1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
9 PP/02 Make -n do something, by making it not do something.
10 When combined with -bP, the name of an option is not output.
12 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
15 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
16 $sender_host_name and config options to manage this, and basic check
19 PP/05 DSCP support for outbound connections and control modifier for inbound.
21 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
22 (Only plugin which currently uses this is kerberos4, which nobody should
23 be using, but we should make it available and other future plugins might
24 conceivably use it, even though it would break NAT; stuff *should* be
25 using channel bindings instead).
27 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
28 name; added for Sendmail compatibility; requires admin caller.
29 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
30 just ignore it); requires trusted caller.
31 Also parse but ignore: -Ac -Am -X<logfile>
34 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
36 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
38 JH/02 Support "G" suffix to numbers in ${if comparisons.
40 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
42 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
48 PP/01 Handle short writes when writing local log-files.
49 In practice, only affects FreeBSD (8 onwards).
50 Bugzilla 1053, with thanks to Dmitry Isaikin.
52 NM/01 Bugzilla 949 - Documentation tweak
54 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
57 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
59 PP/02 Implemented gsasl authenticator.
61 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
63 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
64 `pkg-config foo` for cflags/libs.
66 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
67 with rest of GSASL and with heimdal_gssapi.
69 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
70 `pkg-config foo` for cflags/libs for the TLS implementation.
72 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
73 properties get this fed in as external SSF. A number of robustness
74 and debugging improvements to the cyrus_sasl authenticator.
76 PP/08 cyrus_sasl server now expands the server_realm option.
78 PP/09 Bugzilla 1214 - Log authentication information in reject log.
79 Patch by Jeremy Harris.
81 PP/10 Added dbmjz lookup type.
83 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
85 PP/12 MAIL args handles TAB as well as SP, for better interop with
86 non-compliant senders.
87 Analysis and variant patch by Todd Lyons.
89 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
90 Bug report from Lars Müller <lars@samba.org> (via SUSE),
91 Patch from Dirk Mueller <dmueller@suse.com>
93 PP/13 tls_peerdn now print-escaped for spool files.
94 Observed some $tls_peerdn in wild which contained \n, which resulted
95 in spool file corruption.
97 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
98 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
99 or write after TLS renegotiation, which otherwise led to messages
102 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
103 as a tracking header (ie: a signed header comes before the signature).
104 Patch from Wolfgang Breyha.
106 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
107 comma-sep list; embedded commas doubled.
109 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
111 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
113 Report and patch from Dmitry Banschikov.
115 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
116 Removed SSL_clear() after SSL_new() which led to protocol negotiation
117 failures. We appear to now support TLS1.1+ with Exim.
119 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
120 lets Exim select keys and certificates based upon TLS SNI from client.
121 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
122 before an outbound SMTP session. New log_selector, +tls_sni.
124 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
125 NULL dereference. Report and patch from Alun Jones.
127 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
128 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
129 Not seeing resolver debug output on NetBSD, but suspect this is a
130 resolver implementation change.
132 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
133 Left warnings. Added "eximon gdb" invocation mode.
135 PP/21 Defaulting "accept_8bitmime" to true, not false.
137 PP/22 Added -bw for inetd wait mode support.
139 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
140 locate the relevant includes and libraries. Made this the default.
142 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
143 Bugzilla 1246, report and most of solution from Tomasz Kusy.
145 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
146 This may cause build issues on older platforms.
148 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
149 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
150 gnutls_require_mac & gnutls_require_protocols (no longer supported).
151 Added SNI support via GnuTLS too.
152 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
154 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
156 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
157 Applied second patch from Janne, implementing suggestion to default
158 multiple-strings-in-record handling to match SPF spec.
160 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
162 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
163 read-only, out of scope).
164 Patch from Wolfgang Breyha, report from Stuart Northfield.
166 PP/29 Fix three issues highlighted by clang analyser static analysis.
167 Only crash-plausible issue would require the Cambridge-specific
168 iplookup router and a misconfiguration.
169 Report from Marcin Mirosław.
171 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
173 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
174 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
175 As part of this, removing so much warning spew let me fix some minor
176 real issues in debug logging.
178 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
179 assignment on my part. Fixed.
181 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
182 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
183 Janne Snabb (who went above and beyond: thank you).
185 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
186 string otherwise requires a connection and a bunch more work and it's
187 relatively easy to get wrong. Should also expose TLS library linkage
190 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
191 64-bit ${eval} (JH/03).
193 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
194 GNU libc to support some of the 64-bit stuff, should not lead to
195 conflicts. Defined before os.h is pulled in, so if a given platform
196 needs to override this, it can.
198 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
199 protection layer was required, which is not implemented.
200 Bugzilla 1254, patch from Wolfgang Breyha.
202 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
203 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
204 tls_dhparam take prime identifiers. Also unbreak combination of
205 OpenSSL+DH_params+TLSSNI.
207 PP/39 Disable SSLv2 by default in OpenSSL support.
213 PP/01 Solaris build fix for Oracle's LDAP libraries.
214 Bugzilla 1109, patch from Stephen Usher.
216 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
218 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
221 TF/02 Fix a couple more cases where we did not log the error message
222 when unlink() failed. See also change 4.74-TF/03.
224 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
225 lock up or crash if it happened to be inside a call to libc when it
226 got a SIGUSR1 from exiwhat.
228 The SIGUSR1 handler appends the current process status to the process
229 log which is later printed by exiwhat. It used to use the general
230 purpose logging code to do this, but several functions it calls are
231 not safe for signals.
233 The new output code in the SIGUSR1 handler is specific to the process
234 log, and simple enough that it's easy to inspect for signal safety.
235 Removing some special cases also simplifies the general logging code.
236 Removing the spurious timestamps from the process log simplifies
239 TF/04 Improved ratelimit ACL condition.
241 The /noupdate option has been deprecated in favour of /readonly which
242 has clearer semantics. The /leaky, /strict, and /readonly update modes
243 are mutually exclusive. The update mode is no longer included in the
244 database key; it just determines when the database is updated. (This
245 means that when you upgrde Exim will forget old rate measurements.)
247 Exim now checks that the per_* options are used with an update mode that
248 makes sense for the current ACL. For example, when Exim is processing a
249 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
250 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
251 must specify per_mail/readonly. If you omit the update mode it defaults to
252 /leaky where that makes sense (as before) or /readonly where required.
254 The /noupdate option is now undocumented but still supported for
255 backwards compatibility. It is equivalent to /readonly except that in
256 ACLs where /readonly is required you may specify /leaky/noupdate or
257 /strict/noupdate which are treated the same as /readonly.
259 A useful new feature is the /count= option. This is a generalization
260 of the per_byte option, so that you can measure the throughput of other
261 aggregate values. For example, the per_byte option is now equivalent
262 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
264 The per_rcpt option has been generalized using the /count= mechanism
265 (though it's more complicated than the per_byte equivalence). When it is
266 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
267 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
268 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
269 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
270 Note that using per_rcpt with a non-readonly update mode in more than
271 one ACL will cause the recipients to be double-counted. (The per_mail
272 and per_byte options don't have this problem.)
274 The handling of very low rates has changed slightly. If the computed rate
275 is less than the event's count (usually one) then this event is the first
276 after a long gap. In this case the rate is set to the same as this event's
277 count, so that the first message of a spam run is counted properly.
279 The major new feature is a mechanism for counting the rate of unique
280 events. The new per_addr option counts the number of different
281 recipients that someone has sent messages to in the last time period. It
282 behaves like per_rcpt if all the recipient addresses are different, but
283 duplicate recipient addresses do not increase the measured rate. Like
284 the /count= option this is a general mechanism, so the per_addr option
285 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
286 example, measure the rate that a client uses different sender addresses
287 with the options per_mail/unique=$sender_address. There are further
288 details in the main documentation.
290 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
292 TF/06 Removed a few PCRE remnants.
294 TF/07 Automatically extract Exim's version number from tags in the git
295 repository when doing development or release builds.
297 PP/02 Raise smtp_cmd_buffer_size to 16kB.
298 Bugzilla 879. Patch from Paul Fisher.
300 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
301 Heavily based on revision 40f9a89a from Simon Arlott's tree.
304 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
306 PP/05 Variable $av_failed, true if the AV scanner deferred.
307 Bugzilla 1078. Patch from John Horne.
309 PP/06 Stop make process more reliably on build failure.
310 Bugzilla 1087. Patch from Heiko Schlittermann.
312 PP/07 Make maildir_use_size_file an _expandable_ boolean.
313 Bugzilla 1089. Patch from Heiko Schlittermann.
315 PP/08 Handle ${run} returning more data than OS pipe buffer size.
316 Bugzilla 1131. Patch from Holger Weiß.
318 PP/09 Handle IPv6 addresses with SPF.
319 Bugzilla 860. Patch from Wolfgang Breyha.
321 PP/10 GnuTLS: support TLS 1.2 & 1.1.
323 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
326 PP/11 match_* no longer expand right-hand-side by default.
327 New compile-time build option, EXPAND_LISTMATCH_RHS.
328 New expansion conditions, "inlist", "inlisti".
330 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
332 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
334 PP/14 fix log_write() format string regression from TF/03.
335 Bugzilla 1152. Patch from Dmitry Isaikin.
341 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
343 PP/02 Harmonised TLS library version reporting; only show if debugging.
344 Layout now matches that introduced for other libraries in 4.74 PP/03.
346 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
348 PP/04 New "dns_use_edns0" global option.
350 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
353 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
354 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
356 TK/01 Updated PolarSSL code to 0.14.2.
357 Bugzilla 1097. Patch from Andreas Metzler.
359 PP/07 Catch divide-by-zero in ${eval:...}.
362 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
365 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
366 format-string attack -- SECURITY: remote arbitrary code execution.
368 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
369 time unintentionally subject to list matching rules, letting the header
370 cause arbitrary Exim lookups (of items which can occur in lists, *not*
371 arbitrary string expansion). This allowed for information disclosure.
373 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
374 INT_MIN/-1 -- value coerced to INT_MAX.
380 NM/01 Workround for PCRE version dependency in version reporting
383 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
384 This fixes portability to compilers other than gcc, notably
385 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
387 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
388 makefiles for portability to HP-UX and POSIX correctness.
390 PP/01 Permit LOOKUP_foo enabling on the make command-line.
391 Also via indented variable definition in the Makefile.
392 (Debugging by Oliver Heesakkers).
394 PP/02 Restore caching of spamd results with expanded spamd_address.
395 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
397 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
398 Improves build reliability. Fix from: Frank Elsner
400 NM/02 Fix wide character breakage in the rfc2047 coding
401 Fixes bug 1064. Patch from Andrey N. Oktyabrski
403 NM/03 Allow underscore in dnslist lookups
404 Fixes bug 1026. Patch from Graeme Fowler
406 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
407 Code patches from Adam Ciarcinski of NetBSD.
409 NM/04 Fixed exiqgrep to cope with mailq missing size issue
412 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
413 is logged, to avoid truncation. Patch from John Horne.
415 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
416 Patch from Jakob Hirsch.
418 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
419 SQL string expansion failure details.
420 Patch from Andrey Oktyabrski.
422 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
423 Patch from Simon Arlott.
425 PP/09 New lookups functionality failed to compile on old gcc which rejects
426 extern declarations in function scope.
427 Patch from Oliver Fleischmann
429 PP/10 Use sig_atomic_t for flags set from signal handlers.
430 Check getgroups() return and improve debugging.
431 Fixed developed for diagnosis in bug 927 (which turned out to be
434 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
435 Patch from Mark Zealey.
437 PP/12 Bugzilla 1056: Improved spamd server selection.
438 Patch from Mark Zealey.
440 PP/13 Bugzilla 1086: Deal with maildir quota file races.
441 Based on patch from Heiko Schlittermann.
443 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
444 Patch from Uwe Doering, sign-off by Michael Haardt.
446 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
447 variable declaration deep within a block. Bug and patch from
450 PP/15 lookups-Makefile IRIX compatibilty coercion.
452 PP/16 Make DISABLE_DKIM build knob functional.
454 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
455 Patch by Simon Arlott
457 TF/03 Fix valgrind.h portability to C89 compilers that do not support
458 variable argument macros. Our copy now differs from upstream.
464 TF/01 Failure to get a lock on a hints database can have serious
465 consequences so log it to the panic log.
467 TF/02 Log LMTP confirmation messages in the same way as SMTP,
468 controlled using the smtp_confirmation log selector.
470 TF/03 Include the error message when we fail to unlink a spool file.
472 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
473 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
474 for maintaining out-of-tree patches for some time.
476 PP/01 Bugzilla 139: Documentation and portability issues.
477 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
478 Handle per-OS dynamic-module compilation flags.
480 PP/02 Let /dev/null have normal permissions.
481 The 4.73 fixes were a little too stringent and complained about the
482 permissions on /dev/null. Exempt it from some checks.
483 Reported by Andreas M. Kirchwitz.
485 PP/03 Report version information for many libraries, including
486 Exim version information for dynamically loaded libraries. Created
487 version.h, now support a version extension string for distributors
488 who patch heavily. Dynamic module ABI change.
490 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
491 privilege escalation vulnerability whereby the Exim run-time user
492 can cause root to append content of the attacker's choosing to
495 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
498 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
499 If dropping privileges for untrusted macros, we disabled normal logging
500 on the basis that it would fail; for the Exim run-time user, this is not
501 the case, and it resulted in successful deliveries going unlogged.
502 Fixed. Reported by Andreas Metzler.
508 PP/01 Date: & Message-Id: revert to normally being appended to a message,
509 only prepend for the Resent-* case. Fixes regression introduced in
510 Exim 4.70 by NM/22 for Bugzilla 607.
512 PP/02 Include check_rfc2047_length in configure.default because we're seeing
513 increasing numbers of administrators be bitten by this.
515 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
517 PP/03 Bugzilla 994: added openssl_options main configuration option.
519 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
521 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
523 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
525 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
526 without a peer certificate, leading to a segfault because of an
527 assumption that peers always have certificates. Be a little more
528 paranoid. Problem reported by Martin Tscholak.
530 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
531 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
532 NB: ClamAV planning to remove STREAM in "middle of 2010".
533 CL also introduces -bmalware, various -d+acl logging additions and
534 more caution in buffer sizes.
536 PP/09 Implemented reverse_ip expansion operator.
538 PP/10 Bugzilla 937: provide a "debug" ACL control.
540 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
542 PP/12 Bugzilla 973: Implement --version.
544 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
546 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
548 PP/15 Bugzilla 816: support multiple condition rules on Routers.
550 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
551 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
552 ignore trailing whitespace.
554 JJ/02 prevent non-panic DKIM error from being sent to paniclog
556 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
559 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
560 Notification from Dr Andrew Aitchison.
562 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
563 ExtendedDetectionInfo response format.
564 Notification from John Horne.
566 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
569 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
570 XSL and documented dependency on system catalogs, with examples of how
573 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
576 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
577 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
578 configuration file which is writeable by the Exim user or group.
580 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
581 of configuration files to cover files specified with the -C option if
582 they are going to be used with root privileges, not just the default
585 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
586 option (effectively making it always true).
588 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
589 files to be used while preserving root privileges.
591 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
592 that rogue child processes cannot use them.
594 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
595 run-time user, instead of root.
597 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
598 Exim run-time user without dropping privileges.
600 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
601 result string, instead of calling string_vformat() twice with the same
604 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
605 for other users. Others should always drop root privileges if they use
606 -C on the command line, even for a whitelisted configure file.
608 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
610 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
616 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
617 $data_path, and $header_path variables; fixed documentation bugs and
620 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
621 exipick to access non-standard spools, including the "frozen" queue
624 NM/01 Bugzilla 965: Support mysql stored procedures.
625 Patch from Alain Williams
627 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
629 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
630 Patch from Andreas Metzler
632 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
633 Patch from Kirill Miazine
635 NM/05 Bugzilla 671: Added umask to procmail example.
637 JJ/03 installed exipick 20100323.0, fixing doc bug
639 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
640 directory. Notification and patch from Dan Rosenberg.
642 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
644 TK/02 Improve log output when DKIM signing operation fails.
646 MH/01 Treat the transport option dkim_domain as a colon separated
647 list, not as a single string, and sign the message with each element,
648 omitting multiple occurences of the same signer.
650 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
651 Bugzilla 985, 986. Patch by Simon Arlott
653 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
654 Patch by Simon Arlott
656 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
657 MBX locking. Notification from Dan Rosenberg.
663 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
665 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
667 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
669 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
671 NM/04 Bugzilla 915: Flag broken perl installation during build.
677 TK/01 Added patch by Johannes Berg that expands the main option
678 "spamd_address" if it starts with a dollar sign.
680 TK/02 Write list of recipients to X-Envelope-Sender header when building
681 the mbox-format spool file for content scanning (suggested by Jakob
684 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
685 (http://www.dcc-servers.net/) support via dccifd. Activated by
686 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
688 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
689 by Mark Daniel Reidel <mr@df.eu>.
691 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
692 When building exim an external PCRE library is now needed -
693 PCRE is a system library on the majority of modern systems.
694 See entry on PCRE_LIBS in EDITME file.
696 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
697 conversation. Added nologin parameter to request.
698 Patch contributed by Kirill Miazine.
700 TF/01 Do not log submission mode rewrites if they do not change the address.
702 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
704 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
705 log files in place. Contributed by Roberto Lima.
707 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
709 TF/03 Bugzilla 615: When checking the local_parts router precondition
710 after a local_part_suffix or local_part_prefix option, Exim now
711 does not use the address's named list lookup cache, since this
712 contains cached lookups for the whole local part.
714 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
715 Robert Millan. Documentation is in experimental-spec.txt.
717 TF/04 Bugzilla 668: Fix parallel build (make -j).
719 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
721 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
722 Patch provided by Jan Srzednicki.
724 TF/05 Leading white space used to be stripped from $spam_report which
725 wrecked the formatting. Now it is preserved.
727 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
728 that they are available at delivery time.
730 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
732 TF/08 TLS error reporting now respects the incoming_interface and
733 incoming_port log selectors.
735 TF/09 Produce a more useful error message if an SMTP transport's hosts
736 setting expands to an empty string.
738 NM/06 Bugzilla 744: EXPN did not work under TLS.
739 Patch provided by Phil Pennock.
741 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
742 Patch provided by Richard Godbee.
744 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
745 acl_smtp_notquit, added index entry.
747 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
748 Patch provided by Eugene Bujak.
750 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
751 accept(). Patch provided by Maxim Dounin.
753 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
754 Patch provided by Phil Pennock.
756 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
758 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
759 Patch provided by Brad "anomie" Jorsch.
761 NM/14 Bugzilla 622: Added timeout setting to transport filter.
762 Patch provided by Dean Brooks.
764 TK/05 Add native DKIM support (does not depend on external libraries).
766 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
767 Patch provided by Graeme Fowler.
769 NM/16 Bugzilla 851: Documentation example syntax fix.
771 NM/17 Changed NOTICE file to remove references to embedded PCRE.
773 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
776 NM/19 Bugzilla 745: TLS version reporting.
777 Patch provided by Phil Pennock.
779 NM/20 Bugzilla 167: bool: condition support.
780 Patch provided by Phil Pennock.
782 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
783 clients. Patch provided by Phil Pennock.
785 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
786 Patch provided by Brad "anomie" Jorsch.
788 NM/23 Bugzilla 687: Fix misparses in eximstats.
789 Patch provided by Heiko Schlittermann.
791 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
792 Patch provided by Heiko Schlittermann.
794 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
795 plus update to original patch.
797 NM/26 Bugzilla 799: Documentation correction for ratelimit.
799 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
800 Patch provided by David Brownlee.
802 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
804 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
806 NM/30 Bugzilla 888: TLS documentation bugfixes.
808 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
810 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
811 Unlike the original bugzilla I have changed all shell scripts in src tree.
813 NM/33 Bugzilla 898: Transport filter timeout fix.
814 Patch by Todd Rinaldo.
816 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
817 Patch by Serge Demonchaux.
819 NM/35 Bugzilla 39: Base64 decode bug fixes.
820 Patch by Jakob Hirsch.
822 NM/36 Bugzilla 909: Correct connect() call in dcc code.
824 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
826 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
828 NM/39 Bugzilla 911: Fixed MakeLinks build script.
834 TK/01 Add preliminary DKIM support. Currently requires a forked version of
835 ALT-N's libdkim that I have put here:
836 http://duncanthrax.net/exim-experimental/
838 Note to Michael Haardt: I had to rename some vars in sieve.c. They
839 were called 'true' and it seems that C99 defines that as a reserved
840 keyword to be used with 'bool' variable types. That means you could
841 not include C99-style headers which use bools without triggering
842 build errors in sieve.c.
844 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
845 as mailq or other aliases. Changed the --help handling significantly
846 to do whats expected. exim_usage() emits usage/help information.
848 SC/01 Added the -bylocaldomain option to eximstats.
850 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
852 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
854 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
860 PH/01 Another patch from the Sieve maintainer.
862 PH/02 When an IPv6 address is converted to a string for single-key lookup
863 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
864 dots are used instead of colons so that keys in lsearch files need not
865 contain colons. This was done some time before quoting was made available
866 in lsearch files. However, iplsearch files do require colons in IPv6 keys
867 (notated using the quote facility) so as to distinguish them from IPv4
868 keys. This meant that lookups for IP addresses in host lists did not work
869 for iplsearch lookups.
871 This has been fixed by arranging for IPv6 addresses to be expressed with
872 colons if the lookup type is iplsearch. This is not incompatible, because
873 previously such lookups could never work.
875 The situation is now rather anomolous, since one *can* have colons in
876 ordinary lsearch keys. However, making the change in all cases is
877 incompatible and would probably break a number of configurations.
879 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
882 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
883 conversion specification without a maximum field width, thereby enabling
884 a rogue spamd server to cause a buffer overflow. While nobody in their
885 right mind would setup Exim to query an untrusted spamd server, an
886 attacker that gains access to a server running spamd could potentially
887 exploit this vulnerability to run arbitrary code as the Exim user.
889 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
890 $primary_hostname instead of what libspf2 thinks the hosts name is.
892 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
893 a directory entry by the name of the lookup key. Previously, if a
894 symlink pointed to a non-existing file or a file in a directory that
895 Exim lacked permissions to read, a lookup for a key matching that
896 symlink would fail. Now it is enough that a matching directory entry
897 exists, symlink or not. (Bugzilla 503.)
899 PH/03 The body_linecount and body_zerocount variables are now exported in the
902 PH/04 Added the $dnslist_matched variable.
904 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
905 This means they are set thereafter only if the connection becomes
908 PH/06 Added the client_condition to authenticators so that some can be skipped
909 by clients under certain conditions.
911 PH/07 The error message for a badly-placed control=no_multiline_responses left
912 "_responses" off the end of the name.
914 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
916 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
917 (without spaces) instead of just copying the configuration text.
919 PH/10 Added the /noupdate option to the ratelimit ACL condition.
921 PH/11 Added $max_received_linelength.
923 PH/12 Added +ignore_defer and +include_defer to host lists.
925 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
926 way in which PCRE > 7.0 is built.
928 PH/14 Implemented queue_only_load_latch.
930 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
931 MAIL command. The effect was to mangle the value on 64-bit systems.
933 PH/16 Another patch from the Sieve maintainer.
935 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
937 PH/18 If a system quota error occurred while trying to create the file for
938 a maildir delivery, the message "Mailbox is full" was not appended to the
939 bounce if the delivery eventually timed out. Change 4.67/27 below applied
940 only to a quota excession during the actual writing of the file.
942 PH/19 It seems that peer DN values may contain newlines (and other non-printing
943 characters?) which causes problems in log lines. The DN values are now
944 passed through string_printing() before being added to log lines.
946 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
947 and InterBase are left for another time.)
949 PH/21 Added message_body_newlines option.
951 PH/22 Guard against possible overflow in moan_check_errorcopy().
953 PH/23 POSIX allows open() to be a macro; guard against that.
955 PH/24 If the recipient of an error message contained an @ in the local part
956 (suitably quoted, of course), incorrect values were put in $domain and
957 $local_part during the evaluation of errors_copy.
963 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
964 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
967 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
968 issue a MAIL command.
970 PH/02 In an ACL statement such as
972 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
974 if a client was not listed at all, or was listed with a value other than
975 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
976 the condition was not true (as it should be), so access was not denied.
977 The bug was that the ! inversion was incorrectly passed on to the second
978 item. This has been fixed.
980 PH/03 Added additional dnslists conditions == and =& which are different from
981 = and & when the dns lookup returns more than one IP address.
983 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
984 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
986 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
987 FSYNC, which compiles an option called disable_fsync that allows for
988 bypassing fsync(). The documentation is heavily laced with warnings.
990 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
992 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
993 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
994 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
995 including adding "make clean"; (3) Added -fPIC when compiling the test
996 dynamically loaded module, to get rid of a warning.
998 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
999 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1000 = 0s. The bug is otherwise harmless.
1002 PH/07 There was a bug in the dovecot authenticator such that the value of
1003 $auth1 could be overwritten, and so not correctly preserved, after a
1004 successful authentication. This usually meant that the value preserved by
1005 the server_setid option was incorrect.
1007 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1009 PH/09 Installed PCRE release 7.0.
1011 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1012 run for batched SMTP input. It is now run at the start of every message
1013 in the batch. While fixing this I discovered that the process information
1014 (output by running exiwhat) was not always getting set for -bs and -bS
1015 input. This is fixed, and it now also says "batched" for BSMTP.
1017 PH/11 Added control=no_pipelining.
1019 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1020 patch, slightly modified), and move the expansion of helo_data till after
1021 the connection is made in the smtp transport (so it can use these
1024 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1026 PH/14 Added log_selector = +pid.
1028 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1030 PH/16 Add ${if forany and ${if forall.
1032 PH/17 Added dsn_from option to vary the From: line in DSNs.
1034 PH/18 Flush SMTP output before performing a callout, unless control =
1035 no_callout_flush is set.
1037 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1038 was true (the default) a successful delivery failed to delete the retry
1039 item, thus causing premature timeout of the address. The bug is now
1042 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1044 PH/21 Long custom messages for fakedefer and fakereject are now split up
1045 into multiline reponses in the same way that messages for "deny" and
1046 other ACL rejections are.
1048 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1049 with slight modification.
1051 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1052 draft, changing the syntax and factoring some duplicate code".
1054 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1055 for deliveries of the second and subsequent messages over the same SMTP
1058 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1059 ${reduce, with only minor "tidies".
1061 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1063 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1064 expansion side effects.
1066 PH/27 When a message times out after an over-quota error from an Exim-imposed
1067 quota, the bounce message says "mailbox is full". This message was not
1068 being given when it was a system quota that was exceeded. It now should
1071 MH/03 Made $recipients available in local_scan(). local_scan() already has
1072 better access to the recipient list through recipients_list[], but
1073 $recipients can be useful in postmaster-provided expansion strings.
1075 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1076 in the case of a MAIL command with additional options following the
1077 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1078 were accidentally chopped off.
1080 PH/29 SMTP synchronization checks are implemented when a command is read -
1081 there is a check that no more input is waiting when there shouldn't be
1082 any. However, for some commands, a delay in an ACL can mean that it is
1083 some time before the response is written. In this time, more input might
1084 arrive, invalidly. So now there are extra checks after an ACL has run for
1085 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1086 pipelining has not been advertised.
1088 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1090 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1091 returned in the SMTP response when smtp_return_error_details was false.
1092 This has been fixed.
1094 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1095 instead of the C I/O that was originally supplied, because problems were
1096 reported on Solaris.
1098 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1099 Exim which did not show up earlier: it was assuming that a call to
1100 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1101 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1102 was a macro that became an assignment, so it seemed to work. This has
1103 changed to a proper function call with a void return, hence the compile
1104 error. Exim's code has been fixed.
1106 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1109 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1111 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1113 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1114 because it turns out to be too big or there is a timeout) there is no
1115 "Completed" line in the log. When some messages of this type were
1116 selected by exigrep, they were listed as "not completed". Others were
1117 picked up by some special patterns. I have improved the selection
1118 criteria to be more general.
1120 PH/38 The host_find_failed option in the manualroute router can now be set
1121 to "ignore", to completely ignore a host whose IP address cannot be
1122 found. If all hosts are ignored, the behaviour is controlled by the new
1123 host_all_ignored option.
1125 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1126 homing or because of multiple MX records with /mx) generated more than
1127 one IP address, and the following item turned out to be the local host,
1128 all the secondary addresses of the first item were incorrectly removed
1129 from the list, along with the local host and any following hosts (which
1130 is what is supposed to happen).
1132 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1133 whoever called Exim into the -H file. In the case of the daemon it was
1134 behaving confusingly. When first started, it used values for whoever
1135 started the daemon, but after a SIGHUP it used the Exim user (because it
1136 calls itself on a restart). I have changed the code so that it now always
1139 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1140 message are rejected with the same error (e.g. no authentication or bad
1141 sender address), and a DATA command is nevertheless sent (as can happen
1142 with PIPELINING or a stupid MUA), the error message that was given to the
1143 RCPT commands is included in the rejection of the DATA command. This is
1144 intended to be helpful for MUAs that show only the final error to their
1147 PH/42 Another patch from the Sieve maintainer.
1149 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1150 Eximstats - Fixed some broken HTML links and added missing column headers
1152 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1153 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1155 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1157 PH/43 Yet another patch from the Sieve maintainer.
1159 PH/44 I found a way to check for a TCP/IP connection going away before sending
1160 the response to the final '.' that terminates a message, but only in the
1161 case where the client has not sent further data following the '.'
1162 (unfortunately, this is allowed). However, in many cases there won't be
1163 any further data because there won't be any more messages to send. A call
1164 to select() can be used: if it shows that the input is "ready", there is
1165 either input waiting, or the socket has been closed. An attempt to read
1166 the next input character can distinguish the two cases. Previously, Exim
1167 would have sent an OK response which the client would never have see.
1168 This could lead to message repetition. This fix should cure that, at
1169 least in a lot of common cases.
1171 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1172 advertised in response to EHLO.
1178 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1179 fixed by 4.65/MH/01 (is this a record?) are fixed:
1181 (i) An empty string was always treated as zero by the numeric comparison
1182 operators. This behaviour has been restored.
1184 (ii) It is documented that the numeric comparison operators always treat
1185 their arguments as decimal numbers. This was broken in that numbers
1186 starting with 0 were being interpreted as octal.
1188 While fixing these problems I realized that there was another issue that
1189 hadn't been noticed. Values of message_size_limit (both the global option
1190 and the transport option) were treated as octal if they started with 0.
1191 The documentation was vague. These values are now always treated as
1192 decimal, and I will make that clear in the documentation.
1198 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1199 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1202 MH/01 Don't check that the operands of numeric comparison operators are
1203 integers when their expansion is in "skipping" mode (fixes bug
1204 introduced by 4.64-PH/07).
1206 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1207 child addresses, Exim now panics and dies. Previously, because the count
1208 is held in a short int, deliveries were likely to be lost. As such a
1209 large number of recipients for a single message is ridiculous
1210 (performance will be very, very poor), I have chosen to impose a limit
1211 rather than extend the field.
1217 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1218 leftover -K file (the existence of which was triggered by #402).
1219 While we were at it, introduced process PID as part of the -K
1220 filename. This should rule out race conditions when creating
1223 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1224 processing considerably. Previous code took too long for large mails,
1225 triggering a timeout which in turn triggers #401.
1227 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1228 in the DK code in transports.c. sendfile() is not really portable,
1229 hence the _LINUX specificness.
1231 TF/01 In the add_headers option to the mail command in an Exim filter,
1232 there was a bug that Exim would claim a syntax error in any
1233 header after the first one which had an odd number of characters
1236 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1237 callout verification, Exim cached a "reject" for the entire domain. This
1238 is correct for most verifications, but it is not correct for a recipient
1239 verification with use_sender or use_postmaster set, because in that case
1240 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1241 case of MAIL FROM:<> rejection from other early rejections (e.g.
1242 rejection of HELO). When verifying a recipient using a non-null MAIL
1243 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1244 Whatever the result of the callout, the value of the domain cache is
1245 left unchanged (for any other kind of callout, getting as far as trying
1246 RCPT means that the domain itself is ok).
1248 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1251 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1252 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1253 session, the daemon ignores ECONNECT errors and logs others; it now
1254 ignores EPIPE as well.
1256 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1257 (quoted-printable decoding).
1259 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1260 later the small subsequent patch to fix an introduced bug.
1262 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1264 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1266 PH/08 An error is now given if message_size_limit is specified negative.
1268 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1269 to be given (somewhat) arbitrary names.
1271 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1274 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1275 miscellaneous code fixes
1277 PH/10 Added the log_reject_target ACL modifier to specify where to log
1280 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1281 hostname. This is wrong, because it relates to the incoming message (and
1282 probably the interface on which it is arriving) and not to the outgoing
1283 callout (which could be using a different interface). This has been
1284 changed to use the value of the helo_data option from the smtp transport
1285 instead - this is what is used when a message is actually being sent. If
1286 there is no remote transport (possible with a router that sets up host
1287 addresses), $smtp_active_hostname is used.
1289 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1290 tweaks were necessary in order to get it to work (see also 21 below):
1291 (a) The code assumed that strncpy() returns a negative number on buffer
1292 overflow, which isn't the case. Replaced with Exim's string_format()
1294 (b) There were several signed/unsigned issues. I just did the minimum
1295 hacking in of casts. There is scope for a larger refactoring.
1296 (c) The code used strcasecmp() which is not a standard C function.
1297 Replaced with Exim's strcmpic() function.
1298 (d) The code set only $1; it now sets $auth1 as well.
1299 (e) A simple test gave the error "authentication client didn't specify
1300 service in request". It would seem that Dovecot has changed its
1301 interface. Fortunately there's a specification; I followed it and
1302 changed what the client sends and it appears to be working now.
1304 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1307 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1308 address A is aliased to B and C, where B exists and C does not. Without
1309 -v the output is "A verified" because verification stops after a
1310 successful redirection if more than one address is generated. However,
1311 with -v the child addresses are also verified. Exim was outputting "A
1312 failed to verify" and then showing the successful verification for C,
1313 with its parentage. It now outputs "B failed to verify", showing B's
1314 parentage before showing the successful verification of C.
1316 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1317 look up a TXT record in a specific list after matching in a combined
1320 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1321 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1322 they consult the DNS. I had assumed they would set it the way they
1323 wanted; and indeed my experiments on Linux seem to show that in some
1324 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1325 To be on the safe side, however, I have now made the interface to
1326 host_find_byname() similar to host_find_bydns(), with an argument
1327 containing the DNS resolver options. The host_find_byname() function now
1328 sets these options at its start, just as host_find_bydns() does. The smtp
1329 transport options dns_qualify_single and dns_search_parents are passed to
1330 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1331 of host_find_byname() use the default settings of RES_DEFNAMES
1332 (qualify_single) but not RES_DNSRCH (search_parents).
1334 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1335 spool_read_header() do less string testing, by means of a preliminary
1336 switch on the second character of optional "-foo" lines. (This is
1337 overdue, caused by the large number of possibilities that now exist.
1338 Originally there were few.) While I was there, I also converted the
1339 str(n)cmp tests so they don't re-test the leading "-" and the first
1340 character, in the hope this might squeeze out yet more improvement.
1342 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1343 flag allowing group syntax was set by the header_syntax check but not
1344 turned off, possible causing trouble later; (2) The flag was not being
1345 set at all for the header_verify test, causing "group"-style headers to
1346 be rejected. I have now set it in this case, and also caused header_
1347 verify to ignore an empty address taken from a group. While doing this, I
1348 came across some other cases where the code for allowing group syntax
1349 while scanning a header line wasn't quite right (mostly, not resetting
1350 the flag correctly in the right place). These bugs could have caused
1351 trouble for malformed header lines. I hope it is now all correct.
1353 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1354 with the "reply" argument non-NULL. The code, however (which originally
1355 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1356 but it didn't always do it. This confused somebody who was copying the
1357 code for some other use. I have removed all the tests.
1359 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1360 feature that was used to support insecure browsers during the U.S. crypto
1361 embargo. It requires special client support, and Exim is probably the
1362 only MTA that supported it -- and would never use it because real RSA is
1363 always available. This code has been removed, because it had the bad
1364 effect of slowing Exim down by computing (never used) parameters for the
1365 RSA_EXPORT functionality.
1367 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1368 authenticator to fail if there's a tab character in the incoming data
1369 (there should never be unless someone is messing about, as it's supposed
1370 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1371 if the connection is using TLS or if the remote IP is the same as the
1372 local IP, and the "valid-client-cert option" if a client certificate has
1375 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1376 authenticators. This can be used for authorization after authentication
1377 succeeds. (In the case of plaintext, it servers for both authentication
1380 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1381 if any retry times were supplied.
1383 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1384 connection, where there is no client IP address to check. In this
1385 situation, the verify now always succeeds.
1387 PH/25 Applied John Jetmore's -Mset patch.
1389 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1391 PH/27 In a string expansion for a processed (not raw) header when multiple
1392 headers of the same name were present, leading whitespace was being
1393 removed from all of them, but trailing whitespace was being removed only
1394 from the last one. Now trailing whitespace is removed from each header
1395 before concatenation. Completely empty headers in a concatenation (as
1396 before) are ignored.
1398 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1399 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1401 PH/29 [Removed. This was a change that I later backed out, and forgot to
1402 correct the ChangeLog entry (that I had efficiently created) before
1403 committing the later change.]
1405 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1406 address errors (4xx response to RCPT) over the same connection as other
1407 messages routed to the same hosts. Such deliveries are always "forced",
1408 so retry times are not inspected. This resulted in far too many retries
1409 for the affected addresses. The effect occurred only when there were more
1410 hosts than the hosts_max_try setting in the smtp transport when it had
1411 the 4xx errors. Those hosts that it had tried were not added to the list
1412 of hosts for which the message was waiting, so if all were tried, there
1413 was no problem. Two fixes have been applied:
1415 (i) If there are any address or message errors in an SMTP delivery, none
1416 of the hosts (tried or untried) are now added to the list of hosts
1417 for which the message is waiting, so the message should not be a
1418 candidate for sending over the same connection that was used for a
1419 successful delivery of some other message. This seems entirely
1420 reasonable: after all the message is NOT "waiting for some host".
1421 This is so "obvious" that I'm not sure why it wasn't done
1422 previously. Hope I haven't missed anything, but it can't do any
1423 harm, as the worst effect is to miss an optimization.
1425 (ii) If, despite (i), such a delivery is accidentally attempted, the
1426 routing retry time is respected, so at least it doesn't keep
1427 hammering the server.
1429 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1430 in ${readsocket because some servers need this prod.
1432 PH/32 Added some extra debug output when updating a wait-xxx database.
1434 PH/33 The hint "could be header name not terminated by colon", which has been
1435 given for certain expansion errors for a long time, was not being given
1436 for the ${if def:h_colon_omitted{... case.
1438 PH/34 The spec says: "With one important exception, whenever a domain list is
1439 being scanned, $domain contains the subject domain." There was at least
1440 one case where this was not true.
1442 PH/35 The error "getsockname() failed: connection reset by peer" was being
1443 written to the panic log as well as the main log, but it isn't really
1444 panic-worthy as it just means the connection died rather early on. I have
1445 removed the panic log writing for the ECONNRESET error when getsockname()
1448 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1449 runs only) independently of the message's sender address. This meant
1450 that, if the 4xx error was in fact related to the sender, a different
1451 message to the same recipient with a different sender could confuse
1452 things. In particualar, this can happen when sending to a greylisting
1453 server, but other circumstances could also provoke similar problems.
1454 I have changed the default so that the retry time for these errors is now
1455 based a combination of the sender and recipient addresses. This change
1456 can be overridden by setting address_retry_include_sender=false in the
1459 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1460 remote server are returned as part of bounce messages. This was not
1461 happening for LMTP over a pipe (the lmtp transport), but now it is the
1462 same for both kinds of LMTP.
1464 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1465 in header lines that were in fact CNAMEs. This is no longer the case.
1467 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1468 and queue runs started by the daemon processed all messages. This has
1469 been fixed so that -R and -S can now usefully be given with -q<time>.
1471 PH/40 Import PCRE release 6.7 (fixes some bugs).
1473 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1475 PH/42 Give an error if -q is specified more than once.
1477 PH/43 Renamed the variables $interface_address and $interface_port as
1478 $received_ip_address and $received_port, to make it clear that these
1479 values apply to message reception, and not to the outgoing interface when
1480 a message is delivered. (The old names remain recognized, of course.)
1482 PH/44 There was no timeout on the connect() call when using a Unix domain
1483 socket in the ${readsocket expansion. There now is.
1485 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1486 be meaningful with "accept".
1488 SC/01 Eximstats V1.43
1489 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1491 SC/02 Eximstats V1.44
1492 Use a glob alias rather than an array ref in the generated
1493 parser. This improves both readability and performance.
1495 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1496 Collect SpamAssassin and rejection statistics.
1497 Don't display local sender or destination tables unless
1498 there is data to show.
1499 Added average volumes into the top table text output.
1501 SC/04 Eximstats V1.46
1502 Collect data on the number of addresses (recipients)
1503 as well as the number of messages.
1505 SC/05 Eximstats V1.47
1506 Added 'Message too big' to the list of mail rejection
1507 reasons (thanks to Marco Gaiarin).
1509 SC/06 Eximstats V1.48
1510 Mainlog lines which have GMT offsets and are too short to
1511 have a flag are now skipped.
1513 SC/07 Eximstats V1.49 (Alain Williams)
1514 Added the -emptyok flag.
1516 SC/08 Eximstats V1.50
1517 Fixes for obtaining the IP address from reject messages.
1519 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1520 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1521 whitesspace changes from 4.64-PH/27
1523 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1526 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1527 are found, allow negative numbers in numeric criteria)
1529 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1531 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1532 to match changes made in 4.64-PH/43
1534 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1536 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1537 "domain" is actually an IP address, and force a failure. This locks out
1538 those revolvers/nameservers that support "A-for-A" lookups, in
1539 contravention of the specifications.
1541 PH/48 When a host name was looked up from an IP address, and the subsequent
1542 forward lookup of the name timed out, the host name was left in
1543 $sender_host_name, contrary to the specification.
1545 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1546 restricted to single-key lookups, Exim was not diagnosing an error if
1547 * or *@ was used with a query-style lookup.
1549 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1551 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1552 long ago, but noone interested enough thought of it. Let's just say that
1553 the "1.1" means that there are some new functions that weren't there at
1554 some point in the past.
1556 PH/51 Error processing for expansion failure of helo_data from an smtp
1557 transport during callout processing was broken.
1559 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1560 tested/used via the -bh/-bhc/-bs options.
1562 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1563 bug, fixed in subsequent PCRE releases).
1565 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1566 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1568 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1574 SC/01 Use a glob alias rather than an array ref in eximstats generated
1575 parser. This improves both readability and performance.
1577 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1578 Don't display local sender or destination tables in eximstats unless
1579 there is data to show.
1580 Added average volumes into the eximstats top table text output.
1582 SC/03 Collect data on the number of addresses (recipients) as well
1583 as the number of messages in eximstats.
1585 TF/01 Correct an error in the documentation for the redirect router. Exim
1586 does (usually) call initgroups() when daemonizing.
1588 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1589 with consistent privilege compared to when running as a daemon.
1591 TF/03 Note in the spec that $authenticated_id is not set for local
1592 submissions from trusted users.
1594 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1595 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1597 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1598 by adding some example configuration directives to the default
1599 configuration file. A little bit of work is required to uncomment the
1600 directives and define how usernames and passwords are checked, but
1601 there is now a framework to start from.
1603 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1604 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1605 without this. I don't know how relevant this is to other LDAP libraries.
1607 PH/02 Add the verb name to the "unknown ACL verb" error.
1609 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1611 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1613 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1614 directory not expanded when it should be if an expanded home directory
1615 was set for the address (which is overridden by the transport).
1617 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1620 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1621 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1622 because it is too late at that time, and has no effect.
1624 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1625 security issue with \' (bugzilla #107). I could not use the
1626 PQescapeStringConn() function, because it needs a PGconn value as one of
1629 PH/08 When testing addresses using -bt, indicate those final addresses that
1630 are duplicates that would not cause an additional delivery. At least one
1631 person was confused, thinking that -bt output corresponded to deliveries.
1632 (Suppressing duplicates isn't a good idea as you lose the information
1633 about possibly different redirections that led to the duplicates.)
1635 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1636 systems where poll() doesn't work, in particular OS X.
1638 PH/10 Added more information to debugging output for retry time not reached.
1640 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1641 operations in malware.c.
1643 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1646 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1647 syslog_duplication set false, log lines that would normally be written
1648 both the the main log and to the reject log were not written to syslog at
1651 PH/14 In the default configuration, change the use of "message" in ACL warn
1652 statements to "add_header".
1654 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1655 not followed by a command (e.g. "seen endif").
1657 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1658 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1661 PH/17 Added extra conditions to the default value of delay_warning_condition
1665 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1666 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1667 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1670 The Auto-Submitted: and various List- headers are standardised, whereas I
1671 don't think Precedence: ever was.
1673 PH/18 Refactored debugging code in route_finduser() to show more information,
1674 in particular, the error code if getpwnam() issues one.
1676 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1677 This is apparently needed in addition to the PH/07 change above to avoid
1678 any possible encoding problems.
1680 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1681 but not after initializing Perl.
1683 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1684 output them only if debugging. By default they are written stderr,
1685 apparently, which is not desirable.
1687 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1690 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1693 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1695 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1696 authenticated or an ident call has been made. Suppress the default
1697 values for $authenticated_id and $authenticated_sender (but permit -oMai
1698 and -oMas) when testing with -bh.
1700 PH/24 Re-jigged the order of the tests in the default configuration so that the
1701 tests for valid domains and recipients precede the DNS black list and CSA
1702 tests, on the grounds that those ones are more expensive.
1704 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1705 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1706 This bug exists in every version of Exim that I still have, right back to
1709 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1710 However, an attempt to turn on case-sensitivity in a regex key by
1711 including (?-i) didn't work because the subject string was already
1712 lowercased, and the effects were non-intuitive. It turns out that a
1713 one-line patch can be used to allow (?-i) to work as expected.
1719 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1720 other effects) broke the use of negated acl sub-conditions.
1722 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1725 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1726 "Deny" causes Exim to reject the incoming connection with a 554 error.
1727 Unfortunately, if there is a major crisis, such as a disk failure,
1728 tcp-wrappers gives "deny", whereas what one would like would be some
1729 kind of temporary error. A kludge has been added to help with this.
1730 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1731 554 error is used if errno is still zero or contains ENOENT (which occurs
1732 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1735 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1737 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1738 errors. Otherwise a message that provokes a temporary error (when other
1739 messages do not) can cause a whole host to time out.
1741 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1742 addresses were routed directly to files or pipes from a redirect router.
1743 File deliveries just didn't batch; pipe deliveries might have suffered
1746 PH/06 A failure to get a lock for a hints database would erroneously always say
1747 "Failed to get write lock", even when it was really a read lock.
1749 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1750 of 0600. This has been changed to use the value of the lockfile_mode
1751 option (which defaults to 0600).
1753 PH/08 Applied small patch from the Sieve maintainer.
1755 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1756 folder from quota calculations, a direct delivery into this folder messed
1757 up the contents of the maildirsize file. This was because the regex was
1758 used only to exclude .Trash (or whatever) when the size of the mailbox
1759 was calculated. There was no check that a delivery was happening into an
1760 excluded directory. This bug has been fixed by ignoring all quota
1761 processing for deliveries into excluded directories.
1763 PH/10 Added the maildirfolder_create_regex option to appendfile.
1769 PH/01 The code for finding all the local interface addresses on a FreeBSD
1770 system running IPv6 was broken. This may well have applied to all BSD
1771 systems, as well as to others that have similar system calls. The broken
1772 code found IPv4 interfaces correctly, but gave incorrect values for the
1773 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1774 that it would not match correctly against @[] and not recognize the IPv6
1777 PH/02 The ipliteral router was not recognizing addresses of the form user@
1778 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1780 PH/03 Added disable_ipv6.
1782 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1783 decoded form, because it is most often used to construct To: headers
1784 lines in autoreplies, and the decoded form may well be syntactically
1785 invalid. However, $reply_address has leading white space removed, and all
1786 newlines turned into spaces so that the autoreply transport does not
1789 PH/05 If group was specified without a user on a router, and no group or user
1790 was specified on a transport, the group from the router was ignored.
1792 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1793 for visible compile-time settings that can be used to change these
1794 numbers, for those that want even more. Backwards compatibility with old
1795 spool files has been maintained. However, going back to a previous Exim
1796 release will lost any variables that are in spool files.
1798 PH/07 Two small changes when running in the test harness: increase delay when
1799 passing a TCP/IP connection to a new process, in case the original
1800 process has to generate a bounce, and remove special handling of
1801 127.0.0.2 (sic), which is no longer necessary.
1803 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1804 be the same on different OS.
1806 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1809 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1810 whether --show-vars was specified or not
1812 JJ/02 exipick: Added support for new ACL variable spool format introduced
1815 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1816 syntactically invalid From: or Reply-to: line, and a filter used this to
1817 generate an autoreply, and therefore failed to obtain an address for the
1818 autoreply, Exim could try to deliver to a non-existent relative file
1819 name, causing unrelated and misleading errors. What now happens is that
1820 it logs this as a hard delivery error, but does not attempt to create a
1823 PH/11 The exinext utility has a -C option for testing purposes, but although
1824 the given file was scanned by exinext itself; it wasn't being passed on
1825 when Exim was called.
1827 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1828 an end-of-file indication when reading a command response.
1830 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1831 compiled. In many other places in Exim, IPv6 addresses are always
1832 recognized, so I have changed this. It also means that IPv4 domain
1833 literals of the form [IPV4:n.n.n.n] are now always recognized.
1835 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1836 used if the router is not running as root, for example, when verifying at
1837 ACL time, or when using -bh. The debugging output from this situation was
1838 non-existent - all you got was a failure to exec. I have made two
1841 (a) Failures to set uid/gid, the current directory, or a process leader
1842 in a subprocess such as that created by queryprogram now generate
1843 suitable debugging ouput when -d is set.
1845 (b) The queryprogram router detects when it is not running as root,
1846 outputs suitable debugging information if -d is set, and then runs
1847 the subprocess without attempting to change uid/gid.
1849 PH/15 Minor change to Makefile for building test_host (undocumented testing
1852 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1853 additional section of a DNS packet that returns MX or SRV records.
1854 Instead, it always explicitly searches for A/AAAA records. This avoids
1855 major problems that occur when a DNS server includes only records of one
1856 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1857 fixed another bug: if SRV records were looked up and the corresponding
1858 address records were *not* found in the additional section, the port
1859 values from the SRV records were lost.
1861 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1862 using the correct key (the original address) when searching the retry
1863 rules in order to find which one to use for generating the retry hint.
1865 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1866 adding the default one. Similarly, if it contains a Reply-To: header, the
1867 errors_reply_to option, if set, is not used.
1869 PH/19 When calculating a retry time, Exim used to measure the "time since
1870 failure" by looking at the "first failed" field in the retry record. Now
1871 it does not use this if it is later than than the arrival time of the
1872 message. Instead it uses the arrival time. This makes for better
1873 behaviour in cases where some deliveries succeed, thus re-setting the
1874 "first failed" field. An example is a quota failure for a huge message
1875 when small messages continue to be delivered. Without this change, the
1876 "time since failure" will always be short, possible causing more frequent
1877 delivery attempts for the huge message than are intended.
1878 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1880 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1881 $1, $2, $3) because the numerical variables can be reset during some
1882 expansion items (e.g. "match"), thereby losing the authentication data.
1884 PH/21 Make -bV show the size of off_t variables so that the test suite can
1885 decide whether to run tests for quotas > 2G.
1887 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1888 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1889 filecount value is greater than 2G or if a quota value is greater than 2G
1890 on a system where the size of off_t is not greater than 4, a panic error
1893 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1894 never match. The debug and -bh output now contains an explicit error
1895 message indicating a malformed IPv4 address or mask.
1897 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1898 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1899 PH/23 above applies.
1901 PH/25 Do not write to syslog when running in the test harness. The only
1902 occasion when this arises is a failure to open the main or panic logs
1903 (for which there is an explicit test).
1905 PH/26 Added the /no_tell option to "control=freeze".
1907 PH/27 If a host name lookup failed very early in a connection, for example, if
1908 the IP address matched host_lookup and the reverse lookup yielded a name
1909 that did not have a forward lookup, an error message of the form "no IP
1910 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1911 could be logged. Now it outputs the IP address instead of "NULL".
1913 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1914 allows the sender and the authenticated sender to be set when
1915 submitting a message from within Exim. Since child_open_exim() is
1916 documented for local_scan(), the new function should be too.
1918 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1919 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1920 results in an empty string is now treated as unset.
1922 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1924 PH/31 Added sender_verify_fail logging option.
1926 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1927 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1928 tidied the source and removed it altogether.
1930 PH/33 When a queue run was abandoned because the load average was too high, a
1931 log line was always written; now it is written only if the queue_run log
1932 selector is set. In addition, the log line for abandonment now contains
1933 information about the queue run such as the pid. This is always present
1934 in "start" and "stop" lines but was omitted from the "abandon" line.
1936 PH/34 Omit spaces between a header name and the colon in the error message that
1937 is given when verify = headers_syntax fails (if there are lots of them,
1938 the message gets confusing).
1940 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1941 names, as there are now some PTR records that contain slashes. This check
1942 is only to protect against broken name servers that fall over on strange
1943 characters, so the fact that it applies to all lookups doesn't matter.
1945 PH/36 Now that the new test suite is complete, we can remove some of the
1946 special code in Exim that was needed for the old test suite. For example,
1947 sorting DNS records because real resolvers return them in an arbitrary
1948 order. The new test suite's fake resolver always returns records in the
1951 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1952 bounces) except when queue_only is set, to avoid logging races between
1953 the different processes.
1955 PH/38 Panic-die if .include specifies a non-absolute path.
1957 PH/39 A tweak to the "H" retry rule from its user.
1959 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1960 a label. They prevented compilation on older perls.
1962 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1963 a warning to be raised on newish perls.
1965 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1966 on queue. Changes to match documented behaviour of showing count of
1967 messages matching specified criteria.
1969 PH/40 Changed the default ident timeout from 30s to 5s.
1971 PH/41 Added support for the use of login_cap features, on those BSD systems
1972 that have them, for controlling the resources used by pipe deliveries.
1974 PH/42 The content-scanning code uses fopen() to create files in which to put
1975 message data. Previously it was not paying any attention to the mode of
1976 the files. Exim runs with umask(0) because the rest of the code creates
1977 files with open(), and sets the required mode explicitly. Thus, these
1978 files were ending up world-writeable. This was not a big issue, because,
1979 being within the spool directory, they were not world-accessible. I have
1980 created a function called modefopen, which takes an additional mode
1981 argument. It sets umask(777), creates the file, chmods it to the required
1982 mode, then resets the umask. All the relevant calls to fopen() in the
1983 content scanning code have been changed to use this function.
1985 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1986 to 24 hours. This avoids potential overflow problems when processing G
1987 and H retry rules. I suspect nobody ever tinkers with this value.
1989 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1991 PH/45 When the plaintext authenticator is running as a client, the server's
1992 challenges are checked to ensure they are valid base64 strings. By
1993 default, the authentication attempt is cancelled if an invalid string is
1994 received. Setting client_ignore_invalid_base64 true ignores these errors.
1995 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1996 they are received. Thus, the responses can be made to depend on the
1997 challenges. If an invalid string is ignored, an empty string is placed in
2000 PH/46 Messages that are created by the autoreply transport now contains a
2001 References: header, in accordance with RFCs 2822 and 3834.
2003 PH/47 Added authenticated_sender_force to the smtp transport.
2005 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2007 PH/49 Installed latest patch from the Sieve maintainer.
2009 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2010 was also set, the appendfile transport was unnecessarily scanning a
2011 directory of message files (e.g. for maildir delivery) to find the count
2012 of files (along with the size), even though it did not need this
2013 information. It now does the scan only if it needs to find either the
2014 size of the count of files.
2016 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2018 PH/52 Two bugs concerned with error handling when the smtp transport is
2021 (i) Exim was not creating retry information for temporary errors given
2022 for individual recipients after the DATA command when the smtp transport
2023 was used in LMTP mode. This meant that they could be retried too
2024 frequently, and not timed out correctly.
2026 (ii) Exim was setting the flag that allows error details to be returned
2027 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2028 recipients that were returned after the DATA command.
2030 PH/53 This is related to PH/52, but is more general: for any failing address,
2031 when detailed error information was permitted to be returned to the
2032 sender, but the error was temporary, then after the final timeout, only
2033 "retry timeout exceeded" was returned. Now it returns the full error as
2034 well as "retry timeout exceeded".
2036 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2037 do this, and (what is worse) MTAs that accept it.
2039 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2040 will now be deprecated.
2042 PH/56 New os.c-cygwin from the Cygwin maintainer.
2044 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2045 formats (previously only available in exim formats via -bpr, -bpru,
2046 and -bpra. Now also available in native and exiqgrep formats)
2048 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2049 with very large, slow to parse queues
2051 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2053 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2055 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2056 responses to authentication challenges, though it was showing the
2057 challenges; (ii) I've removed the CR characters from the debug output for
2060 PH/58 Allow for the insertion of a newline as well as a space when a string
2061 is turned into more than one encoded-word during RFC 2047 encoding. The
2062 Sieve code now uses this.
2064 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2065 data_4xx, lost_connection, tls_required.
2067 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2068 message was being sent as an SMTP response.
2070 PH/61 Add -l and -k options to exicyclog.
2072 PH/62 When verifying, if an address was redirected to one new address, so that
2073 verification continued, and the new address failed or deferred after
2074 having set something in $address_data, the value of $address_data was not
2075 passed back to the ACL. This was different to the case when no
2076 redirection occurred. The value is now passed back in both cases.
2078 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2079 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2080 use login_cap.h, so on its own it isn't the distinguishing feature. The
2081 new name refers directly to the setclassresources() function.
2083 PH/65 Added configuration files for NetBSD3.
2085 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2087 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2088 is preferred over IPv4.
2090 PH/68 The bounce_return_message and bounce_return_body options were not being
2091 honoured for bounces generated during the reception of non-SMTP messages.
2092 In particular, this applied to messages rejected by the ACL. This bug has
2093 been fixed. However, if bounce_return_message is true and bounce_return_
2094 body is false, the headers that are returned for a non-SMTP message
2095 include only those that have been read before the error was detected.
2096 (In the case of an ACL rejection, they have all been read.)
2098 PH/69 The HTML version of the specification is now built in a directory called
2099 spec_html instead of spec.html, because the latter looks like a path with
2100 a MIME-type, and this confuses some software.
2102 PH/70 Catch two compiler warnings in sieve.c.
2104 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2105 function verify_get_ident() calls ip_connect() to connect a socket, but
2106 if the "connect()" function timed out, ip_connect() used to close the
2107 socket. However, verify_get_ident() also closes the socket later, and in
2108 between Exim writes to the log, which may get opened at this point. When
2109 the socket was closed in ip_connect(), the log could get the same file
2110 descriptor number as the socket. This naturally causes chaos. The fix is
2111 not to close the socket in ip_connect(); the socket should be closed by
2112 the function that creates it. There was only one place in the code where
2113 this was missing, in the iplookup router, which I don't think anybody now
2114 uses, but I've fixed it anyway.
2116 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2117 well as to direct DNS lookups. Otherwise the handling of names in host
2118 lists is inconsistent and therefore confusing.
2124 PH/01 Two changes to the default runtime configuration:
2126 (1) Move the checks for relay_from_hosts and authenticated clients from
2127 after to before the (commented out) DNS black list checks.
2129 (2) Add control=submission to the relay_from_hosts and authenticated
2130 clients checks, on the grounds that messages accepted by these
2131 statements are most likely to be submissions.
2133 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2135 (1) Generate an error if the third argument for the ${prvs expansion is
2138 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2141 (3) Reset the variables that are obtained from the first argument of
2142 ${prvscheck and used in the second argument before leaving the code,
2143 because their memory is reclaimed, so using them afterwards may do
2146 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2147 one (it's much easier than Tom thought :-).
2149 (5) Because of (4), we can now allow for the use of $prvscheck_result
2150 inside the third argument.
2152 PH/03 For some reason, the default setting of PATH when running a command from
2153 a pipe transport was just "/usr/bin". I have changed it to
2156 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2157 anything to be listed in the output from -bV.
2159 PH/05 When a filter generated an autoreply, the entire To: header line was
2160 quoted in the delivery log line, like this:
2162 => >A.N.Other <ano@some.domain> <original@ddress> ...
2164 This has been changed so that it extracts the operative address. There
2165 may be more than one such address. If so, they are comma-separated, like
2168 => >ano@some.domain,ona@other.domain <original@ddress> ...
2170 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2171 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2172 not being looked up in the rDNS to get a host name, Exim was showing the
2173 IP address twice in Received: lines, even though the IP addresses were
2174 identical. For example:
2176 Received: from [1.2.3.4] (helo=[1.2.3.4])
2178 However, if the real host name was known, it was omitting the HELO data
2179 if it matched the actual IP address. This has been tidied up so that it
2180 doesn't show the same IP address twice.
2182 PH/07 When both +timestamp and +memory debugging was on, the value given by
2183 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2184 called by the debug printing, thereby overwriting the timestamp buffer.
2185 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2187 PH/08 When the original message was included in an autoreply transport, it
2188 always said "this is a copy of the message, including all the headers",
2189 even if body_only or headers_only was set. It now gives an appropriate
2192 PH/09 Applied a patch from the Sieve maintainer which:
2194 o fixes some comments
2195 o adds the (disabled) notify extension core
2196 o adds some debug output for the result of if/elsif tests
2197 o points to the current vacation draft in the documentation
2198 and documents the missing references header update
2202 o fixes a bug in processing the envelope test (when testing
2203 multiple envelope elements, the last element determinted the
2206 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2207 Electronic Mail") by including:
2209 Auto-submitted: auto-generated
2211 in the messages that it generates (bounce messages and others, such as
2212 warnings). In the case of bounce messages for non-SMTP mesages, there was
2213 also a typo: it was using "Auto_submitted" (underscore instead of
2214 hyphen). Since every message generated by Exim is necessarily in response
2215 to another message, thes have all been changed to:
2217 Auto-Submitted: auto-replied
2219 in accordance with these statements in the RFC:
2221 The auto-replied keyword:
2223 - SHOULD be used on messages sent in direct response to another
2224 message by an automatic process,
2226 - MUST NOT be used on manually-generated messages,
2228 - MAY be used on Delivery Status Notifications (DSNs) and Message
2229 Disposition Notifications (MDNs),
2231 - MUST NOT be used on messages generated by automatic or periodic
2232 processes, except for messages which are automatic responses to
2235 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2236 to the default Received: header definition.
2238 PH/12 Added log selector acl_warn_skipped (default on).
2240 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2241 variables because (a) they are in the wrong storage pool and (b) even if
2242 they were copied, it wouldn't work properly because of the caching.
2244 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2245 checking when decoding. Apparently there are clients that generate
2246 overlong encoded strings. Why am I not surprised?
2248 PH/15 If the first argument of "${if match_address" was not empty, but did not
2249 contain an "@" character, Exim crashed. Now it writes a panic log message
2250 and treats the condition as false.
2252 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2254 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2255 extension "envelope-auth". The code is finished and in agreement with
2256 other implementations, but there is no documentation so far and in fact,
2257 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2258 not changing the active code.
2260 Print executed "if" and "elsif" statements when debugging is used. This
2261 helps a great deal to understand what a filter does.
2263 Document more things not specified clearly in RFC3028. I had all this
2264 sorted out, when out of a sudden new issues came to my mind. Oops."
2266 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2269 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2270 canonical form (as documented). However, after a host name lookup from
2271 the IP address, check_host() was doing a simple string comparison with
2272 addresses acquired from the DNS when checking that the found name did
2273 have the original IP as one of its addresses. Since any found IPv6
2274 addresses are likely to be in abbreviated form, the comparison could
2275 fail. Luckily, there already exists a function for doing the comparison
2276 by converting both addresses to binary, so now that is used instead of
2277 the text comparison.
2279 PH/20 There was another similar case to PH/19, when a complete host name was
2280 given in a host list; looking up its IP address could give an abbreviated
2281 form, whereas the current host's name might or might not be abbreviated.
2282 The same fix has been applied.
2288 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2289 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2292 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2293 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2295 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2297 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2298 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2299 Also, for "Auto-Submitted", treat anything other than "no" as
2300 non-personal, in accordance with RFC 3834. (Previously it treated
2301 anything starting "auto-" as non-personal.)
2303 TF/01 The control=submission/name=... option had a problem with syntax
2304 errors if the name included a slash character. The /name= option
2305 now slurps the rest of the string, so it can include any characters
2306 but it must come last in the list of options (after /sender_retain
2309 PH/05 Some modifications to the interface to the fake nameserver for the new
2317 TK/01 Added the "success_on_redirect" address verification option. See
2318 NewStuff for rationale and an example.
2320 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2322 PH/02 Patch to exigrep to allow it to work on syslog lines.
2324 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2325 fread() to skip over the body file's header line, because in Cygwin the
2326 header line is locked and is inaccessible.
2328 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2329 co-exist for some time) to make it clear that it is the Exim ID that is
2330 referenced, not the Message-ID: header line.
2332 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2333 string_format() function, because snprintf() does not exist on all
2336 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2337 ${stat: expansion item.
2339 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2340 protocol synchronization error", to keep the pedants happy.
2342 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2343 well as for IRIX systems, when gcc is being used. See the host.c source
2346 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2348 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2351 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2352 transport and to the smtp transport in LMTP mode.
2354 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2356 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2357 run a filter in a subprocess. This could lead to confusion in subsequent
2358 lookups in the parent process. There should also be a search_tidyup() at
2359 the end of the subprocess.
2361 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2362 only if the host matched helo_try_verify_hosts, which caused the
2363 verification to occur when the EHLO/HELO command was issued. The ACL just
2364 tested the remembered result. Now, if a previous verification attempt has
2365 not happened, "verify = helo" does it there and then.
2367 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2369 TK/03 Fix log output including CR from clamd.
2371 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2372 exist provoked a memory error which could cause a segfault.
2374 PH/15 Installed PCRE 6.2
2376 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2378 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2379 of the problem. Specifically, suggested +O2 rather than +O1 for the
2382 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2384 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2385 record, the port was not correctly logged when the outgoing_port log
2386 selector was set (it logged the transort's default port).
2388 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2389 fallback_hosts, and "hosts" in the smtp transport.
2391 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2392 host errors such as "Connection refused".
2394 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2395 authentication with radiusclient 0.4.9:
2397 - Error returned from rc_read_config was caught wrongly
2398 - Username/password not passed on to radius server due to wrong length.
2400 The presumption is that some radiusclient API changes for 4.51/PH/17
2401 were not taken care of correctly. The code is still untested by me (my
2402 Linux distribution still has 0.3.2 of radiusclient), but it was
2403 contributed by a Radius user.
2405 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2406 expanding the "port" option of the smtp transport.
2408 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2409 while reading a MIME header. Thanks to Tom Hughes for a patch.
2411 PH/24 Include config.h inside local_scan.h so that configuration settings are
2414 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2415 that in an ACL for RCPT (for example), you can examine exactly what was
2418 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2419 commands, but it was not correctly comparing the address with the actual
2420 client host address. Thus, it would show the EHLO address in Received:
2421 header lines when this was not necessary.
2423 PH/27 Added the % operator to ${eval:}.
2425 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2426 it should be ignoring failures (because with -C, for example, it has lost
2427 privilege). It wasn't ignoring creation failures other than "already
2430 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2433 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2434 because an input line was too long, either on its own, or by virtue of
2435 too many continuations, the temporary file was not being removed, and the
2436 return code was incorrect.
2438 PH/31 Missing "BOOL" in function definition in filtertest.c.
2440 PH/32 Applied Sieve patches from the maintainer.
2442 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2444 PH/33 Added "verify = not_blind".
2446 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2447 Local/Makefile (with some defaults set). These are used in built scripts
2448 such as exicyclog, but they have never been used in the exim_install
2449 script (though there are many overriding facilities there). I have
2450 arranged that the exim_install script now takes note of these two
2453 PH/35 Installed configuration files for Dragonfly.
2455 PH/36 When a locally submitted message by a trusted user did not contain a
2456 From: header, and the sender address was obtained from -f or from an SMTP
2457 MAIL command, and the trusted user did not use -F to supply a sender
2458 name, $originator_name was incorrectly used when constructing a From:
2459 header. Furthermore, $originator_name was used for submission mode
2460 messages from external hosts without From: headers in a similar way,
2461 which is clearly wrong.
2463 PH/37 Added control=suppress_local_fixups.
2465 PH/38 When log_selector = +received_sender was set, and the addition of the
2466 sender made the log line's construction buffer exactly full, or one byte
2467 less than full, an overflow happened when the terminating "\n" was
2470 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2471 when the result of a list match is failure because a DNS lookup failed.
2473 PH/40 RM_COMMAND is now used in the building process.
2475 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2476 the "build-* directories that it finds.
2478 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2479 address in a domain literal was a prefix of an interface address.
2481 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2482 when verifying a sender address, unless rewrite_headers is false.
2484 PH/44 Wrote a long comment about why errors_to addresses are verified as
2485 recipients, not senders.
2487 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2488 the ratelimit ACL was added.
2490 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2492 PH/46 Added extra information about PostgreSQL errors to the error string.
2494 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2495 avoiding the need to install special zones in a real server. This is
2496 backwards compatible; if it can't find the fake resolver, it drops back.
2497 Thus, both old and new test suites can be run.
2499 TF/02 Added util/ratelimit.pl
2501 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2502 clock is set back in time.
2504 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2505 Candler <B.Candler@pobox.com>.
2507 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2508 OK for addresses that are the result of redirections.
2510 PH/48 A number of further additions for the benefit of the new test suite,
2511 including a fake gethostbyname() that interfaces to the fake DNS resolver
2514 TF/06 The fix for widen_domains has also been applied to qualify_single and
2515 search_parents which are the other dnslookup options that can cause
2518 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2521 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2523 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2524 service" if it encountered an MX record with an empty target hostname.
2525 The message is now "an MX or SRV record indicated no SMTP service".
2527 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2528 if the DNS of the sending site is misconfigured. This is quite a
2529 common situation. This change restores the behaviour of treating a
2530 helo verification defer as a failure.
2532 PH/51 If self=fail was set on a router, the bounce message did not include the
2533 actual error message.
2539 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2541 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2542 command itself ended in error, the underlying message about the transport
2543 filter timeout was being overwritten with the pipe command error. Now the
2544 underlying error message should be appended to the second error message.
2546 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2548 PH/02 Reduce the amount of output that "make" produces by default. Full output
2549 can still be requested.
2551 PH/03 The warning log line about a condition test deferring for a "warn" verb
2552 was being output only once per connection, rather than after each
2553 occurrence (because it was using the same function as for successful
2554 "warn" verbs). This seems wrong, so I have changed it.
2556 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2557 it should not have, which might have caused a crash in the right
2558 circumstances, but probably never did.
2560 PH/04 Installed a modified version of Tony Finch's patch to make submission
2561 mode fix the return path as well as the Sender: header line, and to
2562 add a /name= option so that you can make the user's friendly name appear
2565 TF/03 Added the control = fakedefer ACL modifier.
2567 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2568 Mark Lowes for thorough testing.
2570 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2572 TK/03 Merged latest SRS patch from Miles Wilton.
2574 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2575 with the definition in sysexits.h (which is #included earlier).
2576 Fortunately, Exim does not actually use EX_OK. The code used to try to
2577 preserve the sysexits.h value, by assumimg that macro definitions were
2578 scanned for macro replacements. I have been disabused of this notion,
2579 so now the code just undefines EX_OK before #including unistd.h.
2581 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2582 in the smtp transport. When a block could not be written in a single
2583 write() function, the timeout was being re-applied to each part-write.
2584 This seems wrong - if the receiver was accepting one byte at a time it
2585 would take for ever. The timeout is now adjusted when this happens. It
2586 doesn't have to be particularly precise.
2588 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2589 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2591 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2592 without a domain if the check to <postmaster@domain> fails.
2594 SC/01 Eximstats: added -xls and the ability to specify output files
2595 (patch written by Frank Heydlauf).
2597 SC/02 Eximstats: use FileHandles for outputing results.
2599 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2601 SC/04 Eximstats: fixed display of large numbers with -nvr option
2603 SC/05 Eximstats: fixed merging of reports with empty tables.
2605 SC/06 Eximstats: added the -include_original_destination flag
2607 SC/07 Eximstats: removed tabs and trailing whitespace.
2609 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2611 TK/06 MBOX spool code: Add real "From " MBOX separator line
2612 so the .eml file is really in mbox format (even though
2613 most programs do not really care). Patch from Alex Miller.
2615 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2616 The latter is generated from $received_to and is only set if the
2617 message has one envelope recipient. SA can use these headers,
2618 obviously out-of-the-box. Patch from Alex Miller.
2620 PH/08 The ${def test on a variable was returning false if the variable's
2621 value was "0", contrary to what the specification has always said!
2622 The result should be true unless the variable is empty.
2624 PH/09 The syntax error of a character other than { following "${if
2625 def:variable_name" (after optional whitespace) was not being diagnosed.
2626 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2627 accidental colon was present, for example, could give incorrect results.
2629 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2630 result is used (not including appendfile, where other changes are about
2633 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2634 This involved changing a lot of size variables from int to off_t. It
2635 should work with maildirs and everything.
2637 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2638 spamd dying while we are connected to it.
2640 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2643 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2644 function for BDB 4.3.
2646 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2648 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2649 the sources for PCRE have been re-arranged and more files are now
2652 PH/15 The code I had for printing potentially long long variables in PH/11
2653 above was not the best (it lost precision). The length of off_t variables
2654 is now inspected at build time, and an appropriate printing format (%ld
2655 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2656 to be "long long int" or "long int". This is needed for the internal
2657 formatting function string_vformat().
2659 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2660 the configuration file to be ":syslog", then the script "guesses" where
2661 the logs files are, rather than using the compiled in default. In our
2662 case the guess is not the same as the compiled default, so the script
2663 suddenly stopped working when I started to use syslog. The patch checks
2664 to see if log_file_path is "". If so, it attempts to read it from exim
2665 with no configuration file to get the compiled in version, before it
2666 falls back to the previous guessing code."
2668 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2669 implementing BATV in an Exim configuration. See NewStuff for the gory
2672 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2673 Makefile that are specific to HP-UX.
2675 PH/18 If the "use_postmaster" option was set for a recipient callout together
2676 with the "random" option, the postmaster address was used as the MAIL
2677 FROM address for the random test, but not for the subsequent recipient
2678 test. It is now used for both.
2680 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2681 patch removes a few documentation additions to RFC 3028, because the
2682 latest draft now contains them. It adds the new en;ascii-case comparator
2683 and a new error check for 8bit text in MIME parts. Comparator and
2684 require names are now matched exactly. I enabled the subaddress
2685 extension, but it is not well tested yet (read: it works for me)."
2687 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2688 rework some of the code of TK/09 above to avoid the hardwired use of
2689 "%lld" and "long long". Replaced the call to snprintf() with a call to
2692 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2693 records point to non-existent hosts", "retry timeout exceeded", and
2694 "retry time not reached for any host after a long failure period".
2696 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2697 experimental DomainKeys support:
2699 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2700 (2) On an error such as an illegally used "control", the wrong name for
2701 the control was given.
2703 These problems did NOT occur unless DomainKeys support was compiled.
2705 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2707 PH/24 Added ${if match_ip condition.
2709 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2710 databases so that it will be absolutely obvious if a crash occurs in the
2711 DB library. This is a regular occurrence (often caused by mis-matched
2714 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2715 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2716 picked up on a user's system that detects such things. There doesn't seem
2717 to be a gcc warning option for this - only an attribute that has to be
2718 put on the function's prototype. It seems that in Fedora Core 4 they have
2719 set this on a number of new functions. No doubt there will be more in due
2722 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2723 specify a transport. However, if an address that was verified by such a
2724 router was the subject of a callout, Exim crashed because it tried to
2725 read the rcpt_include_affixes from the non-existent transport. Now it
2726 just assumes that the setting of that option is false. This bug was
2727 introduced by 4.51/PH/31.
2729 PH/28 Changed -d+all to exclude +memory, because that information is very
2730 rarely of interest, but it makes the output a lot bigger. People tend to
2731 do -d+all out of habit.
2733 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2734 code in os-type was giving problems when libc.so lives in lib64, like on
2737 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2738 aren't the modern standard, and it seems that some systems' include files
2739 don't always have them. Exim was already checking for some of the newer
2740 ones like T_AAAA, and defining it itself. I've added checks for all the
2741 record types that Exim uses.
2743 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2744 not automatically generating a new one, as it is supposed to. This
2745 prevented TLS from working. If the file did exist, but contained invalid
2746 data, a new version was generated, as expected. It was only the case of a
2747 non-existent file that was broken.
2749 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2750 with a change in libdomainkeys > 0.64.
2752 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2753 from DNS. If the selector record carries the flag, it now has
2754 precedence over the domain-wide flag.
2756 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2758 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2759 the use of a port name that isn't defined in /etc/services) occurred, the
2760 message was deferred as in a normal delivery, and thus remained on the
2761 spool, instead of being failed because of the mua_wrapper setting. This
2762 is now fixed, and I tidied up some of the mua_wrapper messages at the
2765 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2766 the messages in a hash of arrays rather than using individual hashes.
2767 This is a bit cleaner and results in dramatic memory savings, albeit
2768 at a slight CPU cost.
2770 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2771 as requested by Marc Sherman.
2773 SC/10 Eximstats: added histograms for user specified patterns as requested
2776 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2778 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2779 fopen() in the content-scanning modules that did not already have it.
2785 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2786 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2788 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2790 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2792 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2793 to test Sieve filters that use "vacation".
2795 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2796 that changes the way the GnuTLS parameters are stored in the cache file.
2797 The new format can be generated externally. For backward compatibility,
2798 if the data in the cache doesn't make sense, Exim assumes it has read an
2799 old-format file, and it generates new data and writes a new file. This
2800 means that you can't go back to an older release without removing the
2803 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2804 work if there are any delivery delays because "one_time" forces the
2805 parent to be marked "delivered", so its unseen clone is never tried
2806 again. For this reason, Exim now forbids the simultaneous setting of
2809 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2810 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2811 entry if you want to know the details. The fix, however, neglected to
2812 consider the case when local delivery batching is involved. The test for
2813 "previously delivered" was not happening when checking to see if an
2814 address could be batched with a previous (undelivered) one; under
2815 certain circumstances this could lead to multiple deliveries to the same
2818 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2819 in its include files, and this causes problems building Exim.
2821 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2822 header_syntax) but Exim was just ignoring anything given after a slash.
2823 In particular, this caused confusion with an attempt to use "verify =
2824 reverse_host_lookup/defer_ok". An error is now given when options are
2825 supplied for verify items that do not have them. (Maybe reverse_host_
2826 lookup should have a defer_ok option, but that's a different point.)
2828 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2829 defined by RFC 821) to 2048, because there were problems with some AUTH
2830 commands, and RFC 1869 says the size should be increased for extended
2831 SMTP commands that take arguments.
2833 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2836 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2837 "unknown" error; now it says that the functionality isn't in the binary.
2839 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2840 an address' error message when a string expansion fails (syntax or
2841 whatever). Otherwise the password may appear in the log. Following change
2842 PH/42 below, there is no longer a chance of it appearing in a bounce
2845 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2847 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2848 of its addresses was ever tried. (Bugzilla bug #2.)
2850 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2851 the result incorrectly in the debug output. (It correctly added a newline
2852 to what was transported.)
2854 TF/01 Added $received_time.
2856 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2857 commented out examples of how to interface to a virus scanner and to
2858 SpamAssassin. Also added commented examples of av_scanner and
2859 spamd_address settings.
2861 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2862 and controls are allowed in which ACLs. There were a couple of minor
2863 errors. Some of the entries in the conditions table (which is a table of
2864 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2865 negation of where the condition IS allowed.
2867 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2869 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2870 header file does not have a version number, so I've had to invent a new
2871 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2872 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2873 radiusclient), but it was contributed by a Radius user.
2875 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2876 files or directories, for OpenSSL.
2878 PH/19 When an Exim process that is running as root has to create an Exim log
2879 file, it does so in a subprocess that runs as exim:exim so as to get the
2880 ownership right at creation (otherwise, other Exim processes might see
2881 the file with the wrong ownership). There was no test for failure of this
2882 fork() call, which would lead to the process getting stuck as it waited
2883 for a non-existent subprocess. Forks do occasionally fail when resources
2884 run out. I reviewed all the other calls to fork(); they all seem to check
2887 PH/20 When checking for unexpected SMTP input at connect time (before writing
2888 the banner), Exim was not dealing correctly with a non-positive return
2889 from the read() function. If the client had disconnected by this time,
2890 the result was a log entry for a synchronization error with an empty
2891 string after "input=" when read() returned zero. If read() returned -1
2892 (an event I could not check), uninitialized data bytes were printed.
2893 There were reports of junk text (parts of files, etc) appearing after
2896 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2898 PH/22 Added support for macro redefinition, and (re)definition in between
2899 driver and ACL definitions.
2901 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2902 forgetting to use the resulting value; it was using the unexpanded value.
2904 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2905 hadn't been configured. The fix is from Juergen Kreileder, who
2906 understands it better than I do:
2908 "Here's what I see happening with three configured cyrus_sasl
2909 authenticators configured (plain, login, cram-md5):
2911 On startup auth_cyrus_sasl_init() gets called for each of these.
2912 This means three calls to sasl_listmech() without a specified mech_list.
2913 => SASL tests which mechs of all available mechs actually work
2914 => three warnings about OTP not working
2915 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2917 With the patch, sasl_listmech() also gets called three times. But now
2918 SASL's mech_list option is set to the server_mech specified in the the
2919 authenticator. Or in other words, the answer from sasl_listmech()
2920 gets limited to just the mech you're testing for (which is different
2922 => the return list contains just 'plain' or 'login', 'cram-md5' or
2923 nothing depending on the value of ob->server_mech.
2925 I've just tested the patch: Authentication still works fine,
2926 unavailable mechs specified in the exim configuration are still
2927 caught, and the auth.log warnings about OTP are gone."
2929 PH/25 When debugging is enabled, the contents of the command line are added
2930 to the debugging output, even when log_selector=+arguments is not
2933 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2934 answer is "GNU", and only if the return is "GNU/something" is the answer
2937 PH/27 $acl_verify_message is now set immediately after the failure of a
2938 verification in an ACL, and so is available in subsequent modifiers. In
2939 particular, the message can be preserved by coding like this:
2941 warn !verify = sender
2942 set acl_m0 = $acl_verify_message
2944 Previously, $acl_verify_message was set only while expanding "message"
2945 and "log_message" when a very denied access.
2947 PH/28 Modified OS/os.c-Linux with
2949 -#ifndef OS_LOAD_AVERAGE
2950 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2952 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2953 nomenclature these days.)
2955 PH/29 Installed patch from the Sieve maintainer that adds the options
2956 sieve_useraddress and sieve_subaddress to the redirect router.
2958 PH/30 In these circumstances:
2959 . Two addresses routed to the same list of hosts;
2960 . First host does not offer TLS;
2961 . First host accepts first address;
2962 . First host gives temporary error to second address;
2963 . Second host offers TLS and a TLS session is established;
2964 . Second host accepts second address.
2965 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2966 and peerdn, if requested) that were in fact used only for the second
2969 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2970 attention to any local part prefix or suffix that was matched by the
2971 router that accepted the address. It now behaves in the same way as it
2972 does for delivery: the affixes are removed from the local part unless
2973 rcpt_include_affixes is set on the transport.
2975 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2976 timeout during the DATA phase of an incoming message.
2978 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2979 applied a patch sanctioned by the Sieve maintainer.
2981 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2982 the uid or gid is negative. A case of a negative gid caused this to be
2983 noticed. The fix allows for either to be negative.
2985 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2986 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2989 PH/36 The change PH/12 above was broken. Fixed it.
2991 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2992 the grounds that routing the same address twice would always produce the
2993 same answer. This might have been true once, but it is certainly no
2994 longer true now. Routing a child address may depend on the previous
2995 routing that produced that child. Some complicated redirection strategies
2996 went wrong when messages had multiple recipients, and made Exim's
2997 behaviour dependent on the order in which the addresses were given.
2999 I have moved the duplicate checking until after the routing is complete.
3000 Exim scans the addresses that are assigned to local and remote
3001 transports, and removes any duplicates. This means that more work will be
3002 done, as duplicates will always all be routed, but duplicates are
3003 presumably rare, so I don't expect this is of any significance.
3005 For deliveries to pipes, files, and autoreplies, the duplicate checking
3006 still happens during the routing process, since they are not going to be
3009 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3010 It corrects a timeout issue with spamd. This is Ian's comment: "The
3011 background is that sometimes spamd either never reads data from a
3012 connection it has accepted, or it never writes response data. The exiscan
3013 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3014 blindly assumes that writes won't block so it may never time out."
3016 PH/39 Allow G after quota size as well as K and M.
3018 PH/40 The value set for $authenticated_id in an authenticator may not contain
3019 binary zeroes or newlines because the value is written to log lines and
3020 to spool files. There was no check on this. Now the value is run through
3021 the string_printing() function so that such characters are converted to
3022 printable escape sequences.
3024 PH/41 $message_linecount is a new variable that contains the total number of
3025 lines in the message. Compare $body_linecount, which is the count for the
3028 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3029 bounce and delay warning messages, except in certain special cases, which
3032 (a) An SMTP error message from a remote host;
3033 (b) A message specified in a :fail: redirection;
3034 (c) A message specified in a "fail" command in a system filter;
3035 (d) A message specified in a FAIL return from the queryprogram router;
3036 (e) A message specified by the cannot_route_message router option.
3038 In these cases only, Exim does include the error details in bounce and
3039 warning messages. There are also a few cases where bland messages such
3040 as "unrouteable address" or "local delivery error" are given.
3042 PH/43 $value is now also set for the "else" part of a ${run expansion.
3044 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3045 being worked on, but at least Exim now implements the latest version to
3048 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3049 process to complete was treated as a delivery failure, a timeout while
3050 writing the message to the pipe was logged, but erroneously treated as a
3051 successful delivery. Such timeouts include transport filter timeouts. For
3052 consistency with the overall process timeout, these timeouts are now
3053 treated as errors, giving rise to delivery failures by default. However,
3054 there is now a new Boolean option for the pipe transport called
3055 timeout_defer, which, if set TRUE, converts the failures into defers for
3056 both kinds of timeout. A transport filter timeout is now identified in
3059 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3060 systems where "make" and "gmake" are different, calling "gmake" at top
3061 level broke things. I've arranged for the value of $(MAKE) to be passed
3062 from the Makefile to this script so that it can call the same version of
3066 A note about Exim versions 4.44 and 4.50
3067 ----------------------------------------
3069 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3070 changes of various kinds. As a consequence, a big documentation update was
3071 needed. This delayed the release for rather longer than seemed good, especially
3072 in the light of a couple of (minor) security issues. Therefore, the changes
3073 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3074 release. So 4.44 and 4.50 are in effect two different branches that both start
3077 I have left the 4.50 change log unchanged; it contains all the changes since
3078 4.43. The change log for 4.44 is below; many of its items are identical to
3079 those for 4.50. This seems to be the most sensible way to preserve the
3080 historical information.
3086 1. Minor wording change to the doc/README.SIEVE file.
3088 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3089 computation of the current number of files was incorrect.
3091 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3092 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3095 4. Give more explanation in the error message when the command for a transport
3096 filter fails to execute.
3098 5. There are several places where Exim runs a non-Exim command in a
3099 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3100 was being done only for the command run by the queryprogram router. It is
3101 now done for all such subprocesses. The other cases are: ${run, transport
3102 filters, and the commands run by the lmtp and pipe transports.
3104 6. Added CONFIGURE_GROUP build-time option.
3106 7. Some older OS have a limit of 256 on the maximum number of file
3107 descriptors. Exim was using setrlimit() to set 1000 as a large value
3108 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3109 systems. I've change it so that if it can't get 1000, it tries for 256.
3111 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3112 was an oversight, and furthermore, ever since the addition of extra
3113 controls (e.g. 4.43/32), the checks on when to allow different forms of
3114 "control" were broken. There should now be diagnostics for all cases when a
3115 control that does not make sense is encountered.
3117 9. Added the /retain_sender option to "control=submission".
3119 10. $recipients is now available in the predata ACL (oversight).
3121 11. Tidy the search cache before the fork to do a delivery from a message
3122 received from the command line. Otherwise the child will trigger a lookup
3123 failure and thereby defer the delivery if it tries to use (for example) a
3124 cached ldap connection that the parent has called unbind on.
3126 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3127 of $address_data from the recipient verification was clobbered by the
3128 sender verification.
3130 13. The value of address_data from a sender verification is now available in
3131 $sender_address_data in subsequent conditions in the ACL statement.
3133 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3135 15. Added a new option "connect=<time>" to callout options, to set a different
3138 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3139 was its contents. (It was OK if the option was not defined at all.)
3141 17. A "Completed" log line is now written for messages that are removed from
3142 the spool by the -Mrm option.
3144 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3145 information about exactly what failed.
3147 19. Added -dd to debug only the daemon process.
3149 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3150 handles timeouts, both on the server side and network timeouts. Renamed the
3151 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3153 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3154 It is now set to "smtps".
3156 22. $host_address is now set to the target address during the checking of
3157 ignore_target_hosts.
3159 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3160 being passed; this would have caused $sender_host_name to have been used if
3161 matching the list had actually called for a host name (not very likely,
3162 since this list is usually IP addresses). A host name is now passed as
3165 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3166 code that specifies a non-restarting handler (typically sigaction() in
3167 modern systems) in an attempt to fix a rare and obscure crash bug.
3169 25. Narrowed the window for a race in the daemon that could cause it to ignore
3170 SIGCHLD signals. This is not a major problem, because they are used only to
3171 wake it up if nothing else does.
3173 26. A malformed maildirsize file could cause Exim to calculate negative values
3174 for the mailbox size or file count. Odd effects could occur as a result.
3175 The maildirsize information is now recalculated if the size or filecount
3178 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3179 support for a long time. Removed HAVE_SYS_VFS_H.
3181 28. Installed the latest version of exipick from John Jetmore.
3183 29. In an address list, if the pattern was not a regular expression, an empty
3184 subject address (from a bounce message) matched only if the pattern was an
3185 empty string. Non-empty patterns were not even tested. This was the wrong
3186 because it is perfectly reasonable to use an empty address as part of a
3187 database query. An empty address is now tested by patterns that are
3188 lookups. However, all the other forms of pattern expect the subject to
3189 contain a local part and a domain, and therefore, for them, an empty
3190 address still always fails if the pattern is not itself empty.
3192 30. Exim went into a mad DNS loop when attempting to do a callout where the
3193 host was specified on an smtp transport, and looking it up yielded more
3194 than one IP address.
3196 31. Re-factored the code for checking spool and log partition space into a
3197 function that finds that data and another that does the check. The former
3198 is then used to implement four new variables: $spool_space, $log_space,
3199 $spool_inodes, and $log_inodes.
3201 32. The RFC2047 encoding function was originally intended for short strings
3202 such as real names; it was not keeping to the 75-character limit for
3203 encoded words that the RFC imposes. It now respects the limit, and
3204 generates multiple encoded words if necessary. To be on the safe side, I
3205 have increased the buffer size for the ${rfc2047: expansion operator from
3208 33. It is now permitted to omit both strings after an "if" condition; if the
3209 condition is true, the result is "true". As before, when the second string
3210 is omitted, a false condition yields an empty string. This makes it less
3211 cumbersome to write custom ACL and router conditions.
3213 34. Failure to deliver a bounce message always caused it to be frozen, even if
3214 there was an errors_to setting on the router. The errors_to setting is now
3217 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3218 canonical form (fully expanded) before being placed in
3219 $sender_host_address.
3221 36. The table in the code that translates DNS record types into text (T_A to
3222 "A" for instance) was missing entries for NS and CNAME. It is just possible
3223 that this could have caused confusion if both these types were looked up
3224 for the same domain, because the text type is used as part of Exim's
3225 per-process caching. But the chance of anyone hitting this buglet seems
3228 37. The dnsdb lookup has been extended in a number of ways.
3230 (1) There is a new type, "zns", which walks up the domain tree until it
3231 finds some nameserver records. It should be used with care.
3233 (2) There is a new type, "mxh", which is like "mx" except that it returns
3234 just the host names, not the priorities.
3236 (3) It is now possible to give a list of domains (or IP addresses) to be
3237 looked up. The behaviour when one of the lookups defers can be
3238 controlled by a keyword.
3240 (4) It is now possible to specify the separator character for use when
3241 multiple records are returned.
3243 38. The dnslists ACL condition has been extended: it is now possible to supply
3244 a list of IP addresses and/or domains to be looked up in a particular DNS
3247 39. Added log_selector=+queue_time_overall.
3249 40. When running the queue in the test harness, wait just a tad after forking a
3250 delivery process, to get repeatability of debugging output.
3252 41. Include certificate and key file names in error message when GnuTLS fails
3253 to set them up, because the GnuTLS error message doesn't include the name
3254 of the failing file when there is a problem reading it.
3256 42. Allow both -bf and -bF in the same test run.
3258 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3260 44. The "Exiscan patch" is now merged into the mainline Exim source.
3262 45. Sometimes the final signoff response after QUIT could fail to get
3263 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3264 < 0 before doing a fflush(). This bug looks as though it goes back to the
3265 introduction of TLS in release 3.20, but "sometimes" must have been rare
3266 because the tests only now provoked it.
3268 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3269 (this can affect the format of dates).
3271 47. exim_tidydb, when checking for the continued existence of a message for
3272 which it has found a message-specific retry record, was not finding
3273 messages that were in split spool directories. Consequently, it was
3274 deleting retry records that should have stayed in existence.
3276 48. Steve fixed some bugs in eximstats.
3278 49. The SPA authentication driver was not abandoning authentication and moving
3279 on to the next authenticator when an expansion was forced to fail,
3280 contradicting the general specification for all authenticators. Instead it
3281 was generating a temporary error. It now behaves as specified.
3283 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3284 (the order specifies the preference for clients). The order is now AES256,
3285 AES128, 3DES, ARCFOUR128.
3287 51. Small patch to Sieve code - explicitly set From: when generating an
3290 52. Exim crashed if a remote delivery caused a very long error message to be
3291 recorded - for instance if somebody sent an entire SpamAssassin report back
3292 as a large number of 550 error lines. This bug was coincidentally fixed by
3293 increasing the size of one of Exim's internal buffers (big_buffer) that
3294 happened as part of the Exiscan merge. However, to be on the safe side, I
3295 have made the code more robust (and fixed the comments that describe what
3298 53. Now that there can be additional text after "Completed" in log lines (if
3299 the queue_time_overall log selector is set), a one-byte patch to exigrep
3300 was needed to allow it to recognize "Completed" as not the last thing in
3303 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3304 patch that reportedly fixes this has been added. I am not expert enough to
3305 create a test for it. This is what the patch creator wrote:
3307 "I found a little strange behaviour of ldap code when working with
3308 Windows 2003 AD Domain, where users was placed in more than one
3309 Organization Units. When I tried to give exim partial DN, the exit code
3310 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3311 But simultaneously result of request was absolutely normal ldap result,
3312 so I produce this patch..."
3314 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3315 REFERENCE, so I have modified the code to exclude the patch when that macro
3318 55. Some experimental protocols are using DNS PTR records for new purposes. The
3319 keys for these records are domain names, not reversed IP addresses. The
3320 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3321 leaves it alone. Component reversal etc. now happens only for IP addresses.
3324 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3326 57. Double the size of the debug message buffer (to 2048) so that more of very
3327 long debug lines gets shown.
3329 58. The exicyclog utility now does better if the number of log files to keep
3330 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3332 59. Two changes related to the smtp_active_hostname option:
3334 (1) $smtp_active_hostname is now available as a variable.
3335 (2) The default for smtp_banner uses $smtp_active_hostname instead
3336 of $primary_hostname.
3338 60. The host_aton() function is supposed to be passed a string that is known
3339 to be a valid IP address. However, in the case of IPv6 addresses, it was
3340 not checking this. This is a hostage to fortune. Exim now panics and dies
3341 if the condition is not met. A case was found where this could be provoked
3342 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3343 components; fortuitously, this particular loophole had already been fixed
3344 by change 4.50/55 above.
3346 If there are any other similar loopholes, the new check in host_aton()
3347 itself should stop them being exploited. The report I received stated that
3348 data on the command line could provoke the exploit when Exim was running as
3349 exim, but did not say which command line option was involved. All I could
3350 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3351 running as the user.
3354 61. There was a buffer overflow vulnerability in the SPA authentication code
3355 (which came originally from the Samba project). I have added a test to the
3356 spa_base64_to_bits() function which I hope fixes it.
3359 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3360 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3362 63. The daemon start-up calls getloadavg() while still root for those OS that
3363 need the first call to be done as root, but it missed one case: when
3364 deliver_queue_load_max is set with deliver_drop_privilege. This is
3365 necessary for the benefit of the queue runner, because there is no re-exec
3366 when deliver_drop_privilege is set.
3368 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3369 This has been fixed.
3371 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3372 was in use, was not putting the data itself into the right store pool;
3373 consequently, it could be overwritten for a subsequent message in the same
3374 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3377 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3379 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3380 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3381 the function were treating the return as a boolean value, which happened to
3382 work because 0=false and not-0=true, but is not correct code.
3384 68. The host_aton() function was not handling scoped IPv6 addresses (those
3385 with, for example, "%eth0" on the end) correctly.
3387 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3388 negated items (that is, ~something) in unsigned ints. Some compilers
3389 apparently mutter when there is no cast.
3391 70. If an address verification called from an ACL failed, and did not produce a
3392 user-specific message (i.e. there was only a "system" message), nothing was
3393 put in $acl_verify_message. In this situation, it now puts the system
3396 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3397 session; change 4.31/43 added the unwanted input to the log line - except
3398 that it did not do this in the start of session case. It now does.
3400 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3401 This is wrong and can cause the other end to generate a synchronization
3402 error if it is another Exim or anything else that does the synchronization
3403 check. A QUIT command is no longer sent after a timeout.
3405 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3406 during host lookups.
3408 74. The defer_ok option of callout verification was not working if it was used
3409 when verifying addresses in header lines, that is, for this case:
3411 verify = header_sender/callout=defer_ok
3413 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3414 those file descriptors could be used for SMTP connections. If anything
3415 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3416 could be sent to the SMTP client, causing chaos. The daemon now opens
3417 stdin, stdout, and stderr to /dev/null when it puts itself into the
3420 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3421 log by default. The user can override this with suitable Perl magic.
3423 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3424 the log message when discard triggers, was not working for the DATA ACL or
3425 for the non-SMTP ACL.
3427 78. Error message wording change in sieve.c.
3429 79. If smtp_accept_max_per_host was set, the number of connections could be
3430 restricted to fewer than expected, because the daemon was trying to set up
3431 a new connection before checking whether the processes handling previous
3432 connections had finished. The check for completed processes is now done
3433 earlier. On busy systems, this bug wouldn't be noticed because something
3434 else would have woken the daemon, and it would have reaped the completed
3437 80. If a message was submitted locally by a user whose login name contained one
3438 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3439 It caused a spool format error. I have fixed the spool reading code. A
3440 related problem was that the "from" clause in the Received: line became
3441 illegal because of the space(s). It is now covered by ${quote_local_part.
3443 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3446 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3448 83. Patch from Sieve maintainer to fix unterminated string problem in
3449 "vacation" handling.
3451 84. Some minor changes to the Linux configuration files to help with other
3452 OS variants using glibc.
3454 85. One more patch for Sieve to update vacation handling to latest spec.
3457 ----------------------------------------------------
3458 See the note above about the 4.44 and 4.50 releases.
3459 ----------------------------------------------------
3465 1. Change 4.43/35 introduced a bug that caused file counts to be
3466 incorrectly computed when quota_filecount was set in an appendfile
3469 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3470 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3473 3. Give more explanation in the error message when the command for a transport
3474 filter fails to execute.
3476 4. There are several places where Exim runs a non-Exim command in a
3477 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3478 was being done only for the command run by the queryprogram router. It is
3479 now done for all such subprocesses. The other cases are: ${run, transport
3480 filters, and the commands run by the lmtp and pipe transports.
3482 5. Some older OS have a limit of 256 on the maximum number of file
3483 descriptors. Exim was using setrlimit() to set 1000 as a large value
3484 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3485 systems. I've change it so that if it can't get 1000, it tries for 256.
3487 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3488 was an oversight, and furthermore, ever since the addition of extra
3489 controls (e.g. 4.43/32), the checks on when to allow different forms of
3490 "control" were broken. There should now be diagnostics for all cases when a
3491 control that does not make sense is encountered.
3493 7. $recipients is now available in the predata ACL (oversight).
3495 8. Tidy the search cache before the fork to do a delivery from a message
3496 received from the command line. Otherwise the child will trigger a lookup
3497 failure and thereby defer the delivery if it tries to use (for example) a
3498 cached ldap connection that the parent has called unbind on.
3500 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3501 of $address_data from the recipient verification was clobbered by the
3502 sender verification.
3504 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3505 was its contents. (It was OK if the option was not defined at all.)
3507 11. A "Completed" log line is now written for messages that are removed from
3508 the spool by the -Mrm option.
3510 12. $host_address is now set to the target address during the checking of
3511 ignore_target_hosts.
3513 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3514 being passed; this would have caused $sender_host_name to have been used if
3515 matching the list had actually called for a host name (not very likely,
3516 since this list is usually IP addresses). A host name is now passed as
3519 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3520 code that specifies a non-restarting handler (typically sigaction() in
3521 modern systems) in an attempt to fix a rare and obscure crash bug.
3523 15. Narrowed the window for a race in the daemon that could cause it to ignore
3524 SIGCHLD signals. This is not a major problem, because they are used only to
3525 wake it up if nothing else does.
3527 16. A malformed maildirsize file could cause Exim to calculate negative values
3528 for the mailbox size or file count. Odd effects could occur as a result.
3529 The maildirsize information is now recalculated if the size or filecount
3532 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3533 support for a long time. Removed HAVE_SYS_VFS_H.
3535 18. Updated exipick to current release from John Jetmore.
3537 19. Allow an empty sender to be matched against a lookup in an address list.
3538 Previously the only cases considered were a regular expression, or an
3541 20. Exim went into a mad DNS lookup loop when doing a callout where the
3542 host was specified on the transport, if the DNS lookup yielded more than
3545 21. The RFC2047 encoding function was originally intended for short strings
3546 such as real names; it was not keeping to the 75-character limit for
3547 encoded words that the RFC imposes. It now respects the limit, and
3548 generates multiple encoded words if necessary. To be on the safe side, I
3549 have increased the buffer size for the ${rfc2047: expansion operator from
3552 22. Failure to deliver a bounce message always caused it to be frozen, even if
3553 there was an errors_to setting on the router. The errors_to setting is now
3556 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3557 canonical form (fully expanded) before being placed in
3558 $sender_host_address.
3560 24. Updated eximstats to version 1.33
3562 25. Include certificate and key file names in error message when GnuTLS fails
3563 to set them up, because the GnuTLS error message doesn't include the name
3564 of the failing file when there is a problem reading it.
3566 26. Expand error message when OpenSSL has problems setting up cert/key files.
3569 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3570 (this can affect the format of dates).
3572 28. exim_tidydb, when checking for the continued existence of a message for
3573 which it has found a message-specific retry record, was not finding
3574 messages that were in split spool directories. Consequently, it was
3575 deleting retry records that should have stayed in existence.
3577 29. eximstats updated to version 1.35
3578 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3579 1.35 - bugfix such that pie charts by volume are generated correctly
3581 30. The SPA authentication driver was not abandoning authentication and moving
3582 on to the next authenticator when an expansion was forced to fail,
3583 contradicting the general specification for all authenticators. Instead it
3584 was generating a temporary error. It now behaves as specified.
3586 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3587 (the order specifies the preference for clients). The order is now AES256,
3588 AES128, 3DES, ARCFOUR128.
3590 31. Small patch to Sieve code - explicitly set From: when generating an
3593 32. Exim crashed if a remote delivery caused a very long error message to be
3594 recorded - for instance if somebody sent an entire SpamAssassin report back
3595 as a large number of 550 error lines. This bug was coincidentally fixed by
3596 increasing the size of one of Exim's internal buffers (big_buffer) that
3597 happened as part of the Exiscan merge. However, to be on the safe side, I
3598 have made the code more robust (and fixed the comments that describe what
3601 33. Some experimental protocols are using DNS PTR records for new purposes. The
3602 keys for these records are domain names, not reversed IP addresses. The
3603 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3604 leaves it alone. Component reversal etc. now happens only for IP addresses.
3607 34. The host_aton() function is supposed to be passed a string that is known
3608 to be a valid IP address. However, in the case of IPv6 addresses, it was
3609 not checking this. This is a hostage to fortune. Exim now panics and dies
3610 if the condition is not met. A case was found where this could be provoked
3611 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3612 components; fortuitously, this particular loophole had already been fixed
3613 by change 4.50/55 or 4.44/33 above.
3615 If there are any other similar loopholes, the new check in host_aton()
3616 itself should stop them being exploited. The report I received stated that
3617 data on the command line could provoke the exploit when Exim was running as
3618 exim, but did not say which command line option was involved. All I could
3619 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3620 running as the user.
3623 35. There was a buffer overflow vulnerability in the SPA authentication code
3624 (which came originally from the Samba project). I have added a test to the
3625 spa_base64_to_bits() function which I hope fixes it.
3628 36. The daemon start-up calls getloadavg() while still root for those OS that
3629 need the first call to be done as root, but it missed one case: when
3630 deliver_queue_load_max is set with deliver_drop_privilege. This is
3631 necessary for the benefit of the queue runner, because there is no re-exec
3632 when deliver_drop_privilege is set.
3634 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3635 was in use, was not putting the data itself into the right store pool;
3636 consequently, it could be overwritten for a subsequent message in the same
3637 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3640 38. Sometimes the final signoff response after QUIT could fail to get
3641 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3642 < 0 before doing a fflush(). This bug looks as though it goes back to the
3643 introduction of TLS in release 3.20, but "sometimes" must have been rare
3644 because the tests only now provoked it.
3650 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3651 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3652 Now it returns an integer. A number of calls were still expecting a T/F
3653 return. Fortuitously, in all cases, the tests worked in OK situations,
3654 which is the norm. However, things would have gone wrong on any write
3655 failures on the smtp file descriptor. This function is used when sending
3656 messages over SMTP and also when doing verify callouts.
3658 2. When Exim is called to do synchronous delivery of a locally submitted
3659 message (the -odf or -odi options), it no longer closes stderr before doing
3662 3. Implemented the mua_wrapper option.
3664 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3666 5. Implemented the functions header_remove(), header_testname(),
3667 header_add_at_position(), and receive_remove_recipient(), and exported them
3670 6. If an ACL "warn" statement specified the addition of headers, Exim already
3671 inserted X-ACL-Warn: at the start if there was no header name. However, it
3672 was not making this test for the second and subsequent header lines if
3673 there were newlines in the string. This meant that an invalid header could
3674 be inserted if Exim was badly configured.
3676 7. Allow an ACL "warn" statement to add header lines at the start or after all
3677 the Received: headers, as well as at the end.
3679 8. Added the rcpt_4xx retry error code.
3681 9. Added postmaster_mailfrom=xxx to callout verification option.
3683 10. Added mailfrom=xxxx to the callout verification option, for verify=
3686 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3687 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3689 12. Inserted some casts to stop certain compilers warning when using pointer
3690 differences as field lengths or precisions in printf-type calls (mostly
3691 affecting debugging statements).
3693 13. Added optional readline() support for -be (dynamically loaded).
3695 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3696 same clock tick as a message's arrival, so that its received time was the
3697 same as the "first fail" time on the retry record, and that message
3698 remained on the queue past the ultimate address timeout, every queue runner
3699 would try a delivery (because it was past the ultimate address timeout) but
3700 after another failure, the ultimate address timeout, which should have then
3701 bounced the address, did not kick in. This was a "< instead of <=" error;
3702 in most cases the first failure would have been in the next clock tick
3703 after the received time, and all would be well.
3705 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3706 being recognized when the domain list was tested by the match_domain
3707 condition in an expansion string.
3709 16. Added the ${str2b64: operator.
3711 17. Exim was always calling setrlimit() to set a large limit for the number of
3712 processes, without checking whether the existing limit was already
3713 adequate. (It did check for the limit on file descriptors.) Furthermore,
3714 errors from getrlimit() and setrlimit() were being ignored. Now they are
3715 logged to the main and panic logs, but Exim does carry on, to try to do its
3716 job under whatever limits there are.
3718 18. Imported PCRE 5.0.
3720 19. Trivial typo in log message " temporarily refused connection" (the leading
3723 20. If the log selector return_path_on_delivery was set and an address was
3724 redirected to /dev/null, the delivery process crashed because it assumed
3725 that a return path would always be set for a "successful" delivery. In this
3726 case, the whole delivery is bypassed as an optimization, and therefore no
3729 21. Internal re-arrangement: the function for sending a challenge and reading
3730 a response while authentication was assuming a zero-terminated challenge
3731 string. It's now changed to take a pointer and a length, to allow for
3732 binary data in such strings.
3734 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3736 23. Exim was not respecting finduser_retries when seeking the login of the
3737 uid under which it was called; it was always trying 10 times. (The default
3738 setting of finduser_retries is zero.) Also, it was sleeping after the final
3739 failure, which is pointless.
3741 24. Implemented tls_on_connect_ports.
3743 25. Implemented acl_smtp_predata.
3745 26. If the domain in control=submission is set empty, Exim assumes that the
3746 authenticated id is a complete email address when it generates From: or
3747 Sender: header lines.
3749 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3750 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3751 chown and chgrp in /bin and hostname in /usr/bin.
3753 28. Exim was keeping the "process log" file open after each use, just as it
3754 does for the main log. This opens the possibility of it remaining open for
3755 long periods when the USR1 signal hits a daemon. Occasional processlog
3756 errors were reported, that could have been caused by this. Anyway, it seems
3757 much more sensible not to leave this file open at all, so that is what now
3760 29. The long-running daemon process does not normally write to the log once it
3761 has entered its main loop, and it closes the log before doing so. This is
3762 so that log files can straightforwardly be renamed and moved. However,
3763 there are a couple of unusual error situations where the daemon does write
3764 log entries, and I had neglected to close the log afterwards.
3766 30. The text of an SMTP error response that was received during a remote
3767 delivery was being truncated at 512 bytes. This is too short for some of
3768 the long messages that one sometimes sees. I've increased the limit to
3771 31. It is now possible to make retry rules that apply only when a message has a
3772 specific sender, in particular, an empty sender.
3774 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3775 it possible to be selective about when SMTP synchronization is enforced.
3777 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3779 32. Implemented hosts_connection_nolog.
3781 33. Added an ACL for QUIT.
3783 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3786 35. Added mailbox_size and mailbox_filecount to appendfile.
3788 36. Added control = no_multiline_responses to ACLs.
3790 37. There was a bug in the logic of the code that waits for the clock to tick
3791 in the case where the clock went backwards by a substantial amount such
3792 that the microsecond fraction of "now" was more than the microsecond
3793 fraction of "then" (but the whole seconds number was less).
3795 38. Added support for the libradius Radius client library this is found on
3796 FreeBSD (previously only the radiusclient library was supported).
3802 1. When certain lookups returned multiple values in the form name=value, the
3803 quoting of the values was not always being done properly. Specifically:
3804 (a) If the value started with a double quote, but contained no whitespace,
3806 (b) If the value contained whitespace other than a space character (i.e.
3807 tabs or newlines or carriage returns) it was not quoted.
3808 This fix has been applied to the mysql and pgsql lookups by writing a
3809 separate quoting function and calling it from the lookup code. The fix
3810 should probably also be applied to nisplus, ibase and oracle lookups, but
3811 since I cannot test any of those, I have not disturbed their existing code.
3813 2. A hit in the callout cache for a specific address caused a log line with no
3814 reason for rejecting RCPT. Now it says "Previous (cached) callout
3815 verification failure".
3817 3. There was an off-by-one bug in the queryprogram router. An over-long
3818 return line was truncated at 256 instead of 255 characters, thereby
3819 overflowing its buffer with the terminating zero. As well as fixing this, I
3820 have increased the buffer size to 1024 (and made a note to document this).
3822 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3823 when Exim is waiting for an SMTP response from a remote server, Exim
3824 restarts its select() call on the socket, thereby resetting its timeout.
3825 This is not a problem when such interrupts are rare. Somebody set up a cron
3826 job to run exiwhat every 2 minutes, which is less than the normal select()
3827 timeout (5 or 10 minutes). This meant that the select() timeout never
3828 kicked in because it was always reset. I have fixed this by comparing the
3829 time when an interrupt arrives with the time at the start of the first call
3830 to select(). If more time than the timeout has elapsed, the interrupt is
3831 treated as a timeout.
3833 5. Some internal re-factoring in preparation for the addition of Sieve
3834 extensions (by MH). In particular, the "personal" test is moved to a
3835 separate function, and given an option for scanning Cc: and Bcc: (which is
3836 not set for Exim filters).
3838 6. When Exim created an email address using the login of the caller as the
3839 local part (e.g. when creating a From: or Sender: header line), it was not
3840 quoting the local part when it contained special characters such as @.
3842 7. Installed new OpenBSD configuration files.
3844 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3845 try to make them clearer.
3847 9. Callout options, other than the timeout value, were being ignored when
3848 verifying sender addresses in header lines. For example, when using
3850 verify = header_sender/callout=no_cache
3852 the cache was (incorrectly) being used.
3854 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3855 only the Cygwin environment.
3857 11. When return_path_on_delivery was set as a log selector, if different remote
3858 addresses in the same message used different return paths and parallel
3859 remote delivery occurred, the wrong values would sometimes be logged.
3860 (Whenever a remote delivery process finished, the return path value from
3861 the most recently started remote delivery process was logged.)
3863 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3864 lines when AUTH and/or TLS are in use. This is the "received protocol"
3865 field. Exim used to use "asmtp" for authenticated SMTP, without any
3866 indication (in the protocol name) for TLS use. Now it follows the RFC and
3867 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3868 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3869 names appear in log lines as well as in Received: header lines.
3871 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3872 extensions, and comparison tests, and to fix some bugs.
3874 14. Changes to the "personal" filter test:
3876 (1) The test was buggy in that it was just doing the equivalent of
3877 "contains" tests on header lines. For example, if a user's address was
3878 anne@some.where, the "personal" test would incorrectly be true for
3880 To: susanne@some.where
3882 This test is now done by extracting each address from the header in turn,
3883 and checking the entire address. Other tests that are part of "personal"
3884 are now done using regular expressions (for example, to check local parts
3885 of addresses in From: header lines).
3887 (2) The list of non-personal local parts in From: addresses has been
3888 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3889 taken from the Sieve specification recommendations.
3891 (3) If the message contains any header line starting with "List-" it is
3892 treated as non-personal.
3894 (4) The test for "circular" in the Subject: header line has been removed
3895 because it now seems ill-conceived.
3897 15. Minor typos in src/EDITME comments corrected.
3899 16. Installed latest exipick from John Jetmore.
3901 17. If headers_add on a router specified a text string that was too long for
3902 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3903 of string_sprintf() is now avoided.
3905 18. $message_body_size was not set (it was always zero) when running the DATA
3906 ACL and the local_scan() function.
3908 19. For the "mail" command in an Exim filter, no default was being set for
3909 the once_repeat time, causing a random time value to be used if "once" was
3910 specified. (If the value happened to be <= 0, no repeat happened.) The
3911 default is now 0s, meaning "never repeat". The "vacation" command was OK
3912 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3913 (I found it when inspecting the code).
3915 20. There is now an overall timeout for performing a callout verification. It
3916 defaults to 4 times the callout timeout, which applies to individual SMTP
3917 commands during the callout. The overall timeout applies when there is more
3918 than one host that can be tried. The timeout is checked before trying the
3919 next host. This prevents very long delays if there are a large number of
3920 hosts and all are timing out (e.g. when the network connections are timing
3921 out). The value of the overall timeout can be changed by specifying an
3922 additional sub-option for "callout", called "maxwait". For example:
3924 verify = sender/callout=5s,maxwait=20s
3926 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3927 to the end before writing, but this should make it even safer).
3929 22. Exim was forgetting that it had advertised PIPELINING for the second and
3930 subsequent messages on an SMTP connection. It was also not resetting its
3931 memory on STARTTLS and an internal HELO.
3933 23. When Exim logs an SMTP synchronization error within a session, it now
3934 records whether PIPELINING has been advertised or not.
3936 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3937 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3938 rather than long int.
3940 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3942 26. Added the never_mail option to autoreply.
3948 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3949 crash if the getsockname() call failed; this can happen if a connection is
3950 closed very soon after it is established. The problem was simply in the
3951 order in which certain operations were done, causing Exim to try to write
3952 to the SMTP stream before it had set up the file descriptor. The bug has
3953 been fixed by making things happen in the correct order.
3959 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3960 before the connection was closed, thus losing the rejection response.
3962 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3963 some early Solaris releases, but causes trouble in current releases where
3964 socklen_t is defined.
3966 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3969 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3972 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3973 that releasing the top of it at the end releases what was used for sub-
3974 expansions (unless the block got too big). However, discard this block if
3975 the first thing is a variable or header, so that we can use its block when
3976 it is dynamic (useful for very large $message_headers, for example).
3978 6. Lookups now cache *every* query, not just the most recent. A new, separate
3979 store pool is used for this. It can be recovered when all lookup caches are
3980 flushed. Lookups now release memory at the end of their result strings.
3981 This has involved some general refactoring of the lookup sources.
3983 7. Some code has been added to the store_xxx() functions to reduce the amount
3984 of flapping under certain conditions.
3986 8. log_incoming_interface used to affect only the <= reception log lines. Now
3987 it causes the local interface and port to be added to several more SMTP log
3988 lines, for example "SMTP connection from", and rejection lines.
3990 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3992 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3994 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3995 could be overwritten at the end of the current message (or the start of a
3996 new message if it was set in a HELO ACL). The value is now preserved for
3997 the duration of the SMTP connection.
3999 12. If a transport had a headers_rewrite setting, and a matching header line
4000 contained an unqualified address, that address was qualified, even if it
4001 did not match any rewriting rules. The underlying bug was that the values
4002 of the flags that permit the existence of unqualified sender and recipient
4003 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4004 non-local messages, and by -bnq for local messages) were not being
4005 preserved with the message after it was received.
4007 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4008 "next input=" as part of the text comprising the host identity instead of
4009 the correct text. The code was using the same buffer for two different
4010 strings. However, depending on which order the printing function evaluated
4011 its arguments, the bug did not always show up. Under Linux, for example, my
4012 test suite worked just fine.
4014 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4015 doesn't work with some older versions of Perl. It has been changed to "my",
4016 which in any case is probably the better facility to use.
4018 15. A really picky compiler found some instances of statements for creating
4019 error messages that either had too many or two few arguments for the format
4022 16. The size of the buffer for calls to the DNS resolver has been increased
4023 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4024 for addresses that have a lot of PTR records. This alleviates a problem; it
4025 does not fully solve it.
4027 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4028 buffer now truncates the list and logs the incident, which is the same
4029 action as happens when Exim is looking up a host name and its aliases.
4030 Previously in this situation something unpredictable would happen;
4031 sometimes it was "internal error: store_reset failed".
4033 18. If a server dropped the connection unexpectedly when an Exim client was
4034 using GnuTLS and trying to read a response, the client delivery process
4035 crashed while trying to generate an error log message.
4037 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4038 string, for example:
4040 warn message = H1: something\nH2: something
4042 the text was added as a single header line from Exim's point of view
4043 though it ended up OK in the delivered message. However, searching for the
4044 second and subsequent header lines using $h_h2: did not work. This has been
4045 fixed. Similarly, if a system filter added multiple headers in this way,
4046 the routers could not see them.
4048 20. Expanded the error message when iplsearch is called with an invalid key to
4049 suggest using net-iplsearch in a host list.
4051 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4052 ACLs are no longer actually imposed (and a message to that effect is
4055 22. If a "gecos" field in a passwd entry contained escaped characters, in
4056 particular, if it contained a \" sequence, Exim got it wrong when building
4057 a From: or a Sender: header from that name. A second bug also caused
4058 incorrect handling when an unquoted " was present following a character
4059 that needed quoting.
4061 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4062 was not being matched caselessly.
4064 24. Arranged for all hyphens in the exim.8 source to be escaped with
4067 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4068 properly. Recipient callout cache records were still being keyed to include
4069 the sender, even when use_sender was set false. This led to far more
4070 callouts that were necessary. The sender is no longer included in the key
4071 when use_sender is false.
4073 26. Added "control = submission" modifier to ACLs.
4075 27. Added the ${base62d: operator to decode base 62 numbers.
4077 28. dnsdb lookups can now access SRV records.
4079 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4080 the configuration file.
4082 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4083 (-v) mode. This makes the output for a verbose queue run more intelligible.
4085 31. Added a use_postmaster feature to recipient callouts.
4087 32. Added the $body_zerocount variable, containing the number of binary zero
4088 bytes in the message body.
4090 33. The time of last modification of the "new" subdirectory is now used as the
4091 "mailbox time last read" when there is a quota error for a maildir
4094 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4096 35. Added +ignore_unknown as a special item in host lists.
4098 36. Code for decoding IPv6 addresses in host lists is now included, even if
4099 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4100 address was recognized as an IP address, but was then not correctly decoded
4101 into binary, causing unexpected and incorrect effects when compared with
4108 1. Very minor rewording of debugging text in manualroute to say "list of
4109 hosts" instead of "hostlist".
4111 2. If verify=header_syntax was set, and a header line with an unqualified
4112 address (no domain) and a large number of spaces between the end of the
4113 name and the colon was received, the reception process suffered a buffer
4114 overflow, and (when I tested it) crashed. This was caused by some obsolete
4115 code that should have been removed. The fix is to remove it!
4117 3. When running in the test harness, delay a bit after writing a bounce
4118 message to get a bit more predictability in the log output.
4120 4. Added a call to search_tidyup() just before forking a reception process. In
4121 theory, someone could use a lookup in the expansion of smtp_accept_max_
4122 per_host which, without the tidyup, could leave open a database connection.
4124 5. Added the variables $recipient_data and $sender_data which get set from a
4125 lookup success in an ACL "recipients" or "senders" condition, or a router
4126 "senders" option, similar to $domain_data and $local_part_data.
4128 6. Moved the writing of debug_print from before to after the "senders" test
4131 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4132 problems for message scanning, either using a data ACL, or using
4133 local_scan() because the Received: header was not generated till after they
4134 were called (in order to set the time as the time of reception completion).
4135 I have revised the way this works. The header is now generated after the
4136 body is received, but before the ACL or local_scan() are called. After they
4137 are run, the timestamp in the header is updated.
4143 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4144 before starting a queue runner without re-exec. This happened only when
4145 deliver_drop_privilege was set or when the Exim user was set to root. The
4146 effect of the bug was that timeouts during subsequent deliveries caused
4147 crashes instead of being properly handled. The handler is now left at its
4148 default (and expected) setting.
4150 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4151 message, again when deliver_drop_privilege is set or Exim is run as root.
4152 The bug described in (1) was not present in this case, but the tidying up
4153 of the other signals was missing. I have made the two cases consistent.
4155 3. The ignore_target_hosts setting on a manualroute router was being ignored
4156 for hosts that were looked up using the /MX notation.
4158 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4161 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4162 operated on the sender address. After changing the $sender_address to <>
4163 for the sender address verify, Exim was re-instated it as the original
4164 (before rewriting) address, but remembering that it had rewritten it, so it
4165 wasn't rewriting it again. This bug also had the effect of breaking the
4166 sender address verification caching when the sender address was rewritten.
4168 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4169 This has been changed so that if the ip literal address matches
4170 ignore_target_hosts, the router declines.
4172 7. Added expansion conditions match_domain, match_address, and match_local_
4173 part (NOT match_host).
4175 8. The placeholder for the Received: header didn't have a length field set.
4177 9. Added code to Exim itself and to exim_lock to test for a specific race
4178 condition that could lead to file corruption when using MBX delivery. The
4179 issue is with the lockfile that is created in /tmp. If this file is removed
4180 after a process has opened it but before that process has acquired a lock,
4181 there is the potential for a second process to recreate the file and also
4182 acquire a lock. This could lead to two Exim processes writing to the file
4183 at the same time. The added code performs the same test as UW imapd; it
4184 checks after acquiring the lock that its file descriptor still refers to
4185 the same named file.
4187 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4188 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4191 11. Added the smtp_active_hostname option. If used, this will typically be made
4192 to depend on the incoming interface address. Because $interface_address is
4193 not set up until the daemon has forked a reception process, error responses
4194 that can happen earlier (such as "too many connections") no longer contain
4197 12. If an expansion in a condition on a "warn" statement fails because a lookup
4198 defers, the "warn" statement is abandoned, and the next ACL statement is
4199 processed. Previously this caused the whole ACL to be aborted.
4201 13. Added the iplsearch lookup type.
4203 14. Added ident_timeout as a log selector.
4205 15. Added tls_certificate_verified as a log selector.
4207 16. Added a global option tls_require_ciphers (compare the smtp transport
4208 option of the same name). This controls incoming TLS connections.
4210 17. I finally figured out how to make tls_require_ciphers do a similar thing
4211 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4212 before starting the TLS session.
4214 18. Tabs are now shown as \t in -bP output.
4216 19. If the log selector return_path_on_delivery was set, Exim crashed when
4217 bouncing a message because it had too many Received: header lines.
4219 20. If two routers both had headers_remove settings, and the first one included
4220 a superfluous trailing colon, the final name in the first list and the
4221 first name in the second list were incorrectly joined into one item (with a
4222 colon in the middle).
4228 1. Added -C and -D options to the exinext utility, mainly to make it easier
4229 to include in the automated testing, but these could be helpful when
4230 multiple configurations are in use.
4232 2. The exinext utility was not formatting the output nicely when there was
4233 an alternate port involved in the retry record key, nor when there was a
4234 message id as well (for retries that were specific to a specific message
4235 and a specific host). It was also confused by IPv6 addresses, because of
4236 the additional colons they contain. I have fixed the IPv4 problem, and
4237 patched it up to do a reasonable job for IPv6.
4239 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4240 delivery, the log line now contains "pipelined" if PIPELINING was used.
4242 4. An SMTP transport process used to panic and die if the bind() call to set
4243 an explicit outgoing interface failed. This has been changed; it is now
4244 treated in the same way as a connect() failure.
4246 5. A reference to $sender_host_name in the part of a conditional expansion
4247 that was being skipped was still causing a DNS lookup. This no longer
4250 6. The def: expansion condition was not recognizing references to header lines
4251 that used bh_ and bheader_.
4253 7. Added the _cache feature to named lists.
4255 8. The code for checking quota_filecount in the appendfile transport was
4256 allowing one more file than it should have been.
4258 9. For compatibility with Sendmail, the command line option
4266 and sets the incoming protocol and host name (for trusted callers). The
4267 host name and its colon can be omitted when only the protocol is to be set.
4268 Note the Exim already has two private options, -pd and -ps, that refer to
4269 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4270 "s", but I don't think that's a major issue.
4272 10. A number of refactoring changes to the code, none of which should affect
4275 (a) The number of logging options was getting close to filling up the
4276 32-bit word that was used as a bit map. I have split them into two classes:
4277 those that are passed in the argument to log_write(), and those that are
4278 only ever tested independently outside of that function. These are now in
4279 separate 32-bit words, so there is plenty of room for expansion again.
4280 There is no change in the user interface or the logging behaviour.
4282 (b) When building, for example, log lines, the code previously used a
4283 macro that called string_cat() twice, in order to add two strings. This is
4284 not really sufficiently general. Furthermore, there was one instance where
4285 it was actually wrong because one of the argument was used twice, and in
4286 one call a function was used. (As it happened, calling the function twice
4287 did not affect the overall behaviour.) The macro has been replaced by a
4288 function that can join an arbitrary number of extra strings onto a growing
4291 (c) The code for expansion conditions now uses a table and a binary chop
4292 instead of a serial search (which was left over from when there were very
4293 few conditions). Also, it now recognizes conditions like "pam" even when
4294 the relevant support is not compiled in: a suitably worded error message is
4295 given if an attempt is made to use such a condition.
4297 11. Added ${time_interval:xxxxx}.
4299 12. A bug was causing one of the ddress fields not to be passed back correctly
4300 from remote delivery subprocesses. The field in question was not being
4301 subsequently used, so this caused to problems in practice.
4303 13. Added new log selectors queue_time and deliver_time.
4305 14. Might have fixed a bug in maildirsizefile handling that threw up
4306 "unexpected character" debug warnings, and recalculated the data
4307 unnecessarily. In any case, I expanded the warning message to give more
4310 15. Added the message "Restricted characters in address" to the statements in
4311 the default ACL that block characters like @ and % in local parts.
4313 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4314 Three changes have been made:
4316 (a) There was a serious bug; a negative response to MAIL caused the whole
4317 recipient domain to be cached as invalid, thereby blocking all messages
4318 to all local parts at the same domain, from all senders. This bug has
4319 been fixed. The domain is no longer cached after a negative response to
4320 MAIL if the sender used is not empty.
4322 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4325 (c) A new callout option, "use_sender" has been added for people who want
4326 the modified behaviour.
4332 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4335 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4336 indeed breaks things for older releases.
4338 3. Added additional logging to the case where there is a problem reading data
4339 from a filter that is running in a subprocess using a pipe, in order to
4340 try to track down a specific problem.
4342 4. Testing facility fudge: when running in the test harness and attempting
4343 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4344 getting "No route to host". Convert this to a timeout.
4346 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4349 6. Some OS don't have socklen_t but use size_t instead. This affects the
4350 fifth argument of getsockopt() amongst other things. This is now
4351 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4352 can be set for individual OS. I have set it for SunOS5, OSF1, and
4353 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4354 some earlier ones do not.
4356 7. Change 4.30/15 was not doing the test caselessly.
4358 8. The standard form for an IPv6 address literal was being rejected by address
4359 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4360 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4361 this, as well as the form without the "IPv6" on the front (but only when
4362 address literals are enabled, of course).
4364 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4366 10. Exim crashed if a message with an empty sender address specified by -f
4367 encountered a router with an errors_to setting. This could be provoked only
4368 by a command such as
4372 where an empty string was supplied; "<>" did not hit this bug.
4374 11. Installed PCRE release 4.5.
4376 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4377 remained set. It is now erased.
4379 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4380 times from message ids (which are base 36 rather than the normal 62).
4382 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4383 were being counted as actual protocol errors, and logged if the log
4384 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4385 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4386 and DATA following a set of rejected RCPTs do not count as protocol errors.
4387 In other words, Exim assumes they were pipelined, though this may not
4388 actually be the case. Of course, in all cases the client gets an
4389 appropriate error code.
4391 15. If a lookup fails in an ACL condition, a message about the failure may
4392 be available; it is used if testing the ACL cannot continue, because most
4393 such messages specify what the cause of the deferral is. However, some
4394 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4395 that caused an old message to be retained and used if a later statement
4396 caused a defer, replacing the real cause of the deferral.
4398 16. If an IP address had so many PTR records that the DNS lookup buffer
4399 was not large enough to hold them, Exim could crash while trying to process
4400 the truncated data. It now detects and logs this case.
4402 17. Further to 4.21/58, another change has been made: if (and only if) the
4403 first line of a message (the first header line) ends with CRLF, a bare LF
4404 in a subsequent header line has a space inserted after it, so as not to
4405 terminate the header.
4407 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4408 unnecessarily, used atoi() instead of strtol(), and didn't check the
4409 termination when getting file sizes from file names by regex.
4411 19. Completely re-implemented the support for maildirsize files, in the light
4412 of a number of problems with the previous contributed implementation
4413 (4.30/29). In particular:
4415 . If the quota is zero, the maildirsize file is maintained, but no quota is
4418 . If the maildir directory does not exist, it is created before any attempt
4419 to write a maildirsize file.
4421 . The quota value in the file is just a cache; if the quota is changed in
4422 the transport, the new value overrides.
4424 . A regular expression is available for excluding directories from the
4427 20. The autoreply transport checks the characters in options that define the
4428 message's headers; it allows continued headers, but it was checking with
4429 isspace() after an embedded newline instead of explicitly looking for a
4432 21. If all the "regular" hosts to which an address was routed had passed their
4433 expiry times, and had not reached their retry times, the address was
4434 bounced, even if fallback hosts were defined. Now Exim should go on to try
4437 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4438 equivalent code in the SMTP transport. Some hosts send humungous responses
4439 to HELO/EHLO, more than 1024 it seems.
4441 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4442 gives compiler warnings in some environments. I've now done it "properly",
4445 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4446 (because of problems with the built-in one) was declared to return uschar *
4447 instead of char *, causing compiler failure.
4449 25. Fixed a file descriptor leak when processing alias/forward files.
4451 26. Fixed a minor format string issue in dbfn.c.
4453 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4455 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4456 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4459 29. When a sender address is verified, it is cached, to save repeating the test
4460 when there is more than one recipient in a message. However, when the
4461 verification involves a callout, it is possible for different callout
4462 options to be set for different recipients. It is too complicated to keep
4463 track of this in the cache, so now Exim always runs a verification when a
4464 callout is required, relying on the callout cache for the optimization.
4465 The overhead is duplication of the address routing, but this should not be
4468 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4469 to be verified with callout=postmaster, and the main callout worked but the
4470 postmaster check failed, the verification correctly failed. However, if a
4471 subsequent RCPT command asked for sender verification *without* the
4472 postmaster check, incorrect caching caused this verification also to fail,
4475 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4476 it was not caching the DNS options (qualify_single, search_parents) that
4477 were used when the lookup failed. A subsequent lookup with different
4478 options therefore always gave the same answer, though there were cases
4479 where it should not have. (Example: a "domains = !$mx_any" option on a
4480 dnslookup router: the "domains" option is always processed without any
4481 widening, but the router might have qualify_single set.) Now Exim uses the
4482 cached value only when the same options are set.
4484 32. Added John Jetmore's "exipick" utility to the distribution.
4486 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4487 than a timeout (e.g. a certificate is required, and is not provided), an
4488 Exim server now closes the connection immediately. Previously it waited for
4489 the client to close - but if the client is SSL, it seems that they each
4490 wait for each other, leading to a delay before one of them times out.
4492 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4493 maintained 0.8.x compatibility because I don't think many are using it, and
4494 it is clearly obsolete.
4496 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4499 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4500 client certificate was expired. A simple patch fixes this, though I don't
4501 understand the full logic of why the verify callback is called multiple
4504 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4505 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4506 which causes problems with some clients (such as the Certicom SSL Plus
4507 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4508 disables the coutermeasure allowing Eudora to connect."
4510 38. Exim was not checking that a write() to a log file succeeded. This could
4511 lead to Bad Things if a log got too big, in particular if it hit a file
4512 size limit. Exim now panics and dies if it cannot write to a log file, just
4513 as it does if it cannot open a log file.
4515 39. Modified OS/Makefile-Linux so that it now contains
4517 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4519 The two -D definitions ensure that Exim is compiled with large file
4520 support, which makes it possible to handle log files that are bigger than
4523 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4524 instance) a domain was checked against a named list that involved a lookup,
4525 causing $domain_data to be set, then another domain was checked against the
4526 same list, then the first domain was re-checked, the value of $domain_data
4527 after the final check could be wrong. In particular, if the second check
4528 failed, it could be set empty. This bug probably also applied to
4531 41. The strip_trailing_dot option was not being applied to the address given
4532 with the -f command-line option.
4534 42. The code for reading a message's header from the spool was incrementing
4535 $received_count, but never initializing it. This meant that the value was
4536 incorrect (doubled) while delivering a message in the same process in which
4537 it was received. In the most common configuration of Exim, this never
4538 happens - a fresh exec is done - but it can happen when
4539 deliver_drop_privilege is set.
4541 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4542 it now includes up to 150 characters of the unexpected data in the log
4545 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4546 and building data strings. The size of both of these buffers was 10 000
4547 bytes - far larger than anybody would *ever* want, thought I. Needless to
4548 say, somebody hit the limit. I have increased the maximum line length to
4549 20 000 and the maximum data length of concatenated lines to 100 000. I have
4550 also fixed two bugs, because there was no checking on these buffers. Tsk,
4551 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4552 buffer is too small.
4554 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4555 lsearch lookups. Now it does.
4557 46. When parsing a route_list item in a manualroute router, a fixed-length
4558 buffer was used for the list of hosts. I made this 1024 bytes long,
4559 thinking that nobody would ever have a list of hosts that long. Wrong.
4560 Somebody had a whole pile of complicated expansion conditions, and the
4561 string was silently truncated, leading to an expansion error. It turns out
4562 that it is easier to change to an unlimited length (owing to other changes
4563 that have happened since this code was originally written) than to build
4564 structure for giving a limitation error. The length of the item that
4565 expands into the list of hosts is now unlimited.
4567 47. The lsearch lookup could not handle data where the length of text line was
4568 more than 4095 characters. Such lines were truncated, leading to shortened
4569 data being returned. It should now handle lines of any length.
4571 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4572 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4575 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4576 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4577 no longer happens in comment lines. A list of replacements is now placed
4578 at the head of all of the source files, except those whose only change is
4579 to replace PERL_COMMAND in the very first #! line.
4581 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4582 messages on the queue, with a bottom-up merge sort, using code contributed
4583 by Michael Haardt. This should make operations like -bp somewhat faster on
4584 large queues. It won't affect queue runners, except when queue_run_in_order
4587 51. Installed eximstats 1.31 in the distribution.
4589 52. Added support for SRV lookups to the dnslookup router.
4591 53. If an ACL referred to $message_body or $message_body_end, the value was not
4592 reset for any messages that followed in the same SMTP session.
4594 54. The store-handling optimization for building very long strings was not
4595 differentiating between the different store pools. I don't think this
4596 actually made any difference in practice, but I've tidied it.
4598 55. While running the routers to verify a sender address, $sender_address
4599 was still set to the sender address. This is wrong, because when routing to
4600 send a bounce to the sender, it would be empty. Therefore, I have changed
4601 it so that, while verifying a sender address, $sender_address is set to <>.
4602 (There is no change to what happens when verifying a recipient address.)
4604 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4605 target A or AAAA records (if not already returned) without resetting the
4606 qualify_single or search_parents options of the DNS resolver. These are
4607 inappropriate in this case because the targets of MX and SRV records must
4608 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4609 target that, when qualified, matched something in the local domain. These
4610 two options are now turned off when doing these lookups.
4612 57. It seems that at least some releases of Reiserfs (which does not have the
4613 concept of a fixed number of inodes) returns zero and not -1 for the
4614 number of available inodes. This interacted badly with check_spool_inodes,
4615 which assumed that -1 was the "no such thing" setting. What I have done is
4616 to check that the total number of inodes is greater than zero before doing
4617 the test of how many are available.
4619 58. When a "warn" ACL statement has a log_message modifier, the message is
4620 remembered, and not repeated. This is to avoid a lot of repetition when a
4621 message has many recipients that cause the same warning to be written.
4622 Howewer, Exim was preserving the list of already written lines for an
4623 entire SMTP session, which doesn't seem right. The memory is now reset if a
4624 new message is started.
4626 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4627 debugging output unless log_rewrite was also set.
4629 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4630 of (int)(handle) when we know that handle contains (void *)(-1).
4632 61. The Exim daemon panic-logs an error return when it closes the incoming
4633 connection. However "connection reset by peer" seems to be common, and
4634 isn't really an error worthy of noting specially, so that particular error
4637 62. When Exim is trying to find all the local interfaces, it used to panic and
4638 die if the ioctl to get the interface flags failed. However, it seems that
4639 on at least one OS (Solaris 9) it is possible to have an interface that is
4640 included in the list of interfaces, but for which you get a failure error
4641 for this call. This happens when the interface is not "plumbed" into a
4642 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4643 failure of the "get flags" call assumes that the interface is down.
4645 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4646 makes life easier for people who are doing arithmetic on fields extracted
4647 from dates, where you often get leading zeros that should not be
4648 interpreted as octal.
4650 64. Added qualify_domain to the redirect router, to override the global
4653 65. If a pathologically long header line contained very many addresses (the
4654 report of this problem mentioned 10 000) and each of them was rewritten,
4655 Exim could use up a very large amount of memory. (It kept on making new
4656 copies of the header line as it rewrote, and never released the old ones.)
4657 At the expense of a bit more processing, the header rewriting function has
4658 been changed so that it no longer eats memory in this way.
4660 66. The generation of the Received: header has been moved from the time that a
4661 message starts to be received, to the time that it finishes. The timestamp
4662 in the Received: header should now be very close to that of the <= log
4663 line. There are two side-effects of this change:
4665 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4666 logged header lines no longer include the local Received: line, because
4667 it has not yet been created. The same applies to a copy of the message
4668 that is returned to a non-SMTP sender when a message is rejected.
4670 (b) When a filter file is tested using -bf, no additional Received: header
4671 is added to the test message. After some thought, I decided that this
4674 This change does not affect the value of $received_for. It is still set
4675 after address rewriting, but before local_scan() is called.
4677 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4679 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4680 gave an unhelpful panic error message, and a defer error. I have managed to
4681 change this behaviour so that it now rejects any supplied certificate,
4682 which seems right, as the list of acceptable certificates is empty.
4684 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4685 gave an unhelpful defer error. I have not managed to make this reject any
4686 supplied certificates, but the error message it gives is "no certificate
4687 supplied", which is not helpful.
4689 70. exigrep's output now also includes lines that are not associated with any
4690 message, but which match the given pattern. Implemented by a patch from
4691 Martin Sluka, which also tidied up the Perl a bit.
4693 71. Recipient callout verification, like sender verification, was using <> in
4694 the MAIL FROM command. This isn't really the right thing, since the actual
4695 sender may affect whether the remote host accepts the recipient or not. I
4696 have changed it to use the actual sender in the callout; this means that
4697 the cache record is now keyed on a recipient/sender pair, not just the
4698 recipient address. There doesn't seem to be a real danger of callout loops,
4699 since a callout by the remote host to check the sender would use <>.
4700 [SEE ABOVE: changed after hitting problems.]
4702 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4703 temporary errors. However, in the case of such a code being given after
4704 the end of a data transmission (i.e. after ".") Exim was failing to write
4705 a retry record for the message. (Yes, there was some broken host that was
4706 actually sending 8xx at this point.)
4708 73. An unknown lookup type in a host list could cause Exim to panic-die when
4709 the list was checked. (An example that provoked this was putting <; in the
4710 middle of a list instead of at the start.) If this happened during a DATA
4711 ACL check, a -D file could be left lying around. This kind of configuration
4712 error no longer causes Exim to die; instead it causes a defer errror. The
4713 incident is still logged to the main and panic logs.
4715 74. Buglet left over from Exim 3 conversion. The message "too many messages
4716 in one connection" was written to the rejectlog but not the mainlog, except
4717 when address rewriting (yes!) was being logged.
4719 75. Added write_rejectlog option.
4721 76. When a system filter was run not as root (that is, when system_filter_user
4722 was set), the values of the $n variables were not being returned to the
4723 main process; thus, they were not subsequently available in the $sn
4726 77. Added +return_path_on_delivery log selector.
4728 78. A connection timeout was being treated differently from recipients deferred
4729 when testing hosts_max_try with a message that was older than the host's
4730 retry timeout. (The host should not be counted, thus allowing all hosts to
4731 be tried at least once before bouncing.) This may have been the cause of an
4732 occasionally reported bug whereby a message would remain on the queue
4733 longer than the retry timeout, but would be bounced if a delivery was
4734 forced. I say "may" because I never totally pinned down the problem;
4735 setting up timeout/retry tests is difficult. See also the next item.
4737 79. The ultimate address timeout was not being applied to errors that involved
4738 a combination of host plus message (for example, a timeout on a MAIL
4739 command). When an address resolved to a number of possible hosts, and they
4740 were not all tried for each delivery (e.g. because of hosts_max_try), a
4741 message could remain on the queue longer than the retry timeout.
4743 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4746 81. Fixed an obscure SMTP outgoing bug which required at least the following
4747 conditions: (a) there was another message waiting for the same server;
4748 (b) the server returned 5xx to all RCPT commands in the first message so
4749 that the message was not completed; (c) the server dropped the connection
4750 or gave a negative response to the RSET that Exim sends to abort the
4751 transaction. The observed case was a dropped connection after DATA that had
4752 been sent in pipelining mode. That is, the server had advertised PIPELINING
4753 but was not implementing it correctly. The effect of the bug was incorrect
4754 behaviour, such as trying another host, and this could lead to a crash.
4760 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4761 and daemon.c were passed as pointers to ints; they should have been
4762 pointers to socklen_t variables (which are typically unsigned ints).
4764 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4767 3. Fixed a really odd bug that affected only the testing scheme; patching a
4768 certain fixed string in the binary changed the value of another string that
4769 happened to be identical to the end of the original first string.
4771 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4772 name", it returns that address as the IP address. On some operating
4773 systems (e.g. Solaris), it also passes back the IP address string as the
4774 "host name". However, on others (e.g. Linux), it passes back an empty
4775 string. Exim wasn't checking for this, and was changing the host name to an
4776 empty string, assuming it had been canonicized.
4778 5. Although rare, it is permitted to have more than one PTR record for a given
4779 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4780 all the names associated with an address, because they do in Solaris.
4781 However, it seems that they do not in Linux for data that comes from the
4782 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4783 I found this out when I moved to a new Linux workstation and tried to run
4784 the Exim test suite.
4786 To get round this problem I have changed the code so that it now does its
4787 own call to the DNS to look up PTR records when searching for a host name.
4788 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4789 addresses that are only in /etc/hosts are still found.
4791 This behaviour is, however, controlled by an option called host_lookup_
4792 order, which defaults to "bydns:byaddr". If people want to use the other
4793 order, or indeed, just use one or the other means of lookup, they can
4794 specify it in this variable.
4796 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4797 some operating systems, this comes back from gethostbyaddr() as an empty
4798 string, and this is what Exim used to test for. However, it seems that in
4799 other systems, "." is yielded. Exim now tests for this case too.
4801 7. The values of check_spool_space and check_log_space are now held internally
4802 as a number of kilobytes instead of an absolute number of bytes. If a
4803 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4804 kilobyte. This means that much larger values can be stored.
4806 8. Exim monitor: an attempt to get the action menu when not actually pointing
4807 at a message produces an empty menu entitled "No message selected". This
4808 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4809 no entries in it ("Shell widget menu has zero width and/or height"). So I
4810 have added a single, blank menu entry in this case.
4812 9. Added ${quote_local_part.
4814 10. MIME decoding is now applied to the contents of Subject: header lines when
4817 11. Now that a reference to $sender_host_address automatically causes a reverse
4818 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4819 host lookup before query-style lookups in lists that might use this
4820 variable. This has therefore been abolished, and the "net-" prefix is no
4821 longer necessary for query-style lookups.
4823 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4824 have been a typo for LFLAGS, so it has been changed.
4826 13. The install script calls Exim with "-C /dev/null" in order to find the
4827 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4828 to be output. Howeve, since Exim outputs its version number before the
4829 error, it didn't break the script. It just looked ugly. I fixed this by
4830 always allowing "-C /dev/null" if the caller is root.
4832 14. Ignore overlarge ACL variable number when reading spool file - insurance
4833 against a later release with more variables having written the file.
4835 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4836 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4837 this, as well as the form without the "IPv6" on the front.
4839 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4840 OS/Makefile-Darwin file.
4842 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4843 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4845 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4846 variable. However, some OS use ints for inodes. I've added cast to long int
4847 to get rid of the compiler warning.
4849 19. I had forgotten to lock out "/../" in configuration file names when
4850 ALT_CONFIG_PREFIX was set.
4852 20. Routers used for verification do not need to specify transports. However,
4853 if such a router generated a host list, and callout was configured, Exim
4854 crashed, because it could not find a port number from the (non-existent)
4855 transport. It now assumes port 25 in this circumstance.
4857 21. Added the -t option to exigrep.
4859 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4860 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4861 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4862 them is set, LOOKUP_LSEARCH is forced.
4864 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4866 24. Added sender and host information to the "rejected by local_scan()" log
4867 line; previously there was no indication of these.
4869 25. Added .include_if_exists.
4871 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4872 Linux. It turns out that not all file systems support this. Apparently some
4873 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4874 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4875 not supported on the file descriptor, is now ignored when Exim is trying to
4876 sync a directory. This applies only to Linux.
4878 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4880 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4881 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4884 29. Added support for maildirsize files from supplied patch (modified a bit).
4886 30. The use of :fail: followed by an empty string could lead Exim to respond to
4887 sender verification failures with (e.g.):
4889 550 Verification failed for <xxx>
4890 550 Sender verify failed
4892 where the first response line was missing the '-' that indicates it is not
4893 the final line of the response.
4895 31. The loop for finding the name of the user that called Exim had a hardwired
4896 limit of 10; it now uses the value of finduser_retries, which is used for
4897 all other user lookups.
4899 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4902 33. Exim was neglecting to zero errno before one call of strtol() when
4903 expanding a string and expecting an integer value. On some systems this
4904 resulted in spurious "integer overflow" errors. Also, it was casting the
4905 result into an int without checking.
4907 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4908 did not work. The code looks as if it has *never* worked, though it appears
4909 to have been documented since at least releast 1.62. I have made it work.
4911 35. The "timeout_DNS" error in retry rules, also documented since at least
4912 1.62, also never worked. As it isn't clear exactly what this means, and
4913 clearly it isn't a major issue, I have abolished the feature by treating it
4914 as "timeout", and writing a warning to the main and panic logs.
4916 36. The display of retry rules for -brt wasn't always showing the error code
4919 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4920 timeout_connect_A, timeout_connect_MX.
4922 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4923 to the empty sender.
4925 39. The daemon was not analysing the content of -oX till after it had closed
4926 stderr and disconnected from the controlling terminal. This meant that any
4927 syntax errors were only noted on the panic log, and the return code from
4928 the command was 0. By re-arranging the code a little, I've made the
4929 decoding happen first, so such errors now appear on stderr, and the return
4930 code is 1. However, the actual setting up of the sockets still happens in
4931 the disconnected process, so errors there are still only recorded on the
4934 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4935 connections (as happens on some IP stacks) was logged at start up time as
4936 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4937 it from "IPv6 and IPv4", which means that two separate sockets are being
4940 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4941 says whether AF_INET or AF_INET6 was passed as an argument.
4943 42. Exiwhat output was messed up when time zones were included in log
4946 43. Exiwhat now gives more information about the daemon's listening ports,
4947 and whether -tls-on-connect was used.
4949 44. The "port" option of the smtp transport is now expanded.
4951 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4952 silently ignored. Now an error message is written to the main and panic
4955 46. There's a new ACL modifier called "logwrite" which writes to a log file
4956 as soon as it is encountered.
4958 47. Added $local_user_uid and $local_user_gid at routing time.
4960 48. Exim crashed when trying to verify a sender address that was being
4963 49. Exim was recognizing only a space character after ".include". It now also
4964 recognizes a tab character.
4966 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4967 extracting the relevant information from the specification. The man page no
4968 longer contains scrambled data for the -d option, and I've added a section
4969 at the front about calling Exim under different names.
4971 51. Added "extra_headers" argument to the "mail" command in filter files.
4973 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4976 53. Installed eximstats 1.29.
4978 54. Added transport_filter_timeout as a generic transport option.
4980 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4981 Cc: header lines. This was required by RFC 822, but it not required by RFC
4984 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4985 incoming messages that did not have them. Now it does so only if the
4986 message originates locally, that is, if there is no associated remote host
4987 address. When Resent- header lines are present, this applies to the Resent-
4988 lines rather than the non-Resent- lines.
4990 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4991 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4993 58. Messages for configuration errors now include the name of the main
4994 configuration file - useful now that there may be more than one file in a
4995 list (.included file names were always shown).
4997 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4998 for those rare installations that do not start the daemon as root or run it
4999 setuid root. I've cut out the call to initgroups() if the daemon is not
5002 60. The Exim user and group can now be bound into the binary as text strings
5003 that are looked up at the start of Exim's processing.
5005 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5007 62. Added $mailstore_basename variable.
5009 63. Installed patch to sieve.c from Michael Haardt.
5011 64. When Exim failed to open the panic log after failing to open the main log,
5012 the original message it was trying to log was written to stderr and debug
5013 output, but if they were not available (the usual case in production), it
5014 was lost. Now it is written to syslog before the two lines that record the
5015 failures to open the logs.
5017 65. Users' Exim filters run in subprocesses under the user's uid. It is
5018 possible for a "deliver" command or an alias in a "personal" command to
5019 provoke an address rewrite. If logging of address rewriting is configured,
5020 this fails because the process is not running as root or exim. There may be
5021 a better way of dealing with this, but for the moment (because 4.30 needs
5022 to be released), I have disabled address rewrite logging when running a
5023 filter in a non-root, non-exim process.
5029 1. The buildconfig auxiliary program wasn't quoting the value set for
5030 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5031 not defined. This bug was masked in 4.22 by the effect that was fixed in
5034 2. Some messages that were rejected after a message id was allocated were
5035 shown as "incomplete" by exigrep. It no longer does this for messages that
5036 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5038 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5039 have allow_domain_literals set, the ID did not get logged in the <= line.
5040 Domain literals are now always recognized in Message-ID: header lines.
5042 4. The first argument for a ${extract expansion item is the key name or field
5043 number. Leading and trailing spaces in this item were not being ignored,
5044 causing some misleading effects.
5046 5. When deliver_drop_privilege was set, single queue runner processes started
5047 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5048 same command in the process it spins off) were not dropping privilege.
5050 6. When the daemon running as "exim" started a queue runner, it always
5051 re-executed Exim in the spun-off process. This is a waste of effort when
5052 deliver_drop_privilege is set. The new process now just calls the
5053 queue-runner function directly.
5059 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5062 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5063 ignored. Though the use of -oP was forcing the writing of a pid file, it
5064 was always written to the default place.
5066 3. If the message "no IP address found for host xxxx" is generated during
5067 incoming verification, it is now followed by identification of the incoming
5068 connection (so you can more easily find what provoked it).
5070 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5072 5. Added some features to "harden" Exim a bit more against certain attacks:
5074 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5075 be put in Local/Makefile. This is like the never_users runtime option,
5076 but it cannot be overridden. The default setting is "root".
5078 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5079 prefix string with which any file named in a -C command line option
5082 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5083 is retained for -C and -D only if the caller of Exim is root. Without
5084 it, the exim user may also use -C and -D and retain privilege.
5086 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5087 command line option is disabled.
5089 6. Macro names set by the -D option must start with an upper case letter, just
5090 like macro names defined in the configuration file.
5092 7. Added "dereference=" facility to LDAP.
5094 8. Two instances of the typo "uknown" in the source files are fixed.
5096 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5097 the Configure-Makefile script screwed up while processing it.
5099 10. Incorporated PCRE 4.4.
5101 11. The SMTP synchronization check was not operating right at the start of an
5102 SMTP session. For example, it could not catch a HELO sent before the client
5103 waited for the greeting. There is now a check for outstanding input at the
5104 point when the greeting is written. Because of the duplex, asynchronous
5105 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5106 way, but not yet received, when the check is performed.
5108 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5109 on TCP/IP sockets, because this apparently causes some broken clients to
5112 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5113 unchanged) from the Cygwin maintainer.
5115 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5116 was supported instead of testing for mbx. Effectively a typo.
5118 15. The spa authenticator server code was not checking that the input it
5119 received was valid base64.
5121 16. The debug output line for the "set" modifier in ACLs was not showing the
5122 name of the variable that was being set.
5124 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5126 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5127 lookup on its own. Something else was needed to trigger the lookup. For
5128 example, a match in host_lookup or the need for a host name in a host list.
5129 Now, if $sender_host_name is referenced and the host name has not yet been
5130 looked up, a lookup is performed. If the lookup fails, the variable remains
5131 empty, and $host_lookup_failed is set to "1".
5133 19. Added "eqi" as a case-independent comparison operator.
5135 20. The saslauthd authentication condition could segfault if neither service
5136 nor realm was specified.
5138 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5139 error message that was logged was misleading, and incoming SMTP
5140 connections were dropped. The message is now more accurate, and temporary
5141 errors are given to SMTP connections.
5143 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5144 (except RSET) with a 421 error, until QUIT is received. However, it was
5145 failing to send a response to QUIT.
5147 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5148 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5149 ACL is now run after the helo_try_verify_hosts code.
5151 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5152 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5153 case-independent, so other case variants are also recognized). Apparently
5154 some systems use these upper case variants.
5156 25. If more than two messages were waiting for the same host, and a transport
5157 filter was specified for the transport, Exim sent two messages over the
5158 same TCP/IP connection, and then failed with "socket operation on non-
5159 socket" when it tried to send the third.
5161 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5163 27. The extern definition of crypt16() in expand.c was not being excluded when
5164 the OS had its own crypt16() function.
5166 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5167 as a preferred synonym for return_size_limit, both as an option and as an
5170 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5172 30. Changed the default configuration ACL to relax the local part checking rule
5173 for addresses that are not in any local domains. For these addresses,
5174 slashes and pipe symbols are allowed within local parts, but the sequence
5175 /../ is explicitly forbidden.
5177 31. SPA server authentication was not clearing the challenge buffer before
5180 32. log_message in a "warn" ACL statement was writing to the reject log as
5181 well as to the main log, which contradicts the documentation and doesn't
5182 seem right (because no rejection is happening). So I have stopped it.
5184 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5185 However, I am unable to do any testing of this.
5187 34. Fixed an infelicity in the appendfile transport. When checking directories
5188 for a mailbox, to see if any needed to be created, it was accidentally
5189 using path names with one or more superfluous leading slashes; tracing
5190 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5192 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5193 contents are added to the log line that is written for every discarded
5194 recipient. (Previously a log_message setting was ignored.)
5196 36. The ${quote: operator now quotes the string if it is empty.
5198 37. The install script runs exim in order to find its version number. If for
5199 some reason other than non-existence or emptiness, which it checks, it
5200 could not run './exim', it was installing it with an empty version number,
5201 i.e. as "exim-". This error state is now caught, and the installation is
5204 38. An argument was missing from the function that creates an error message
5205 when Exim fails to connect to the socket for saslauthd authentication.
5206 This could cause Exim to crash, or give a corrupted message.
5208 39. Added isip, isip4, and isip6 to ${if conditions.
5210 40. The ACL variables $acl_xx are now saved with the message, and can be
5211 accessed later in routers, transports, and filters.
5213 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5214 strings in the file are not string-expanded.
5216 42. If a MAIL command specified a SIZE value that was too large to fit into an
5217 int variable, the check against message_size_limit failed. Such values are
5218 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5219 day this will have to be increased, but I don't think I want to be around
5220 when emails are that large.
5227 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5228 iconv() is not standard in FreeBSD.
5230 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5231 IPv6 enabled. The observed symptom was a segmentation fault on return from
5232 the function os_common_find_running_interfaces() in src/os.c.
5234 3. In the check_special_case() function in daemon.c I had used "errno" as an
5235 argument name, which causes warnings on some systems. This was basically a
5236 typo, since it was named "eno" in the comments!
5238 4. The code that waits for the clock to tick (at a resolution of some fraction
5239 of a second) so as to ensure message-id uniqueness was always waiting for
5240 at least one whole tick, when it could have waited for less. [This is
5241 almost certainly not relevant at current processor speeds, where it is
5242 unlikely to ever wait at all. But we try to future-proof.]
5244 5. The function that sleeps for a time interval that includes fractions of a
5245 second contained a race. It did not block SIGALRM between setting the
5246 timer, and suspending (a couple of lines later). If the interval was short
5247 and the sigsuspend() was delayed until after it had expired, the suspension
5248 never ended. On busy systems this could lead to processes getting stuck for
5251 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5252 process, before it forks any delivery processes. The open lookup caching
5253 mechanism meant that the open file or database connection was passed into
5254 the delivery process. The problem was that delivery processes always tidy
5255 up cached lookup data. This could cause a problem for the next delivery
5256 process started by the queue runner, because the external queue runner
5257 process does not know about the closure. So the next delivery process
5258 still has data in the lookup cache. In the case of a file lookup, there was
5259 no problem because closing a file descriptor in a subprocess doesn't affect
5260 the parent. However, if the lookup was caching a connection to a database,
5261 the connection was closed, and the second delivery process was likely to
5262 see errors such as "PGSQL: query failed: server closed the connection
5263 unexpectedly". The problem has been fixed by closing all cached lookups
5264 in a queue runner before running a delivery process.
5266 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5267 seem to have the "const" qualifier which it has on other OS. I've
5270 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5271 *of the same type* (client or server) with the same public name that an
5272 error should be diagnosed.
5274 9. When Exim looked up a host name for an IP address, but failed to find the
5275 original IP address when looking up the host name (a safety check), it
5276 output the message "<ip address> does not match any IP for NULL", which was
5277 confusing, to say the least. The bug was that the host name should have
5278 appeared instead of "NULL".
5280 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5281 user that is built into the binary, and the -C or -D options is used, root
5282 privilege is dropped before the configuration file is read. In addition,
5283 logging is switched to stderr instead of the normal log files. If the
5284 configuration then re-defines the Exim user, the unprivileged environment
5285 is probably not what is expected, so Exim logs a panic warning message (but
5288 However, if deliver_drop_privilege is set, the unprivileged state may well
5289 be exactly what is intended, so the warning has been cut out in that case,
5290 and Exim is allowed to try to write to its normal log files.
5296 1. smtp_return_error_details was not giving details for temporary sender
5297 or receiver verification errors.
5299 2. Diagnose a configuration error if two authenticators have the same public
5302 3. Exim used not to create the message log file for a message until the first
5303 delivery attempt. This could be confusing when incoming messages were held
5304 for policy or load reasons. The message log file is now created at the time
5305 the message is received, and an initial "Received" line is written to it.
5307 4. The automatically generated man page for command line options had a minor
5308 bug that caused no ill effects; however, a more serious problem was that
5309 the procedure for building the man page automatically didn't always
5310 operate. Consequently, release 4.20 contains an out-of-date version. This
5311 shouldn't happen again.
5313 5. When building Exim with embedded Perl support, the script that builds the
5314 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5315 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5317 6. The freeze_tell option was not being used for messages that were frozen on
5318 arrival, either by an ACL or by local_scan().
5320 7. Added the smtp_incomplete_transaction log selector.
5322 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5323 was accepting AUTH without a new EHLO.
5325 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5326 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5329 10. Logging of TCP/IP connections (when configured) now happens in the main
5330 daemon process instead of the child process, so that the TCP/IP connection
5331 count is more accurate (but it can never be perfect).
5333 11. The use of "drop" in a nested ACL was not being handled correctly in the
5334 outer ACL. Now, if condition failure induced by the nested "drop" causes
5335 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5336 or "require"), the connection is dropped.
5338 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5339 that yield "discard" can now be used with an "accept" or a "discard" verb,
5340 but an error is generated for any others (because I can't see a useful way
5341 to define what should happen).
5343 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5344 are now processed in the same way as lines in the Exim configuration file.
5345 In particular, continuation lines are supported.
5347 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5349 15. Added -ti meaning -t -i.
5351 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5352 domains, and warn by logging if others are found.
5354 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5355 returned from GIFCONF when Exim is trying to find the list of interfaces on
5356 a host. The code in os.c has been modified to copy each ifreq to an aligned
5357 structure in all cases.
5359 Also, in some cases, the returned ifreq's were being copied to a 'struct
5360 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5361 means the last couple of bytes of an IPv6 address could be chopped if the
5362 ifreq contained only a normal sockaddr (14 bytes storage).
5364 18. Named domain lists were not supported in the hosts_treat_as_local option.
5365 An entry such as +xxxx was not recognized, and was treated as a literal
5368 19. Ensure that header lines added by a DATA ACL are included in the reject log
5369 if the ACL subsequently rejects the message.
5371 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5372 MD5 (which is deprecated).
5374 21. When testing a filter file using -bf, Exim was writing a message when it
5375 took the sender from a "From " line in the message, but it was not doing so
5376 when it took $return_path from a Return-Path: header line. It now does.
5378 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5379 with a valid header line field name (a series of printing characters
5380 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5382 23. Changed "disc" in the source to "disk" to conform to the documentation and
5383 the book and for uniformity.
5385 24. Ignore Sendmail's -Ooption=value command line item.
5387 25. When execve() failed while trying to run a command in a pipe transport,
5388 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5389 could be confused with a return value of 69 from the command itself. This
5390 has been changed to 127, the value the shell returns if it is asked to run
5391 a non-existent command. The wording for the related log line suggests a
5392 non-existent command as the problem.
5394 26. If received_header_text expands to an empty string, do not add a Received:
5395 header line to the message. (Well, it adds a token one on the spool, but
5396 marks it "old" so that it doesn't get used or transmitted.)
5398 27. Installed eximstats 1.28 (addition of -nt option).
5400 28. There was no check for failure on the call to getsockname() in the daemon
5401 code. This can fail if there is a shortage of resources on the system, with
5402 ENOMEM, for example. A temporary error is now given on failure.
5404 29. Contrary to the C standard, it seems that in some environments, the
5405 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5406 program. Exim now does this explicitly; it affects the formatting of
5407 timestamps using strftime().
5409 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5410 complaints. I've now initialized all the variables, to avoid this.
5412 32. Header lines added by a system filter were not being "seen" during
5413 transport-time rewrites.
5415 33. The info_callback() function passed to OpenSSL is set up with type void
5416 (*)(SSL *, int, int), as described somewhere. However, when calling the
5417 function (actually a macro) that sets it up, the type void(*)() is
5418 expected. I've put in a cast to prevent warnings from picky compilers.
5420 34. If a DNS black list lookup found a CNAME record, but there were no A
5421 records associated with the domain it pointed at, Exim crashed.
5423 35. If a DNS black list lookup returned more than one A record, Exim ignored
5424 all but the first. It now scans all returned addresses if a particular IP
5425 value is being sought. In this situation, the contents of the
5426 $dnslist_value variable are a list of all the addresses, separated by a
5429 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5430 to accept a host name and all its aliases if the forward lookup for any of
5431 them yielded the IP address of the incoming connection. Now it accepts only
5432 those names whose forward lookup yields the correct IP address. Any other
5433 names are discarded. This closes a loophole whereby a rogue DNS
5434 administrator could create reverse DNS records to break through a
5435 wildcarded host restriction in an ACL.
5437 37. If a user filter or a system filter that ran in a subprocess used any of
5438 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5439 the wrong values were passed to the pipe command ($thisaddress had the
5440 value of $0, $0 had the value of $1, etc). This bug was introduced by
5441 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5443 38. Improved the line breaking for long SMTP error messages from ACLs.
5444 Previously, if there was no break point between 40 and 75 characters, Exim
5445 left the rest of the message alone. Two changes have been made: (a) I've
5446 reduced the minimum length to 35 characters; (b) if it can't find a break
5447 point between 35 and 75 characters, it looks ahead and uses the first one
5448 that it finds. This may give the occasional overlong line, but at least the
5449 remaining text gets split now.
5451 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5452 file descriptors might be low, and that setting 1000 would always raise it.
5453 It turns out that in some environments, the limit is already over 1000 and
5454 that lowering it causes trouble. So now Exim takes care not to decrease it.
5456 40. When delivering a message, the value of $return_path is set to $sender_
5457 address at the start of routing (routers may change the value). By an
5458 oversight, this default was not being set up when an address was tested by
5459 -bt or -bv, which affected the outcome if any router or filter referred to
5462 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5463 log only when "message" or "log_message" are set. However, if one of the
5464 conditions was an address verification, or a call to a nested ACL, the
5465 messages generated by the underlying test were being passed through. This
5466 no longer happens. The underlying message is available in $acl_verify_
5467 message for both "message" and "log_message" expansions, so it can be
5468 passed through if needed.
5470 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5471 new expansion $bh_ to give the encoded byte string without charset
5472 translation. Translation happens only if iconv() is available; HAVE_ICONV
5473 indicates this at build time. HEADERS_CHARSET gives the charset to
5474 translate to; headers_charset can change it in the configuration, and
5475 "headers charset" can change it in an individual filter file.
5477 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5478 that creates RFC 2047 encoded "words" labels them as that charset instead
5479 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5480 expansion operator; (ii) when Exim creates a From: line for a local
5481 message; (iii) when a header line is rewritten to include a "phrase" part.
5483 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5484 buggy, causing it to skip the first lines of messages whose message ID
5485 ended in 'D'. This would not have bitten before Exim release 4.14, because
5486 message IDs were unlikely to end in 'D' before then. The effect was to have
5487 incorrect size information for certain domains.
5489 45. #include "config.h" was missing at the start of the crypt16.c module. This
5490 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5493 46. If there was a timeout during a "random" callout check, Exim treated it as
5494 a failure of the random address, and carried on sending RSET and the real
5495 address. If the delay was just some slowness somewhere, the response to the
5496 original RCPT would be taken as a response to RSET and so on, causing
5497 mayhem of various kinds.
5499 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5500 when I implemented it. It didn't allow for the fact that some option values
5501 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5502 the right test for positive values.
5504 48. Domain names in DNS records are case-independent. Exim always looks them up
5505 in lower case. Some resolvers return domain names in exactly the case they
5506 appear in the zone file, that is, they may contain uppercase letters. Not
5507 all resolvers do this - some return always lower case. Exim was treating a
5508 change of case by a resolver as a change of domain, similar to a widening
5509 of a domain abbreviation. This triggered its re-routing code and so it was
5510 trying to route what was effectively the same domain again. This normally
5511 caused routing to fail (because the router wouldn't handle the domain
5512 twice). Now Exim checks for this case specially, and just changes the
5513 casing of the domain that it ultimately uses when it transmits the message
5516 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5519 50. If a filter generated a file delivery with a non-absolute name (possible if
5520 no home directory exists for the router), the forbid_file option was not
5523 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5524 the existing equality matching.
5526 52. Exim was using ints instead of ino_t variables in some places where it was
5527 dealing with inode numbers.
5529 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5530 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5531 called TMPDIR, and if it finds it is different, it changes its value.
5533 54. The smtp_printf() function is now made available to local_scan() so
5534 additional output lines can be written before returning. There is also an
5535 smtp_fflush() function to enable the detection of a dropped connection.
5536 The variables smtp_input and smtp_batched_input are exported to
5539 55. Changed the default runtime configuration: the message "Unknown user"
5540 has been removed from the ACL, and instead placed on the localuser router,
5541 using the cannot_route_message feature. This means that any verification
5542 failures that generate their own messages won't get overridden. Similarly,
5543 the "Unrouteable address" message that was in the ACL for unverifiable
5544 relay addresses has also been removed.
5546 56. Added hosts_avoid_esmtp to the smtp transport.
5548 57. The exicyclog script was not checking for the esoteric option
5549 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5550 will work only if exicyclog is run under the appropriate euid.
5552 58. Following a discussion on the list, the rules by which Exim recognises line
5553 endings on incoming messages have been changed. The -dropcr and drop_cr
5554 options are now no-ops, retained only for backwards compatibility. The
5555 following line terminators are recognized: LF CRLF CR. However, special
5556 processing applies to CR:
5558 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5559 nor a local message in the state where . is a terminator.
5561 (ii) If a bare CR is encountered in a header line, an extra space is added
5562 after the line terminator so as not to end the header. The reasoning
5563 behind this is that bare CRs in header lines are most likely either
5564 to be mistakes, or people trying to play silly games.
5566 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5567 was being incorrectly given as 18 bytes larger than it should have been.
5568 This is a VOB (very old bug).
5570 60. This may never have affected anything current, but just in case it has:
5571 When the local host is found other than at the start of a list of hosts,
5572 the local host, those with the same MX, and any that follow, are discarded.
5573 When the list in question was part of a longer list of hosts, the following
5574 hosts (not currently being processed) were also being discarded. This no
5575 longer happens. I'm not sure if this situation could ever has previously
5578 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5581 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5582 header. This is something that is recommended in a new Internet Draft, and
5583 is something that is documented as being done by Sendmail. There are two
5584 possible values. For messages generated by the autoreply transport, Exim
5587 Auto-Submitted: auto-replied
5589 whereas for all other generated messages (e.g. bounces) it adds
5591 Auto-Submitted: auto-generated
5593 63. The "personal" condition in filters now includes a test for the
5594 Auto-Submitted: header. If it contains the string "auto-" the message it
5595 not considered personal.
5597 64. Added rcpt_include_affixes as a generic transport option.
5599 65. Added queue_only_override (default true).
5601 66. Added the syslog_duplication option.
5603 67. If what should have been the first header line of a message consisted of
5604 a space followed by a colon, Exim was mis-interpreting it as a header line.
5605 It isn't of course - it is syntactically invalid and should therefore be
5606 treated as the start of the message body. The misbehaviour could have
5607 caused a number of strange effects, including loss of data in subsequent
5608 header lines, and spool format errors.
5610 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5611 client host had authenticated. This control can now be exercised by an ACL
5612 for more flexibility.
5614 69. By default, callouts do not happen when testing with -bh. There is now a
5615 variant, -bhc, which does actually run the callout code, including
5616 consulting and updating the callout cache.
5618 70. Added support for saslauthd authentication, courtesy of Alexander
5621 71. If statvfs() failed on the spool or log directories while checking their
5622 size for availability, Exim confusingly gave the error "space shortage".
5623 Furthermore, in debugging mode it crashed with a floating point exception.
5624 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5625 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5626 serious problem, Exim now writes to the main and panic logs when this
5627 happens, with details of the failure. It then refuses to accept the
5628 incoming message, giving the message "spool directory problem" or "log
5629 directory problem" with a 421 code for SMTP messages.
5631 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5632 0, 1, and 2 exist, because some OS complain for execs without them (see
5633 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5634 descriptors. However, the code omitted to check that the open succeeded,
5635 causing mysterious errors if for some reason the permissions on /dev/null
5636 got screwed. Now Exim writes a message to the main and panic logs, and
5637 bombs out if it can't open /dev/null.
5639 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5640 interact so that it is all more flexible. It is supposed to remain
5641 backwards compatible. Also added extra_local_interfaces.
5643 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5644 to bomb out with an assertion failure - to the client this appears as a
5645 connection drop. This problem occurs in the part of the code that was taken
5646 from the Samba project. Fortunately, the assertion is in a very simple
5647 function, so I have fixed this by reproducing the function inline in the
5648 one place where it is called, and arranging for authentication to fail
5649 instead of killing the process with assert().
5651 75. The SPA client code was not working when the server requested OEM rather
5652 than Unicode encoding.
5654 76. Added code to make require_files with a specific uid setting more usable in
5655 the case where statting the file as root fails - usually a non-root-mounted
5656 NFS file system. When this happens and the failure is EACCES, Exim now
5657 forks a subprocess and does the per-uid checking as the relevant uid.
5659 77. Added process_log_path.
5661 78. If log_file_path was not explicitly set, a setting of check_log_space or
5662 check_log_inodes was ignored.
5664 79. If a space check for the spool or log partitions fails, the incident is now
5665 logged. Of course, in the latter case the data may get lost...
5667 80. Added the %p formatting code to string_format() so that it can be used to
5668 print addresses in debug_print(). Adjusted all the address printing in the
5669 debugging in store.c to use %p rather than %d.
5671 81. There was a concern that a line of code in smtp_in.c could overflow a
5672 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5673 initially expressed, the concern was not well-founded, because trailing
5674 spaces are removed early. However, if the trailing spaces were followed by
5675 a NULL, they did not get removed, so the overflow was possible. Two fixes
5678 (a) I re-wrote the offending code in a cleaner fashion.
5679 (b) If an incoming SMTP command contains a NULL character, it is rejected
5682 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5683 runs initgroups(), so that if the Exim user is in any additional groups,
5684 they will be used during message reception.
5690 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob