have made the code more robust (and fixed the comments that describe what
3296 53. Now that there can be additional text after "Completed" in log lines (if
3297 the queue_time_overall log selector is set), a one-byte patch to exigrep
3298 was needed to allow it to recognize "Completed" as not the last thing in
3301 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3302 patch that reportedly fixes this has been added. I am not expert enough to
3303 create a test for it. This is what the patch creator wrote:
3305 "I found a little strange behaviour of ldap code when working with
3306 Windows 2003 AD Domain, where users was placed in more than one
3307 Organization Units. When I tried to give exim partial DN, the exit code
3308 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3309 But simultaneously result of request was absolutely normal ldap result,
3310 so I produce this patch..."
3312 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3313 REFERENCE, so I have modified the code to exclude the patch when that macro
3316 55. Some experimental protocols are using DNS PTR records for new purposes. The
3317 keys for these records are domain names, not reversed IP addresses. The
3318 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3319 leaves it alone. Component reversal etc. now happens only for IP addresses.
3322 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3324 57. Double the size of the debug message buffer (to 2048) so that more of very
3325 long debug lines gets shown.
3327 58. The exicyclog utility now does better if the number of log files to keep
3328 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3330 59. Two changes related to the smtp_active_hostname option:
3332 (1) $smtp_active_hostname is now available as a variable.
3333 (2) The default for smtp_banner uses $smtp_active_hostname instead
3334 of $primary_hostname.
3336 60. The host_aton() function is supposed to be passed a string that is known
3337 to be a valid IP address. However, in the case of IPv6 addresses, it was
3338 not checking this. This is a hostage to fortune. Exim now panics and dies
3339 if the condition is not met. A case was found where this could be provoked
3340 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3341 components; fortuitously, this particular loophole had already been fixed
3342 by change 4.50/55 above.
3344 If there are any other similar loopholes, the new check in host_aton()
3345 itself should stop them being exploited. The report I received stated that
3346 data on the command line could provoke the exploit when Exim was running as
3347 exim, but did not say which command line option was involved. All I could
3348 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3349 running as the user.
3352 61. There was a buffer overflow vulnerability in the SPA authentication code
3353 (which came originally from the Samba project). I have added a test to the
3354 spa_base64_to_bits() function which I hope fixes it.
3357 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3358 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3360 63. The daemon start-up calls getloadavg() while still root for those OS that
3361 need the first call to be done as root, but it missed one case: when
3362 deliver_queue_load_max is set with deliver_drop_privilege. This is
3363 necessary for the benefit of the queue runner, because there is no re-exec
3364 when deliver_drop_privilege is set.
3366 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3367 This has been fixed.
3369 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3370 was in use, was not putting the data itself into the right store pool;
3371 consequently, it could be overwritten for a subsequent message in the same
3372 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3375 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3377 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3378 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3379 the function were treating the return as a boolean value, which happened to
3380 work because 0=false and not-0=true, but is not correct code.
3382 68. The host_aton() function was not handling scoped IPv6 addresses (those
3383 with, for example, "%eth0" on the end) correctly.
3385 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3386 negated items (that is, ~something) in unsigned ints. Some compilers
3387 apparently mutter when there is no cast.
3389 70. If an address verification called from an ACL failed, and did not produce a
3390 user-specific message (i.e. there was only a "system" message), nothing was
3391 put in $acl_verify_message. In this situation, it now puts the system
3394 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3395 session; change 4.31/43 added the unwanted input to the log line - except
3396 that it did not do this in the start of session case. It now does.
3398 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3399 This is wrong and can cause the other end to generate a synchronization
3400 error if it is another Exim or anything else that does the synchronization
3401 check. A QUIT command is no longer sent after a timeout.
3403 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3404 during host lookups.
3406 74. The defer_ok option of callout verification was not working if it was used
3407 when verifying addresses in header lines, that is, for this case:
3409 verify = header_sender/callout=defer_ok
3411 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3412 those file descriptors could be used for SMTP connections. If anything
3413 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3414 could be sent to the SMTP client, causing chaos. The daemon now opens
3415 stdin, stdout, and stderr to /dev/null when it puts itself into the
3418 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3419 log by default. The user can override this with suitable Perl magic.
3421 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3422 the log message when discard triggers, was not working for the DATA ACL or
3423 for the non-SMTP ACL.
3425 78. Error message wording change in sieve.c.
3427 79. If smtp_accept_max_per_host was set, the number of connections could be
3428 restricted to fewer than expected, because the daemon was trying to set up
3429 a new connection before checking whether the processes handling previous
3430 connections had finished. The check for completed processes is now done
3431 earlier. On busy systems, this bug wouldn't be noticed because something
3432 else would have woken the daemon, and it would have reaped the completed
3435 80. If a message was submitted locally by a user whose login name contained one
3436 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3437 It caused a spool format error. I have fixed the spool reading code. A
3438 related problem was that the "from" clause in the Received: line became
3439 illegal because of the space(s). It is now covered by ${quote_local_part.
3441 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3444 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3446 83. Patch from Sieve maintainer to fix unterminated string problem in
3447 "vacation" handling.
3449 84. Some minor changes to the Linux configuration files to help with other
3450 OS variants using glibc.
3452 85. One more patch for Sieve to update vacation handling to latest spec.
3455 ----------------------------------------------------
3456 See the note above about the 4.44 and 4.50 releases.
3457 ----------------------------------------------------
3463 1. Change 4.43/35 introduced a bug that caused file counts to be
3464 incorrectly computed when quota_filecount was set in an appendfile
3467 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3468 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3471 3. Give more explanation in the error message when the command for a transport
3472 filter fails to execute.
3474 4. There are several places where Exim runs a non-Exim command in a
3475 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3476 was being done only for the command run by the queryprogram router. It is
3477 now done for all such subprocesses. The other cases are: ${run, transport
3478 filters, and the commands run by the lmtp and pipe transports.
3480 5. Some older OS have a limit of 256 on the maximum number of file
3481 descriptors. Exim was using setrlimit() to set 1000 as a large value
3482 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3483 systems. I've change it so that if it can't get 1000, it tries for 256.
3485 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3486 was an oversight, and furthermore, ever since the addition of extra
3487 controls (e.g. 4.43/32), the checks on when to allow different forms of
3488 "control" were broken. There should now be diagnostics for all cases when a
3489 control that does not make sense is encountered.
3491 7. $recipients is now available in the predata ACL (oversight).
3493 8. Tidy the search cache before the fork to do a delivery from a message
3494 received from the command line. Otherwise the child will trigger a lookup
3495 failure and thereby defer the delivery if it tries to use (for example) a
3496 cached ldap connection that the parent has called unbind on.
3498 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3499 of $address_data from the recipient verification was clobbered by the
3500 sender verification.
3502 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3503 was its contents. (It was OK if the option was not defined at all.)
3505 11. A "Completed" log line is now written for messages that are removed from
3506 the spool by the -Mrm option.
3508 12. $host_address is now set to the target address during the checking of
3509 ignore_target_hosts.
3511 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3512 being passed; this would have caused $sender_host_name to have been used if
3513 matching the list had actually called for a host name (not very likely,
3514 since this list is usually IP addresses). A host name is now passed as
3517 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3518 code that specifies a non-restarting handler (typically sigaction() in
3519 modern systems) in an attempt to fix a rare and obscure crash bug.
3521 15. Narrowed the window for a race in the daemon that could cause it to ignore
3522 SIGCHLD signals. This is not a major problem, because they are used only to
3523 wake it up if nothing else does.
3525 16. A malformed maildirsize file could cause Exim to calculate negative values
3526 for the mailbox size or file count. Odd effects could occur as a result.
3527 The maildirsize information is now recalculated if the size or filecount
3530 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3531 support for a long time. Removed HAVE_SYS_VFS_H.
3533 18. Updated exipick to current release from John Jetmore.
3535 19. Allow an empty sender to be matched against a lookup in an address list.
3536 Previously the only cases considered were a regular expression, or an
3539 20. Exim went into a mad DNS lookup loop when doing a callout where the
3540 host was specified on the transport, if the DNS lookup yielded more than
3543 21. The RFC2047 encoding function was originally intended for short strings
3544 such as real names; it was not keeping to the 75-character limit for
3545 encoded words that the RFC imposes. It now respects the limit, and
3546 generates multiple encoded words if necessary. To be on the safe side, I
3547 have increased the buffer size for the ${rfc2047: expansion operator from
3550 22. Failure to deliver a bounce message always caused it to be frozen, even if
3551 there was an errors_to setting on the router. The errors_to setting is now
3554 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3555 canonical form (fully expanded) before being placed in
3556 $sender_host_address.
3558 24. Updated eximstats to version 1.33
3560 25. Include certificate and key file names in error message when GnuTLS fails
3561 to set them up, because the GnuTLS error message doesn't include the name
3562 of the failing file when there is a problem reading it.
3564 26. Expand error message when OpenSSL has problems setting up cert/key files.
3567 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3568 (this can affect the format of dates).
3570 28. exim_tidydb, when checking for the continued existence of a message for
3571 which it has found a message-specific retry record, was not finding
3572 messages that were in split spool directories. Consequently, it was
3573 deleting retry records that should have stayed in existence.
3575 29. eximstats updated to version 1.35
3576 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3577 1.35 - bugfix such that pie charts by volume are generated correctly
3579 30. The SPA authentication driver was not abandoning authentication and moving
3580 on to the next authenticator when an expansion was forced to fail,
3581 contradicting the general specification for all authenticators. Instead it
3582 was generating a temporary error. It now behaves as specified.
3584 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3585 (the order specifies the preference for clients). The order is now AES256,
3586 AES128, 3DES, ARCFOUR128.
3588 31. Small patch to Sieve code - explicitly set From: when generating an
3591 32. Exim crashed if a remote delivery caused a very long error message to be
3592 recorded - for instance if somebody sent an entire SpamAssassin report back
3593 as a large number of 550 error lines. This bug was coincidentally fixed by
3594 increasing the size of one of Exim's internal buffers (big_buffer) that
3595 happened as part of the Exiscan merge. However, to be on the safe side, I
3596 have made the code more robust (and fixed the comments that describe what
3599 33. Some experimental protocols are using DNS PTR records for new purposes. The
3600 keys for these records are domain names, not reversed IP addresses. The
3601 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3602 leaves it alone. Component reversal etc. now happens only for IP addresses.
3605 34. The host_aton() function is supposed to be passed a string that is known
3606 to be a valid IP address. However, in the case of IPv6 addresses, it was
3607 not checking this. This is a hostage to fortune. Exim now panics and dies
3608 if the condition is not met. A case was found where this could be provoked
3609 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3610 components; fortuitously, this particular loophole had already been fixed
3611 by change 4.50/55 or 4.44/33 above.
3613 If there are any other similar loopholes, the new check in host_aton()
3614 itself should stop them being exploited. The report I received stated that
3615 data on the command line could provoke the exploit when Exim was running as
3616 exim, but did not say which command line option was involved. All I could
3617 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3618 running as the user.
3621 35. There was a buffer overflow vulnerability in the SPA authentication code
3622 (which came originally from the Samba project). I have added a test to the
3623 spa_base64_to_bits() function which I hope fixes it.
3626 36. The daemon start-up calls getloadavg() while still root for those OS that
3627 need the first call to be done as root, but it missed one case: when
3628 deliver_queue_load_max is set with deliver_drop_privilege. This is
3629 necessary for the benefit of the queue runner, because there is no re-exec
3630 when deliver_drop_privilege is set.
3632 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3633 was in use, was not putting the data itself into the right store pool;
3634 consequently, it could be overwritten for a subsequent message in the same
3635 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3638 38. Sometimes the final signoff response after QUIT could fail to get
3639 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3640 < 0 before doing a fflush(). This bug looks as though it goes back to the
3641 introduction of TLS in release 3.20, but "sometimes" must have been rare
3642 because the tests only now provoked it.
3648 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3649 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3650 Now it returns an integer. A number of calls were still expecting a T/F
3651 return. Fortuitously, in all cases, the tests worked in OK situations,
3652 which is the norm. However, things would have gone wrong on any write
3653 failures on the smtp file descriptor. This function is used when sending
3654 messages over SMTP and also when doing verify callouts.
3656 2. When Exim is called to do synchronous delivery of a locally submitted
3657 message (the -odf or -odi options), it no longer closes stderr before doing
3660 3. Implemented the mua_wrapper option.
3662 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3664 5. Implemented the functions header_remove(), header_testname(),
3665 header_add_at_position(), and receive_remove_recipient(), and exported them
3668 6. If an ACL "warn" statement specified the addition of headers, Exim already
3669 inserted X-ACL-Warn: at the start if there was no header name. However, it
3670 was not making this test for the second and subsequent header lines if
3671 there were newlines in the string. This meant that an invalid header could
3672 be inserted if Exim was badly configured.
3674 7. Allow an ACL "warn" statement to add header lines at the start or after all
3675 the Received: headers, as well as at the end.
3677 8. Added the rcpt_4xx retry error code.
3679 9. Added postmaster_mailfrom=xxx to callout verification option.
3681 10. Added mailfrom=xxxx to the callout verification option, for verify=
3684 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3685 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3687 12. Inserted some casts to stop certain compilers warning when using pointer
3688 differences as field lengths or precisions in printf-type calls (mostly
3689 affecting debugging statements).
3691 13. Added optional readline() support for -be (dynamically loaded).
3693 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3694 same clock tick as a message's arrival, so that its received time was the
3695 same as the "first fail" time on the retry record, and that message
3696 remained on the queue past the ultimate address timeout, every queue runner
3697 would try a delivery (because it was past the ultimate address timeout) but
3698 after another failure, the ultimate address timeout, which should have then
3699 bounced the address, did not kick in. This was a "< instead of <=" error;
3700 in most cases the first failure would have been in the next clock tick
3701 after the received time, and all would be well.
3703 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3704 being recognized when the domain list was tested by the match_domain
3705 condition in an expansion string.
3707 16. Added the ${str2b64: operator.
3709 17. Exim was always calling setrlimit() to set a large limit for the number of
3710 processes, without checking whether the existing limit was already
3711 adequate. (It did check for the limit on file descriptors.) Furthermore,
3712 errors from getrlimit() and setrlimit() were being ignored. Now they are
3713 logged to the main and panic logs, but Exim does carry on, to try to do its
3714 job under whatever limits there are.
3716 18. Imported PCRE 5.0.
3718 19. Trivial typo in log message " temporarily refused connection" (the leading
3721 20. If the log selector return_path_on_delivery was set and an address was
3722 redirected to /dev/null, the delivery process crashed because it assumed
3723 that a return path would always be set for a "successful" delivery. In this
3724 case, the whole delivery is bypassed as an optimization, and therefore no
3727 21. Internal re-arrangement: the function for sending a challenge and reading
3728 a response while authentication was assuming a zero-terminated challenge
3729 string. It's now changed to take a pointer and a length, to allow for
3730 binary data in such strings.
3732 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3734 23. Exim was not respecting finduser_retries when seeking the login of the
3735 uid under which it was called; it was always trying 10 times. (The default
3736 setting of finduser_retries is zero.) Also, it was sleeping after the final
3737 failure, which is pointless.
3739 24. Implemented tls_on_connect_ports.
3741 25. Implemented acl_smtp_predata.
3743 26. If the domain in control=submission is set empty, Exim assumes that the
3744 authenticated id is a complete email address when it generates From: or
3745 Sender: header lines.
3747 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3748 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3749 chown and chgrp in /bin and hostname in /usr/bin.
3751 28. Exim was keeping the "process log" file open after each use, just as it
3752 does for the main log. This opens the possibility of it remaining open for
3753 long periods when the USR1 signal hits a daemon. Occasional processlog
3754 errors were reported, that could have been caused by this. Anyway, it seems
3755 much more sensible not to leave this file open at all, so that is what now
3758 29. The long-running daemon process does not normally write to the log once it
3759 has entered its main loop, and it closes the log before doing so. This is
3760 so that log files can straightforwardly be renamed and moved. However,
3761 there are a couple of unusual error situations where the daemon does write
3762 log entries, and I had neglected to close the log afterwards.
3764 30. The text of an SMTP error response that was received during a remote
3765 delivery was being truncated at 512 bytes. This is too short for some of
3766 the long messages that one sometimes sees. I've increased the limit to
3769 31. It is now possible to make retry rules that apply only when a message has a
3770 specific sender, in particular, an empty sender.
3772 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3773 it possible to be selective about when SMTP synchronization is enforced.
3775 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3777 32. Implemented hosts_connection_nolog.
3779 33. Added an ACL for QUIT.
3781 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3784 35. Added mailbox_size and mailbox_filecount to appendfile.
3786 36. Added control = no_multiline_responses to ACLs.
3788 37. There was a bug in the logic of the code that waits for the clock to tick
3789 in the case where the clock went backwards by a substantial amount such
3790 that the microsecond fraction of "now" was more than the microsecond
3791 fraction of "then" (but the whole seconds number was less).
3793 38. Added support for the libradius Radius client library this is found on
3794 FreeBSD (previously only the radiusclient library was supported).
3800 1. When certain lookups returned multiple values in the form name=value, the
3801 quoting of the values was not always being done properly. Specifically:
3802 (a) If the value started with a double quote, but contained no whitespace,
3804 (b) If the value contained whitespace other than a space character (i.e.
3805 tabs or newlines or carriage returns) it was not quoted.
3806 This fix has been applied to the mysql and pgsql lookups by writing a
3807 separate quoting function and calling it from the lookup code. The fix
3808 should probably also be applied to nisplus, ibase and oracle lookups, but
3809 since I cannot test any of those, I have not disturbed their existing code.
3811 2. A hit in the callout cache for a specific address caused a log line with no
3812 reason for rejecting RCPT. Now it says "Previous (cached) callout
3813 verification failure".
3815 3. There was an off-by-one bug in the queryprogram router. An over-long
3816 return line was truncated at 256 instead of 255 characters, thereby
3817 overflowing its buffer with the terminating zero. As well as fixing this, I
3818 have increased the buffer size to 1024 (and made a note to document this).
3820 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3821 when Exim is waiting for an SMTP response from a remote server, Exim
3822 restarts its select() call on the socket, thereby resetting its timeout.
3823 This is not a problem when such interrupts are rare. Somebody set up a cron
3824 job to run exiwhat every 2 minutes, which is less than the normal select()
3825 timeout (5 or 10 minutes). This meant that the select() timeout never
3826 kicked in because it was always reset. I have fixed this by comparing the
3827 time when an interrupt arrives with the time at the start of the first call
3828 to select(). If more time than the timeout has elapsed, the interrupt is
3829 treated as a timeout.
3831 5. Some internal re-factoring in preparation for the addition of Sieve
3832 extensions (by MH). In particular, the "personal" test is moved to a
3833 separate function, and given an option for scanning Cc: and Bcc: (which is
3834 not set for Exim filters).
3836 6. When Exim created an email address using the login of the caller as the
3837 local part (e.g. when creating a From: or Sender: header line), it was not
3838 quoting the local part when it contained special characters such as @.
3840 7. Installed new OpenBSD configuration files.
3842 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3843 try to make them clearer.
3845 9. Callout options, other than the timeout value, were being ignored when
3846 verifying sender addresses in header lines. For example, when using
3848 verify = header_sender/callout=no_cache
3850 the cache was (incorrectly) being used.
3852 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3853 only the Cygwin environment.
3855 11. When return_path_on_delivery was set as a log selector, if different remote
3856 addresses in the same message used different return paths and parallel
3857 remote delivery occurred, the wrong values would sometimes be logged.
3858 (Whenever a remote delivery process finished, the return path value from
3859 the most recently started remote delivery process was logged.)
3861 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3862 lines when AUTH and/or TLS are in use. This is the "received protocol"
3863 field. Exim used to use "asmtp" for authenticated SMTP, without any
3864 indication (in the protocol name) for TLS use. Now it follows the RFC and
3865 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3866 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3867 names appear in log lines as well as in Received: header lines.
3869 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3870 extensions, and comparison tests, and to fix some bugs.
3872 14. Changes to the "personal" filter test:
3874 (1) The test was buggy in that it was just doing the equivalent of
3875 "contains" tests on header lines. For example, if a user's address was
3876 anne@some.where, the "personal" test would incorrectly be true for
3878 To: susanne@some.where
3880 This test is now done by extracting each address from the header in turn,
3881 and checking the entire address. Other tests that are part of "personal"
3882 are now done using regular expressions (for example, to check local parts
3883 of addresses in From: header lines).
3885 (2) The list of non-personal local parts in From: addresses has been
3886 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3887 taken from the Sieve specification recommendations.
3889 (3) If the message contains any header line starting with "List-" it is
3890 treated as non-personal.
3892 (4) The test for "circular" in the Subject: header line has been removed
3893 because it now seems ill-conceived.
3895 15. Minor typos in src/EDITME comments corrected.
3897 16. Installed latest exipick from John Jetmore.
3899 17. If headers_add on a router specified a text string that was too long for
3900 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3901 of string_sprintf() is now avoided.
3903 18. $message_body_size was not set (it was always zero) when running the DATA
3904 ACL and the local_scan() function.
3906 19. For the "mail" command in an Exim filter, no default was being set for
3907 the once_repeat time, causing a random time value to be used if "once" was
3908 specified. (If the value happened to be <= 0, no repeat happened.) The
3909 default is now 0s, meaning "never repeat". The "vacation" command was OK
3910 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3911 (I found it when inspecting the code).
3913 20. There is now an overall timeout for performing a callout verification. It
3914 defaults to 4 times the callout timeout, which applies to individual SMTP
3915 commands during the callout. The overall timeout applies when there is more
3916 than one host that can be tried. The timeout is checked before trying the
3917 next host. This prevents very long delays if there are a large number of
3918 hosts and all are timing out (e.g. when the network connections are timing
3919 out). The value of the overall timeout can be changed by specifying an
3920 additional sub-option for "callout", called "maxwait". For example:
3922 verify = sender/callout=5s,maxwait=20s
3924 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3925 to the end before writing, but this should make it even safer).
3927 22. Exim was forgetting that it had advertised PIPELINING for the second and
3928 subsequent messages on an SMTP connection. It was also not resetting its
3929 memory on STARTTLS and an internal HELO.
3931 23. When Exim logs an SMTP synchronization error within a session, it now
3932 records whether PIPELINING has been advertised or not.
3934 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3935 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3936 rather than long int.
3938 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3940 26. Added the never_mail option to autoreply.
3946 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3947 crash if the getsockname() call failed; this can happen if a connection is
3948 closed very soon after it is established. The problem was simply in the
3949 order in which certain operations were done, causing Exim to try to write
3950 to the SMTP stream before it had set up the file descriptor. The bug has
3951 been fixed by making things happen in the correct order.
3957 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3958 before the connection was closed, thus losing the rejection response.
3960 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3961 some early Solaris releases, but causes trouble in current releases where
3962 socklen_t is defined.
3964 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3967 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3970 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3971 that releasing the top of it at the end releases what was used for sub-
3972 expansions (unless the block got too big). However, discard this block if
3973 the first thing is a variable or header, so that we can use its block when
3974 it is dynamic (useful for very large $message_headers, for example).
3976 6. Lookups now cache *every* query, not just the most recent. A new, separate
3977 store pool is used for this. It can be recovered when all lookup caches are
3978 flushed. Lookups now release memory at the end of their result strings.
3979 This has involved some general refactoring of the lookup sources.
3981 7. Some code has been added to the store_xxx() functions to reduce the amount
3982 of flapping under certain conditions.
3984 8. log_incoming_interface used to affect only the <= reception log lines. Now
3985 it causes the local interface and port to be added to several more SMTP log
3986 lines, for example "SMTP connection from", and rejection lines.
3988 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3990 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3992 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3993 could be overwritten at the end of the current message (or the start of a
3994 new message if it was set in a HELO ACL). The value is now preserved for
3995 the duration of the SMTP connection.
3997 12. If a transport had a headers_rewrite setting, and a matching header line
3998 contained an unqualified address, that address was qualified, even if it
3999 did not match any rewriting rules. The underlying bug was that the values
4000 of the flags that permit the existence of unqualified sender and recipient
4001 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4002 non-local messages, and by -bnq for local messages) were not being
4003 preserved with the message after it was received.
4005 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4006 "next input=" as part of the text comprising the host identity instead of
4007 the correct text. The code was using the same buffer for two different
4008 strings. However, depending on which order the printing function evaluated
4009 its arguments, the bug did not always show up. Under Linux, for example, my
4010 test suite worked just fine.
4012 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4013 doesn't work with some older versions of Perl. It has been changed to "my",
4014 which in any case is probably the better facility to use.
4016 15. A really picky compiler found some instances of statements for creating
4017 error messages that either had too many or two few arguments for the format
4020 16. The size of the buffer for calls to the DNS resolver has been increased
4021 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4022 for addresses that have a lot of PTR records. This alleviates a problem; it
4023 does not fully solve it.
4025 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4026 buffer now truncates the list and logs the incident, which is the same
4027 action as happens when Exim is looking up a host name and its aliases.
4028 Previously in this situation something unpredictable would happen;
4029 sometimes it was "internal error: store_reset failed".
4031 18. If a server dropped the connection unexpectedly when an Exim client was
4032 using GnuTLS and trying to read a response, the client delivery process
4033 crashed while trying to generate an error log message.
4035 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4036 string, for example:
4038 warn message = H1: something\nH2: something
4040 the text was added as a single header line from Exim's point of view
4041 though it ended up OK in the delivered message. However, searching for the
4042 second and subsequent header lines using $h_h2: did not work. This has been
4043 fixed. Similarly, if a system filter added multiple headers in this way,
4044 the routers could not see them.
4046 20. Expanded the error message when iplsearch is called with an invalid key to
4047 suggest using net-iplsearch in a host list.
4049 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4050 ACLs are no longer actually imposed (and a message to that effect is
4053 22. If a "gecos" field in a passwd entry contained escaped characters, in
4054 particular, if it contained a \" sequence, Exim got it wrong when building
4055 a From: or a Sender: header from that name. A second bug also caused
4056 incorrect handling when an unquoted " was present following a character
4057 that needed quoting.
4059 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4060 was not being matched caselessly.
4062 24. Arranged for all hyphens in the exim.8 source to be escaped with
4065 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4066 properly. Recipient callout cache records were still being keyed to include
4067 the sender, even when use_sender was set false. This led to far more
4068 callouts that were necessary. The sender is no longer included in the key
4069 when use_sender is false.
4071 26. Added "control = submission" modifier to ACLs.
4073 27. Added the ${base62d: operator to decode base 62 numbers.
4075 28. dnsdb lookups can now access SRV records.
4077 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4078 the configuration file.
4080 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4081 (-v) mode. This makes the output for a verbose queue run more intelligible.
4083 31. Added a use_postmaster feature to recipient callouts.
4085 32. Added the $body_zerocount variable, containing the number of binary zero
4086 bytes in the message body.
4088 33. The time of last modification of the "new" subdirectory is now used as the
4089 "mailbox time last read" when there is a quota error for a maildir
4092 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4094 35. Added +ignore_unknown as a special item in host lists.
4096 36. Code for decoding IPv6 addresses in host lists is now included, even if
4097 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4098 address was recognized as an IP address, but was then not correctly decoded
4099 into binary, causing unexpected and incorrect effects when compared with
4106 1. Very minor rewording of debugging text in manualroute to say "list of
4107 hosts" instead of "hostlist".
4109 2. If verify=header_syntax was set, and a header line with an unqualified
4110 address (no domain) and a large number of spaces between the end of the
4111 name and the colon was received, the reception process suffered a buffer
4112 overflow, and (when I tested it) crashed. This was caused by some obsolete
4113 code that should have been removed. The fix is to remove it!
4115 3. When running in the test harness, delay a bit after writing a bounce
4116 message to get a bit more predictability in the log output.
4118 4. Added a call to search_tidyup() just before forking a reception process. In
4119 theory, someone could use a lookup in the expansion of smtp_accept_max_
4120 per_host which, without the tidyup, could leave open a database connection.
4122 5. Added the variables $recipient_data and $sender_data which get set from a
4123 lookup success in an ACL "recipients" or "senders" condition, or a router
4124 "senders" option, similar to $domain_data and $local_part_data.
4126 6. Moved the writing of debug_print from before to after the "senders" test
4129 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4130 problems for message scanning, either using a data ACL, or using
4131 local_scan() because the Received: header was not generated till after they
4132 were called (in order to set the time as the time of reception completion).
4133 I have revised the way this works. The header is now generated after the
4134 body is received, but before the ACL or local_scan() are called. After they
4135 are run, the timestamp in the header is updated.
4141 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4142 before starting a queue runner without re-exec. This happened only when
4143 deliver_drop_privilege was set or when the Exim user was set to root. The
4144 effect of the bug was that timeouts during subsequent deliveries caused
4145 crashes instead of being properly handled. The handler is now left at its
4146 default (and expected) setting.
4148 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4149 message, again when deliver_drop_privilege is set or Exim is run as root.
4150 The bug described in (1) was not present in this case, but the tidying up
4151 of the other signals was missing. I have made the two cases consistent.
4153 3. The ignore_target_hosts setting on a manualroute router was being ignored
4154 for hosts that were looked up using the /MX notation.
4156 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4159 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4160 operated on the sender address. After changing the $sender_address to <>
4161 for the sender address verify, Exim was re-instated it as the original
4162 (before rewriting) address, but remembering that it had rewritten it, so it
4163 wasn't rewriting it again. This bug also had the effect of breaking the
4164 sender address verification caching when the sender address was rewritten.
4166 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4167 This has been changed so that if the ip literal address matches
4168 ignore_target_hosts, the router declines.
4170 7. Added expansion conditions match_domain, match_address, and match_local_
4171 part (NOT match_host).
4173 8. The placeholder for the Received: header didn't have a length field set.
4175 9. Added code to Exim itself and to exim_lock to test for a specific race
4176 condition that could lead to file corruption when using MBX delivery. The
4177 issue is with the lockfile that is created in /tmp. If this file is removed
4178 after a process has opened it but before that process has acquired a lock,
4179 there is the potential for a second process to recreate the file and also
4180 acquire a lock. This could lead to two Exim processes writing to the file
4181 at the same time. The added code performs the same test as UW imapd; it
4182 checks after acquiring the lock that its file descriptor still refers to
4183 the same named file.
4185 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4186 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4189 11. Added the smtp_active_hostname option. If used, this will typically be made
4190 to depend on the incoming interface address. Because $interface_address is
4191 not set up until the daemon has forked a reception process, error responses
4192 that can happen earlier (such as "too many connections") no longer contain
4195 12. If an expansion in a condition on a "warn" statement fails because a lookup
4196 defers, the "warn" statement is abandoned, and the next ACL statement is
4197 processed. Previously this caused the whole ACL to be aborted.
4199 13. Added the iplsearch lookup type.
4201 14. Added ident_timeout as a log selector.
4203 15. Added tls_certificate_verified as a log selector.
4205 16. Added a global option tls_require_ciphers (compare the smtp transport
4206 option of the same name). This controls incoming TLS connections.
4208 17. I finally figured out how to make tls_require_ciphers do a similar thing
4209 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4210 before starting the TLS session.
4212 18. Tabs are now shown as \t in -bP output.
4214 19. If the log selector return_path_on_delivery was set, Exim crashed when
4215 bouncing a message because it had too many Received: header lines.
4217 20. If two routers both had headers_remove settings, and the first one included
4218 a superfluous trailing colon, the final name in the first list and the
4219 first name in the second list were incorrectly joined into one item (with a
4220 colon in the middle).
4226 1. Added -C and -D options to the exinext utility, mainly to make it easier
4227 to include in the automated testing, but these could be helpful when
4228 multiple configurations are in use.
4230 2. The exinext utility was not formatting the output nicely when there was
4231 an alternate port involved in the retry record key, nor when there was a
4232 message id as well (for retries that were specific to a specific message
4233 and a specific host). It was also confused by IPv6 addresses, because of
4234 the additional colons they contain. I have fixed the IPv4 problem, and
4235 patched it up to do a reasonable job for IPv6.
4237 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4238 delivery, the log line now contains "pipelined" if PIPELINING was used.
4240 4. An SMTP transport process used to panic and die if the bind() call to set
4241 an explicit outgoing interface failed. This has been changed; it is now
4242 treated in the same way as a connect() failure.
4244 5. A reference to $sender_host_name in the part of a conditional expansion
4245 that was being skipped was still causing a DNS lookup. This no longer
4248 6. The def: expansion condition was not recognizing references to header lines
4249 that used bh_ and bheader_.
4251 7. Added the _cache feature to named lists.
4253 8. The code for checking quota_filecount in the appendfile transport was
4254 allowing one more file than it should have been.
4256 9. For compatibility with Sendmail, the command line option
4264 and sets the incoming protocol and host name (for trusted callers). The
4265 host name and its colon can be omitted when only the protocol is to be set.
4266 Note the Exim already has two private options, -pd and -ps, that refer to
4267 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4268 "s", but I don't think that's a major issue.
4270 10. A number of refactoring changes to the code, none of which should affect
4273 (a) The number of logging options was getting close to filling up the
4274 32-bit word that was used as a bit map. I have split them into two classes:
4275 those that are passed in the argument to log_write(), and those that are
4276 only ever tested independently outside of that function. These are now in
4277 separate 32-bit words, so there is plenty of room for expansion again.
4278 There is no change in the user interface or the logging behaviour.
4280 (b) When building, for example, log lines, the code previously used a
4281 macro that called string_cat() twice, in order to add two strings. This is
4282 not really sufficiently general. Furthermore, there was one instance where
4283 it was actually wrong because one of the argument was used twice, and in
4284 one call a function was used. (As it happened, calling the function twice
4285 did not affect the overall behaviour.) The macro has been replaced by a
4286 function that can join an arbitrary number of extra strings onto a growing
4289 (c) The code for expansion conditions now uses a table and a binary chop
4290 instead of a serial search (which was left over from when there were very
4291 few conditions). Also, it now recognizes conditions like "pam" even when
4292 the relevant support is not compiled in: a suitably worded error message is
4293 given if an attempt is made to use such a condition.
4295 11. Added ${time_interval:xxxxx}.
4297 12. A bug was causing one of the ddress fields not to be passed back correctly
4298 from remote delivery subprocesses. The field in question was not being
4299 subsequently used, so this caused to problems in practice.
4301 13. Added new log selectors queue_time and deliver_time.
4303 14. Might have fixed a bug in maildirsizefile handling that threw up
4304 "unexpected character" debug warnings, and recalculated the data
4305 unnecessarily. In any case, I expanded the warning message to give more
4308 15. Added the message "Restricted characters in address" to the statements in
4309 the default ACL that block characters like @ and % in local parts.
4311 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4312 Three changes have been made:
4314 (a) There was a serious bug; a negative response to MAIL caused the whole
4315 recipient domain to be cached as invalid, thereby blocking all messages
4316 to all local parts at the same domain, from all senders. This bug has
4317 been fixed. The domain is no longer cached after a negative response to
4318 MAIL if the sender used is not empty.
4320 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4323 (c) A new callout option, "use_sender" has been added for people who want
4324 the modified behaviour.
4330 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4333 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4334 indeed breaks things for older releases.
4336 3. Added additional logging to the case where there is a problem reading data
4337 from a filter that is running in a subprocess using a pipe, in order to
4338 try to track down a specific problem.
4340 4. Testing facility fudge: when running in the test harness and attempting
4341 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4342 getting "No route to host". Convert this to a timeout.
4344 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4347 6. Some OS don't have socklen_t but use size_t instead. This affects the
4348 fifth argument of getsockopt() amongst other things. This is now
4349 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4350 can be set for individual OS. I have set it for SunOS5, OSF1, and
4351 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4352 some earlier ones do not.
4354 7. Change 4.30/15 was not doing the test caselessly.
4356 8. The standard form for an IPv6 address literal was being rejected by address
4357 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4358 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4359 this, as well as the form without the "IPv6" on the front (but only when
4360 address literals are enabled, of course).
4362 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4364 10. Exim crashed if a message with an empty sender address specified by -f
4365 encountered a router with an errors_to setting. This could be provoked only
4366 by a command such as
4370 where an empty string was supplied; "<>" did not hit this bug.
4372 11. Installed PCRE release 4.5.
4374 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4375 remained set. It is now erased.
4377 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4378 times from message ids (which are base 36 rather than the normal 62).
4380 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4381 were being counted as actual protocol errors, and logged if the log
4382 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4383 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4384 and DATA following a set of rejected RCPTs do not count as protocol errors.
4385 In other words, Exim assumes they were pipelined, though this may not
4386 actually be the case. Of course, in all cases the client gets an
4387 appropriate error code.
4389 15. If a lookup fails in an ACL condition, a message about the failure may
4390 be available; it is used if testing the ACL cannot continue, because most
4391 such messages specify what the cause of the deferral is. However, some
4392 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4393 that caused an old message to be retained and used if a later statement
4394 caused a defer, replacing the real cause of the deferral.
4396 16. If an IP address had so many PTR records that the DNS lookup buffer
4397 was not large enough to hold them, Exim could crash while trying to process
4398 the truncated data. It now detects and logs this case.
4400 17. Further to 4.21/58, another change has been made: if (and only if) the
4401 first line of a message (the first header line) ends with CRLF, a bare LF
4402 in a subsequent header line has a space inserted after it, so as not to
4403 terminate the header.
4405 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4406 unnecessarily, used atoi() instead of strtol(), and didn't check the
4407 termination when getting file sizes from file names by regex.
4409 19. Completely re-implemented the support for maildirsize files, in the light
4410 of a number of problems with the previous contributed implementation
4411 (4.30/29). In particular:
4413 . If the quota is zero, the maildirsize file is maintained, but no quota is
4416 . If the maildir directory does not exist, it is created before any attempt
4417 to write a maildirsize file.
4419 . The quota value in the file is just a cache; if the quota is changed in
4420 the transport, the new value overrides.
4422 . A regular expression is available for excluding directories from the
4425 20. The autoreply transport checks the characters in options that define the
4426 message's headers; it allows continued headers, but it was checking with
4427 isspace() after an embedded newline instead of explicitly looking for a
4430 21. If all the "regular" hosts to which an address was routed had passed their
4431 expiry times, and had not reached their retry times, the address was
4432 bounced, even if fallback hosts were defined. Now Exim should go on to try
4435 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4436 equivalent code in the SMTP transport. Some hosts send humungous responses
4437 to HELO/EHLO, more than 1024 it seems.
4439 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4440 gives compiler warnings in some environments. I've now done it "properly",
4443 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4444 (because of problems with the built-in one) was declared to return uschar *
4445 instead of char *, causing compiler failure.
4447 25. Fixed a file descriptor leak when processing alias/forward files.
4449 26. Fixed a minor format string issue in dbfn.c.
4451 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4453 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4454 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4457 29. When a sender address is verified, it is cached, to save repeating the test
4458 when there is more than one recipient in a message. However, when the
4459 verification involves a callout, it is possible for different callout
4460 options to be set for different recipients. It is too complicated to keep
4461 track of this in the cache, so now Exim always runs a verification when a
4462 callout is required, relying on the callout cache for the optimization.
4463 The overhead is duplication of the address routing, but this should not be
4466 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4467 to be verified with callout=postmaster, and the main callout worked but the
4468 postmaster check failed, the verification correctly failed. However, if a
4469 subsequent RCPT command asked for sender verification *without* the
4470 postmaster check, incorrect caching caused this verification also to fail,
4473 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4474 it was not caching the DNS options (qualify_single, search_parents) that
4475 were used when the lookup failed. A subsequent lookup with different
4476 options therefore always gave the same answer, though there were cases
4477 where it should not have. (Example: a "domains = !$mx_any" option on a
4478 dnslookup router: the "domains" option is always processed without any
4479 widening, but the router might have qualify_single set.) Now Exim uses the
4480 cached value only when the same options are set.
4482 32. Added John Jetmore's "exipick" utility to the distribution.
4484 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4485 than a timeout (e.g. a certificate is required, and is not provided), an
4486 Exim server now closes the connection immediately. Previously it waited for
4487 the client to close - but if the client is SSL, it seems that they each
4488 wait for each other, leading to a delay before one of them times out.
4490 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4491 maintained 0.8.x compatibility because I don't think many are using it, and
4492 it is clearly obsolete.
4494 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4497 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4498 client certificate was expired. A simple patch fixes this, though I don't
4499 understand the full logic of why the verify callback is called multiple
4502 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4503 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4504 which causes problems with some clients (such as the Certicom SSL Plus
4505 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4506 disables the coutermeasure allowing Eudora to connect."
4508 38. Exim was not checking that a write() to a log file succeeded. This could
4509 lead to Bad Things if a log got too big, in particular if it hit a file
4510 size limit. Exim now panics and dies if it cannot write to a log file, just
4511 as it does if it cannot open a log file.
4513 39. Modified OS/Makefile-Linux so that it now contains
4515 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4517 The two -D definitions ensure that Exim is compiled with large file
4518 support, which makes it possible to handle log files that are bigger than
4521 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4522 instance) a domain was checked against a named list that involved a lookup,
4523 causing $domain_data to be set, then another domain was checked against the
4524 same list, then the first domain was re-checked, the value of $domain_data
4525 after the final check could be wrong. In particular, if the second check
4526 failed, it could be set empty. This bug probably also applied to
4529 41. The strip_trailing_dot option was not being applied to the address given
4530 with the -f command-line option.
4532 42. The code for reading a message's header from the spool was incrementing
4533 $received_count, but never initializing it. This meant that the value was
4534 incorrect (doubled) while delivering a message in the same process in which
4535 it was received. In the most common configuration of Exim, this never
4536 happens - a fresh exec is done - but it can happen when
4537 deliver_drop_privilege is set.
4539 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4540 it now includes up to 150 characters of the unexpected data in the log
4543 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4544 and building data strings. The size of both of these buffers was 10 000
4545 bytes - far larger than anybody would *ever* want, thought I. Needless to
4546 say, somebody hit the limit. I have increased the maximum line length to
4547 20 000 and the maximum data length of concatenated lines to 100 000. I have
4548 also fixed two bugs, because there was no checking on these buffers. Tsk,
4549 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4550 buffer is too small.
4552 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4553 lsearch lookups. Now it does.
4555 46. When parsing a route_list item in a manualroute router, a fixed-length
4556 buffer was used for the list of hosts. I made this 1024 bytes long,
4557 thinking that nobody would ever have a list of hosts that long. Wrong.
4558 Somebody had a whole pile of complicated expansion conditions, and the
4559 string was silently truncated, leading to an expansion error. It turns out
4560 that it is easier to change to an unlimited length (owing to other changes
4561 that have happened since this code was originally written) than to build
4562 structure for giving a limitation error. The length of the item that
4563 expands into the list of hosts is now unlimited.
4565 47. The lsearch lookup could not handle data where the length of text line was
4566 more than 4095 characters. Such lines were truncated, leading to shortened
4567 data being returned. It should now handle lines of any length.
4569 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4570 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4573 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4574 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4575 no longer happens in comment lines. A list of replacements is now placed
4576 at the head of all of the source files, except those whose only change is
4577 to replace PERL_COMMAND in the very first #! line.
4579 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4580 messages on the queue, with a bottom-up merge sort, using code contributed
4581 by Michael Haardt. This should make operations like -bp somewhat faster on
4582 large queues. It won't affect queue runners, except when queue_run_in_order
4585 51. Installed eximstats 1.31 in the distribution.
4587 52. Added support for SRV lookups to the dnslookup router.
4589 53. If an ACL referred to $message_body or $message_body_end, the value was not
4590 reset for any messages that followed in the same SMTP session.
4592 54. The store-handling optimization for building very long strings was not
4593 differentiating between the different store pools. I don't think this
4594 actually made any difference in practice, but I've tidied it.
4596 55. While running the routers to verify a sender address, $sender_address
4597 was still set to the sender address. This is wrong, because when routing to
4598 send a bounce to the sender, it would be empty. Therefore, I have changed
4599 it so that, while verifying a sender address, $sender_address is set to <>.
4600 (There is no change to what happens when verifying a recipient address.)
4602 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4603 target A or AAAA records (if not already returned) without resetting the
4604 qualify_single or search_parents options of the DNS resolver. These are
4605 inappropriate in this case because the targets of MX and SRV records must
4606 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4607 target that, when qualified, matched something in the local domain. These
4608 two options are now turned off when doing these lookups.
4610 57. It seems that at least some releases of Reiserfs (which does not have the
4611 concept of a fixed number of inodes) returns zero and not -1 for the
4612 number of available inodes. This interacted badly with check_spool_inodes,
4613 which assumed that -1 was the "no such thing" setting. What I have done is
4614 to check that the total number of inodes is greater than zero before doing
4615 the test of how many are available.
4617 58. When a "warn" ACL statement has a log_message modifier, the message is
4618 remembered, and not repeated. This is to avoid a lot of repetition when a
4619 message has many recipients that cause the same warning to be written.
4620 Howewer, Exim was preserving the list of already written lines for an
4621 entire SMTP session, which doesn't seem right. The memory is now reset if a
4622 new message is started.
4624 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4625 debugging output unless log_rewrite was also set.
4627 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4628 of (int)(handle) when we know that handle contains (void *)(-1).
4630 61. The Exim daemon panic-logs an error return when it closes the incoming
4631 connection. However "connection reset by peer" seems to be common, and
4632 isn't really an error worthy of noting specially, so that particular error
4635 62. When Exim is trying to find all the local interfaces, it used to panic and
4636 die if the ioctl to get the interface flags failed. However, it seems that
4637 on at least one OS (Solaris 9) it is possible to have an interface that is
4638 included in the list of interfaces, but for which you get a failure error
4639 for this call. This happens when the interface is not "plumbed" into a
4640 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4641 failure of the "get flags" call assumes that the interface is down.
4643 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4644 makes life easier for people who are doing arithmetic on fields extracted
4645 from dates, where you often get leading zeros that should not be
4646 interpreted as octal.
4648 64. Added qualify_domain to the redirect router, to override the global
4651 65. If a pathologically long header line contained very many addresses (the
4652 report of this problem mentioned 10 000) and each of them was rewritten,
4653 Exim could use up a very large amount of memory. (It kept on making new
4654 copies of the header line as it rewrote, and never released the old ones.)
4655 At the expense of a bit more processing, the header rewriting function has
4656 been changed so that it no longer eats memory in this way.
4658 66. The generation of the Received: header has been moved from the time that a
4659 message starts to be received, to the time that it finishes. The timestamp
4660 in the Received: header should now be very close to that of the <= log
4661 line. There are two side-effects of this change:
4663 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4664 logged header lines no longer include the local Received: line, because
4665 it has not yet been created. The same applies to a copy of the message
4666 that is returned to a non-SMTP sender when a message is rejected.
4668 (b) When a filter file is tested using -bf, no additional Received: header
4669 is added to the test message. After some thought, I decided that this
4672 This change does not affect the value of $received_for. It is still set
4673 after address rewriting, but before local_scan() is called.
4675 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4677 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4678 gave an unhelpful panic error message, and a defer error. I have managed to
4679 change this behaviour so that it now rejects any supplied certificate,
4680 which seems right, as the list of acceptable certificates is empty.
4682 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4683 gave an unhelpful defer error. I have not managed to make this reject any
4684 supplied certificates, but the error message it gives is "no certificate
4685 supplied", which is not helpful.
4687 70. exigrep's output now also includes lines that are not associated with any
4688 message, but which match the given pattern. Implemented by a patch from
4689 Martin Sluka, which also tidied up the Perl a bit.
4691 71. Recipient callout verification, like sender verification, was using <> in
4692 the MAIL FROM command. This isn't really the right thing, since the actual
4693 sender may affect whether the remote host accepts the recipient or not. I
4694 have changed it to use the actual sender in the callout; this means that
4695 the cache record is now keyed on a recipient/sender pair, not just the
4696 recipient address. There doesn't seem to be a real danger of callout loops,
4697 since a callout by the remote host to check the sender would use <>.
4698 [SEE ABOVE: changed after hitting problems.]
4700 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4701 temporary errors. However, in the case of such a code being given after
4702 the end of a data transmission (i.e. after ".") Exim was failing to write
4703 a retry record for the message. (Yes, there was some broken host that was
4704 actually sending 8xx at this point.)
4706 73. An unknown lookup type in a host list could cause Exim to panic-die when
4707 the list was checked. (An example that provoked this was putting <; in the
4708 middle of a list instead of at the start.) If this happened during a DATA
4709 ACL check, a -D file could be left lying around. This kind of configuration
4710 error no longer causes Exim to die; instead it causes a defer errror. The
4711 incident is still logged to the main and panic logs.
4713 74. Buglet left over from Exim 3 conversion. The message "too many messages
4714 in one connection" was written to the rejectlog but not the mainlog, except
4715 when address rewriting (yes!) was being logged.
4717 75. Added write_rejectlog option.
4719 76. When a system filter was run not as root (that is, when system_filter_user
4720 was set), the values of the $n variables were not being returned to the
4721 main process; thus, they were not subsequently available in the $sn
4724 77. Added +return_path_on_delivery log selector.
4726 78. A connection timeout was being treated differently from recipients deferred
4727 when testing hosts_max_try with a message that was older than the host's
4728 retry timeout. (The host should not be counted, thus allowing all hosts to
4729 be tried at least once before bouncing.) This may have been the cause of an
4730 occasionally reported bug whereby a message would remain on the queue
4731 longer than the retry timeout, but would be bounced if a delivery was
4732 forced. I say "may" because I never totally pinned down the problem;
4733 setting up timeout/retry tests is difficult. See also the next item.
4735 79. The ultimate address timeout was not being applied to errors that involved
4736 a combination of host plus message (for example, a timeout on a MAIL
4737 command). When an address resolved to a number of possible hosts, and they
4738 were not all tried for each delivery (e.g. because of hosts_max_try), a
4739 message could remain on the queue longer than the retry timeout.
4741 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4744 81. Fixed an obscure SMTP outgoing bug which required at least the following
4745 conditions: (a) there was another message waiting for the same server;
4746 (b) the server returned 5xx to all RCPT commands in the first message so
4747 that the message was not completed; (c) the server dropped the connection
4748 or gave a negative response to the RSET that Exim sends to abort the
4749 transaction. The observed case was a dropped connection after DATA that had
4750 been sent in pipelining mode. That is, the server had advertised PIPELINING
4751 but was not implementing it correctly. The effect of the bug was incorrect
4752 behaviour, such as trying another host, and this could lead to a crash.
4758 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4759 and daemon.c were passed as pointers to ints; they should have been
4760 pointers to socklen_t variables (which are typically unsigned ints).
4762 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4765 3. Fixed a really odd bug that affected only the testing scheme; patching a
4766 certain fixed string in the binary changed the value of another string that
4767 happened to be identical to the end of the original first string.
4769 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4770 name", it returns that address as the IP address. On some operating
4771 systems (e.g. Solaris), it also passes back the IP address string as the
4772 "host name". However, on others (e.g. Linux), it passes back an empty
4773 string. Exim wasn't checking for this, and was changing the host name to an
4774 empty string, assuming it had been canonicized.
4776 5. Although rare, it is permitted to have more than one PTR record for a given
4777 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4778 all the names associated with an address, because they do in Solaris.
4779 However, it seems that they do not in Linux for data that comes from the
4780 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4781 I found this out when I moved to a new Linux workstation and tried to run
4782 the Exim test suite.
4784 To get round this problem I have changed the code so that it now does its
4785 own call to the DNS to look up PTR records when searching for a host name.
4786 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4787 addresses that are only in /etc/hosts are still found.
4789 This behaviour is, however, controlled by an option called host_lookup_
4790 order, which defaults to "bydns:byaddr". If people want to use the other
4791 order, or indeed, just use one or the other means of lookup, they can
4792 specify it in this variable.
4794 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4795 some operating systems, this comes back from gethostbyaddr() as an empty
4796 string, and this is what Exim used to test for. However, it seems that in
4797 other systems, "." is yielded. Exim now tests for this case too.
4799 7. The values of check_spool_space and check_log_space are now held internally
4800 as a number of kilobytes instead of an absolute number of bytes. If a
4801 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4802 kilobyte. This means that much larger values can be stored.
4804 8. Exim monitor: an attempt to get the action menu when not actually pointing
4805 at a message produces an empty menu entitled "No message selected". This
4806 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4807 no entries in it ("Shell widget menu has zero width and/or height"). So I
4808 have added a single, blank menu entry in this case.
4810 9. Added ${quote_local_part.
4812 10. MIME decoding is now applied to the contents of Subject: header lines when
4815 11. Now that a reference to $sender_host_address automatically causes a reverse
4816 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4817 host lookup before query-style lookups in lists that might use this
4818 variable. This has therefore been abolished, and the "net-" prefix is no
4819 longer necessary for query-style lookups.
4821 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4822 have been a typo for LFLAGS, so it has been changed.
4824 13. The install script calls Exim with "-C /dev/null" in order to find the
4825 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4826 to be output. Howeve, since Exim outputs its version number before the
4827 error, it didn't break the script. It just looked ugly. I fixed this by
4828 always allowing "-C /dev/null" if the caller is root.
4830 14. Ignore overlarge ACL variable number when reading spool file - insurance
4831 against a later release with more variables having written the file.
4833 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4834 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4835 this, as well as the form without the "IPv6" on the front.
4837 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4838 OS/Makefile-Darwin file.
4840 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4841 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4843 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4844 variable. However, some OS use ints for inodes. I've added cast to long int
4845 to get rid of the compiler warning.
4847 19. I had forgotten to lock out "/../" in configuration file names when
4848 ALT_CONFIG_PREFIX was set.
4850 20. Routers used for verification do not need to specify transports. However,
4851 if such a router generated a host list, and callout was configured, Exim
4852 crashed, because it could not find a port number from the (non-existent)
4853 transport. It now assumes port 25 in this circumstance.
4855 21. Added the -t option to exigrep.
4857 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4858 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4859 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4860 them is set, LOOKUP_LSEARCH is forced.
4862 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4864 24. Added sender and host information to the "rejected by local_scan()" log
4865 line; previously there was no indication of these.
4867 25. Added .include_if_exists.
4869 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4870 Linux. It turns out that not all file systems support this. Apparently some
4871 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4872 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4873 not supported on the file descriptor, is now ignored when Exim is trying to
4874 sync a directory. This applies only to Linux.
4876 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4878 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4879 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4882 29. Added support for maildirsize files from supplied patch (modified a bit).
4884 30. The use of :fail: followed by an empty string could lead Exim to respond to
4885 sender verification failures with (e.g.):
4887 550 Verification failed for <xxx>
4888 550 Sender verify failed
4890 where the first response line was missing the '-' that indicates it is not
4891 the final line of the response.
4893 31. The loop for finding the name of the user that called Exim had a hardwired
4894 limit of 10; it now uses the value of finduser_retries, which is used for
4895 all other user lookups.
4897 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4900 33. Exim was neglecting to zero errno before one call of strtol() when
4901 expanding a string and expecting an integer value. On some systems this
4902 resulted in spurious "integer overflow" errors. Also, it was casting the
4903 result into an int without checking.
4905 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4906 did not work. The code looks as if it has *never* worked, though it appears
4907 to have been documented since at least releast 1.62. I have made it work.
4909 35. The "timeout_DNS" error in retry rules, also documented since at least
4910 1.62, also never worked. As it isn't clear exactly what this means, and
4911 clearly it isn't a major issue, I have abolished the feature by treating it
4912 as "timeout", and writing a warning to the main and panic logs.
4914 36. The display of retry rules for -brt wasn't always showing the error code
4917 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4918 timeout_connect_A, timeout_connect_MX.
4920 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4921 to the empty sender.
4923 39. The daemon was not analysing the content of -oX till after it had closed
4924 stderr and disconnected from the controlling terminal. This meant that any
4925 syntax errors were only noted on the panic log, and the return code from
4926 the command was 0. By re-arranging the code a little, I've made the
4927 decoding happen first, so such errors now appear on stderr, and the return
4928 code is 1. However, the actual setting up of the sockets still happens in
4929 the disconnected process, so errors there are still only recorded on the
4932 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4933 connections (as happens on some IP stacks) was logged at start up time as
4934 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4935 it from "IPv6 and IPv4", which means that two separate sockets are being
4938 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4939 says whether AF_INET or AF_INET6 was passed as an argument.
4941 42. Exiwhat output was messed up when time zones were included in log
4944 43. Exiwhat now gives more information about the daemon's listening ports,
4945 and whether -tls-on-connect was used.
4947 44. The "port" option of the smtp transport is now expanded.
4949 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4950 silently ignored. Now an error message is written to the main and panic
4953 46. There's a new ACL modifier called "logwrite" which writes to a log file
4954 as soon as it is encountered.
4956 47. Added $local_user_uid and $local_user_gid at routing time.
4958 48. Exim crashed when trying to verify a sender address that was being
4961 49. Exim was recognizing only a space character after ".include". It now also
4962 recognizes a tab character.
4964 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4965 extracting the relevant information from the specification. The man page no
4966 longer contains scrambled data for the -d option, and I've added a section
4967 at the front about calling Exim under different names.
4969 51. Added "extra_headers" argument to the "mail" command in filter files.
4971 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4974 53. Installed eximstats 1.29.
4976 54. Added transport_filter_timeout as a generic transport option.
4978 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4979 Cc: header lines. This was required by RFC 822, but it not required by RFC
4982 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4983 incoming messages that did not have them. Now it does so only if the
4984 message originates locally, that is, if there is no associated remote host
4985 address. When Resent- header lines are present, this applies to the Resent-
4986 lines rather than the non-Resent- lines.
4988 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4989 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4991 58. Messages for configuration errors now include the name of the main
4992 configuration file - useful now that there may be more than one file in a
4993 list (.included file names were always shown).
4995 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4996 for those rare installations that do not start the daemon as root or run it
4997 setuid root. I've cut out the call to initgroups() if the daemon is not
5000 60. The Exim user and group can now be bound into the binary as text strings
5001 that are looked up at the start of Exim's processing.
5003 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5005 62. Added $mailstore_basename variable.
5007 63. Installed patch to sieve.c from Michael Haardt.
5009 64. When Exim failed to open the panic log after failing to open the main log,
5010 the original message it was trying to log was written to stderr and debug
5011 output, but if they were not available (the usual case in production), it
5012 was lost. Now it is written to syslog before the two lines that record the
5013 failures to open the logs.
5015 65. Users' Exim filters run in subprocesses under the user's uid. It is
5016 possible for a "deliver" command or an alias in a "personal" command to
5017 provoke an address rewrite. If logging of address rewriting is configured,
5018 this fails because the process is not running as root or exim. There may be
5019 a better way of dealing with this, but for the moment (because 4.30 needs
5020 to be released), I have disabled address rewrite logging when running a
5021 filter in a non-root, non-exim process.
5027 1. The buildconfig auxiliary program wasn't quoting the value set for
5028 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5029 not defined. This bug was masked in 4.22 by the effect that was fixed in
5032 2. Some messages that were rejected after a message id was allocated were
5033 shown as "incomplete" by exigrep. It no longer does this for messages that
5034 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5036 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5037 have allow_domain_literals set, the ID did not get logged in the <= line.
5038 Domain literals are now always recognized in Message-ID: header lines.
5040 4. The first argument for a ${extract expansion item is the key name or field
5041 number. Leading and trailing spaces in this item were not being ignored,
5042 causing some misleading effects.
5044 5. When deliver_drop_privilege was set, single queue runner processes started
5045 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5046 same command in the process it spins off) were not dropping privilege.
5048 6. When the daemon running as "exim" started a queue runner, it always
5049 re-executed Exim in the spun-off process. This is a waste of effort when
5050 deliver_drop_privilege is set. The new process now just calls the
5051 queue-runner function directly.
5057 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5060 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5061 ignored. Though the use of -oP was forcing the writing of a pid file, it
5062 was always written to the default place.
5064 3. If the message "no IP address found for host xxxx" is generated during
5065 incoming verification, it is now followed by identification of the incoming
5066 connection (so you can more easily find what provoked it).
5068 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5070 5. Added some features to "harden" Exim a bit more against certain attacks:
5072 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5073 be put in Local/Makefile. This is like the never_users runtime option,
5074 but it cannot be overridden. The default setting is "root".
5076 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5077 prefix string with which any file named in a -C command line option
5080 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5081 is retained for -C and -D only if the caller of Exim is root. Without
5082 it, the exim user may also use -C and -D and retain privilege.
5084 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5085 command line option is disabled.
5087 6. Macro names set by the -D option must start with an upper case letter, just
5088 like macro names defined in the configuration file.
5090 7. Added "dereference=" facility to LDAP.
5092 8. Two instances of the typo "uknown" in the source files are fixed.
5094 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5095 the Configure-Makefile script screwed up while processing it.
5097 10. Incorporated PCRE 4.4.
5099 11. The SMTP synchronization check was not operating right at the start of an
5100 SMTP session. For example, it could not catch a HELO sent before the client
5101 waited for the greeting. There is now a check for outstanding input at the
5102 point when the greeting is written. Because of the duplex, asynchronous
5103 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5104 way, but not yet received, when the check is performed.
5106 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5107 on TCP/IP sockets, because this apparently causes some broken clients to
5110 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5111 unchanged) from the Cygwin maintainer.
5113 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5114 was supported instead of testing for mbx. Effectively a typo.
5116 15. The spa authenticator server code was not checking that the input it
5117 received was valid base64.
5119 16. The debug output line for the "set" modifier in ACLs was not showing the
5120 name of the variable that was being set.
5122 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5124 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5125 lookup on its own. Something else was needed to trigger the lookup. For
5126 example, a match in host_lookup or the need for a host name in a host list.
5127 Now, if $sender_host_name is referenced and the host name has not yet been
5128 looked up, a lookup is performed. If the lookup fails, the variable remains
5129 empty, and $host_lookup_failed is set to "1".
5131 19. Added "eqi" as a case-independent comparison operator.
5133 20. The saslauthd authentication condition could segfault if neither service
5134 nor realm was specified.
5136 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5137 error message that was logged was misleading, and incoming SMTP
5138 connections were dropped. The message is now more accurate, and temporary
5139 errors are given to SMTP connections.
5141 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5142 (except RSET) with a 421 error, until QUIT is received. However, it was
5143 failing to send a response to QUIT.
5145 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5146 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5147 ACL is now run after the helo_try_verify_hosts code.
5149 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5150 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5151 case-independent, so other case variants are also recognized). Apparently
5152 some systems use these upper case variants.
5154 25. If more than two messages were waiting for the same host, and a transport
5155 filter was specified for the transport, Exim sent two messages over the
5156 same TCP/IP connection, and then failed with "socket operation on non-
5157 socket" when it tried to send the third.
5159 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5161 27. The extern definition of crypt16() in expand.c was not being excluded when
5162 the OS had its own crypt16() function.
5164 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5165 as a preferred synonym for return_size_limit, both as an option and as an
5168 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5170 30. Changed the default configuration ACL to relax the local part checking rule
5171 for addresses that are not in any local domains. For these addresses,
5172 slashes and pipe symbols are allowed within local parts, but the sequence
5173 /../ is explicitly forbidden.
5175 31. SPA server authentication was not clearing the challenge buffer before
5178 32. log_message in a "warn" ACL statement was writing to the reject log as
5179 well as to the main log, which contradicts the documentation and doesn't
5180 seem right (because no rejection is happening). So I have stopped it.
5182 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5183 However, I am unable to do any testing of this.
5185 34. Fixed an infelicity in the appendfile transport. When checking directories
5186 for a mailbox, to see if any needed to be created, it was accidentally
5187 using path names with one or more superfluous leading slashes; tracing
5188 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5190 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5191 contents are added to the log line that is written for every discarded
5192 recipient. (Previously a log_message setting was ignored.)
5194 36. The ${quote: operator now quotes the string if it is empty.
5196 37. The install script runs exim in order to find its version number. If for
5197 some reason other than non-existence or emptiness, which it checks, it
5198 could not run './exim', it was installing it with an empty version number,
5199 i.e. as "exim-". This error state is now caught, and the installation is
5202 38. An argument was missing from the function that creates an error message
5203 when Exim fails to connect to the socket for saslauthd authentication.
5204 This could cause Exim to crash, or give a corrupted message.
5206 39. Added isip, isip4, and isip6 to ${if conditions.
5208 40. The ACL variables $acl_xx are now saved with the message, and can be
5209 accessed later in routers, transports, and filters.
5211 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5212 strings in the file are not string-expanded.
5214 42. If a MAIL command specified a SIZE value that was too large to fit into an
5215 int variable, the check against message_size_limit failed. Such values are
5216 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5217 day this will have to be increased, but I don't think I want to be around
5218 when emails are that large.
5225 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5226 iconv() is not standard in FreeBSD.
5228 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5229 IPv6 enabled. The observed symptom was a segmentation fault on return from
5230 the function os_common_find_running_interfaces() in src/os.c.
5232 3. In the check_special_case() function in daemon.c I had used "errno" as an
5233 argument name, which causes warnings on some systems. This was basically a
5234 typo, since it was named "eno" in the comments!
5236 4. The code that waits for the clock to tick (at a resolution of some fraction
5237 of a second) so as to ensure message-id uniqueness was always waiting for
5238 at least one whole tick, when it could have waited for less. [This is
5239 almost certainly not relevant at current processor speeds, where it is
5240 unlikely to ever wait at all. But we try to future-proof.]
5242 5. The function that sleeps for a time interval that includes fractions of a
5243 second contained a race. It did not block SIGALRM between setting the
5244 timer, and suspending (a couple of lines later). If the interval was short
5245 and the sigsuspend() was delayed until after it had expired, the suspension
5246 never ended. On busy systems this could lead to processes getting stuck for
5249 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5250 process, before it forks any delivery processes. The open lookup caching
5251 mechanism meant that the open file or database connection was passed into
5252 the delivery process. The problem was that delivery processes always tidy
5253 up cached lookup data. This could cause a problem for the next delivery
5254 process started by the queue runner, because the external queue runner
5255 process does not know about the closure. So the next delivery process
5256 still has data in the lookup cache. In the case of a file lookup, there was
5257 no problem because closing a file descriptor in a subprocess doesn't affect
5258 the parent. However, if the lookup was caching a connection to a database,
5259 the connection was closed, and the second delivery process was likely to
5260 see errors such as "PGSQL: query failed: server closed the connection
5261 unexpectedly". The problem has been fixed by closing all cached lookups
5262 in a queue runner before running a delivery process.
5264 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5265 seem to have the "const" qualifier which it has on other OS. I've
5268 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5269 *of the same type* (client or server) with the same public name that an
5270 error should be diagnosed.
5272 9. When Exim looked up a host name for an IP address, but failed to find the
5273 original IP address when looking up the host name (a safety check), it
5274 output the message "<ip address> does not match any IP for NULL", which was
5275 confusing, to say the least. The bug was that the host name should have
5276 appeared instead of "NULL".
5278 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5279 user that is built into the binary, and the -C or -D options is used, root
5280 privilege is dropped before the configuration file is read. In addition,
5281 logging is switched to stderr instead of the normal log files. If the
5282 configuration then re-defines the Exim user, the unprivileged environment
5283 is probably not what is expected, so Exim logs a panic warning message (but
5286 However, if deliver_drop_privilege is set, the unprivileged state may well
5287 be exactly what is intended, so the warning has been cut out in that case,
5288 and Exim is allowed to try to write to its normal log files.
5294 1. smtp_return_error_details was not giving details for temporary sender
5295 or receiver verification errors.
5297 2. Diagnose a configuration error if two authenticators have the same public
5300 3. Exim used not to create the message log file for a message until the first
5301 delivery attempt. This could be confusing when incoming messages were held
5302 for policy or load reasons. The message log file is now created at the time
5303 the message is received, and an initial "Received" line is written to it.
5305 4. The automatically generated man page for command line options had a minor
5306 bug that caused no ill effects; however, a more serious problem was that
5307 the procedure for building the man page automatically didn't always
5308 operate. Consequently, release 4.20 contains an out-of-date version. This
5309 shouldn't happen again.
5311 5. When building Exim with embedded Perl support, the script that builds the
5312 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5313 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5315 6. The freeze_tell option was not being used for messages that were frozen on
5316 arrival, either by an ACL or by local_scan().
5318 7. Added the smtp_incomplete_transaction log selector.
5320 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5321 was accepting AUTH without a new EHLO.
5323 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5324 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5327 10. Logging of TCP/IP connections (when configured) now happens in the main
5328 daemon process instead of the child process, so that the TCP/IP connection
5329 count is more accurate (but it can never be perfect).
5331 11. The use of "drop" in a nested ACL was not being handled correctly in the
5332 outer ACL. Now, if condition failure induced by the nested "drop" causes
5333 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5334 or "require"), the connection is dropped.
5336 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5337 that yield "discard" can now be used with an "accept" or a "discard" verb,
5338 but an error is generated for any others (because I can't see a useful way
5339 to define what should happen).
5341 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5342 are now processed in the same way as lines in the Exim configuration file.
5343 In particular, continuation lines are supported.
5345 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5347 15. Added -ti meaning -t -i.
5349 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5350 domains, and warn by logging if others are found.
5352 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5353 returned from GIFCONF when Exim is trying to find the list of interfaces on
5354 a host. The code in os.c has been modified to copy each ifreq to an aligned
5355 structure in all cases.
5357 Also, in some cases, the returned ifreq's were being copied to a 'struct
5358 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5359 means the last couple of bytes of an IPv6 address could be chopped if the
5360 ifreq contained only a normal sockaddr (14 bytes storage).
5362 18. Named domain lists were not supported in the hosts_treat_as_local option.
5363 An entry such as +xxxx was not recognized, and was treated as a literal
5366 19. Ensure that header lines added by a DATA ACL are included in the reject log
5367 if the ACL subsequently rejects the message.
5369 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5370 MD5 (which is deprecated).
5372 21. When testing a filter file using -bf, Exim was writing a message when it
5373 took the sender from a "From " line in the message, but it was not doing so
5374 when it took $return_path from a Return-Path: header line. It now does.
5376 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5377 with a valid header line field name (a series of printing characters
5378 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5380 23. Changed "disc" in the source to "disk" to conform to the documentation and
5381 the book and for uniformity.
5383 24. Ignore Sendmail's -Ooption=value command line item.
5385 25. When execve() failed while trying to run a command in a pipe transport,
5386 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5387 could be confused with a return value of 69 from the command itself. This
5388 has been changed to 127, the value the shell returns if it is asked to run
5389 a non-existent command. The wording for the related log line suggests a
5390 non-existent command as the problem.
5392 26. If received_header_text expands to an empty string, do not add a Received:
5393 header line to the message. (Well, it adds a token one on the spool, but
5394 marks it "old" so that it doesn't get used or transmitted.)
5396 27. Installed eximstats 1.28 (addition of -nt option).
5398 28. There was no check for failure on the call to getsockname() in the daemon
5399 code. This can fail if there is a shortage of resources on the system, with
5400 ENOMEM, for example. A temporary error is now given on failure.
5402 29. Contrary to the C standard, it seems that in some environments, the
5403 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5404 program. Exim now does this explicitly; it affects the formatting of
5405 timestamps using strftime().
5407 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5408 complaints. I've now initialized all the variables, to avoid this.
5410 32. Header lines added by a system filter were not being "seen" during
5411 transport-time rewrites.
5413 33. The info_callback() function passed to OpenSSL is set up with type void
5414 (*)(SSL *, int, int), as described somewhere. However, when calling the
5415 function (actually a macro) that sets it up, the type void(*)() is
5416 expected. I've put in a cast to prevent warnings from picky compilers.
5418 34. If a DNS black list lookup found a CNAME record, but there were no A
5419 records associated with the domain it pointed at, Exim crashed.
5421 35. If a DNS black list lookup returned more than one A record, Exim ignored
5422 all but the first. It now scans all returned addresses if a particular IP
5423 value is being sought. In this situation, the contents of the
5424 $dnslist_value variable are a list of all the addresses, separated by a
5427 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5428 to accept a host name and all its aliases if the forward lookup for any of
5429 them yielded the IP address of the incoming connection. Now it accepts only
5430 those names whose forward lookup yields the correct IP address. Any other
5431 names are discarded. This closes a loophole whereby a rogue DNS
5432 administrator could create reverse DNS records to break through a
5433 wildcarded host restriction in an ACL.
5435 37. If a user filter or a system filter that ran in a subprocess used any of
5436 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5437 the wrong values were passed to the pipe command ($thisaddress had the
5438 value of $0, $0 had the value of $1, etc). This bug was introduced by
5439 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5441 38. Improved the line breaking for long SMTP error messages from ACLs.
5442 Previously, if there was no break point between 40 and 75 characters, Exim
5443 left the rest of the message alone. Two changes have been made: (a) I've
5444 reduced the minimum length to 35 characters; (b) if it can't find a break
5445 point between 35 and 75 characters, it looks ahead and uses the first one
5446 that it finds. This may give the occasional overlong line, but at least the
5447 remaining text gets split now.
5449 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5450 file descriptors might be low, and that setting 1000 would always raise it.
5451 It turns out that in some environments, the limit is already over 1000 and
5452 that lowering it causes trouble. So now Exim takes care not to decrease it.
5454 40. When delivering a message, the value of $return_path is set to $sender_
5455 address at the start of routing (routers may change the value). By an
5456 oversight, this default was not being set up when an address was tested by
5457 -bt or -bv, which affected the outcome if any router or filter referred to
5460 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5461 log only when "message" or "log_message" are set. However, if one of the
5462 conditions was an address verification, or a call to a nested ACL, the
5463 messages generated by the underlying test were being passed through. This
5464 no longer happens. The underlying message is available in $acl_verify_
5465 message for both "message" and "log_message" expansions, so it can be
5466 passed through if needed.
5468 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5469 new expansion $bh_ to give the encoded byte string without charset
5470 translation. Translation happens only if iconv() is available; HAVE_ICONV
5471 indicates this at build time. HEADERS_CHARSET gives the charset to
5472 translate to; headers_charset can change it in the configuration, and
5473 "headers charset" can change it in an individual filter file.
5475 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5476 that creates RFC 2047 encoded "words" labels them as that charset instead
5477 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5478 expansion operator; (ii) when Exim creates a From: line for a local
5479 message; (iii) when a header line is rewritten to include a "phrase" part.
5481 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5482 buggy, causing it to skip the first lines of messages whose message ID
5483 ended in 'D'. This would not have bitten before Exim release 4.14, because
5484 message IDs were unlikely to end in 'D' before then. The effect was to have
5485 incorrect size information for certain domains.
5487 45. #include "config.h" was missing at the start of the crypt16.c module. This
5488 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5491 46. If there was a timeout during a "random" callout check, Exim treated it as
5492 a failure of the random address, and carried on sending RSET and the real
5493 address. If the delay was just some slowness somewhere, the response to the
5494 original RCPT would be taken as a response to RSET and so on, causing
5495 mayhem of various kinds.
5497 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5498 when I implemented it. It didn't allow for the fact that some option values
5499 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5500 the right test for positive values.
5502 48. Domain names in DNS records are case-independent. Exim always looks them up
5503 in lower case. Some resolvers return domain names in exactly the case they
5504 appear in the zone file, that is, they may contain uppercase letters. Not
5505 all resolvers do this - some return always lower case. Exim was treating a
5506 change of case by a resolver as a change of domain, similar to a widening
5507 of a domain abbreviation. This triggered its re-routing code and so it was
5508 trying to route what was effectively the same domain again. This normally
5509 caused routing to fail (because the router wouldn't handle the domain
5510 twice). Now Exim checks for this case specially, and just changes the
5511 casing of the domain that it ultimately uses when it transmits the message
5514 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5517 50. If a filter generated a file delivery with a non-absolute name (possible if
5518 no home directory exists for the router), the forbid_file option was not
5521 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5522 the existing equality matching.
5524 52. Exim was using ints instead of ino_t variables in some places where it was
5525 dealing with inode numbers.
5527 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5528 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5529 called TMPDIR, and if it finds it is different, it changes its value.
5531 54. The smtp_printf() function is now made available to local_scan() so
5532 additional output lines can be written before returning. There is also an
5533 smtp_fflush() function to enable the detection of a dropped connection.
5534 The variables smtp_input and smtp_batched_input are exported to
5537 55. Changed the default runtime configuration: the message "Unknown user"
5538 has been removed from the ACL, and instead placed on the localuser router,
5539 using the cannot_route_message feature. This means that any verification
5540 failures that generate their own messages won't get overridden. Similarly,
5541 the "Unrouteable address" message that was in the ACL for unverifiable
5542 relay addresses has also been removed.
5544 56. Added hosts_avoid_esmtp to the smtp transport.
5546 57. The exicyclog script was not checking for the esoteric option
5547 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5548 will work only if exicyclog is run under the appropriate euid.
5550 58. Following a discussion on the list, the rules by which Exim recognises line
5551 endings on incoming messages have been changed. The -dropcr and drop_cr
5552 options are now no-ops, retained only for backwards compatibility. The
5553 following line terminators are recognized: LF CRLF CR. However, special
5554 processing applies to CR:
5556 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5557 nor a local message in the state where . is a terminator.
5559 (ii) If a bare CR is encountered in a header line, an extra space is added
5560 after the line terminator so as not to end the header. The reasoning
5561 behind this is that bare CRs in header lines are most likely either
5562 to be mistakes, or people trying to play silly games.
5564 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5565 was being incorrectly given as 18 bytes larger than it should have been.
5566 This is a VOB (very old bug).
5568 60. This may never have affected anything current, but just in case it has:
5569 When the local host is found other than at the start of a list of hosts,
5570 the local host, those with the same MX, and any that follow, are discarded.
5571 When the list in question was part of a longer list of hosts, the following
5572 hosts (not currently being processed) were also being discarded. This no
5573 longer happens. I'm not sure if this situation could ever has previously
5576 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5579 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5580 header. This is something that is recommended in a new Internet Draft, and
5581 is something that is documented as being done by Sendmail. There are two
5582 possible values. For messages generated by the autoreply transport, Exim
5585 Auto-Submitted: auto-replied
5587 whereas for all other generated messages (e.g. bounces) it adds
5589 Auto-Submitted: auto-generated
5591 63. The "personal" condition in filters now includes a test for the
5592 Auto-Submitted: header. If it contains the string "auto-" the message it
5593 not considered personal.
5595 64. Added rcpt_include_affixes as a generic transport option.
5597 65. Added queue_only_override (default true).
5599 66. Added the syslog_duplication option.
5601 67. If what should have been the first header line of a message consisted of
5602 a space followed by a colon, Exim was mis-interpreting it as a header line.
5603 It isn't of course - it is syntactically invalid and should therefore be
5604 treated as the start of the message body. The misbehaviour could have
5605 caused a number of strange effects, including loss of data in subsequent
5606 header lines, and spool format errors.
5608 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5609 client host had authenticated. This control can now be exercised by an ACL
5610 for more flexibility.
5612 69. By default, callouts do not happen when testing with -bh. There is now a
5613 variant, -bhc, which does actually run the callout code, including
5614 consulting and updating the callout cache.
5616 70. Added support for saslauthd authentication, courtesy of Alexander
5619 71. If statvfs() failed on the spool or log directories while checking their
5620 size for availability, Exim confusingly gave the error "space shortage".
5621 Furthermore, in debugging mode it crashed with a floating point exception.
5622 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5623 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5624 serious problem, Exim now writes to the main and panic logs when this
5625 happens, with details of the failure. It then refuses to accept the
5626 incoming message, giving the message "spool directory problem" or "log
5627 directory problem" with a 421 code for SMTP messages.
5629 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5630 0, 1, and 2 exist, because some OS complain for execs without them (see
5631 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5632 descriptors. However, the code omitted to check that the open succeeded,
5633 causing mysterious errors if for some reason the permissions on /dev/null
5634 got screwed. Now Exim writes a message to the main and panic logs, and
5635 bombs out if it can't open /dev/null.
5637 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5638 interact so that it is all more flexible. It is supposed to remain
5639 backwards compatible. Also added extra_local_interfaces.
5641 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5642 to bomb out with an assertion failure - to the client this appears as a
5643 connection drop. This problem occurs in the part of the code that was taken
5644 from the Samba project. Fortunately, the assertion is in a very simple
5645 function, so I have fixed this by reproducing the function inline in the
5646 one place where it is called, and arranging for authentication to fail
5647 instead of killing the process with assert().
5649 75. The SPA client code was not working when the server requested OEM rather
5650 than Unicode encoding.
5652 76. Added code to make require_files with a specific uid setting more usable in
5653 the case where statting the file as root fails - usually a non-root-mounted
5654 NFS file system. When this happens and the failure is EACCES, Exim now
5655 forks a subprocess and does the per-uid checking as the relevant uid.
5657 77. Added process_log_path.
5659 78. If log_file_path was not explicitly set, a setting of check_log_space or
5660 check_log_inodes was ignored.
5662 79. If a space check for the spool or log partitions fails, the incident is now
5663 logged. Of course, in the latter case the data may get lost...
5665 80. Added the %p formatting code to string_format() so that it can be used to
5666 print addresses in debug_print(). Adjusted all the address printing in the
5667 debugging in store.c to use %p rather than %d.
5669 81. There was a concern that a line of code in smtp_in.c could overflow a
5670 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5671 initially expressed, the concern was not well-founded, because trailing
5672 spaces are removed early. However, if the trailing spaces were followed by
5673 a NULL, they did not get removed, so the overflow was possible. Two fixes
5676 (a) I re-wrote the offending code in a cleaner fashion.
5677 (b) If an incoming SMTP command contains a NULL character, it is rejected
5680 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5681 runs initgroups(), so that if the Exim user is in any additional groups,
5682 they will be used during message reception.
5688 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob